{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.881844380403458,
  "eval_steps": 3000,
  "global_step": 30000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.909158945083618,
      "epoch": 0.0004803073967339097,
      "grad_norm": 18.0,
      "learning_rate": 2e-06,
      "loss": 14.0493,
      "mean_token_accuracy": 0.0,
      "num_tokens": 10821.0,
      "step": 5
    },
    {
      "entropy": 4.897940301895142,
      "epoch": 0.0009606147934678194,
      "grad_norm": 23.0,
      "learning_rate": 4.5e-06,
      "loss": 14.0706,
      "mean_token_accuracy": 0.0,
      "num_tokens": 21838.0,
      "step": 10
    },
    {
      "entropy": 5.05525016784668,
      "epoch": 0.001440922190201729,
      "grad_norm": 26.25,
      "learning_rate": 7e-06,
      "loss": 13.7388,
      "mean_token_accuracy": 0.0,
      "num_tokens": 33483.0,
      "step": 15
    },
    {
      "entropy": 5.374046516418457,
      "epoch": 0.0019212295869356388,
      "grad_norm": 36.0,
      "learning_rate": 9.5e-06,
      "loss": 13.2353,
      "mean_token_accuracy": 0.0,
      "num_tokens": 45150.0,
      "step": 20
    },
    {
      "entropy": 6.746120595932007,
      "epoch": 0.0024015369836695487,
      "grad_norm": 26.25,
      "learning_rate": 1.2e-05,
      "loss": 12.2263,
      "mean_token_accuracy": 0.0,
      "num_tokens": 55862.0,
      "step": 25
    },
    {
      "entropy": 9.457833766937256,
      "epoch": 0.002881844380403458,
      "grad_norm": 4.59375,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 11.1521,
      "mean_token_accuracy": 0.0,
      "num_tokens": 66766.0,
      "step": 30
    },
    {
      "entropy": 10.62428560256958,
      "epoch": 0.0033621517771373678,
      "grad_norm": 2.71875,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.7428,
      "mean_token_accuracy": 0.0,
      "num_tokens": 77012.0,
      "step": 35
    },
    {
      "entropy": 10.70099697113037,
      "epoch": 0.0038424591738712775,
      "grad_norm": 2.46875,
      "learning_rate": 1.95e-05,
      "loss": 10.5653,
      "mean_token_accuracy": 0.0068274421093519775,
      "num_tokens": 89034.0,
      "step": 40
    },
    {
      "entropy": 10.700504493713378,
      "epoch": 0.004322766570605188,
      "grad_norm": 2.484375,
      "learning_rate": 2.2e-05,
      "loss": 10.264,
      "mean_token_accuracy": 0.03058484271168709,
      "num_tokens": 100341.0,
      "step": 45
    },
    {
      "entropy": 10.682637310028076,
      "epoch": 0.004803073967339097,
      "grad_norm": 2.28125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.9901,
      "mean_token_accuracy": 0.028485930524766445,
      "num_tokens": 111997.0,
      "step": 50
    },
    {
      "entropy": 10.652040386199952,
      "epoch": 0.005283381364073006,
      "grad_norm": 2.1875,
      "learning_rate": 2.7e-05,
      "loss": 9.8134,
      "mean_token_accuracy": 0.030870671570301055,
      "num_tokens": 123705.0,
      "step": 55
    },
    {
      "entropy": 10.639647006988525,
      "epoch": 0.005763688760806916,
      "grad_norm": 2.125,
      "learning_rate": 2.95e-05,
      "loss": 9.6274,
      "mean_token_accuracy": 0.029013283923268317,
      "num_tokens": 134091.0,
      "step": 60
    },
    {
      "entropy": 10.625792407989502,
      "epoch": 0.006243996157540826,
      "grad_norm": 1.953125,
      "learning_rate": 3.2e-05,
      "loss": 9.5799,
      "mean_token_accuracy": 0.035698648542165756,
      "num_tokens": 145731.0,
      "step": 65
    },
    {
      "entropy": 10.611747932434081,
      "epoch": 0.0067243035542747355,
      "grad_norm": 2.109375,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.4796,
      "mean_token_accuracy": 0.029607368633151054,
      "num_tokens": 156979.0,
      "step": 70
    },
    {
      "entropy": 10.60178451538086,
      "epoch": 0.007204610951008645,
      "grad_norm": 2.015625,
      "learning_rate": 3.7e-05,
      "loss": 9.391,
      "mean_token_accuracy": 0.03300900887697935,
      "num_tokens": 167021.0,
      "step": 75
    },
    {
      "entropy": 10.588104724884033,
      "epoch": 0.007684918347742555,
      "grad_norm": 2.015625,
      "learning_rate": 3.95e-05,
      "loss": 9.3298,
      "mean_token_accuracy": 0.03809518590569496,
      "num_tokens": 179269.0,
      "step": 80
    },
    {
      "entropy": 10.573075771331787,
      "epoch": 0.008165225744476465,
      "grad_norm": 1.8828125,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.2226,
      "mean_token_accuracy": 0.034396625123918056,
      "num_tokens": 190328.0,
      "step": 85
    },
    {
      "entropy": 10.551105213165282,
      "epoch": 0.008645533141210375,
      "grad_norm": 2.03125,
      "learning_rate": 4.45e-05,
      "loss": 9.0564,
      "mean_token_accuracy": 0.032392379455268384,
      "num_tokens": 201522.0,
      "step": 90
    },
    {
      "entropy": 10.511863613128662,
      "epoch": 0.009125840537944284,
      "grad_norm": 1.8671875,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 8.995,
      "mean_token_accuracy": 0.036259174533188346,
      "num_tokens": 212963.0,
      "step": 95
    },
    {
      "entropy": 10.454308891296387,
      "epoch": 0.009606147934678195,
      "grad_norm": 1.578125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 8.9701,
      "mean_token_accuracy": 0.0323725325986743,
      "num_tokens": 224446.0,
      "step": 100
    },
    {
      "entropy": 10.384486865997314,
      "epoch": 0.010086455331412104,
      "grad_norm": 1.765625,
      "learning_rate": 5.2e-05,
      "loss": 8.7698,
      "mean_token_accuracy": 0.03222609367221594,
      "num_tokens": 235334.0,
      "step": 105
    },
    {
      "entropy": 10.287021160125732,
      "epoch": 0.010566762728146013,
      "grad_norm": 1.5546875,
      "learning_rate": 5.45e-05,
      "loss": 8.6568,
      "mean_token_accuracy": 0.033952686190605166,
      "num_tokens": 247024.0,
      "step": 110
    },
    {
      "entropy": 10.163359928131104,
      "epoch": 0.011047070124879923,
      "grad_norm": 1.578125,
      "learning_rate": 5.7e-05,
      "loss": 8.4531,
      "mean_token_accuracy": 0.03682863060384989,
      "num_tokens": 258354.0,
      "step": 115
    },
    {
      "entropy": 10.03564281463623,
      "epoch": 0.011527377521613832,
      "grad_norm": 1.4296875,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.3556,
      "mean_token_accuracy": 0.039085553586483,
      "num_tokens": 271202.0,
      "step": 120
    },
    {
      "entropy": 9.844954490661621,
      "epoch": 0.012007684918347743,
      "grad_norm": 1.2890625,
      "learning_rate": 6.2e-05,
      "loss": 8.2264,
      "mean_token_accuracy": 0.03886361531913281,
      "num_tokens": 282405.0,
      "step": 125
    },
    {
      "entropy": 9.633767795562743,
      "epoch": 0.012487992315081652,
      "grad_norm": 1.203125,
      "learning_rate": 6.450000000000001e-05,
      "loss": 8.0588,
      "mean_token_accuracy": 0.04458533097058535,
      "num_tokens": 294211.0,
      "step": 130
    },
    {
      "entropy": 9.427376461029052,
      "epoch": 0.012968299711815562,
      "grad_norm": 1.1015625,
      "learning_rate": 6.7e-05,
      "loss": 8.0067,
      "mean_token_accuracy": 0.044877220317721365,
      "num_tokens": 305420.0,
      "step": 135
    },
    {
      "entropy": 9.113702964782714,
      "epoch": 0.013448607108549471,
      "grad_norm": 0.96875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.8472,
      "mean_token_accuracy": 0.04761340096592903,
      "num_tokens": 316430.0,
      "step": 140
    },
    {
      "entropy": 8.837570667266846,
      "epoch": 0.013928914505283382,
      "grad_norm": 1.03125,
      "learning_rate": 7.2e-05,
      "loss": 7.7606,
      "mean_token_accuracy": 0.05185309983789921,
      "num_tokens": 328406.0,
      "step": 145
    },
    {
      "entropy": 8.593095207214356,
      "epoch": 0.01440922190201729,
      "grad_norm": 1.0,
      "learning_rate": 7.45e-05,
      "loss": 7.6877,
      "mean_token_accuracy": 0.05381656400859356,
      "num_tokens": 339051.0,
      "step": 150
    },
    {
      "entropy": 8.415284156799316,
      "epoch": 0.014889529298751201,
      "grad_norm": 1.0625,
      "learning_rate": 7.7e-05,
      "loss": 7.6235,
      "mean_token_accuracy": 0.055242300406098364,
      "num_tokens": 350308.0,
      "step": 155
    },
    {
      "entropy": 8.24738597869873,
      "epoch": 0.01536983669548511,
      "grad_norm": 0.80078125,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.5935,
      "mean_token_accuracy": 0.054396551474928855,
      "num_tokens": 362159.0,
      "step": 160
    },
    {
      "entropy": 8.126520538330078,
      "epoch": 0.01585014409221902,
      "grad_norm": 1.015625,
      "learning_rate": 8.2e-05,
      "loss": 7.5837,
      "mean_token_accuracy": 0.05628910064697266,
      "num_tokens": 375157.0,
      "step": 165
    },
    {
      "entropy": 8.08217649459839,
      "epoch": 0.01633045148895293,
      "grad_norm": 0.84375,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.4497,
      "mean_token_accuracy": 0.06717152446508408,
      "num_tokens": 387318.0,
      "step": 170
    },
    {
      "entropy": 8.051177406311036,
      "epoch": 0.01681075888568684,
      "grad_norm": 1.1796875,
      "learning_rate": 8.7e-05,
      "loss": 7.4622,
      "mean_token_accuracy": 0.07725611329078674,
      "num_tokens": 398933.0,
      "step": 175
    },
    {
      "entropy": 7.939669942855835,
      "epoch": 0.01729106628242075,
      "grad_norm": 0.9296875,
      "learning_rate": 8.95e-05,
      "loss": 7.4158,
      "mean_token_accuracy": 0.0700771912932396,
      "num_tokens": 410460.0,
      "step": 180
    },
    {
      "entropy": 7.9294243335723875,
      "epoch": 0.01777137367915466,
      "grad_norm": 1.1171875,
      "learning_rate": 9.2e-05,
      "loss": 7.39,
      "mean_token_accuracy": 0.07831323072314263,
      "num_tokens": 421894.0,
      "step": 185
    },
    {
      "entropy": 7.895280361175537,
      "epoch": 0.01825168107588857,
      "grad_norm": 0.953125,
      "learning_rate": 9.45e-05,
      "loss": 7.3806,
      "mean_token_accuracy": 0.07451360076665878,
      "num_tokens": 432976.0,
      "step": 190
    },
    {
      "entropy": 7.967076063156128,
      "epoch": 0.018731988472622477,
      "grad_norm": 1.015625,
      "learning_rate": 9.7e-05,
      "loss": 7.4475,
      "mean_token_accuracy": 0.07888153940439224,
      "num_tokens": 444859.0,
      "step": 195
    },
    {
      "entropy": 7.939164543151856,
      "epoch": 0.01921229586935639,
      "grad_norm": 1.390625,
      "learning_rate": 9.95e-05,
      "loss": 7.3302,
      "mean_token_accuracy": 0.08403046280145646,
      "num_tokens": 456508.0,
      "step": 200
    },
    {
      "entropy": 7.868430519104004,
      "epoch": 0.0196926032660903,
      "grad_norm": 1.125,
      "learning_rate": 0.000102,
      "loss": 7.3456,
      "mean_token_accuracy": 0.07998873889446259,
      "num_tokens": 469064.0,
      "step": 205
    },
    {
      "entropy": 7.868760108947754,
      "epoch": 0.020172910662824207,
      "grad_norm": 1.828125,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.2899,
      "mean_token_accuracy": 0.08207204826176166,
      "num_tokens": 480743.0,
      "step": 210
    },
    {
      "entropy": 7.7555779933929445,
      "epoch": 0.020653218059558116,
      "grad_norm": 1.265625,
      "learning_rate": 0.000107,
      "loss": 7.3069,
      "mean_token_accuracy": 0.08280007690191268,
      "num_tokens": 493378.0,
      "step": 215
    },
    {
      "entropy": 7.839309644699097,
      "epoch": 0.021133525456292025,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0001095,
      "loss": 7.2462,
      "mean_token_accuracy": 0.0859088484197855,
      "num_tokens": 505641.0,
      "step": 220
    },
    {
      "entropy": 7.805751466751099,
      "epoch": 0.021613832853025938,
      "grad_norm": 1.15625,
      "learning_rate": 0.000112,
      "loss": 7.3079,
      "mean_token_accuracy": 0.0868067942559719,
      "num_tokens": 517783.0,
      "step": 225
    },
    {
      "entropy": 7.756006574630737,
      "epoch": 0.022094140249759846,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0001145,
      "loss": 7.1612,
      "mean_token_accuracy": 0.09106182381510734,
      "num_tokens": 528407.0,
      "step": 230
    },
    {
      "entropy": 7.80887017250061,
      "epoch": 0.022574447646493755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.3031,
      "mean_token_accuracy": 0.08784078434109688,
      "num_tokens": 539974.0,
      "step": 235
    },
    {
      "entropy": 7.641842365264893,
      "epoch": 0.023054755043227664,
      "grad_norm": 1.359375,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.1391,
      "mean_token_accuracy": 0.08666003718972207,
      "num_tokens": 551127.0,
      "step": 240
    },
    {
      "entropy": 7.679393720626831,
      "epoch": 0.023535062439961577,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000122,
      "loss": 7.1554,
      "mean_token_accuracy": 0.09684930071234703,
      "num_tokens": 561980.0,
      "step": 245
    },
    {
      "entropy": 7.723794889450073,
      "epoch": 0.024015369836695485,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0001245,
      "loss": 7.1898,
      "mean_token_accuracy": 0.0910400390625,
      "num_tokens": 573840.0,
      "step": 250
    },
    {
      "entropy": 7.562839365005493,
      "epoch": 0.024495677233429394,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000127,
      "loss": 7.0923,
      "mean_token_accuracy": 0.09587239101529121,
      "num_tokens": 585110.0,
      "step": 255
    },
    {
      "entropy": 7.556887006759643,
      "epoch": 0.024975984630163303,
      "grad_norm": 1.046875,
      "learning_rate": 0.0001295,
      "loss": 6.9932,
      "mean_token_accuracy": 0.0963286928832531,
      "num_tokens": 596902.0,
      "step": 260
    },
    {
      "entropy": 7.55956130027771,
      "epoch": 0.025456292026897216,
      "grad_norm": 1.265625,
      "learning_rate": 0.000132,
      "loss": 7.1566,
      "mean_token_accuracy": 0.0950677141547203,
      "num_tokens": 608738.0,
      "step": 265
    },
    {
      "entropy": 7.5788225650787355,
      "epoch": 0.025936599423631124,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.0778,
      "mean_token_accuracy": 0.0930832102894783,
      "num_tokens": 621078.0,
      "step": 270
    },
    {
      "entropy": 7.6280433177948,
      "epoch": 0.026416906820365033,
      "grad_norm": 1.140625,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.1451,
      "mean_token_accuracy": 0.09341271966695786,
      "num_tokens": 632333.0,
      "step": 275
    },
    {
      "entropy": 7.559076023101807,
      "epoch": 0.026897214217098942,
      "grad_norm": 1.578125,
      "learning_rate": 0.0001395,
      "loss": 6.9556,
      "mean_token_accuracy": 0.10151683166623116,
      "num_tokens": 644474.0,
      "step": 280
    },
    {
      "entropy": 7.4703763961792,
      "epoch": 0.027377521613832854,
      "grad_norm": 1.171875,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.0352,
      "mean_token_accuracy": 0.10359716191887855,
      "num_tokens": 655608.0,
      "step": 285
    },
    {
      "entropy": 7.540377807617188,
      "epoch": 0.027857829010566763,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001445,
      "loss": 7.0476,
      "mean_token_accuracy": 0.09810357466340065,
      "num_tokens": 667371.0,
      "step": 290
    },
    {
      "entropy": 7.448433876037598,
      "epoch": 0.028338136407300672,
      "grad_norm": 1.5625,
      "learning_rate": 0.000147,
      "loss": 7.0591,
      "mean_token_accuracy": 0.09780933484435081,
      "num_tokens": 679152.0,
      "step": 295
    },
    {
      "entropy": 7.4532647132873535,
      "epoch": 0.02881844380403458,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0001495,
      "loss": 6.9632,
      "mean_token_accuracy": 0.10561580210924149,
      "num_tokens": 690359.0,
      "step": 300
    },
    {
      "entropy": 7.422236824035645,
      "epoch": 0.029298751200768493,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000152,
      "loss": 6.9359,
      "mean_token_accuracy": 0.10709809437394142,
      "num_tokens": 701819.0,
      "step": 305
    },
    {
      "entropy": 7.465102815628052,
      "epoch": 0.029779058597502402,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.9961,
      "mean_token_accuracy": 0.10205374658107758,
      "num_tokens": 711048.0,
      "step": 310
    },
    {
      "entropy": 7.381311082839966,
      "epoch": 0.03025936599423631,
      "grad_norm": 1.421875,
      "learning_rate": 0.000157,
      "loss": 6.8809,
      "mean_token_accuracy": 0.10968201458454133,
      "num_tokens": 721399.0,
      "step": 315
    },
    {
      "entropy": 7.40810341835022,
      "epoch": 0.03073967339097022,
      "grad_norm": 1.171875,
      "learning_rate": 0.0001595,
      "loss": 6.891,
      "mean_token_accuracy": 0.10995450168848038,
      "num_tokens": 732144.0,
      "step": 320
    },
    {
      "entropy": 7.357445764541626,
      "epoch": 0.03121998078770413,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000162,
      "loss": 6.8626,
      "mean_token_accuracy": 0.10403823405504227,
      "num_tokens": 743962.0,
      "step": 325
    },
    {
      "entropy": 7.388841199874878,
      "epoch": 0.03170028818443804,
      "grad_norm": 1.640625,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.0076,
      "mean_token_accuracy": 0.10685212463140488,
      "num_tokens": 755641.0,
      "step": 330
    },
    {
      "entropy": 7.286140584945679,
      "epoch": 0.03218059558117195,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.8385,
      "mean_token_accuracy": 0.1151984341442585,
      "num_tokens": 766798.0,
      "step": 335
    },
    {
      "entropy": 7.339591550827026,
      "epoch": 0.03266090297790586,
      "grad_norm": 1.265625,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.9311,
      "mean_token_accuracy": 0.10881740972399712,
      "num_tokens": 778097.0,
      "step": 340
    },
    {
      "entropy": 7.252063226699829,
      "epoch": 0.03314121037463977,
      "grad_norm": 1.21875,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.8496,
      "mean_token_accuracy": 0.11310645863413811,
      "num_tokens": 789775.0,
      "step": 345
    },
    {
      "entropy": 7.39683985710144,
      "epoch": 0.03362151777137368,
      "grad_norm": 1.96875,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.9062,
      "mean_token_accuracy": 0.10776558816432953,
      "num_tokens": 800155.0,
      "step": 350
    },
    {
      "entropy": 7.379228162765503,
      "epoch": 0.034101825168107586,
      "grad_norm": 1.9765625,
      "learning_rate": 0.000177,
      "loss": 6.9239,
      "mean_token_accuracy": 0.11197539716959,
      "num_tokens": 811213.0,
      "step": 355
    },
    {
      "entropy": 7.30075511932373,
      "epoch": 0.0345821325648415,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0001795,
      "loss": 6.903,
      "mean_token_accuracy": 0.10408640876412392,
      "num_tokens": 822832.0,
      "step": 360
    },
    {
      "entropy": 7.287055206298828,
      "epoch": 0.03506243996157541,
      "grad_norm": 1.5625,
      "learning_rate": 0.000182,
      "loss": 6.8727,
      "mean_token_accuracy": 0.11483326479792595,
      "num_tokens": 834824.0,
      "step": 365
    },
    {
      "entropy": 7.292969608306885,
      "epoch": 0.03554274735830932,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0001845,
      "loss": 6.9053,
      "mean_token_accuracy": 0.10474938601255417,
      "num_tokens": 846969.0,
      "step": 370
    },
    {
      "entropy": 7.357876491546631,
      "epoch": 0.03602305475504323,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000187,
      "loss": 6.9773,
      "mean_token_accuracy": 0.10199039503931999,
      "num_tokens": 859271.0,
      "step": 375
    },
    {
      "entropy": 7.247849559783935,
      "epoch": 0.03650336215177714,
      "grad_norm": 1.265625,
      "learning_rate": 0.0001895,
      "loss": 6.8723,
      "mean_token_accuracy": 0.10891828313469887,
      "num_tokens": 869754.0,
      "step": 380
    },
    {
      "entropy": 7.2068620204925535,
      "epoch": 0.036983669548511046,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000192,
      "loss": 6.819,
      "mean_token_accuracy": 0.11546091064810753,
      "num_tokens": 880519.0,
      "step": 385
    },
    {
      "entropy": 7.166879415512085,
      "epoch": 0.037463976945244955,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0001945,
      "loss": 6.775,
      "mean_token_accuracy": 0.10799357295036316,
      "num_tokens": 890765.0,
      "step": 390
    },
    {
      "entropy": 7.240225791931152,
      "epoch": 0.037944284341978864,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.7757,
      "mean_token_accuracy": 0.10960216820240021,
      "num_tokens": 901717.0,
      "step": 395
    },
    {
      "entropy": 7.299621725082398,
      "epoch": 0.03842459173871278,
      "grad_norm": 1.4375,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.8428,
      "mean_token_accuracy": 0.11034838110208511,
      "num_tokens": 913334.0,
      "step": 400
    },
    {
      "entropy": 7.204387092590332,
      "epoch": 0.03890489913544669,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000202,
      "loss": 6.8557,
      "mean_token_accuracy": 0.11134258806705474,
      "num_tokens": 924753.0,
      "step": 405
    },
    {
      "entropy": 7.204525709152222,
      "epoch": 0.0393852065321806,
      "grad_norm": 1.421875,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.7857,
      "mean_token_accuracy": 0.11340023577213287,
      "num_tokens": 935692.0,
      "step": 410
    },
    {
      "entropy": 7.154750633239746,
      "epoch": 0.039865513928914506,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000207,
      "loss": 6.8747,
      "mean_token_accuracy": 0.11008304730057716,
      "num_tokens": 947902.0,
      "step": 415
    },
    {
      "entropy": 7.122686529159546,
      "epoch": 0.040345821325648415,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002095,
      "loss": 6.8128,
      "mean_token_accuracy": 0.11032308787107467,
      "num_tokens": 959437.0,
      "step": 420
    },
    {
      "entropy": 7.040357637405395,
      "epoch": 0.040826128722382324,
      "grad_norm": 1.65625,
      "learning_rate": 0.000212,
      "loss": 6.6189,
      "mean_token_accuracy": 0.12019914388656616,
      "num_tokens": 970191.0,
      "step": 425
    },
    {
      "entropy": 7.0764172077178955,
      "epoch": 0.04130643611911623,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002145,
      "loss": 6.7458,
      "mean_token_accuracy": 0.11545630767941475,
      "num_tokens": 981811.0,
      "step": 430
    },
    {
      "entropy": 7.159797048568725,
      "epoch": 0.04178674351585014,
      "grad_norm": 1.125,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.8529,
      "mean_token_accuracy": 0.10769774615764618,
      "num_tokens": 994034.0,
      "step": 435
    },
    {
      "entropy": 7.207771253585816,
      "epoch": 0.04226705091258405,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002195,
      "loss": 6.7989,
      "mean_token_accuracy": 0.11341427490115166,
      "num_tokens": 1005506.0,
      "step": 440
    },
    {
      "entropy": 7.045540857315063,
      "epoch": 0.042747358309317966,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000222,
      "loss": 6.7647,
      "mean_token_accuracy": 0.10924506187438965,
      "num_tokens": 1017103.0,
      "step": 445
    },
    {
      "entropy": 7.124048948287964,
      "epoch": 0.043227665706051875,
      "grad_norm": 1.390625,
      "learning_rate": 0.0002245,
      "loss": 6.8523,
      "mean_token_accuracy": 0.10999714508652687,
      "num_tokens": 1029606.0,
      "step": 450
    },
    {
      "entropy": 7.184269189834595,
      "epoch": 0.043707973102785784,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.7961,
      "mean_token_accuracy": 0.11004141569137574,
      "num_tokens": 1041899.0,
      "step": 455
    },
    {
      "entropy": 7.053015661239624,
      "epoch": 0.04418828049951969,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.7001,
      "mean_token_accuracy": 0.11952895596623421,
      "num_tokens": 1053577.0,
      "step": 460
    },
    {
      "entropy": 7.075217962265015,
      "epoch": 0.0446685878962536,
      "grad_norm": 1.1875,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.7525,
      "mean_token_accuracy": 0.12150197252631187,
      "num_tokens": 1065541.0,
      "step": 465
    },
    {
      "entropy": 7.141125965118408,
      "epoch": 0.04514889529298751,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.7007,
      "mean_token_accuracy": 0.11925816610455513,
      "num_tokens": 1076138.0,
      "step": 470
    },
    {
      "entropy": 7.022414493560791,
      "epoch": 0.04562920268972142,
      "grad_norm": 1.203125,
      "learning_rate": 0.000237,
      "loss": 6.7675,
      "mean_token_accuracy": 0.11674837097525596,
      "num_tokens": 1087620.0,
      "step": 475
    },
    {
      "entropy": 7.041283512115479,
      "epoch": 0.04610951008645533,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002395,
      "loss": 6.6396,
      "mean_token_accuracy": 0.12078191563487053,
      "num_tokens": 1099493.0,
      "step": 480
    },
    {
      "entropy": 7.031822586059571,
      "epoch": 0.046589817483189244,
      "grad_norm": 1.359375,
      "learning_rate": 0.000242,
      "loss": 6.7,
      "mean_token_accuracy": 0.11970095857977867,
      "num_tokens": 1110559.0,
      "step": 485
    },
    {
      "entropy": 6.978209018707275,
      "epoch": 0.04707012487992315,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002445,
      "loss": 6.6632,
      "mean_token_accuracy": 0.12008017003536224,
      "num_tokens": 1121051.0,
      "step": 490
    },
    {
      "entropy": 6.980980300903321,
      "epoch": 0.04755043227665706,
      "grad_norm": 1.203125,
      "learning_rate": 0.000247,
      "loss": 6.6395,
      "mean_token_accuracy": 0.12107920199632645,
      "num_tokens": 1132305.0,
      "step": 495
    },
    {
      "entropy": 7.038042116165161,
      "epoch": 0.04803073967339097,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0002495,
      "loss": 6.6505,
      "mean_token_accuracy": 0.1215340994298458,
      "num_tokens": 1142979.0,
      "step": 500
    },
    {
      "entropy": 6.921855020523071,
      "epoch": 0.04851104707012488,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000252,
      "loss": 6.646,
      "mean_token_accuracy": 0.12661107257008553,
      "num_tokens": 1153889.0,
      "step": 505
    },
    {
      "entropy": 6.98356146812439,
      "epoch": 0.04899135446685879,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002545,
      "loss": 6.6417,
      "mean_token_accuracy": 0.1176763840019703,
      "num_tokens": 1166030.0,
      "step": 510
    },
    {
      "entropy": 7.046093416213989,
      "epoch": 0.0494716618635927,
      "grad_norm": 1.328125,
      "learning_rate": 0.000257,
      "loss": 6.8173,
      "mean_token_accuracy": 0.11064628511667252,
      "num_tokens": 1177545.0,
      "step": 515
    },
    {
      "entropy": 7.001708602905273,
      "epoch": 0.049951969260326606,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002595,
      "loss": 6.6353,
      "mean_token_accuracy": 0.12478243634104728,
      "num_tokens": 1188944.0,
      "step": 520
    },
    {
      "entropy": 6.840310907363891,
      "epoch": 0.05043227665706052,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000262,
      "loss": 6.696,
      "mean_token_accuracy": 0.1172410435974598,
      "num_tokens": 1199517.0,
      "step": 525
    },
    {
      "entropy": 7.022114849090576,
      "epoch": 0.05091258405379443,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.6212,
      "mean_token_accuracy": 0.11661999747157097,
      "num_tokens": 1211536.0,
      "step": 530
    },
    {
      "entropy": 6.830185031890869,
      "epoch": 0.05139289145052834,
      "grad_norm": 1.3125,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.5521,
      "mean_token_accuracy": 0.12921451553702354,
      "num_tokens": 1224159.0,
      "step": 535
    },
    {
      "entropy": 6.94548978805542,
      "epoch": 0.05187319884726225,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.614,
      "mean_token_accuracy": 0.1301414117217064,
      "num_tokens": 1235226.0,
      "step": 540
    },
    {
      "entropy": 6.777507877349853,
      "epoch": 0.05235350624399616,
      "grad_norm": 1.296875,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.5321,
      "mean_token_accuracy": 0.13609066978096962,
      "num_tokens": 1245862.0,
      "step": 545
    },
    {
      "entropy": 6.856529331207275,
      "epoch": 0.052833813640730067,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002745,
      "loss": 6.5508,
      "mean_token_accuracy": 0.1289893075823784,
      "num_tokens": 1257512.0,
      "step": 550
    },
    {
      "entropy": 6.894668006896973,
      "epoch": 0.053314121037463975,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000277,
      "loss": 6.6539,
      "mean_token_accuracy": 0.1206201434135437,
      "num_tokens": 1269317.0,
      "step": 555
    },
    {
      "entropy": 6.811044359207154,
      "epoch": 0.053794428434197884,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002795,
      "loss": 6.551,
      "mean_token_accuracy": 0.11646598875522614,
      "num_tokens": 1281878.0,
      "step": 560
    },
    {
      "entropy": 6.738573932647705,
      "epoch": 0.05427473583093179,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.5612,
      "mean_token_accuracy": 0.12958591282367707,
      "num_tokens": 1293178.0,
      "step": 565
    },
    {
      "entropy": 6.851106452941894,
      "epoch": 0.05475504322766571,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0002845,
      "loss": 6.5421,
      "mean_token_accuracy": 0.12348497658967972,
      "num_tokens": 1304969.0,
      "step": 570
    },
    {
      "entropy": 6.937162446975708,
      "epoch": 0.05523535062439962,
      "grad_norm": 1.34375,
      "learning_rate": 0.000287,
      "loss": 6.5765,
      "mean_token_accuracy": 0.12681149393320085,
      "num_tokens": 1317301.0,
      "step": 575
    },
    {
      "entropy": 6.686863088607788,
      "epoch": 0.05571565802113353,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002895,
      "loss": 6.5505,
      "mean_token_accuracy": 0.12601301521062852,
      "num_tokens": 1327688.0,
      "step": 580
    },
    {
      "entropy": 6.901493978500366,
      "epoch": 0.056195965417867436,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000292,
      "loss": 6.4949,
      "mean_token_accuracy": 0.1258933037519455,
      "num_tokens": 1338750.0,
      "step": 585
    },
    {
      "entropy": 6.832601118087768,
      "epoch": 0.056676272814601344,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002945,
      "loss": 6.6508,
      "mean_token_accuracy": 0.1182264193892479,
      "num_tokens": 1350899.0,
      "step": 590
    },
    {
      "entropy": 6.818350124359131,
      "epoch": 0.05715658021133525,
      "grad_norm": 1.7890625,
      "learning_rate": 0.000297,
      "loss": 6.5006,
      "mean_token_accuracy": 0.1271736852824688,
      "num_tokens": 1363184.0,
      "step": 595
    },
    {
      "entropy": 6.7492002010345455,
      "epoch": 0.05763688760806916,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002995,
      "loss": 6.4969,
      "mean_token_accuracy": 0.12888728752732276,
      "num_tokens": 1374192.0,
      "step": 600
    },
    {
      "entropy": 6.7368873119354244,
      "epoch": 0.05811719500480307,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000302,
      "loss": 6.5138,
      "mean_token_accuracy": 0.12501192465424538,
      "num_tokens": 1384888.0,
      "step": 605
    },
    {
      "entropy": 6.81034517288208,
      "epoch": 0.05859750240153699,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003045,
      "loss": 6.6139,
      "mean_token_accuracy": 0.12904802039265634,
      "num_tokens": 1395139.0,
      "step": 610
    },
    {
      "entropy": 6.729338312149048,
      "epoch": 0.059077809798270896,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000307,
      "loss": 6.4134,
      "mean_token_accuracy": 0.12848665565252304,
      "num_tokens": 1407219.0,
      "step": 615
    },
    {
      "entropy": 6.779170131683349,
      "epoch": 0.059558117195004805,
      "grad_norm": 1.125,
      "learning_rate": 0.0003095,
      "loss": 6.5328,
      "mean_token_accuracy": 0.12432346194982528,
      "num_tokens": 1418917.0,
      "step": 620
    },
    {
      "entropy": 6.744690942764282,
      "epoch": 0.060038424591738714,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000312,
      "loss": 6.4838,
      "mean_token_accuracy": 0.12512711808085442,
      "num_tokens": 1430792.0,
      "step": 625
    },
    {
      "entropy": 6.667728471755981,
      "epoch": 0.06051873198847262,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003145,
      "loss": 6.504,
      "mean_token_accuracy": 0.13070958107709885,
      "num_tokens": 1442825.0,
      "step": 630
    },
    {
      "entropy": 6.764918041229248,
      "epoch": 0.06099903938520653,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000317,
      "loss": 6.5837,
      "mean_token_accuracy": 0.1336795650422573,
      "num_tokens": 1454078.0,
      "step": 635
    },
    {
      "entropy": 6.783257913589478,
      "epoch": 0.06147934678194044,
      "grad_norm": 1.453125,
      "learning_rate": 0.0003195,
      "loss": 6.5293,
      "mean_token_accuracy": 0.12106153219938279,
      "num_tokens": 1465164.0,
      "step": 640
    },
    {
      "entropy": 6.80000376701355,
      "epoch": 0.06195965417867435,
      "grad_norm": 1.09375,
      "learning_rate": 0.000322,
      "loss": 6.5039,
      "mean_token_accuracy": 0.13234740868210793,
      "num_tokens": 1476813.0,
      "step": 645
    },
    {
      "entropy": 6.68659987449646,
      "epoch": 0.06243996157540826,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.5251,
      "mean_token_accuracy": 0.1282544143497944,
      "num_tokens": 1488547.0,
      "step": 650
    },
    {
      "entropy": 6.814279413223266,
      "epoch": 0.06292026897214217,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.531,
      "mean_token_accuracy": 0.12601352855563164,
      "num_tokens": 1500714.0,
      "step": 655
    },
    {
      "entropy": 6.687133073806763,
      "epoch": 0.06340057636887608,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.5182,
      "mean_token_accuracy": 0.13100821301341056,
      "num_tokens": 1512212.0,
      "step": 660
    },
    {
      "entropy": 6.712434768676758,
      "epoch": 0.06388088376560999,
      "grad_norm": 1.203125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.4878,
      "mean_token_accuracy": 0.13093362599611283,
      "num_tokens": 1523610.0,
      "step": 665
    },
    {
      "entropy": 6.596017265319825,
      "epoch": 0.0643611911623439,
      "grad_norm": 1.125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.4492,
      "mean_token_accuracy": 0.12683645486831666,
      "num_tokens": 1535543.0,
      "step": 670
    },
    {
      "entropy": 6.848142671585083,
      "epoch": 0.06484149855907781,
      "grad_norm": 1.203125,
      "learning_rate": 0.000337,
      "loss": 6.6626,
      "mean_token_accuracy": 0.12445207461714744,
      "num_tokens": 1547183.0,
      "step": 675
    },
    {
      "entropy": 6.626632308959961,
      "epoch": 0.06532180595581172,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0003395,
      "loss": 6.4179,
      "mean_token_accuracy": 0.136246507614851,
      "num_tokens": 1559265.0,
      "step": 680
    },
    {
      "entropy": 6.692934179306031,
      "epoch": 0.06580211335254563,
      "grad_norm": 1.1875,
      "learning_rate": 0.000342,
      "loss": 6.4384,
      "mean_token_accuracy": 0.1327574633061886,
      "num_tokens": 1569774.0,
      "step": 685
    },
    {
      "entropy": 6.530060815811157,
      "epoch": 0.06628242074927954,
      "grad_norm": 1.25,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.3719,
      "mean_token_accuracy": 0.1348055936396122,
      "num_tokens": 1579408.0,
      "step": 690
    },
    {
      "entropy": 6.638763380050659,
      "epoch": 0.06676272814601344,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000347,
      "loss": 6.4333,
      "mean_token_accuracy": 0.1340454787015915,
      "num_tokens": 1591849.0,
      "step": 695
    },
    {
      "entropy": 6.610886335372925,
      "epoch": 0.06724303554274735,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003495,
      "loss": 6.5007,
      "mean_token_accuracy": 0.12831750214099885,
      "num_tokens": 1603269.0,
      "step": 700
    },
    {
      "entropy": 6.77165641784668,
      "epoch": 0.06772334293948126,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000352,
      "loss": 6.5134,
      "mean_token_accuracy": 0.1261325843632221,
      "num_tokens": 1613639.0,
      "step": 705
    },
    {
      "entropy": 6.6583836555480955,
      "epoch": 0.06820365033621517,
      "grad_norm": 1.375,
      "learning_rate": 0.0003545,
      "loss": 6.542,
      "mean_token_accuracy": 0.12910353168845176,
      "num_tokens": 1625393.0,
      "step": 710
    },
    {
      "entropy": 6.6097071170806885,
      "epoch": 0.0686839577329491,
      "grad_norm": 1.265625,
      "learning_rate": 0.000357,
      "loss": 6.3672,
      "mean_token_accuracy": 0.13228188902139665,
      "num_tokens": 1636355.0,
      "step": 715
    },
    {
      "entropy": 6.604225015640258,
      "epoch": 0.069164265129683,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003595,
      "loss": 6.409,
      "mean_token_accuracy": 0.1348489813506603,
      "num_tokens": 1646157.0,
      "step": 720
    },
    {
      "entropy": 6.58693528175354,
      "epoch": 0.06964457252641691,
      "grad_norm": 1.28125,
      "learning_rate": 0.000362,
      "loss": 6.4101,
      "mean_token_accuracy": 0.12884832546114922,
      "num_tokens": 1657827.0,
      "step": 725
    },
    {
      "entropy": 6.573770189285279,
      "epoch": 0.07012487992315082,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003645,
      "loss": 6.4863,
      "mean_token_accuracy": 0.13203563466668128,
      "num_tokens": 1669023.0,
      "step": 730
    },
    {
      "entropy": 6.614386749267578,
      "epoch": 0.07060518731988473,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000367,
      "loss": 6.4313,
      "mean_token_accuracy": 0.1316665753722191,
      "num_tokens": 1680048.0,
      "step": 735
    },
    {
      "entropy": 6.710096597671509,
      "epoch": 0.07108549471661864,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003695,
      "loss": 6.6213,
      "mean_token_accuracy": 0.11867887079715729,
      "num_tokens": 1690907.0,
      "step": 740
    },
    {
      "entropy": 6.723827362060547,
      "epoch": 0.07156580211335255,
      "grad_norm": 1.1875,
      "learning_rate": 0.000372,
      "loss": 6.4584,
      "mean_token_accuracy": 0.13038890957832336,
      "num_tokens": 1702337.0,
      "step": 745
    },
    {
      "entropy": 6.607548427581787,
      "epoch": 0.07204610951008646,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003745,
      "loss": 6.3811,
      "mean_token_accuracy": 0.1292431376874447,
      "num_tokens": 1713242.0,
      "step": 750
    },
    {
      "entropy": 6.5345042705535885,
      "epoch": 0.07252641690682037,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000377,
      "loss": 6.3688,
      "mean_token_accuracy": 0.13297409936785698,
      "num_tokens": 1724939.0,
      "step": 755
    },
    {
      "entropy": 6.485815048217773,
      "epoch": 0.07300672430355427,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003795,
      "loss": 6.3247,
      "mean_token_accuracy": 0.14120787829160691,
      "num_tokens": 1737311.0,
      "step": 760
    },
    {
      "entropy": 6.638444805145264,
      "epoch": 0.07348703170028818,
      "grad_norm": 1.171875,
      "learning_rate": 0.000382,
      "loss": 6.4629,
      "mean_token_accuracy": 0.13194502517580986,
      "num_tokens": 1749528.0,
      "step": 765
    },
    {
      "entropy": 6.529104328155517,
      "epoch": 0.07396733909702209,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003845,
      "loss": 6.3293,
      "mean_token_accuracy": 0.13749989122152328,
      "num_tokens": 1760304.0,
      "step": 770
    },
    {
      "entropy": 6.502080202102661,
      "epoch": 0.074447646493756,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.3039,
      "mean_token_accuracy": 0.13969450294971467,
      "num_tokens": 1772014.0,
      "step": 775
    },
    {
      "entropy": 6.4714720249176025,
      "epoch": 0.07492795389048991,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.3781,
      "mean_token_accuracy": 0.13299143686890602,
      "num_tokens": 1784585.0,
      "step": 780
    },
    {
      "entropy": 6.529697895050049,
      "epoch": 0.07540826128722382,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.3155,
      "mean_token_accuracy": 0.13666464909911155,
      "num_tokens": 1795246.0,
      "step": 785
    },
    {
      "entropy": 6.44914927482605,
      "epoch": 0.07588856868395773,
      "grad_norm": 1.265625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.2962,
      "mean_token_accuracy": 0.13686830475926398,
      "num_tokens": 1806826.0,
      "step": 790
    },
    {
      "entropy": 6.512354946136474,
      "epoch": 0.07636887608069164,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.4276,
      "mean_token_accuracy": 0.13477274253964425,
      "num_tokens": 1818766.0,
      "step": 795
    },
    {
      "entropy": 6.510392951965332,
      "epoch": 0.07684918347742556,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003995,
      "loss": 6.2729,
      "mean_token_accuracy": 0.14873504042625427,
      "num_tokens": 1830824.0,
      "step": 800
    },
    {
      "entropy": 6.404947662353516,
      "epoch": 0.07732949087415947,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000402,
      "loss": 6.3783,
      "mean_token_accuracy": 0.13591941818594933,
      "num_tokens": 1841792.0,
      "step": 805
    },
    {
      "entropy": 6.476539421081543,
      "epoch": 0.07780979827089338,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004045,
      "loss": 6.3281,
      "mean_token_accuracy": 0.1313629515469074,
      "num_tokens": 1854683.0,
      "step": 810
    },
    {
      "entropy": 6.562305116653443,
      "epoch": 0.07829010566762729,
      "grad_norm": 1.25,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.4129,
      "mean_token_accuracy": 0.13369727060198783,
      "num_tokens": 1865826.0,
      "step": 815
    },
    {
      "entropy": 6.482418060302734,
      "epoch": 0.0787704130643612,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004095,
      "loss": 6.3598,
      "mean_token_accuracy": 0.14122529625892638,
      "num_tokens": 1876759.0,
      "step": 820
    },
    {
      "entropy": 6.471727514266968,
      "epoch": 0.0792507204610951,
      "grad_norm": 1.03125,
      "learning_rate": 0.000412,
      "loss": 6.4061,
      "mean_token_accuracy": 0.13444310426712036,
      "num_tokens": 1888013.0,
      "step": 825
    },
    {
      "entropy": 6.579637718200684,
      "epoch": 0.07973102785782901,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004145,
      "loss": 6.4009,
      "mean_token_accuracy": 0.14101578444242477,
      "num_tokens": 1898964.0,
      "step": 830
    },
    {
      "entropy": 6.426772117614746,
      "epoch": 0.08021133525456292,
      "grad_norm": 1.171875,
      "learning_rate": 0.000417,
      "loss": 6.3216,
      "mean_token_accuracy": 0.14054880663752556,
      "num_tokens": 1909728.0,
      "step": 835
    },
    {
      "entropy": 6.411238527297973,
      "epoch": 0.08069164265129683,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004195,
      "loss": 6.2457,
      "mean_token_accuracy": 0.14817485064268113,
      "num_tokens": 1919895.0,
      "step": 840
    },
    {
      "entropy": 6.340784978866577,
      "epoch": 0.08117195004803074,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000422,
      "loss": 6.1755,
      "mean_token_accuracy": 0.1461382269859314,
      "num_tokens": 1930839.0,
      "step": 845
    },
    {
      "entropy": 6.435613965988159,
      "epoch": 0.08165225744476465,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004245,
      "loss": 6.3812,
      "mean_token_accuracy": 0.13786052912473679,
      "num_tokens": 1942605.0,
      "step": 850
    },
    {
      "entropy": 6.454065656661987,
      "epoch": 0.08213256484149856,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000427,
      "loss": 6.3534,
      "mean_token_accuracy": 0.13335032984614373,
      "num_tokens": 1954863.0,
      "step": 855
    },
    {
      "entropy": 6.487230014801026,
      "epoch": 0.08261287223823247,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004295,
      "loss": 6.4437,
      "mean_token_accuracy": 0.13347432389855385,
      "num_tokens": 1966088.0,
      "step": 860
    },
    {
      "entropy": 6.391985607147217,
      "epoch": 0.08309317963496637,
      "grad_norm": 1.171875,
      "learning_rate": 0.000432,
      "loss": 6.2804,
      "mean_token_accuracy": 0.13734670206904412,
      "num_tokens": 1978081.0,
      "step": 865
    },
    {
      "entropy": 6.432765245437622,
      "epoch": 0.08357348703170028,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004345,
      "loss": 6.3338,
      "mean_token_accuracy": 0.1310129150748253,
      "num_tokens": 1991212.0,
      "step": 870
    },
    {
      "entropy": 6.510468673706055,
      "epoch": 0.08405379442843419,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000437,
      "loss": 6.3222,
      "mean_token_accuracy": 0.14048025831580163,
      "num_tokens": 2002074.0,
      "step": 875
    },
    {
      "entropy": 6.308026313781738,
      "epoch": 0.0845341018251681,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004395,
      "loss": 6.256,
      "mean_token_accuracy": 0.14530111104249954,
      "num_tokens": 2013039.0,
      "step": 880
    },
    {
      "entropy": 6.424022388458252,
      "epoch": 0.08501440922190202,
      "grad_norm": 0.984375,
      "learning_rate": 0.000442,
      "loss": 6.3833,
      "mean_token_accuracy": 0.13611895516514777,
      "num_tokens": 2024744.0,
      "step": 885
    },
    {
      "entropy": 6.35896577835083,
      "epoch": 0.08549471661863593,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004445,
      "loss": 6.2463,
      "mean_token_accuracy": 0.14010558724403382,
      "num_tokens": 2035587.0,
      "step": 890
    },
    {
      "entropy": 6.4153601169586185,
      "epoch": 0.08597502401536984,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000447,
      "loss": 6.3276,
      "mean_token_accuracy": 0.1342465154826641,
      "num_tokens": 2047924.0,
      "step": 895
    },
    {
      "entropy": 6.351702404022217,
      "epoch": 0.08645533141210375,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.2714,
      "mean_token_accuracy": 0.14150312915444374,
      "num_tokens": 2060161.0,
      "step": 900
    },
    {
      "entropy": 6.34501953125,
      "epoch": 0.08693563880883766,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.2744,
      "mean_token_accuracy": 0.1404205270111561,
      "num_tokens": 2071674.0,
      "step": 905
    },
    {
      "entropy": 6.472716283798218,
      "epoch": 0.08741594620557157,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.414,
      "mean_token_accuracy": 0.13350878208875655,
      "num_tokens": 2084025.0,
      "step": 910
    },
    {
      "entropy": 6.409608030319214,
      "epoch": 0.08789625360230548,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.3529,
      "mean_token_accuracy": 0.1360016256570816,
      "num_tokens": 2094501.0,
      "step": 915
    },
    {
      "entropy": 6.451559734344483,
      "epoch": 0.08837656099903939,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.3329,
      "mean_token_accuracy": 0.14456773400306702,
      "num_tokens": 2106207.0,
      "step": 920
    },
    {
      "entropy": 6.352231740951538,
      "epoch": 0.0888568683957733,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000462,
      "loss": 6.2281,
      "mean_token_accuracy": 0.15309673249721528,
      "num_tokens": 2116879.0,
      "step": 925
    },
    {
      "entropy": 6.339853239059448,
      "epoch": 0.0893371757925072,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004645,
      "loss": 6.2992,
      "mean_token_accuracy": 0.14001021683216094,
      "num_tokens": 2128884.0,
      "step": 930
    },
    {
      "entropy": 6.280442619323731,
      "epoch": 0.08981748318924111,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000467,
      "loss": 6.2508,
      "mean_token_accuracy": 0.1376420110464096,
      "num_tokens": 2140810.0,
      "step": 935
    },
    {
      "entropy": 6.40587887763977,
      "epoch": 0.09029779058597502,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004695,
      "loss": 6.2755,
      "mean_token_accuracy": 0.13705328851938248,
      "num_tokens": 2152312.0,
      "step": 940
    },
    {
      "entropy": 6.29297366142273,
      "epoch": 0.09077809798270893,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000472,
      "loss": 6.242,
      "mean_token_accuracy": 0.13348551020026206,
      "num_tokens": 2164408.0,
      "step": 945
    },
    {
      "entropy": 6.320023441314698,
      "epoch": 0.09125840537944284,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004745,
      "loss": 6.2521,
      "mean_token_accuracy": 0.14301476031541824,
      "num_tokens": 2176050.0,
      "step": 950
    },
    {
      "entropy": 6.2552725315094,
      "epoch": 0.09173871277617675,
      "grad_norm": 1.0,
      "learning_rate": 0.000477,
      "loss": 6.1363,
      "mean_token_accuracy": 0.14384374693036078,
      "num_tokens": 2187354.0,
      "step": 955
    },
    {
      "entropy": 6.197836017608642,
      "epoch": 0.09221902017291066,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004795,
      "loss": 6.1652,
      "mean_token_accuracy": 0.14129099994897842,
      "num_tokens": 2198686.0,
      "step": 960
    },
    {
      "entropy": 6.174200105667114,
      "epoch": 0.09269932756964457,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000482,
      "loss": 6.1982,
      "mean_token_accuracy": 0.14176447689533234,
      "num_tokens": 2209780.0,
      "step": 965
    },
    {
      "entropy": 6.389877271652222,
      "epoch": 0.09317963496637849,
      "grad_norm": 1.0,
      "learning_rate": 0.0004845,
      "loss": 6.2955,
      "mean_token_accuracy": 0.14297716021537782,
      "num_tokens": 2221760.0,
      "step": 970
    },
    {
      "entropy": 6.263871335983277,
      "epoch": 0.0936599423631124,
      "grad_norm": 1.15625,
      "learning_rate": 0.000487,
      "loss": 6.3112,
      "mean_token_accuracy": 0.13698594346642495,
      "num_tokens": 2233767.0,
      "step": 975
    },
    {
      "entropy": 6.315644836425781,
      "epoch": 0.0941402497598463,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004895,
      "loss": 6.2336,
      "mean_token_accuracy": 0.14062682241201402,
      "num_tokens": 2244841.0,
      "step": 980
    },
    {
      "entropy": 6.291112852096558,
      "epoch": 0.09462055715658022,
      "grad_norm": 1.078125,
      "learning_rate": 0.000492,
      "loss": 6.0876,
      "mean_token_accuracy": 0.14949826523661613,
      "num_tokens": 2255620.0,
      "step": 985
    },
    {
      "entropy": 6.233056163787841,
      "epoch": 0.09510086455331412,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004945,
      "loss": 6.2873,
      "mean_token_accuracy": 0.14527306705713272,
      "num_tokens": 2266345.0,
      "step": 990
    },
    {
      "entropy": 6.2935254096984865,
      "epoch": 0.09558117195004803,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000497,
      "loss": 6.2219,
      "mean_token_accuracy": 0.14342693910002707,
      "num_tokens": 2277190.0,
      "step": 995
    },
    {
      "entropy": 6.163033580780029,
      "epoch": 0.09606147934678194,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995,
      "loss": 6.0935,
      "mean_token_accuracy": 0.14876532852649688,
      "num_tokens": 2288241.0,
      "step": 1000
    },
    {
      "entropy": 6.305257749557495,
      "epoch": 0.09654178674351585,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999999983283737,
      "loss": 6.2413,
      "mean_token_accuracy": 0.14763403087854385,
      "num_tokens": 2301092.0,
      "step": 1005
    },
    {
      "entropy": 6.181101942062378,
      "epoch": 0.09702209414024976,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999999915373924,
      "loss": 6.1601,
      "mean_token_accuracy": 0.14240228086709977,
      "num_tokens": 2313622.0,
      "step": 1010
    },
    {
      "entropy": 6.148915004730225,
      "epoch": 0.09750240153698367,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999999795225793,
      "loss": 6.1146,
      "mean_token_accuracy": 0.14425086975097656,
      "num_tokens": 2325699.0,
      "step": 1015
    },
    {
      "entropy": 6.282630729675293,
      "epoch": 0.09798270893371758,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999999622839347,
      "loss": 6.181,
      "mean_token_accuracy": 0.14359124302864074,
      "num_tokens": 2337816.0,
      "step": 1020
    },
    {
      "entropy": 6.371134090423584,
      "epoch": 0.09846301633045149,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999999398214593,
      "loss": 6.306,
      "mean_token_accuracy": 0.13660114109516144,
      "num_tokens": 2348793.0,
      "step": 1025
    },
    {
      "entropy": 6.207452297210693,
      "epoch": 0.0989433237271854,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999999121351532,
      "loss": 6.2091,
      "mean_token_accuracy": 0.14510286301374437,
      "num_tokens": 2360209.0,
      "step": 1030
    },
    {
      "entropy": 6.3725972175598145,
      "epoch": 0.0994236311239193,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999998792250173,
      "loss": 6.2188,
      "mean_token_accuracy": 0.14163973331451415,
      "num_tokens": 2370597.0,
      "step": 1035
    },
    {
      "entropy": 6.260314416885376,
      "epoch": 0.09990393852065321,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004999998410910524,
      "loss": 6.2691,
      "mean_token_accuracy": 0.13516954630613326,
      "num_tokens": 2381466.0,
      "step": 1040
    },
    {
      "entropy": 6.353367328643799,
      "epoch": 0.10038424591738712,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999997977332592,
      "loss": 6.3668,
      "mean_token_accuracy": 0.13565330430865288,
      "num_tokens": 2394507.0,
      "step": 1045
    },
    {
      "entropy": 6.277764844894409,
      "epoch": 0.10086455331412104,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999997491516389,
      "loss": 6.2132,
      "mean_token_accuracy": 0.14350937008857728,
      "num_tokens": 2406367.0,
      "step": 1050
    },
    {
      "entropy": 6.184229898452759,
      "epoch": 0.10134486071085495,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004999996953461925,
      "loss": 6.1792,
      "mean_token_accuracy": 0.14410098493099213,
      "num_tokens": 2416962.0,
      "step": 1055
    },
    {
      "entropy": 6.294152784347534,
      "epoch": 0.10182516810758886,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999996363169212,
      "loss": 6.2309,
      "mean_token_accuracy": 0.13224372863769532,
      "num_tokens": 2429441.0,
      "step": 1060
    },
    {
      "entropy": 6.178936433792114,
      "epoch": 0.10230547550432277,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999995720638266,
      "loss": 6.1068,
      "mean_token_accuracy": 0.15178997665643693,
      "num_tokens": 2441360.0,
      "step": 1065
    },
    {
      "entropy": 6.242135000228882,
      "epoch": 0.10278578290105668,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049999950258691,
      "loss": 6.1192,
      "mean_token_accuracy": 0.1425897553563118,
      "num_tokens": 2452928.0,
      "step": 1070
    },
    {
      "entropy": 6.086752128601074,
      "epoch": 0.10326609029779059,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999994278861731,
      "loss": 6.0528,
      "mean_token_accuracy": 0.15216556191444397,
      "num_tokens": 2464218.0,
      "step": 1075
    },
    {
      "entropy": 6.244682741165161,
      "epoch": 0.1037463976945245,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999993479616175,
      "loss": 6.1495,
      "mean_token_accuracy": 0.15013783127069474,
      "num_tokens": 2475084.0,
      "step": 1080
    },
    {
      "entropy": 6.161567115783692,
      "epoch": 0.1042267050912584,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004999992628132451,
      "loss": 6.1502,
      "mean_token_accuracy": 0.142550940066576,
      "num_tokens": 2486845.0,
      "step": 1085
    },
    {
      "entropy": 6.204703760147095,
      "epoch": 0.10470701248799232,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999991724410582,
      "loss": 6.2282,
      "mean_token_accuracy": 0.13688293248414993,
      "num_tokens": 2498049.0,
      "step": 1090
    },
    {
      "entropy": 6.171539258956909,
      "epoch": 0.10518731988472622,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999990768450583,
      "loss": 6.1512,
      "mean_token_accuracy": 0.14735339730978012,
      "num_tokens": 2509511.0,
      "step": 1095
    },
    {
      "entropy": 6.291477632522583,
      "epoch": 0.10566762728146013,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999989760252482,
      "loss": 6.1154,
      "mean_token_accuracy": 0.15118677616119386,
      "num_tokens": 2521518.0,
      "step": 1100
    },
    {
      "entropy": 5.977184057235718,
      "epoch": 0.10614793467819404,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999988699816299,
      "loss": 5.9663,
      "mean_token_accuracy": 0.15758592933416365,
      "num_tokens": 2533141.0,
      "step": 1105
    },
    {
      "entropy": 6.063448810577393,
      "epoch": 0.10662824207492795,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999987587142058,
      "loss": 5.9728,
      "mean_token_accuracy": 0.14997318163514137,
      "num_tokens": 2542907.0,
      "step": 1110
    },
    {
      "entropy": 6.109343767166138,
      "epoch": 0.10710854947166186,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999986422229789,
      "loss": 6.1184,
      "mean_token_accuracy": 0.1399646908044815,
      "num_tokens": 2554497.0,
      "step": 1115
    },
    {
      "entropy": 6.291832542419433,
      "epoch": 0.10758885686839577,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999985205079514,
      "loss": 6.2366,
      "mean_token_accuracy": 0.14908648729324342,
      "num_tokens": 2567035.0,
      "step": 1120
    },
    {
      "entropy": 6.043786144256591,
      "epoch": 0.10806916426512968,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999983935691265,
      "loss": 6.007,
      "mean_token_accuracy": 0.15258836150169372,
      "num_tokens": 2578003.0,
      "step": 1125
    },
    {
      "entropy": 6.293131971359253,
      "epoch": 0.10854947166186359,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499998261406507,
      "loss": 6.3085,
      "mean_token_accuracy": 0.14156671091914177,
      "num_tokens": 2590272.0,
      "step": 1130
    },
    {
      "entropy": 6.125889730453491,
      "epoch": 0.10902977905859751,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999981240200958,
      "loss": 6.1307,
      "mean_token_accuracy": 0.14405303746461867,
      "num_tokens": 2602874.0,
      "step": 1135
    },
    {
      "entropy": 6.140783977508545,
      "epoch": 0.10951008645533142,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999979814098966,
      "loss": 6.0883,
      "mean_token_accuracy": 0.15575037002563477,
      "num_tokens": 2613835.0,
      "step": 1140
    },
    {
      "entropy": 6.244704246520996,
      "epoch": 0.10999039385206533,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999978335759121,
      "loss": 6.1962,
      "mean_token_accuracy": 0.14173132926225662,
      "num_tokens": 2624986.0,
      "step": 1145
    },
    {
      "entropy": 6.1988218307495115,
      "epoch": 0.11047070124879924,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999976805181461,
      "loss": 6.1939,
      "mean_token_accuracy": 0.1403311349451542,
      "num_tokens": 2637654.0,
      "step": 1150
    },
    {
      "entropy": 6.199357604980468,
      "epoch": 0.11095100864553314,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499997522236602,
      "loss": 6.1037,
      "mean_token_accuracy": 0.14716718941926957,
      "num_tokens": 2648947.0,
      "step": 1155
    },
    {
      "entropy": 6.130415296554565,
      "epoch": 0.11143131604226705,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999973587312837,
      "loss": 6.0963,
      "mean_token_accuracy": 0.14842916876077653,
      "num_tokens": 2659573.0,
      "step": 1160
    },
    {
      "entropy": 6.0729498863220215,
      "epoch": 0.11191162343900096,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999971900021947,
      "loss": 6.0778,
      "mean_token_accuracy": 0.15214993208646774,
      "num_tokens": 2671491.0,
      "step": 1165
    },
    {
      "entropy": 6.157928133010865,
      "epoch": 0.11239193083573487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999970160493391,
      "loss": 6.0571,
      "mean_token_accuracy": 0.15085275322198868,
      "num_tokens": 2682705.0,
      "step": 1170
    },
    {
      "entropy": 5.988412666320801,
      "epoch": 0.11287223823246878,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999968368727209,
      "loss": 5.9857,
      "mean_token_accuracy": 0.1512448124587536,
      "num_tokens": 2693539.0,
      "step": 1175
    },
    {
      "entropy": 6.143959999084473,
      "epoch": 0.11335254562920269,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999966524723442,
      "loss": 6.0359,
      "mean_token_accuracy": 0.15259774029254913,
      "num_tokens": 2705403.0,
      "step": 1180
    },
    {
      "entropy": 6.179290437698365,
      "epoch": 0.1138328530259366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999964628482135,
      "loss": 6.1553,
      "mean_token_accuracy": 0.14431207478046418,
      "num_tokens": 2717320.0,
      "step": 1185
    },
    {
      "entropy": 6.080302906036377,
      "epoch": 0.1143131604226705,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999962680003328,
      "loss": 6.1372,
      "mean_token_accuracy": 0.1427521824836731,
      "num_tokens": 2730056.0,
      "step": 1190
    },
    {
      "entropy": 6.139111661911011,
      "epoch": 0.11479346781940442,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499996067928707,
      "loss": 5.9903,
      "mean_token_accuracy": 0.15253911167383194,
      "num_tokens": 2741399.0,
      "step": 1195
    },
    {
      "entropy": 6.1936037063598635,
      "epoch": 0.11527377521613832,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999958626333406,
      "loss": 6.2196,
      "mean_token_accuracy": 0.14051759988069534,
      "num_tokens": 2752277.0,
      "step": 1200
    },
    {
      "entropy": 6.179323339462281,
      "epoch": 0.11575408261287223,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999956521142383,
      "loss": 6.1245,
      "mean_token_accuracy": 0.1430825725197792,
      "num_tokens": 2763172.0,
      "step": 1205
    },
    {
      "entropy": 6.169216156005859,
      "epoch": 0.11623439000960614,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999954363714051,
      "loss": 6.1048,
      "mean_token_accuracy": 0.1494622513651848,
      "num_tokens": 2775096.0,
      "step": 1210
    },
    {
      "entropy": 6.178955316543579,
      "epoch": 0.11671469740634005,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999952154048459,
      "loss": 6.208,
      "mean_token_accuracy": 0.14561985433101654,
      "num_tokens": 2786007.0,
      "step": 1215
    },
    {
      "entropy": 6.087857627868653,
      "epoch": 0.11719500480307397,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000499994989214566,
      "loss": 6.0045,
      "mean_token_accuracy": 0.15556134507060052,
      "num_tokens": 2797265.0,
      "step": 1220
    },
    {
      "entropy": 6.131901121139526,
      "epoch": 0.11767531219980788,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999947578005705,
      "loss": 6.1802,
      "mean_token_accuracy": 0.14725500270724295,
      "num_tokens": 2809427.0,
      "step": 1225
    },
    {
      "entropy": 6.175329542160034,
      "epoch": 0.11815561959654179,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999945211628648,
      "loss": 6.1007,
      "mean_token_accuracy": 0.15144748464226723,
      "num_tokens": 2822880.0,
      "step": 1230
    },
    {
      "entropy": 6.134575986862183,
      "epoch": 0.1186359269932757,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999942793014544,
      "loss": 6.1706,
      "mean_token_accuracy": 0.14877447858452797,
      "num_tokens": 2834760.0,
      "step": 1235
    },
    {
      "entropy": 6.1685347080230715,
      "epoch": 0.11911623439000961,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499994032216345,
      "loss": 6.107,
      "mean_token_accuracy": 0.14922730922698973,
      "num_tokens": 2846517.0,
      "step": 1240
    },
    {
      "entropy": 6.030606937408447,
      "epoch": 0.11959654178674352,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999937799075422,
      "loss": 5.9855,
      "mean_token_accuracy": 0.16014222726225852,
      "num_tokens": 2857280.0,
      "step": 1245
    },
    {
      "entropy": 6.123331689834595,
      "epoch": 0.12007684918347743,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499993522375052,
      "loss": 6.0458,
      "mean_token_accuracy": 0.15252126306295394,
      "num_tokens": 2869735.0,
      "step": 1250
    },
    {
      "entropy": 6.090895652770996,
      "epoch": 0.12055715658021134,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999932596188802,
      "loss": 6.0477,
      "mean_token_accuracy": 0.15124670863151551,
      "num_tokens": 2881681.0,
      "step": 1255
    },
    {
      "entropy": 6.028911304473877,
      "epoch": 0.12103746397694524,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999929916390331,
      "loss": 6.0556,
      "mean_token_accuracy": 0.1572458490729332,
      "num_tokens": 2893828.0,
      "step": 1260
    },
    {
      "entropy": 5.935788679122925,
      "epoch": 0.12151777137367915,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999927184355169,
      "loss": 6.0172,
      "mean_token_accuracy": 0.14968452900648116,
      "num_tokens": 2905804.0,
      "step": 1265
    },
    {
      "entropy": 6.311611938476562,
      "epoch": 0.12199807877041306,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999924400083377,
      "loss": 6.1599,
      "mean_token_accuracy": 0.1448735289275646,
      "num_tokens": 2918709.0,
      "step": 1270
    },
    {
      "entropy": 6.015513372421265,
      "epoch": 0.12247838616714697,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999921563575022,
      "loss": 6.079,
      "mean_token_accuracy": 0.1476942054927349,
      "num_tokens": 2930589.0,
      "step": 1275
    },
    {
      "entropy": 6.166260385513306,
      "epoch": 0.12295869356388088,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999918674830169,
      "loss": 6.1131,
      "mean_token_accuracy": 0.15547159016132356,
      "num_tokens": 2940635.0,
      "step": 1280
    },
    {
      "entropy": 6.008727741241455,
      "epoch": 0.12343900096061479,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999915733848886,
      "loss": 6.0079,
      "mean_token_accuracy": 0.15664828717708587,
      "num_tokens": 2952731.0,
      "step": 1285
    },
    {
      "entropy": 6.062463045120239,
      "epoch": 0.1239193083573487,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499991274063124,
      "loss": 6.0144,
      "mean_token_accuracy": 0.1488832116127014,
      "num_tokens": 2964544.0,
      "step": 1290
    },
    {
      "entropy": 6.022071123123169,
      "epoch": 0.12439961575408261,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999909695177301,
      "loss": 5.9402,
      "mean_token_accuracy": 0.1597997650504112,
      "num_tokens": 2974996.0,
      "step": 1295
    },
    {
      "entropy": 5.994995880126953,
      "epoch": 0.12487992315081652,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499990659748714,
      "loss": 5.9467,
      "mean_token_accuracy": 0.15251300781965255,
      "num_tokens": 2986496.0,
      "step": 1300
    },
    {
      "entropy": 6.049901390075684,
      "epoch": 0.12536023054755044,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999903447560828,
      "loss": 6.037,
      "mean_token_accuracy": 0.1527161717414856,
      "num_tokens": 2997693.0,
      "step": 1305
    },
    {
      "entropy": 6.033747673034668,
      "epoch": 0.12584053794428435,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999900245398439,
      "loss": 5.9853,
      "mean_token_accuracy": 0.15558816790580748,
      "num_tokens": 3009155.0,
      "step": 1310
    },
    {
      "entropy": 5.998337697982788,
      "epoch": 0.12632084534101826,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999896991000047,
      "loss": 5.9141,
      "mean_token_accuracy": 0.1596623882651329,
      "num_tokens": 3020944.0,
      "step": 1315
    },
    {
      "entropy": 5.977126979827881,
      "epoch": 0.12680115273775217,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999893684365729,
      "loss": 5.977,
      "mean_token_accuracy": 0.15605373829603195,
      "num_tokens": 3032893.0,
      "step": 1320
    },
    {
      "entropy": 6.0546129703521725,
      "epoch": 0.12728146013448607,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999890325495559,
      "loss": 6.0419,
      "mean_token_accuracy": 0.1505540929734707,
      "num_tokens": 3044937.0,
      "step": 1325
    },
    {
      "entropy": 5.99818902015686,
      "epoch": 0.12776176753121998,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999886914389617,
      "loss": 5.9324,
      "mean_token_accuracy": 0.15481951385736464,
      "num_tokens": 3056071.0,
      "step": 1330
    },
    {
      "entropy": 6.046021556854248,
      "epoch": 0.1282420749279539,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004999883451047981,
      "loss": 6.1056,
      "mean_token_accuracy": 0.15031077489256858,
      "num_tokens": 3067724.0,
      "step": 1335
    },
    {
      "entropy": 6.037922430038452,
      "epoch": 0.1287223823246878,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999879935470733,
      "loss": 5.9961,
      "mean_token_accuracy": 0.1605654314160347,
      "num_tokens": 3080779.0,
      "step": 1340
    },
    {
      "entropy": 6.017280960083008,
      "epoch": 0.1292026897214217,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999876367657954,
      "loss": 5.9936,
      "mean_token_accuracy": 0.15067695081233978,
      "num_tokens": 3091991.0,
      "step": 1345
    },
    {
      "entropy": 5.997524261474609,
      "epoch": 0.12968299711815562,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999872747609725,
      "loss": 5.9778,
      "mean_token_accuracy": 0.15370408892631532,
      "num_tokens": 3103713.0,
      "step": 1350
    },
    {
      "entropy": 6.033483362197876,
      "epoch": 0.13016330451488953,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999869075326132,
      "loss": 6.0116,
      "mean_token_accuracy": 0.15471786111593247,
      "num_tokens": 3115891.0,
      "step": 1355
    },
    {
      "entropy": 5.958883380889892,
      "epoch": 0.13064361191162344,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499986535080726,
      "loss": 5.8745,
      "mean_token_accuracy": 0.16241799741983415,
      "num_tokens": 3126223.0,
      "step": 1360
    },
    {
      "entropy": 6.039581346511841,
      "epoch": 0.13112391930835735,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999861574053196,
      "loss": 6.0623,
      "mean_token_accuracy": 0.14615725725889206,
      "num_tokens": 3137537.0,
      "step": 1365
    },
    {
      "entropy": 6.128436279296875,
      "epoch": 0.13160422670509125,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999857745064027,
      "loss": 6.0553,
      "mean_token_accuracy": 0.1591821163892746,
      "num_tokens": 3150126.0,
      "step": 1370
    },
    {
      "entropy": 6.011444139480591,
      "epoch": 0.13208453410182516,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499985386383984,
      "loss": 6.0267,
      "mean_token_accuracy": 0.1561764433979988,
      "num_tokens": 3161684.0,
      "step": 1375
    },
    {
      "entropy": 5.9896073818206785,
      "epoch": 0.13256484149855907,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999849930380729,
      "loss": 5.9039,
      "mean_token_accuracy": 0.1553078919649124,
      "num_tokens": 3172342.0,
      "step": 1380
    },
    {
      "entropy": 6.075509262084961,
      "epoch": 0.13304514889529298,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999845944686781,
      "loss": 6.0467,
      "mean_token_accuracy": 0.15783883035182952,
      "num_tokens": 3183681.0,
      "step": 1385
    },
    {
      "entropy": 5.887626361846924,
      "epoch": 0.1335254562920269,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999841906758093,
      "loss": 5.8602,
      "mean_token_accuracy": 0.16416978389024733,
      "num_tokens": 3194122.0,
      "step": 1390
    },
    {
      "entropy": 5.952710771560669,
      "epoch": 0.1340057636887608,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999837816594757,
      "loss": 5.9189,
      "mean_token_accuracy": 0.1564559370279312,
      "num_tokens": 3205719.0,
      "step": 1395
    },
    {
      "entropy": 5.924874639511108,
      "epoch": 0.1344860710854947,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999833674196865,
      "loss": 5.9381,
      "mean_token_accuracy": 0.15794704854488373,
      "num_tokens": 3216262.0,
      "step": 1400
    },
    {
      "entropy": 6.069398736953735,
      "epoch": 0.13496637848222862,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999829479564518,
      "loss": 6.0079,
      "mean_token_accuracy": 0.14638098031282426,
      "num_tokens": 3228035.0,
      "step": 1405
    },
    {
      "entropy": 5.878259229660034,
      "epoch": 0.13544668587896252,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499982523269781,
      "loss": 5.8583,
      "mean_token_accuracy": 0.1624181777238846,
      "num_tokens": 3239643.0,
      "step": 1410
    },
    {
      "entropy": 6.007868909835816,
      "epoch": 0.13592699327569643,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999820933596842,
      "loss": 5.9453,
      "mean_token_accuracy": 0.15799377262592315,
      "num_tokens": 3250293.0,
      "step": 1415
    },
    {
      "entropy": 6.07816219329834,
      "epoch": 0.13640730067243034,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499981658226171,
      "loss": 6.0542,
      "mean_token_accuracy": 0.14943305402994156,
      "num_tokens": 3262448.0,
      "step": 1420
    },
    {
      "entropy": 5.908995771408081,
      "epoch": 0.13688760806916425,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499981217869252,
      "loss": 5.929,
      "mean_token_accuracy": 0.15811057537794113,
      "num_tokens": 3274041.0,
      "step": 1425
    },
    {
      "entropy": 6.054948139190674,
      "epoch": 0.1373679154658982,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499980772288937,
      "loss": 5.9561,
      "mean_token_accuracy": 0.15184218436479568,
      "num_tokens": 3285063.0,
      "step": 1430
    },
    {
      "entropy": 5.976988172531128,
      "epoch": 0.1378482228626321,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999803214852367,
      "loss": 5.977,
      "mean_token_accuracy": 0.1544366106390953,
      "num_tokens": 3297031.0,
      "step": 1435
    },
    {
      "entropy": 6.070099973678589,
      "epoch": 0.138328530259366,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999798654581613,
      "loss": 5.947,
      "mean_token_accuracy": 0.15437746345996856,
      "num_tokens": 3308187.0,
      "step": 1440
    },
    {
      "entropy": 5.916920280456543,
      "epoch": 0.13880883765609991,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999794042077214,
      "loss": 5.9652,
      "mean_token_accuracy": 0.15356913805007935,
      "num_tokens": 3319439.0,
      "step": 1445
    },
    {
      "entropy": 6.065345001220703,
      "epoch": 0.13928914505283382,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999789377339279,
      "loss": 5.9672,
      "mean_token_accuracy": 0.15113933831453324,
      "num_tokens": 3331715.0,
      "step": 1450
    },
    {
      "entropy": 5.8917687892913815,
      "epoch": 0.13976945244956773,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999784660367915,
      "loss": 5.9209,
      "mean_token_accuracy": 0.1623266249895096,
      "num_tokens": 3342281.0,
      "step": 1455
    },
    {
      "entropy": 6.142686176300049,
      "epoch": 0.14024975984630164,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999779891163231,
      "loss": 6.0476,
      "mean_token_accuracy": 0.1508173242211342,
      "num_tokens": 3353530.0,
      "step": 1460
    },
    {
      "entropy": 5.919660520553589,
      "epoch": 0.14073006724303555,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999775069725339,
      "loss": 6.0189,
      "mean_token_accuracy": 0.15143757462501525,
      "num_tokens": 3365083.0,
      "step": 1465
    },
    {
      "entropy": 6.090314149856567,
      "epoch": 0.14121037463976946,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499977019605435,
      "loss": 5.972,
      "mean_token_accuracy": 0.15416960716247557,
      "num_tokens": 3378143.0,
      "step": 1470
    },
    {
      "entropy": 5.997537088394165,
      "epoch": 0.14169068203650337,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999765270150378,
      "loss": 5.9514,
      "mean_token_accuracy": 0.15883016735315322,
      "num_tokens": 3388438.0,
      "step": 1475
    },
    {
      "entropy": 5.8716082096099855,
      "epoch": 0.14217098943323728,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999760292013536,
      "loss": 5.95,
      "mean_token_accuracy": 0.15198695808649063,
      "num_tokens": 3400150.0,
      "step": 1480
    },
    {
      "entropy": 5.939674329757691,
      "epoch": 0.14265129682997119,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999755261643941,
      "loss": 5.827,
      "mean_token_accuracy": 0.16298246681690215,
      "num_tokens": 3411073.0,
      "step": 1485
    },
    {
      "entropy": 5.8928299903869625,
      "epoch": 0.1431316042267051,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999750179041709,
      "loss": 5.8905,
      "mean_token_accuracy": 0.1664283663034439,
      "num_tokens": 3422196.0,
      "step": 1490
    },
    {
      "entropy": 6.0256095886230465,
      "epoch": 0.143611911623439,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999745044206959,
      "loss": 5.9115,
      "mean_token_accuracy": 0.15011470019817352,
      "num_tokens": 3432568.0,
      "step": 1495
    },
    {
      "entropy": 5.954646492004395,
      "epoch": 0.1440922190201729,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999739857139809,
      "loss": 5.9423,
      "mean_token_accuracy": 0.1548205330967903,
      "num_tokens": 3444096.0,
      "step": 1500
    },
    {
      "entropy": 5.87907772064209,
      "epoch": 0.14457252641690682,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499973461784038,
      "loss": 5.8361,
      "mean_token_accuracy": 0.16604594886302948,
      "num_tokens": 3455502.0,
      "step": 1505
    },
    {
      "entropy": 5.99258599281311,
      "epoch": 0.14505283381364073,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999729326308792,
      "loss": 5.9705,
      "mean_token_accuracy": 0.15101340860128404,
      "num_tokens": 3467753.0,
      "step": 1510
    },
    {
      "entropy": 5.8902027130126955,
      "epoch": 0.14553314121037464,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499972398254517,
      "loss": 5.835,
      "mean_token_accuracy": 0.16128894686698914,
      "num_tokens": 3479206.0,
      "step": 1515
    },
    {
      "entropy": 5.944278573989868,
      "epoch": 0.14601344860710855,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499971858654964,
      "loss": 5.8652,
      "mean_token_accuracy": 0.15937050879001619,
      "num_tokens": 3490914.0,
      "step": 1520
    },
    {
      "entropy": 5.884905815124512,
      "epoch": 0.14649375600384246,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999713138322321,
      "loss": 5.9293,
      "mean_token_accuracy": 0.15429736226797103,
      "num_tokens": 3502932.0,
      "step": 1525
    },
    {
      "entropy": 5.872306871414184,
      "epoch": 0.14697406340057637,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999707637863346,
      "loss": 5.836,
      "mean_token_accuracy": 0.16672665476799012,
      "num_tokens": 3515071.0,
      "step": 1530
    },
    {
      "entropy": 6.054158544540405,
      "epoch": 0.14745437079731027,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999702085172838,
      "loss": 5.9631,
      "mean_token_accuracy": 0.15781875029206277,
      "num_tokens": 3526845.0,
      "step": 1535
    },
    {
      "entropy": 5.88759765625,
      "epoch": 0.14793467819404418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999696480250929,
      "loss": 5.9392,
      "mean_token_accuracy": 0.15117181837558746,
      "num_tokens": 3538810.0,
      "step": 1540
    },
    {
      "entropy": 5.999429655075073,
      "epoch": 0.1484149855907781,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999690823097747,
      "loss": 5.9088,
      "mean_token_accuracy": 0.16723537892103196,
      "num_tokens": 3549570.0,
      "step": 1545
    },
    {
      "entropy": 5.722354888916016,
      "epoch": 0.148895292987512,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999685113713426,
      "loss": 5.8273,
      "mean_token_accuracy": 0.1671121299266815,
      "num_tokens": 3560937.0,
      "step": 1550
    },
    {
      "entropy": 5.976478576660156,
      "epoch": 0.1493756003842459,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999679352098096,
      "loss": 5.8734,
      "mean_token_accuracy": 0.15577627420425416,
      "num_tokens": 3573603.0,
      "step": 1555
    },
    {
      "entropy": 5.879854822158814,
      "epoch": 0.14985590778097982,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999673538251891,
      "loss": 5.8848,
      "mean_token_accuracy": 0.15590892136096954,
      "num_tokens": 3586034.0,
      "step": 1560
    },
    {
      "entropy": 5.8598157405853275,
      "epoch": 0.15033621517771373,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999667672174947,
      "loss": 5.8176,
      "mean_token_accuracy": 0.1654493823647499,
      "num_tokens": 3597305.0,
      "step": 1565
    },
    {
      "entropy": 6.031460905075074,
      "epoch": 0.15081652257444764,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00049996617538674,
      "loss": 5.9745,
      "mean_token_accuracy": 0.1476095847785473,
      "num_tokens": 3608758.0,
      "step": 1570
    },
    {
      "entropy": 5.989815330505371,
      "epoch": 0.15129682997118155,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999655783329386,
      "loss": 5.9659,
      "mean_token_accuracy": 0.1517655596137047,
      "num_tokens": 3621508.0,
      "step": 1575
    },
    {
      "entropy": 5.9538108825683596,
      "epoch": 0.15177713736791545,
      "grad_norm": 0.875,
      "learning_rate": 0.0004999649760561046,
      "loss": 5.9111,
      "mean_token_accuracy": 0.15990381091833114,
      "num_tokens": 3634826.0,
      "step": 1580
    },
    {
      "entropy": 5.782315063476562,
      "epoch": 0.15225744476464936,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999643685562519,
      "loss": 5.7586,
      "mean_token_accuracy": 0.17986451387405394,
      "num_tokens": 3646427.0,
      "step": 1585
    },
    {
      "entropy": 5.871938896179199,
      "epoch": 0.15273775216138327,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999637558333945,
      "loss": 5.8846,
      "mean_token_accuracy": 0.15845945328474045,
      "num_tokens": 3658157.0,
      "step": 1590
    },
    {
      "entropy": 5.965056943893432,
      "epoch": 0.15321805955811718,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999631378875467,
      "loss": 5.9283,
      "mean_token_accuracy": 0.15535358637571334,
      "num_tokens": 3670695.0,
      "step": 1595
    },
    {
      "entropy": 5.852532148361206,
      "epoch": 0.15369836695485112,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999625147187228,
      "loss": 5.847,
      "mean_token_accuracy": 0.15907091200351714,
      "num_tokens": 3683379.0,
      "step": 1600
    },
    {
      "entropy": 5.896727418899536,
      "epoch": 0.15417867435158503,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999618863269373,
      "loss": 5.7427,
      "mean_token_accuracy": 0.1712520807981491,
      "num_tokens": 3694411.0,
      "step": 1605
    },
    {
      "entropy": 5.7830735206604,
      "epoch": 0.15465898174831894,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999612527122049,
      "loss": 5.7999,
      "mean_token_accuracy": 0.164157934486866,
      "num_tokens": 3703833.0,
      "step": 1610
    },
    {
      "entropy": 5.98201847076416,
      "epoch": 0.15513928914505284,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999606138745402,
      "loss": 5.8637,
      "mean_token_accuracy": 0.17183852046728135,
      "num_tokens": 3714702.0,
      "step": 1615
    },
    {
      "entropy": 5.91125226020813,
      "epoch": 0.15561959654178675,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999599698139581,
      "loss": 5.9601,
      "mean_token_accuracy": 0.16254139840602874,
      "num_tokens": 3726000.0,
      "step": 1620
    },
    {
      "entropy": 5.942967700958252,
      "epoch": 0.15609990393852066,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999593205304734,
      "loss": 5.8441,
      "mean_token_accuracy": 0.15695785135030746,
      "num_tokens": 3737681.0,
      "step": 1625
    },
    {
      "entropy": 5.980777454376221,
      "epoch": 0.15658021133525457,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999586660241012,
      "loss": 5.9471,
      "mean_token_accuracy": 0.1590610146522522,
      "num_tokens": 3749477.0,
      "step": 1630
    },
    {
      "entropy": 5.983596611022949,
      "epoch": 0.15706051873198848,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999580062948569,
      "loss": 5.9269,
      "mean_token_accuracy": 0.15622665733098984,
      "num_tokens": 3759946.0,
      "step": 1635
    },
    {
      "entropy": 5.82830491065979,
      "epoch": 0.1575408261287224,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999573413427556,
      "loss": 5.8301,
      "mean_token_accuracy": 0.1631380781531334,
      "num_tokens": 3770147.0,
      "step": 1640
    },
    {
      "entropy": 5.879538917541504,
      "epoch": 0.1580211335254563,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999566711678128,
      "loss": 5.789,
      "mean_token_accuracy": 0.16764923334121704,
      "num_tokens": 3781028.0,
      "step": 1645
    },
    {
      "entropy": 5.877831649780274,
      "epoch": 0.1585014409221902,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999559957700442,
      "loss": 5.8577,
      "mean_token_accuracy": 0.15628671497106553,
      "num_tokens": 3792529.0,
      "step": 1650
    },
    {
      "entropy": 5.9889665126800535,
      "epoch": 0.15898174831892412,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999553151494653,
      "loss": 5.9658,
      "mean_token_accuracy": 0.16207647919654847,
      "num_tokens": 3804630.0,
      "step": 1655
    },
    {
      "entropy": 5.756636714935302,
      "epoch": 0.15946205571565802,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999546293060919,
      "loss": 5.7305,
      "mean_token_accuracy": 0.17414763569831848,
      "num_tokens": 3815019.0,
      "step": 1660
    },
    {
      "entropy": 5.884789562225341,
      "epoch": 0.15994236311239193,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00049995393823994,
      "loss": 5.8346,
      "mean_token_accuracy": 0.16163911521434784,
      "num_tokens": 3825648.0,
      "step": 1665
    },
    {
      "entropy": 5.865713882446289,
      "epoch": 0.16042267050912584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999532419510255,
      "loss": 5.8267,
      "mean_token_accuracy": 0.16119918525218963,
      "num_tokens": 3836779.0,
      "step": 1670
    },
    {
      "entropy": 5.831444215774536,
      "epoch": 0.16090297790585975,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499952540439365,
      "loss": 5.8686,
      "mean_token_accuracy": 0.1589963838458061,
      "num_tokens": 3848350.0,
      "step": 1675
    },
    {
      "entropy": 5.968743467330933,
      "epoch": 0.16138328530259366,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999518337049743,
      "loss": 5.8602,
      "mean_token_accuracy": 0.16284065395593644,
      "num_tokens": 3859792.0,
      "step": 1680
    },
    {
      "entropy": 5.835683107376099,
      "epoch": 0.16186359269932757,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049995112174787,
      "loss": 5.7993,
      "mean_token_accuracy": 0.16204592287540437,
      "num_tokens": 3871539.0,
      "step": 1685
    },
    {
      "entropy": 5.7748315811157225,
      "epoch": 0.16234390009606148,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999504045680687,
      "loss": 5.8424,
      "mean_token_accuracy": 0.1638094499707222,
      "num_tokens": 3885025.0,
      "step": 1690
    },
    {
      "entropy": 5.921676111221314,
      "epoch": 0.1628242074927954,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999496821655869,
      "loss": 5.9563,
      "mean_token_accuracy": 0.1556147813796997,
      "num_tokens": 3896519.0,
      "step": 1695
    },
    {
      "entropy": 5.8414802074432375,
      "epoch": 0.1633045148895293,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999489545404414,
      "loss": 5.7111,
      "mean_token_accuracy": 0.16769966781139373,
      "num_tokens": 3907636.0,
      "step": 1700
    },
    {
      "entropy": 5.826439046859742,
      "epoch": 0.1637848222862632,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999482216926493,
      "loss": 5.8457,
      "mean_token_accuracy": 0.1643659219145775,
      "num_tokens": 3917746.0,
      "step": 1705
    },
    {
      "entropy": 5.835819673538208,
      "epoch": 0.1642651296829971,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999474836222273,
      "loss": 5.7635,
      "mean_token_accuracy": 0.16848202645778657,
      "num_tokens": 3928901.0,
      "step": 1710
    },
    {
      "entropy": 5.7875734806060795,
      "epoch": 0.16474543707973102,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999467403291928,
      "loss": 5.8911,
      "mean_token_accuracy": 0.16683052182197572,
      "num_tokens": 3940094.0,
      "step": 1715
    },
    {
      "entropy": 5.896417474746704,
      "epoch": 0.16522574447646493,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999459918135628,
      "loss": 5.7946,
      "mean_token_accuracy": 0.1682782679796219,
      "num_tokens": 3952714.0,
      "step": 1720
    },
    {
      "entropy": 5.809536647796631,
      "epoch": 0.16570605187319884,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499945238075355,
      "loss": 5.7855,
      "mean_token_accuracy": 0.1620207592844963,
      "num_tokens": 3963107.0,
      "step": 1725
    },
    {
      "entropy": 5.8475425243377686,
      "epoch": 0.16618635926993275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999444791145865,
      "loss": 5.7234,
      "mean_token_accuracy": 0.16508166193962098,
      "num_tokens": 3973454.0,
      "step": 1730
    },
    {
      "entropy": 5.773418426513672,
      "epoch": 0.16666666666666666,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999437149312754,
      "loss": 5.838,
      "mean_token_accuracy": 0.1589142844080925,
      "num_tokens": 3985352.0,
      "step": 1735
    },
    {
      "entropy": 5.918348503112793,
      "epoch": 0.16714697406340057,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499942945525439,
      "loss": 5.8059,
      "mean_token_accuracy": 0.16445475369691848,
      "num_tokens": 3997088.0,
      "step": 1740
    },
    {
      "entropy": 5.87802152633667,
      "epoch": 0.16762728146013448,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999421708970954,
      "loss": 5.8404,
      "mean_token_accuracy": 0.1627303183078766,
      "num_tokens": 4008551.0,
      "step": 1745
    },
    {
      "entropy": 5.842701101303101,
      "epoch": 0.16810758885686838,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999413910462625,
      "loss": 5.8705,
      "mean_token_accuracy": 0.1607220634818077,
      "num_tokens": 4020886.0,
      "step": 1750
    },
    {
      "entropy": 5.901450777053833,
      "epoch": 0.1685878962536023,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999406059729586,
      "loss": 5.8082,
      "mean_token_accuracy": 0.16304053068161012,
      "num_tokens": 4032811.0,
      "step": 1755
    },
    {
      "entropy": 5.838007640838623,
      "epoch": 0.1690682036503362,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999398156772016,
      "loss": 5.795,
      "mean_token_accuracy": 0.17031055241823195,
      "num_tokens": 4044102.0,
      "step": 1760
    },
    {
      "entropy": 5.838218402862549,
      "epoch": 0.16954851104707014,
      "grad_norm": 1.046875,
      "learning_rate": 0.00049993902015901,
      "loss": 5.8305,
      "mean_token_accuracy": 0.16805302500724792,
      "num_tokens": 4056849.0,
      "step": 1765
    },
    {
      "entropy": 5.8025048732757565,
      "epoch": 0.17002881844380405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999382194184023,
      "loss": 5.8535,
      "mean_token_accuracy": 0.16151585429906845,
      "num_tokens": 4068392.0,
      "step": 1770
    },
    {
      "entropy": 5.792164945602417,
      "epoch": 0.17050912584053796,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999374134553972,
      "loss": 5.6781,
      "mean_token_accuracy": 0.1671793356537819,
      "num_tokens": 4078949.0,
      "step": 1775
    },
    {
      "entropy": 5.893219375610352,
      "epoch": 0.17098943323727187,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999366022700131,
      "loss": 5.8373,
      "mean_token_accuracy": 0.1613660380244255,
      "num_tokens": 4091432.0,
      "step": 1780
    },
    {
      "entropy": 5.718272161483765,
      "epoch": 0.17146974063400577,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999357858622691,
      "loss": 5.6977,
      "mean_token_accuracy": 0.1697030022740364,
      "num_tokens": 4103001.0,
      "step": 1785
    },
    {
      "entropy": 5.845357942581177,
      "epoch": 0.17195004803073968,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999349642321842,
      "loss": 5.8151,
      "mean_token_accuracy": 0.16591036021709443,
      "num_tokens": 4113570.0,
      "step": 1790
    },
    {
      "entropy": 5.829782104492187,
      "epoch": 0.1724303554274736,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999341373797772,
      "loss": 5.8416,
      "mean_token_accuracy": 0.15971099138259887,
      "num_tokens": 4125658.0,
      "step": 1795
    },
    {
      "entropy": 5.787124729156494,
      "epoch": 0.1729106628242075,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999333053050675,
      "loss": 5.7384,
      "mean_token_accuracy": 0.16838262975215912,
      "num_tokens": 4137424.0,
      "step": 1800
    },
    {
      "entropy": 5.815556621551513,
      "epoch": 0.1733909702209414,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999324680080744,
      "loss": 5.7858,
      "mean_token_accuracy": 0.16323755234479903,
      "num_tokens": 4149772.0,
      "step": 1805
    },
    {
      "entropy": 5.791418552398682,
      "epoch": 0.17387127761767532,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999316254888172,
      "loss": 5.7849,
      "mean_token_accuracy": 0.16932481080293654,
      "num_tokens": 4162246.0,
      "step": 1810
    },
    {
      "entropy": 5.744324827194214,
      "epoch": 0.17435158501440923,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999307777473157,
      "loss": 5.7563,
      "mean_token_accuracy": 0.17080532908439636,
      "num_tokens": 4173449.0,
      "step": 1815
    },
    {
      "entropy": 5.768918180465699,
      "epoch": 0.17483189241114314,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999299247835893,
      "loss": 5.773,
      "mean_token_accuracy": 0.16689184457063674,
      "num_tokens": 4185560.0,
      "step": 1820
    },
    {
      "entropy": 5.852003479003907,
      "epoch": 0.17531219980787704,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000499929066597658,
      "loss": 5.8456,
      "mean_token_accuracy": 0.15657656863331795,
      "num_tokens": 4197843.0,
      "step": 1825
    },
    {
      "entropy": 5.869591426849365,
      "epoch": 0.17579250720461095,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999282031895418,
      "loss": 5.6906,
      "mean_token_accuracy": 0.16604516059160232,
      "num_tokens": 4210009.0,
      "step": 1830
    },
    {
      "entropy": 5.797318696975708,
      "epoch": 0.17627281460134486,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999273345592604,
      "loss": 5.9445,
      "mean_token_accuracy": 0.15455603450536728,
      "num_tokens": 4222213.0,
      "step": 1835
    },
    {
      "entropy": 5.9059834480285645,
      "epoch": 0.17675312199807877,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999264607068343,
      "loss": 5.7864,
      "mean_token_accuracy": 0.16164651066064833,
      "num_tokens": 4234159.0,
      "step": 1840
    },
    {
      "entropy": 5.803954219818115,
      "epoch": 0.17723342939481268,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999255816322837,
      "loss": 5.8304,
      "mean_token_accuracy": 0.1548845611512661,
      "num_tokens": 4246358.0,
      "step": 1845
    },
    {
      "entropy": 5.9636155128479,
      "epoch": 0.1777137367915466,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499924697335629,
      "loss": 5.9285,
      "mean_token_accuracy": 0.15726198703050615,
      "num_tokens": 4256698.0,
      "step": 1850
    },
    {
      "entropy": 5.774808931350708,
      "epoch": 0.1781940441882805,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999238078168906,
      "loss": 5.7062,
      "mean_token_accuracy": 0.17923928201198577,
      "num_tokens": 4268007.0,
      "step": 1855
    },
    {
      "entropy": 5.802064657211304,
      "epoch": 0.1786743515850144,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999229130760894,
      "loss": 5.808,
      "mean_token_accuracy": 0.1669637754559517,
      "num_tokens": 4278583.0,
      "step": 1860
    },
    {
      "entropy": 5.939214849472046,
      "epoch": 0.17915465898174832,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499922013113246,
      "loss": 5.8335,
      "mean_token_accuracy": 0.16115633994340897,
      "num_tokens": 4290267.0,
      "step": 1865
    },
    {
      "entropy": 5.753700637817383,
      "epoch": 0.17963496637848222,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999211079283814,
      "loss": 5.7483,
      "mean_token_accuracy": 0.16866709738969804,
      "num_tokens": 4302177.0,
      "step": 1870
    },
    {
      "entropy": 5.758403539657593,
      "epoch": 0.18011527377521613,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999201975215164,
      "loss": 5.6937,
      "mean_token_accuracy": 0.16991591304540635,
      "num_tokens": 4312420.0,
      "step": 1875
    },
    {
      "entropy": 5.855390214920044,
      "epoch": 0.18059558117195004,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999192818926725,
      "loss": 5.8369,
      "mean_token_accuracy": 0.16301838606595992,
      "num_tokens": 4324050.0,
      "step": 1880
    },
    {
      "entropy": 5.798628568649292,
      "epoch": 0.18107588856868395,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999183610418706,
      "loss": 5.6856,
      "mean_token_accuracy": 0.16928814202547074,
      "num_tokens": 4335336.0,
      "step": 1885
    },
    {
      "entropy": 5.6459338665008545,
      "epoch": 0.18155619596541786,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999174349691322,
      "loss": 5.7078,
      "mean_token_accuracy": 0.1670696869492531,
      "num_tokens": 4346239.0,
      "step": 1890
    },
    {
      "entropy": 5.83158860206604,
      "epoch": 0.18203650336215177,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999165036744788,
      "loss": 5.7856,
      "mean_token_accuracy": 0.16568428725004197,
      "num_tokens": 4357195.0,
      "step": 1895
    },
    {
      "entropy": 5.748570966720581,
      "epoch": 0.18251681075888568,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999155671579322,
      "loss": 5.7083,
      "mean_token_accuracy": 0.16901581287384032,
      "num_tokens": 4368905.0,
      "step": 1900
    },
    {
      "entropy": 5.84704818725586,
      "epoch": 0.1829971181556196,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499914625419514,
      "loss": 5.7747,
      "mean_token_accuracy": 0.16582313925027847,
      "num_tokens": 4381708.0,
      "step": 1905
    },
    {
      "entropy": 5.680188226699829,
      "epoch": 0.1834774255523535,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999136784592459,
      "loss": 5.6413,
      "mean_token_accuracy": 0.18005160689353944,
      "num_tokens": 4393199.0,
      "step": 1910
    },
    {
      "entropy": 5.790105772018433,
      "epoch": 0.1839577329490874,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999127262771502,
      "loss": 5.8594,
      "mean_token_accuracy": 0.1600307747721672,
      "num_tokens": 4403986.0,
      "step": 1915
    },
    {
      "entropy": 5.8221275806427,
      "epoch": 0.1844380403458213,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999117688732487,
      "loss": 5.7076,
      "mean_token_accuracy": 0.16819000989198685,
      "num_tokens": 4415574.0,
      "step": 1920
    },
    {
      "entropy": 5.743313217163086,
      "epoch": 0.18491834774255522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999108062475638,
      "loss": 5.7407,
      "mean_token_accuracy": 0.17155899852514267,
      "num_tokens": 4427579.0,
      "step": 1925
    },
    {
      "entropy": 5.785640573501587,
      "epoch": 0.18539865513928913,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499909838400118,
      "loss": 5.7394,
      "mean_token_accuracy": 0.16696204692125322,
      "num_tokens": 4440045.0,
      "step": 1930
    },
    {
      "entropy": 5.786698246002198,
      "epoch": 0.18587896253602307,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999088653309334,
      "loss": 5.828,
      "mean_token_accuracy": 0.1624658688902855,
      "num_tokens": 4450129.0,
      "step": 1935
    },
    {
      "entropy": 5.755189228057861,
      "epoch": 0.18635926993275698,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999078870400329,
      "loss": 5.6705,
      "mean_token_accuracy": 0.1707078069448471,
      "num_tokens": 4461378.0,
      "step": 1940
    },
    {
      "entropy": 5.773273372650147,
      "epoch": 0.18683957732949089,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999069035274391,
      "loss": 5.7133,
      "mean_token_accuracy": 0.17081280052661896,
      "num_tokens": 4472249.0,
      "step": 1945
    },
    {
      "entropy": 5.776774835586548,
      "epoch": 0.1873198847262248,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999059147931747,
      "loss": 5.7758,
      "mean_token_accuracy": 0.16512089669704438,
      "num_tokens": 4484243.0,
      "step": 1950
    },
    {
      "entropy": 5.780983924865723,
      "epoch": 0.1878001921229587,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999049208372629,
      "loss": 5.6782,
      "mean_token_accuracy": 0.16991802006959916,
      "num_tokens": 4495776.0,
      "step": 1955
    },
    {
      "entropy": 5.752791500091552,
      "epoch": 0.1882804995196926,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999039216597267,
      "loss": 5.7767,
      "mean_token_accuracy": 0.1656827986240387,
      "num_tokens": 4508581.0,
      "step": 1960
    },
    {
      "entropy": 5.813444137573242,
      "epoch": 0.18876080691642652,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999029172605892,
      "loss": 5.7015,
      "mean_token_accuracy": 0.168874491751194,
      "num_tokens": 4519753.0,
      "step": 1965
    },
    {
      "entropy": 5.7711827754974365,
      "epoch": 0.18924111431316043,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999019076398738,
      "loss": 5.7721,
      "mean_token_accuracy": 0.1760159507393837,
      "num_tokens": 4529957.0,
      "step": 1970
    },
    {
      "entropy": 5.800232076644898,
      "epoch": 0.18972142170989434,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499900892797604,
      "loss": 5.7455,
      "mean_token_accuracy": 0.16272206157445906,
      "num_tokens": 4541181.0,
      "step": 1975
    },
    {
      "entropy": 5.730713272094727,
      "epoch": 0.19020172910662825,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998998727338031,
      "loss": 5.7044,
      "mean_token_accuracy": 0.16977352052927017,
      "num_tokens": 4552235.0,
      "step": 1980
    },
    {
      "entropy": 5.76020998954773,
      "epoch": 0.19068203650336216,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998988474484952,
      "loss": 5.7425,
      "mean_token_accuracy": 0.17006415724754334,
      "num_tokens": 4564045.0,
      "step": 1985
    },
    {
      "entropy": 5.718191051483155,
      "epoch": 0.19116234390009607,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998978169417038,
      "loss": 5.6126,
      "mean_token_accuracy": 0.17550907731056214,
      "num_tokens": 4575141.0,
      "step": 1990
    },
    {
      "entropy": 5.641534996032715,
      "epoch": 0.19164265129682997,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998967812134529,
      "loss": 5.739,
      "mean_token_accuracy": 0.16591082215309144,
      "num_tokens": 4586491.0,
      "step": 1995
    },
    {
      "entropy": 5.800497817993164,
      "epoch": 0.19212295869356388,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998957402637664,
      "loss": 5.7131,
      "mean_token_accuracy": 0.17069052755832673,
      "num_tokens": 4598421.0,
      "step": 2000
    },
    {
      "entropy": 5.761542940139771,
      "epoch": 0.1926032660902978,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998946940926687,
      "loss": 5.634,
      "mean_token_accuracy": 0.1821053981781006,
      "num_tokens": 4609688.0,
      "step": 2005
    },
    {
      "entropy": 5.751686668395996,
      "epoch": 0.1930835734870317,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499893642700184,
      "loss": 5.7993,
      "mean_token_accuracy": 0.16548632085323334,
      "num_tokens": 4621169.0,
      "step": 2010
    },
    {
      "entropy": 5.7683569431304935,
      "epoch": 0.1935638808837656,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998925860863368,
      "loss": 5.733,
      "mean_token_accuracy": 0.17027136385440828,
      "num_tokens": 4632512.0,
      "step": 2015
    },
    {
      "entropy": 5.796252727508545,
      "epoch": 0.19404418828049952,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998915242511516,
      "loss": 5.8012,
      "mean_token_accuracy": 0.16176183074712752,
      "num_tokens": 4644356.0,
      "step": 2020
    },
    {
      "entropy": 5.854101514816284,
      "epoch": 0.19452449567723343,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998904571946528,
      "loss": 5.7745,
      "mean_token_accuracy": 0.16625221967697143,
      "num_tokens": 4655601.0,
      "step": 2025
    },
    {
      "entropy": 5.79165997505188,
      "epoch": 0.19500480307396734,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004998893849168655,
      "loss": 5.7786,
      "mean_token_accuracy": 0.16630641520023345,
      "num_tokens": 4666813.0,
      "step": 2030
    },
    {
      "entropy": 5.824037742614746,
      "epoch": 0.19548511047070125,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998883074178144,
      "loss": 5.7745,
      "mean_token_accuracy": 0.16555878072977065,
      "num_tokens": 4679123.0,
      "step": 2035
    },
    {
      "entropy": 5.764426136016846,
      "epoch": 0.19596541786743515,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998872246975247,
      "loss": 5.6658,
      "mean_token_accuracy": 0.17124608904123306,
      "num_tokens": 4690485.0,
      "step": 2040
    },
    {
      "entropy": 5.614950132369995,
      "epoch": 0.19644572526416906,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004998861367560213,
      "loss": 5.6973,
      "mean_token_accuracy": 0.1706745520234108,
      "num_tokens": 4702573.0,
      "step": 2045
    },
    {
      "entropy": 5.8513716697692875,
      "epoch": 0.19692603266090297,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998850435933296,
      "loss": 5.7228,
      "mean_token_accuracy": 0.16488247960805893,
      "num_tokens": 4715653.0,
      "step": 2050
    },
    {
      "entropy": 5.668337106704712,
      "epoch": 0.19740634005763688,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998839452094749,
      "loss": 5.7278,
      "mean_token_accuracy": 0.1701984152197838,
      "num_tokens": 4726051.0,
      "step": 2055
    },
    {
      "entropy": 5.768125200271607,
      "epoch": 0.1978866474543708,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998828416044829,
      "loss": 5.8616,
      "mean_token_accuracy": 0.1592288315296173,
      "num_tokens": 4738176.0,
      "step": 2060
    },
    {
      "entropy": 5.79635853767395,
      "epoch": 0.1983669548511047,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499881732778379,
      "loss": 5.761,
      "mean_token_accuracy": 0.1653179332613945,
      "num_tokens": 4749547.0,
      "step": 2065
    },
    {
      "entropy": 5.7004289627075195,
      "epoch": 0.1988472622478386,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499880618731189,
      "loss": 5.673,
      "mean_token_accuracy": 0.16624890714883805,
      "num_tokens": 4761967.0,
      "step": 2070
    },
    {
      "entropy": 5.728426837921143,
      "epoch": 0.19932756964457252,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998794994629388,
      "loss": 5.6654,
      "mean_token_accuracy": 0.17760985493659973,
      "num_tokens": 4772847.0,
      "step": 2075
    },
    {
      "entropy": 5.7485308170318605,
      "epoch": 0.19980787704130643,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004998783749736545,
      "loss": 5.7332,
      "mean_token_accuracy": 0.1665830835700035,
      "num_tokens": 4785049.0,
      "step": 2080
    },
    {
      "entropy": 5.779238176345825,
      "epoch": 0.20028818443804033,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998772452633619,
      "loss": 5.7618,
      "mean_token_accuracy": 0.1638600841164589,
      "num_tokens": 4797484.0,
      "step": 2085
    },
    {
      "entropy": 5.727624607086182,
      "epoch": 0.20076849183477424,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998761103320876,
      "loss": 5.6787,
      "mean_token_accuracy": 0.16071517616510392,
      "num_tokens": 4809422.0,
      "step": 2090
    },
    {
      "entropy": 5.798519659042358,
      "epoch": 0.20124879923150815,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998749701798577,
      "loss": 5.8628,
      "mean_token_accuracy": 0.164070263504982,
      "num_tokens": 4822340.0,
      "step": 2095
    },
    {
      "entropy": 5.751769924163819,
      "epoch": 0.2017291066282421,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998738248066986,
      "loss": 5.6535,
      "mean_token_accuracy": 0.17146731466054915,
      "num_tokens": 4833382.0,
      "step": 2100
    },
    {
      "entropy": 5.6563629627227785,
      "epoch": 0.202209414024976,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998726742126372,
      "loss": 5.6462,
      "mean_token_accuracy": 0.1777672901749611,
      "num_tokens": 4843916.0,
      "step": 2105
    },
    {
      "entropy": 5.783173704147339,
      "epoch": 0.2026897214217099,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998715183976999,
      "loss": 5.6659,
      "mean_token_accuracy": 0.17276445925235748,
      "num_tokens": 4855420.0,
      "step": 2110
    },
    {
      "entropy": 5.747212839126587,
      "epoch": 0.20317002881844382,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998703573619137,
      "loss": 5.7408,
      "mean_token_accuracy": 0.16771334260702134,
      "num_tokens": 4866540.0,
      "step": 2115
    },
    {
      "entropy": 5.756590795516968,
      "epoch": 0.20365033621517772,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004998691911053056,
      "loss": 5.6757,
      "mean_token_accuracy": 0.16594192534685134,
      "num_tokens": 4878860.0,
      "step": 2120
    },
    {
      "entropy": 5.843172788619995,
      "epoch": 0.20413064361191163,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998680196279026,
      "loss": 5.8017,
      "mean_token_accuracy": 0.16326622664928436,
      "num_tokens": 4889692.0,
      "step": 2125
    },
    {
      "entropy": 5.814892292022705,
      "epoch": 0.20461095100864554,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998668429297319,
      "loss": 5.745,
      "mean_token_accuracy": 0.1690722644329071,
      "num_tokens": 4901337.0,
      "step": 2130
    },
    {
      "entropy": 5.6207257270812985,
      "epoch": 0.20509125840537945,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998656610108208,
      "loss": 5.6647,
      "mean_token_accuracy": 0.175962170958519,
      "num_tokens": 4912340.0,
      "step": 2135
    },
    {
      "entropy": 5.7022956848144535,
      "epoch": 0.20557156580211336,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998644738711969,
      "loss": 5.633,
      "mean_token_accuracy": 0.17715211808681489,
      "num_tokens": 4923429.0,
      "step": 2140
    },
    {
      "entropy": 5.744864225387573,
      "epoch": 0.20605187319884727,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998632815108874,
      "loss": 5.7004,
      "mean_token_accuracy": 0.16438920944929122,
      "num_tokens": 4934826.0,
      "step": 2145
    },
    {
      "entropy": 5.779660034179687,
      "epoch": 0.20653218059558118,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998620839299203,
      "loss": 5.7914,
      "mean_token_accuracy": 0.16467878818511963,
      "num_tokens": 4945189.0,
      "step": 2150
    },
    {
      "entropy": 5.814841079711914,
      "epoch": 0.2070124879923151,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998608811283233,
      "loss": 5.6626,
      "mean_token_accuracy": 0.17581747621297836,
      "num_tokens": 4956060.0,
      "step": 2155
    },
    {
      "entropy": 5.692324304580689,
      "epoch": 0.207492795389049,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998596731061244,
      "loss": 5.6785,
      "mean_token_accuracy": 0.17047660052776337,
      "num_tokens": 4967710.0,
      "step": 2160
    },
    {
      "entropy": 5.760519409179688,
      "epoch": 0.2079731027857829,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998584598633516,
      "loss": 5.7035,
      "mean_token_accuracy": 0.16868236809968948,
      "num_tokens": 4978716.0,
      "step": 2165
    },
    {
      "entropy": 5.747573614120483,
      "epoch": 0.2084534101825168,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998572414000329,
      "loss": 5.7636,
      "mean_token_accuracy": 0.16640378534793854,
      "num_tokens": 4990742.0,
      "step": 2170
    },
    {
      "entropy": 5.754537057876587,
      "epoch": 0.20893371757925072,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998560177161969,
      "loss": 5.7229,
      "mean_token_accuracy": 0.1728250965476036,
      "num_tokens": 5001217.0,
      "step": 2175
    },
    {
      "entropy": 5.7619335651397705,
      "epoch": 0.20941402497598463,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998547888118718,
      "loss": 5.7089,
      "mean_token_accuracy": 0.16733649522066116,
      "num_tokens": 5012750.0,
      "step": 2180
    },
    {
      "entropy": 5.703377151489258,
      "epoch": 0.20989433237271854,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998535546870862,
      "loss": 5.7162,
      "mean_token_accuracy": 0.17622778564691544,
      "num_tokens": 5023818.0,
      "step": 2185
    },
    {
      "entropy": 5.8201076030731205,
      "epoch": 0.21037463976945245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998523153418687,
      "loss": 5.7395,
      "mean_token_accuracy": 0.16917874813079833,
      "num_tokens": 5035060.0,
      "step": 2190
    },
    {
      "entropy": 5.7587230682373045,
      "epoch": 0.21085494716618636,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998510707762481,
      "loss": 5.7859,
      "mean_token_accuracy": 0.16076416671276092,
      "num_tokens": 5046780.0,
      "step": 2195
    },
    {
      "entropy": 5.703002882003784,
      "epoch": 0.21133525456292027,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998498209902533,
      "loss": 5.7498,
      "mean_token_accuracy": 0.1657275453209877,
      "num_tokens": 5059152.0,
      "step": 2200
    },
    {
      "entropy": 5.767661762237549,
      "epoch": 0.21181556195965417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998485659839134,
      "loss": 5.7165,
      "mean_token_accuracy": 0.16900938898324966,
      "num_tokens": 5069971.0,
      "step": 2205
    },
    {
      "entropy": 5.757136583328247,
      "epoch": 0.21229586935638808,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998473057572575,
      "loss": 5.6553,
      "mean_token_accuracy": 0.17132542431354522,
      "num_tokens": 5080640.0,
      "step": 2210
    },
    {
      "entropy": 5.691873693466187,
      "epoch": 0.212776176753122,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998460403103146,
      "loss": 5.6686,
      "mean_token_accuracy": 0.17301307767629623,
      "num_tokens": 5091240.0,
      "step": 2215
    },
    {
      "entropy": 5.691802358627319,
      "epoch": 0.2132564841498559,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998447696431146,
      "loss": 5.7072,
      "mean_token_accuracy": 0.16933174431324005,
      "num_tokens": 5102920.0,
      "step": 2220
    },
    {
      "entropy": 5.7080395221710205,
      "epoch": 0.2137367915465898,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998434937556865,
      "loss": 5.714,
      "mean_token_accuracy": 0.16653985381126404,
      "num_tokens": 5114240.0,
      "step": 2225
    },
    {
      "entropy": 5.8630828857421875,
      "epoch": 0.21421709894332372,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998422126480602,
      "loss": 5.9345,
      "mean_token_accuracy": 0.1564141757786274,
      "num_tokens": 5126653.0,
      "step": 2230
    },
    {
      "entropy": 5.923197984695435,
      "epoch": 0.21469740634005763,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998409263202653,
      "loss": 5.7951,
      "mean_token_accuracy": 0.1645052909851074,
      "num_tokens": 5138778.0,
      "step": 2235
    },
    {
      "entropy": 5.697419023513794,
      "epoch": 0.21517771373679154,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998396347723318,
      "loss": 5.6409,
      "mean_token_accuracy": 0.17507038712501527,
      "num_tokens": 5148670.0,
      "step": 2240
    },
    {
      "entropy": 5.607378005981445,
      "epoch": 0.21565802113352545,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998383380042895,
      "loss": 5.6838,
      "mean_token_accuracy": 0.1660774126648903,
      "num_tokens": 5159617.0,
      "step": 2245
    },
    {
      "entropy": 5.793371343612671,
      "epoch": 0.21613832853025935,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998370360161688,
      "loss": 5.7412,
      "mean_token_accuracy": 0.16822775602340698,
      "num_tokens": 5171225.0,
      "step": 2250
    },
    {
      "entropy": 5.766514635086059,
      "epoch": 0.21661863592699326,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998357288079996,
      "loss": 5.7291,
      "mean_token_accuracy": 0.16446396410465242,
      "num_tokens": 5184586.0,
      "step": 2255
    },
    {
      "entropy": 5.653764152526856,
      "epoch": 0.21709894332372717,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998344163798125,
      "loss": 5.6851,
      "mean_token_accuracy": 0.16855508089065552,
      "num_tokens": 5196089.0,
      "step": 2260
    },
    {
      "entropy": 5.756007528305053,
      "epoch": 0.21757925072046108,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998330987316379,
      "loss": 5.7378,
      "mean_token_accuracy": 0.16411580741405488,
      "num_tokens": 5207581.0,
      "step": 2265
    },
    {
      "entropy": 5.717509746551514,
      "epoch": 0.21805955811719502,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998317758635062,
      "loss": 5.6875,
      "mean_token_accuracy": 0.17126210331916808,
      "num_tokens": 5219108.0,
      "step": 2270
    },
    {
      "entropy": 5.717676448822021,
      "epoch": 0.21853986551392893,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998304477754484,
      "loss": 5.6361,
      "mean_token_accuracy": 0.17278024703264236,
      "num_tokens": 5231137.0,
      "step": 2275
    },
    {
      "entropy": 5.830537891387939,
      "epoch": 0.21902017291066284,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998291144674952,
      "loss": 5.7006,
      "mean_token_accuracy": 0.17262052595615388,
      "num_tokens": 5243528.0,
      "step": 2280
    },
    {
      "entropy": 5.570743799209595,
      "epoch": 0.21950048030739674,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998277759396776,
      "loss": 5.6544,
      "mean_token_accuracy": 0.17672596722841263,
      "num_tokens": 5254869.0,
      "step": 2285
    },
    {
      "entropy": 5.6864776611328125,
      "epoch": 0.21998078770413065,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998264321920265,
      "loss": 5.6374,
      "mean_token_accuracy": 0.17273785322904586,
      "num_tokens": 5265370.0,
      "step": 2290
    },
    {
      "entropy": 5.709369754791259,
      "epoch": 0.22046109510086456,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998250832245734,
      "loss": 5.6108,
      "mean_token_accuracy": 0.17232156842947005,
      "num_tokens": 5277812.0,
      "step": 2295
    },
    {
      "entropy": 5.591331195831299,
      "epoch": 0.22094140249759847,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998237290373494,
      "loss": 5.5895,
      "mean_token_accuracy": 0.17628773897886277,
      "num_tokens": 5289107.0,
      "step": 2300
    },
    {
      "entropy": 5.745526599884033,
      "epoch": 0.22142170989433238,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499822369630386,
      "loss": 5.659,
      "mean_token_accuracy": 0.16332831233739853,
      "num_tokens": 5301489.0,
      "step": 2305
    },
    {
      "entropy": 5.575994968414307,
      "epoch": 0.2219020172910663,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998210050037148,
      "loss": 5.625,
      "mean_token_accuracy": 0.17718679308891297,
      "num_tokens": 5313273.0,
      "step": 2310
    },
    {
      "entropy": 5.763540124893188,
      "epoch": 0.2223823246878002,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998196351573674,
      "loss": 5.6748,
      "mean_token_accuracy": 0.16899870485067367,
      "num_tokens": 5324227.0,
      "step": 2315
    },
    {
      "entropy": 5.596045875549317,
      "epoch": 0.2228626320845341,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998182600913757,
      "loss": 5.6037,
      "mean_token_accuracy": 0.1776882916688919,
      "num_tokens": 5335138.0,
      "step": 2320
    },
    {
      "entropy": 5.692679834365845,
      "epoch": 0.22334293948126802,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998168798057715,
      "loss": 5.6801,
      "mean_token_accuracy": 0.17184858918190002,
      "num_tokens": 5345652.0,
      "step": 2325
    },
    {
      "entropy": 5.842924690246582,
      "epoch": 0.22382324687800192,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499815494300587,
      "loss": 5.7807,
      "mean_token_accuracy": 0.16769531071186067,
      "num_tokens": 5357999.0,
      "step": 2330
    },
    {
      "entropy": 5.695201444625854,
      "epoch": 0.22430355427473583,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998141035758542,
      "loss": 5.6655,
      "mean_token_accuracy": 0.16747171431779861,
      "num_tokens": 5369032.0,
      "step": 2335
    },
    {
      "entropy": 5.685867786407471,
      "epoch": 0.22478386167146974,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998127076316054,
      "loss": 5.5858,
      "mean_token_accuracy": 0.17415026128292083,
      "num_tokens": 5380692.0,
      "step": 2340
    },
    {
      "entropy": 5.738334608078003,
      "epoch": 0.22526416906820365,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998113064678734,
      "loss": 5.7279,
      "mean_token_accuracy": 0.16383333057165145,
      "num_tokens": 5392320.0,
      "step": 2345
    },
    {
      "entropy": 5.717440557479859,
      "epoch": 0.22574447646493756,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004998099000846901,
      "loss": 5.7129,
      "mean_token_accuracy": 0.17294058203697205,
      "num_tokens": 5404374.0,
      "step": 2350
    },
    {
      "entropy": 5.743234443664551,
      "epoch": 0.22622478386167147,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998084884820887,
      "loss": 5.7044,
      "mean_token_accuracy": 0.16908725798130037,
      "num_tokens": 5415471.0,
      "step": 2355
    },
    {
      "entropy": 5.572421836853027,
      "epoch": 0.22670509125840538,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004998070716601016,
      "loss": 5.5376,
      "mean_token_accuracy": 0.1757811814546585,
      "num_tokens": 5425971.0,
      "step": 2360
    },
    {
      "entropy": 5.761317300796509,
      "epoch": 0.2271853986551393,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998056496187618,
      "loss": 5.6628,
      "mean_token_accuracy": 0.17541870921850206,
      "num_tokens": 5437638.0,
      "step": 2365
    },
    {
      "entropy": 5.695319604873657,
      "epoch": 0.2276657060518732,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998042223581025,
      "loss": 5.688,
      "mean_token_accuracy": 0.17006312757730485,
      "num_tokens": 5449463.0,
      "step": 2370
    },
    {
      "entropy": 5.75418791770935,
      "epoch": 0.2281460134486071,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998027898781565,
      "loss": 5.7672,
      "mean_token_accuracy": 0.16644203364849092,
      "num_tokens": 5460972.0,
      "step": 2375
    },
    {
      "entropy": 5.873421955108642,
      "epoch": 0.228626320845341,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998013521789574,
      "loss": 5.7659,
      "mean_token_accuracy": 0.1616092711687088,
      "num_tokens": 5471575.0,
      "step": 2380
    },
    {
      "entropy": 5.647591924667358,
      "epoch": 0.22910662824207492,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004997999092605384,
      "loss": 5.6174,
      "mean_token_accuracy": 0.17019406706094742,
      "num_tokens": 5483041.0,
      "step": 2385
    },
    {
      "entropy": 5.656691837310791,
      "epoch": 0.22958693563880883,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499798461122933,
      "loss": 5.6548,
      "mean_token_accuracy": 0.17062922716140747,
      "num_tokens": 5494040.0,
      "step": 2390
    },
    {
      "entropy": 5.595247745513916,
      "epoch": 0.23006724303554274,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997970077661748,
      "loss": 5.5368,
      "mean_token_accuracy": 0.18294250816106797,
      "num_tokens": 5504217.0,
      "step": 2395
    },
    {
      "entropy": 5.761266231536865,
      "epoch": 0.23054755043227665,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997955491902977,
      "loss": 5.7086,
      "mean_token_accuracy": 0.16774140298366547,
      "num_tokens": 5517180.0,
      "step": 2400
    },
    {
      "entropy": 5.650377893447876,
      "epoch": 0.23102785782901056,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997940853953354,
      "loss": 5.6279,
      "mean_token_accuracy": 0.1671198323369026,
      "num_tokens": 5529626.0,
      "step": 2405
    },
    {
      "entropy": 5.7746195793151855,
      "epoch": 0.23150816522574447,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499792616381322,
      "loss": 5.6646,
      "mean_token_accuracy": 0.1768963173031807,
      "num_tokens": 5540844.0,
      "step": 2410
    },
    {
      "entropy": 5.665439796447754,
      "epoch": 0.23198847262247838,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997911421482914,
      "loss": 5.5842,
      "mean_token_accuracy": 0.1775585025548935,
      "num_tokens": 5551732.0,
      "step": 2415
    },
    {
      "entropy": 5.7055745124816895,
      "epoch": 0.23246878001921228,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499789662696278,
      "loss": 5.7224,
      "mean_token_accuracy": 0.17121947556734085,
      "num_tokens": 5562692.0,
      "step": 2420
    },
    {
      "entropy": 5.830398082733154,
      "epoch": 0.2329490874159462,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997881780253162,
      "loss": 5.7108,
      "mean_token_accuracy": 0.16949707716703416,
      "num_tokens": 5574884.0,
      "step": 2425
    },
    {
      "entropy": 5.5266454219818115,
      "epoch": 0.2334293948126801,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997866881354403,
      "loss": 5.5626,
      "mean_token_accuracy": 0.17615221440792084,
      "num_tokens": 5585335.0,
      "step": 2430
    },
    {
      "entropy": 5.79819769859314,
      "epoch": 0.23390970220941404,
      "grad_norm": 1.0,
      "learning_rate": 0.000499785193026685,
      "loss": 5.7122,
      "mean_token_accuracy": 0.16447871327400207,
      "num_tokens": 5597621.0,
      "step": 2435
    },
    {
      "entropy": 5.555981016159057,
      "epoch": 0.23439000960614795,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997836926990851,
      "loss": 5.567,
      "mean_token_accuracy": 0.1729709431529045,
      "num_tokens": 5609795.0,
      "step": 2440
    },
    {
      "entropy": 5.7423711776733395,
      "epoch": 0.23487031700288186,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997821871526752,
      "loss": 5.6678,
      "mean_token_accuracy": 0.17034737765789032,
      "num_tokens": 5621473.0,
      "step": 2445
    },
    {
      "entropy": 5.803370666503906,
      "epoch": 0.23535062439961577,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997806763874905,
      "loss": 5.7509,
      "mean_token_accuracy": 0.1693995103240013,
      "num_tokens": 5635064.0,
      "step": 2450
    },
    {
      "entropy": 5.740002870559692,
      "epoch": 0.23583093179634967,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997791604035659,
      "loss": 5.7351,
      "mean_token_accuracy": 0.1653935804963112,
      "num_tokens": 5647382.0,
      "step": 2455
    },
    {
      "entropy": 5.6946430683135985,
      "epoch": 0.23631123919308358,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997776392009366,
      "loss": 5.6227,
      "mean_token_accuracy": 0.17229534089565277,
      "num_tokens": 5658417.0,
      "step": 2460
    },
    {
      "entropy": 5.633540248870849,
      "epoch": 0.2367915465898175,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997761127796381,
      "loss": 5.5924,
      "mean_token_accuracy": 0.1780801773071289,
      "num_tokens": 5672190.0,
      "step": 2465
    },
    {
      "entropy": 5.630267906188965,
      "epoch": 0.2372718539865514,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997745811397056,
      "loss": 5.6343,
      "mean_token_accuracy": 0.17877777218818663,
      "num_tokens": 5684259.0,
      "step": 2470
    },
    {
      "entropy": 5.684125900268555,
      "epoch": 0.2377521613832853,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997730442811748,
      "loss": 5.6829,
      "mean_token_accuracy": 0.1745256558060646,
      "num_tokens": 5696505.0,
      "step": 2475
    },
    {
      "entropy": 5.632245492935181,
      "epoch": 0.23823246878001922,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997715022040814,
      "loss": 5.5737,
      "mean_token_accuracy": 0.17421737164258957,
      "num_tokens": 5707638.0,
      "step": 2480
    },
    {
      "entropy": 5.642940044403076,
      "epoch": 0.23871277617675313,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499769954908461,
      "loss": 5.5426,
      "mean_token_accuracy": 0.1817597672343254,
      "num_tokens": 5719098.0,
      "step": 2485
    },
    {
      "entropy": 5.824435234069824,
      "epoch": 0.23919308357348704,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997684023943498,
      "loss": 5.7816,
      "mean_token_accuracy": 0.15434520691633224,
      "num_tokens": 5731345.0,
      "step": 2490
    },
    {
      "entropy": 5.546219301223755,
      "epoch": 0.23967339097022095,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997668446617837,
      "loss": 5.5331,
      "mean_token_accuracy": 0.17521776705980302,
      "num_tokens": 5742604.0,
      "step": 2495
    },
    {
      "entropy": 5.592561054229736,
      "epoch": 0.24015369836695485,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997652817107989,
      "loss": 5.5647,
      "mean_token_accuracy": 0.1803002640604973,
      "num_tokens": 5754620.0,
      "step": 2500
    },
    {
      "entropy": 5.711943244934082,
      "epoch": 0.24063400576368876,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997637135414315,
      "loss": 5.6407,
      "mean_token_accuracy": 0.17505031377077102,
      "num_tokens": 5764764.0,
      "step": 2505
    },
    {
      "entropy": 5.5460649013519285,
      "epoch": 0.24111431316042267,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997621401537183,
      "loss": 5.5314,
      "mean_token_accuracy": 0.1876803979277611,
      "num_tokens": 5776341.0,
      "step": 2510
    },
    {
      "entropy": 5.738549900054932,
      "epoch": 0.24159462055715658,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997605615476955,
      "loss": 5.7606,
      "mean_token_accuracy": 0.16073987782001495,
      "num_tokens": 5788604.0,
      "step": 2515
    },
    {
      "entropy": 5.74634370803833,
      "epoch": 0.2420749279538905,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997589777234,
      "loss": 5.6553,
      "mean_token_accuracy": 0.17228216230869292,
      "num_tokens": 5800194.0,
      "step": 2520
    },
    {
      "entropy": 5.6268805980682375,
      "epoch": 0.2425552353506244,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997573886808684,
      "loss": 5.6295,
      "mean_token_accuracy": 0.17930565178394317,
      "num_tokens": 5812978.0,
      "step": 2525
    },
    {
      "entropy": 5.637392854690551,
      "epoch": 0.2430355427473583,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997557944201375,
      "loss": 5.5556,
      "mean_token_accuracy": 0.18395199477672577,
      "num_tokens": 5824591.0,
      "step": 2530
    },
    {
      "entropy": 5.614971017837524,
      "epoch": 0.24351585014409222,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997541949412445,
      "loss": 5.6467,
      "mean_token_accuracy": 0.1709784209728241,
      "num_tokens": 5835970.0,
      "step": 2535
    },
    {
      "entropy": 5.781296730041504,
      "epoch": 0.24399615754082613,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997525902442266,
      "loss": 5.6819,
      "mean_token_accuracy": 0.1686740532517433,
      "num_tokens": 5848687.0,
      "step": 2540
    },
    {
      "entropy": 5.643764638900757,
      "epoch": 0.24447646493756003,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997509803291207,
      "loss": 5.6604,
      "mean_token_accuracy": 0.17258683294057847,
      "num_tokens": 5860820.0,
      "step": 2545
    },
    {
      "entropy": 5.67121114730835,
      "epoch": 0.24495677233429394,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997493651959647,
      "loss": 5.5907,
      "mean_token_accuracy": 0.1715891554951668,
      "num_tokens": 5872172.0,
      "step": 2550
    },
    {
      "entropy": 5.692111682891846,
      "epoch": 0.24543707973102785,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997477448447955,
      "loss": 5.6336,
      "mean_token_accuracy": 0.17338863462209703,
      "num_tokens": 5883574.0,
      "step": 2555
    },
    {
      "entropy": 5.726970052719116,
      "epoch": 0.24591738712776176,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997461192756512,
      "loss": 5.7262,
      "mean_token_accuracy": 0.17206473648548126,
      "num_tokens": 5895223.0,
      "step": 2560
    },
    {
      "entropy": 5.618826961517334,
      "epoch": 0.24639769452449567,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997444884885694,
      "loss": 5.5582,
      "mean_token_accuracy": 0.17600789666175842,
      "num_tokens": 5907448.0,
      "step": 2565
    },
    {
      "entropy": 5.683489036560059,
      "epoch": 0.24687800192122958,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997428524835879,
      "loss": 5.6606,
      "mean_token_accuracy": 0.17142950296401976,
      "num_tokens": 5918906.0,
      "step": 2570
    },
    {
      "entropy": 5.6601954936981205,
      "epoch": 0.2473583093179635,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997412112607446,
      "loss": 5.5257,
      "mean_token_accuracy": 0.1818102404475212,
      "num_tokens": 5930619.0,
      "step": 2575
    },
    {
      "entropy": 5.708859300613403,
      "epoch": 0.2478386167146974,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997395648200778,
      "loss": 5.728,
      "mean_token_accuracy": 0.16916778534650803,
      "num_tokens": 5941554.0,
      "step": 2580
    },
    {
      "entropy": 5.779385662078857,
      "epoch": 0.2483189241114313,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997379131616257,
      "loss": 5.7036,
      "mean_token_accuracy": 0.17098585814237593,
      "num_tokens": 5953309.0,
      "step": 2585
    },
    {
      "entropy": 5.650564956665039,
      "epoch": 0.24879923150816521,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004997362562854266,
      "loss": 5.645,
      "mean_token_accuracy": 0.17450834512710572,
      "num_tokens": 5965286.0,
      "step": 2590
    },
    {
      "entropy": 5.632644748687744,
      "epoch": 0.24927953890489912,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997345941915187,
      "loss": 5.6166,
      "mean_token_accuracy": 0.17439770251512526,
      "num_tokens": 5976558.0,
      "step": 2595
    },
    {
      "entropy": 5.684119987487793,
      "epoch": 0.24975984630163303,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997329268799412,
      "loss": 5.662,
      "mean_token_accuracy": 0.16537658944725991,
      "num_tokens": 5988793.0,
      "step": 2600
    },
    {
      "entropy": 5.709760332107544,
      "epoch": 0.25024015369836694,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997312543507322,
      "loss": 5.6874,
      "mean_token_accuracy": 0.16319903284311293,
      "num_tokens": 5999825.0,
      "step": 2605
    },
    {
      "entropy": 5.706351470947266,
      "epoch": 0.2507204610951009,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997295766039309,
      "loss": 5.6318,
      "mean_token_accuracy": 0.170637284219265,
      "num_tokens": 6010408.0,
      "step": 2610
    },
    {
      "entropy": 5.758199882507324,
      "epoch": 0.25120076849183476,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997278936395761,
      "loss": 5.6601,
      "mean_token_accuracy": 0.17168794125318526,
      "num_tokens": 6021161.0,
      "step": 2615
    },
    {
      "entropy": 5.557199287414551,
      "epoch": 0.2516810758885687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997262054577071,
      "loss": 5.5178,
      "mean_token_accuracy": 0.181975619494915,
      "num_tokens": 6032646.0,
      "step": 2620
    },
    {
      "entropy": 5.698000192642212,
      "epoch": 0.2521613832853026,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997245120583627,
      "loss": 5.6845,
      "mean_token_accuracy": 0.17188096046447754,
      "num_tokens": 6045019.0,
      "step": 2625
    },
    {
      "entropy": 5.527618551254273,
      "epoch": 0.2526416906820365,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997228134415825,
      "loss": 5.5628,
      "mean_token_accuracy": 0.17528536915779114,
      "num_tokens": 6055644.0,
      "step": 2630
    },
    {
      "entropy": 5.727271509170532,
      "epoch": 0.2531219980787704,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997211096074059,
      "loss": 5.6052,
      "mean_token_accuracy": 0.17573324143886565,
      "num_tokens": 6066488.0,
      "step": 2635
    },
    {
      "entropy": 5.630525588989258,
      "epoch": 0.25360230547550433,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997194005558722,
      "loss": 5.6224,
      "mean_token_accuracy": 0.17289988249540328,
      "num_tokens": 6078300.0,
      "step": 2640
    },
    {
      "entropy": 5.664487886428833,
      "epoch": 0.2540826128722382,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997176862870216,
      "loss": 5.5989,
      "mean_token_accuracy": 0.1705970972776413,
      "num_tokens": 6089180.0,
      "step": 2645
    },
    {
      "entropy": 5.684300136566162,
      "epoch": 0.25456292026897215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997159668008933,
      "loss": 5.6268,
      "mean_token_accuracy": 0.18017285764217378,
      "num_tokens": 6099093.0,
      "step": 2650
    },
    {
      "entropy": 5.627106094360352,
      "epoch": 0.25504322766570603,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004997142420975277,
      "loss": 5.6645,
      "mean_token_accuracy": 0.17633569836616517,
      "num_tokens": 6109161.0,
      "step": 2655
    },
    {
      "entropy": 5.826577854156494,
      "epoch": 0.25552353506243997,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997125121769647,
      "loss": 5.7585,
      "mean_token_accuracy": 0.1654739573597908,
      "num_tokens": 6120031.0,
      "step": 2660
    },
    {
      "entropy": 5.65047516822815,
      "epoch": 0.25600384245917385,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997107770392444,
      "loss": 5.6178,
      "mean_token_accuracy": 0.17260300368070602,
      "num_tokens": 6132398.0,
      "step": 2665
    },
    {
      "entropy": 5.605051326751709,
      "epoch": 0.2564841498559078,
      "grad_norm": 1.09375,
      "learning_rate": 0.000499709036684407,
      "loss": 5.5961,
      "mean_token_accuracy": 0.1707433581352234,
      "num_tokens": 6143099.0,
      "step": 2670
    },
    {
      "entropy": 5.630539417266846,
      "epoch": 0.25696445725264166,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997072911124932,
      "loss": 5.567,
      "mean_token_accuracy": 0.18127228319644928,
      "num_tokens": 6154815.0,
      "step": 2675
    },
    {
      "entropy": 5.673951101303101,
      "epoch": 0.2574447646493756,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997055403235432,
      "loss": 5.6335,
      "mean_token_accuracy": 0.1791025683283806,
      "num_tokens": 6165696.0,
      "step": 2680
    },
    {
      "entropy": 5.711018228530884,
      "epoch": 0.2579250720461095,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997037843175978,
      "loss": 5.6631,
      "mean_token_accuracy": 0.1713305965065956,
      "num_tokens": 6177042.0,
      "step": 2685
    },
    {
      "entropy": 5.691460752487183,
      "epoch": 0.2584053794428434,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997020230946978,
      "loss": 5.6437,
      "mean_token_accuracy": 0.17093220055103303,
      "num_tokens": 6188691.0,
      "step": 2690
    },
    {
      "entropy": 5.5285240650177006,
      "epoch": 0.25888568683957736,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004997002566548841,
      "loss": 5.4647,
      "mean_token_accuracy": 0.18098385334014894,
      "num_tokens": 6199931.0,
      "step": 2695
    },
    {
      "entropy": 5.6233972072601315,
      "epoch": 0.25936599423631124,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996984849981976,
      "loss": 5.6121,
      "mean_token_accuracy": 0.17830558866262436,
      "num_tokens": 6211493.0,
      "step": 2700
    },
    {
      "entropy": 5.7481287002563475,
      "epoch": 0.2598463016330452,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996967081246794,
      "loss": 5.5374,
      "mean_token_accuracy": 0.17606928646564485,
      "num_tokens": 6221505.0,
      "step": 2705
    },
    {
      "entropy": 5.519698286056519,
      "epoch": 0.26032660902977905,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996949260343711,
      "loss": 5.5081,
      "mean_token_accuracy": 0.18392858058214187,
      "num_tokens": 6233058.0,
      "step": 2710
    },
    {
      "entropy": 5.681969499588012,
      "epoch": 0.260806916426513,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996931387273137,
      "loss": 5.6666,
      "mean_token_accuracy": 0.16781300455331802,
      "num_tokens": 6244713.0,
      "step": 2715
    },
    {
      "entropy": 5.6033467769622805,
      "epoch": 0.2612872238232469,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996913462035487,
      "loss": 5.6218,
      "mean_token_accuracy": 0.17073221057653426,
      "num_tokens": 6256624.0,
      "step": 2720
    },
    {
      "entropy": 5.610658645629883,
      "epoch": 0.2617675312199808,
      "grad_norm": 0.84375,
      "learning_rate": 0.000499689548463118,
      "loss": 5.6027,
      "mean_token_accuracy": 0.17021558731794356,
      "num_tokens": 6271134.0,
      "step": 2725
    },
    {
      "entropy": 5.752146673202515,
      "epoch": 0.2622478386167147,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996877455060631,
      "loss": 5.6993,
      "mean_token_accuracy": 0.1727217897772789,
      "num_tokens": 6282082.0,
      "step": 2730
    },
    {
      "entropy": 5.672946786880493,
      "epoch": 0.2627281460134486,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996859373324259,
      "loss": 5.6348,
      "mean_token_accuracy": 0.17068509608507157,
      "num_tokens": 6293869.0,
      "step": 2735
    },
    {
      "entropy": 5.69234881401062,
      "epoch": 0.2632084534101825,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996841239422485,
      "loss": 5.6785,
      "mean_token_accuracy": 0.16386467665433885,
      "num_tokens": 6304858.0,
      "step": 2740
    },
    {
      "entropy": 5.6755890369415285,
      "epoch": 0.26368876080691644,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996823053355729,
      "loss": 5.6002,
      "mean_token_accuracy": 0.176376673579216,
      "num_tokens": 6316738.0,
      "step": 2745
    },
    {
      "entropy": 5.579497671127319,
      "epoch": 0.2641690682036503,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996804815124413,
      "loss": 5.5573,
      "mean_token_accuracy": 0.17610080689191818,
      "num_tokens": 6327691.0,
      "step": 2750
    },
    {
      "entropy": 5.666668939590454,
      "epoch": 0.26464937560038426,
      "grad_norm": 1.125,
      "learning_rate": 0.0004996786524728962,
      "loss": 5.5749,
      "mean_token_accuracy": 0.173881895840168,
      "num_tokens": 6337964.0,
      "step": 2755
    },
    {
      "entropy": 5.595135927200317,
      "epoch": 0.26512968299711814,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996768182169797,
      "loss": 5.6286,
      "mean_token_accuracy": 0.17362290471792222,
      "num_tokens": 6349764.0,
      "step": 2760
    },
    {
      "entropy": 5.6827469825744625,
      "epoch": 0.2656099903938521,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996749787447349,
      "loss": 5.5855,
      "mean_token_accuracy": 0.17588018625974655,
      "num_tokens": 6360538.0,
      "step": 2765
    },
    {
      "entropy": 5.644002532958984,
      "epoch": 0.26609029779058596,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499673134056204,
      "loss": 5.6181,
      "mean_token_accuracy": 0.17757804244756697,
      "num_tokens": 6372489.0,
      "step": 2770
    },
    {
      "entropy": 5.499239540100097,
      "epoch": 0.2665706051873199,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996712841514303,
      "loss": 5.4985,
      "mean_token_accuracy": 0.18251207172870637,
      "num_tokens": 6383080.0,
      "step": 2775
    },
    {
      "entropy": 5.724687099456787,
      "epoch": 0.2670509125840538,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996694290304563,
      "loss": 5.6232,
      "mean_token_accuracy": 0.17356717586517334,
      "num_tokens": 6394033.0,
      "step": 2780
    },
    {
      "entropy": 5.651960182189941,
      "epoch": 0.2675312199807877,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996675686933255,
      "loss": 5.632,
      "mean_token_accuracy": 0.17403780966997145,
      "num_tokens": 6404990.0,
      "step": 2785
    },
    {
      "entropy": 5.583127403259278,
      "epoch": 0.2680115273775216,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996657031400807,
      "loss": 5.4925,
      "mean_token_accuracy": 0.18252827525138854,
      "num_tokens": 6416337.0,
      "step": 2790
    },
    {
      "entropy": 5.583510828018189,
      "epoch": 0.26849183477425553,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996638323707655,
      "loss": 5.6999,
      "mean_token_accuracy": 0.17072573751211167,
      "num_tokens": 6427892.0,
      "step": 2795
    },
    {
      "entropy": 5.663077068328858,
      "epoch": 0.2689721421709894,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996619563854232,
      "loss": 5.5987,
      "mean_token_accuracy": 0.17658236026763915,
      "num_tokens": 6440300.0,
      "step": 2800
    },
    {
      "entropy": 5.589122152328491,
      "epoch": 0.26945244956772335,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996600751840974,
      "loss": 5.5071,
      "mean_token_accuracy": 0.17740865498781205,
      "num_tokens": 6451499.0,
      "step": 2805
    },
    {
      "entropy": 5.65373477935791,
      "epoch": 0.26993275696445723,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996581887668317,
      "loss": 5.5993,
      "mean_token_accuracy": 0.17863193303346633,
      "num_tokens": 6462882.0,
      "step": 2810
    },
    {
      "entropy": 5.653873777389526,
      "epoch": 0.27041306436119117,
      "grad_norm": 1.0625,
      "learning_rate": 0.00049965629713367,
      "loss": 5.5799,
      "mean_token_accuracy": 0.1730649381875992,
      "num_tokens": 6475042.0,
      "step": 2815
    },
    {
      "entropy": 5.557312488555908,
      "epoch": 0.27089337175792505,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996544002846561,
      "loss": 5.5842,
      "mean_token_accuracy": 0.1783270537853241,
      "num_tokens": 6486245.0,
      "step": 2820
    },
    {
      "entropy": 5.599606609344482,
      "epoch": 0.271373679154659,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996524982198343,
      "loss": 5.4857,
      "mean_token_accuracy": 0.1768743082880974,
      "num_tokens": 6497705.0,
      "step": 2825
    },
    {
      "entropy": 5.572207498550415,
      "epoch": 0.27185398655139287,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004996505909392485,
      "loss": 5.5271,
      "mean_token_accuracy": 0.17951610833406448,
      "num_tokens": 6508193.0,
      "step": 2830
    },
    {
      "entropy": 5.67252025604248,
      "epoch": 0.2723342939481268,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996486784429429,
      "loss": 5.6391,
      "mean_token_accuracy": 0.17500174790620804,
      "num_tokens": 6519228.0,
      "step": 2835
    },
    {
      "entropy": 5.625978231430054,
      "epoch": 0.2728146013448607,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004996467607309622,
      "loss": 5.6012,
      "mean_token_accuracy": 0.1761829063296318,
      "num_tokens": 6530743.0,
      "step": 2840
    },
    {
      "entropy": 5.612495994567871,
      "epoch": 0.2732949087415946,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996448378033507,
      "loss": 5.4899,
      "mean_token_accuracy": 0.1722333922982216,
      "num_tokens": 6542588.0,
      "step": 2845
    },
    {
      "entropy": 5.707457685470581,
      "epoch": 0.2737752161383285,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996429096601532,
      "loss": 5.6893,
      "mean_token_accuracy": 0.16250439882278442,
      "num_tokens": 6554002.0,
      "step": 2850
    },
    {
      "entropy": 5.732671689987183,
      "epoch": 0.27425552353506244,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004996409763014144,
      "loss": 5.71,
      "mean_token_accuracy": 0.17036043256521224,
      "num_tokens": 6563358.0,
      "step": 2855
    },
    {
      "entropy": 5.6183513641357425,
      "epoch": 0.2747358309317964,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996390377271791,
      "loss": 5.6116,
      "mean_token_accuracy": 0.17085390090942382,
      "num_tokens": 6574986.0,
      "step": 2860
    },
    {
      "entropy": 5.605986452102661,
      "epoch": 0.27521613832853026,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996370939374924,
      "loss": 5.5603,
      "mean_token_accuracy": 0.1796693056821823,
      "num_tokens": 6587407.0,
      "step": 2865
    },
    {
      "entropy": 5.561212491989136,
      "epoch": 0.2756964457252642,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996351449323994,
      "loss": 5.5098,
      "mean_token_accuracy": 0.18226875960826874,
      "num_tokens": 6598483.0,
      "step": 2870
    },
    {
      "entropy": 5.56893515586853,
      "epoch": 0.2761767531219981,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996331907119455,
      "loss": 5.6098,
      "mean_token_accuracy": 0.1776777282357216,
      "num_tokens": 6610749.0,
      "step": 2875
    },
    {
      "entropy": 5.687351655960083,
      "epoch": 0.276657060518732,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996312312761758,
      "loss": 5.5796,
      "mean_token_accuracy": 0.16841792613267897,
      "num_tokens": 6621762.0,
      "step": 2880
    },
    {
      "entropy": 5.589012336730957,
      "epoch": 0.2771373679154659,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499629266625136,
      "loss": 5.6511,
      "mean_token_accuracy": 0.16858434230089187,
      "num_tokens": 6632296.0,
      "step": 2885
    },
    {
      "entropy": 5.639417600631714,
      "epoch": 0.27761767531219983,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996272967588715,
      "loss": 5.5973,
      "mean_token_accuracy": 0.17394536584615708,
      "num_tokens": 6644413.0,
      "step": 2890
    },
    {
      "entropy": 5.641393375396729,
      "epoch": 0.2780979827089337,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996253216774283,
      "loss": 5.5458,
      "mean_token_accuracy": 0.17925278544425965,
      "num_tokens": 6655634.0,
      "step": 2895
    },
    {
      "entropy": 5.6084517478942875,
      "epoch": 0.27857829010566765,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996233413808521,
      "loss": 5.4927,
      "mean_token_accuracy": 0.18065261244773864,
      "num_tokens": 6665977.0,
      "step": 2900
    },
    {
      "entropy": 5.606080198287964,
      "epoch": 0.27905859750240153,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996213558691889,
      "loss": 5.548,
      "mean_token_accuracy": 0.17679802030324937,
      "num_tokens": 6677165.0,
      "step": 2905
    },
    {
      "entropy": 5.642763233184814,
      "epoch": 0.27953890489913547,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996193651424848,
      "loss": 5.6721,
      "mean_token_accuracy": 0.1694797120988369,
      "num_tokens": 6688182.0,
      "step": 2910
    },
    {
      "entropy": 5.626778411865234,
      "epoch": 0.28001921229586935,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499617369200786,
      "loss": 5.5704,
      "mean_token_accuracy": 0.18135443776845933,
      "num_tokens": 6699259.0,
      "step": 2915
    },
    {
      "entropy": 5.623982906341553,
      "epoch": 0.2804995196926033,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996153680441389,
      "loss": 5.5901,
      "mean_token_accuracy": 0.1713014855980873,
      "num_tokens": 6710071.0,
      "step": 2920
    },
    {
      "entropy": 5.646922492980957,
      "epoch": 0.28097982708933716,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049961336167259,
      "loss": 5.6166,
      "mean_token_accuracy": 0.17556966245174407,
      "num_tokens": 6720267.0,
      "step": 2925
    },
    {
      "entropy": 5.64747257232666,
      "epoch": 0.2814601344860711,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996113500861857,
      "loss": 5.5968,
      "mean_token_accuracy": 0.17084414660930633,
      "num_tokens": 6730919.0,
      "step": 2930
    },
    {
      "entropy": 5.558165216445923,
      "epoch": 0.281940441882805,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996093332849729,
      "loss": 5.5952,
      "mean_token_accuracy": 0.17768187820911407,
      "num_tokens": 6741480.0,
      "step": 2935
    },
    {
      "entropy": 5.70050163269043,
      "epoch": 0.2824207492795389,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996073112689983,
      "loss": 5.5396,
      "mean_token_accuracy": 0.1811237409710884,
      "num_tokens": 6751816.0,
      "step": 2940
    },
    {
      "entropy": 5.51968994140625,
      "epoch": 0.2829010566762728,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004996052840383088,
      "loss": 5.5537,
      "mean_token_accuracy": 0.17758982628583908,
      "num_tokens": 6763645.0,
      "step": 2945
    },
    {
      "entropy": 5.535739755630493,
      "epoch": 0.28338136407300674,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996032515929516,
      "loss": 5.5363,
      "mean_token_accuracy": 0.17651461958885192,
      "num_tokens": 6774795.0,
      "step": 2950
    },
    {
      "entropy": 5.6613250255584715,
      "epoch": 0.2838616714697406,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996012139329738,
      "loss": 5.4491,
      "mean_token_accuracy": 0.18220943808555604,
      "num_tokens": 6786833.0,
      "step": 2955
    },
    {
      "entropy": 5.492148590087891,
      "epoch": 0.28434197886647455,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995991710584228,
      "loss": 5.471,
      "mean_token_accuracy": 0.18226066827774048,
      "num_tokens": 6798134.0,
      "step": 2960
    },
    {
      "entropy": 5.634646844863892,
      "epoch": 0.28482228626320844,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004995971229693459,
      "loss": 5.5818,
      "mean_token_accuracy": 0.18246834278106688,
      "num_tokens": 6807918.0,
      "step": 2965
    },
    {
      "entropy": 5.643370199203491,
      "epoch": 0.28530259365994237,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995950696657909,
      "loss": 5.6627,
      "mean_token_accuracy": 0.17691846936941147,
      "num_tokens": 6818689.0,
      "step": 2970
    },
    {
      "entropy": 5.596081829071045,
      "epoch": 0.28578290105667625,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995930111478051,
      "loss": 5.5265,
      "mean_token_accuracy": 0.17872194498777388,
      "num_tokens": 6830871.0,
      "step": 2975
    },
    {
      "entropy": 5.5695876121521,
      "epoch": 0.2862632084534102,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995909474154365,
      "loss": 5.5218,
      "mean_token_accuracy": 0.18281462639570237,
      "num_tokens": 6842857.0,
      "step": 2980
    },
    {
      "entropy": 5.561906528472901,
      "epoch": 0.28674351585014407,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004995888784687331,
      "loss": 5.5439,
      "mean_token_accuracy": 0.17641340792179108,
      "num_tokens": 6853727.0,
      "step": 2985
    },
    {
      "entropy": 5.6508629322052,
      "epoch": 0.287223823246878,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995868043077428,
      "loss": 5.5712,
      "mean_token_accuracy": 0.1839816763997078,
      "num_tokens": 6864728.0,
      "step": 2990
    },
    {
      "entropy": 5.6089515209198,
      "epoch": 0.2877041306436119,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995847249325137,
      "loss": 5.6382,
      "mean_token_accuracy": 0.17240980714559556,
      "num_tokens": 6875001.0,
      "step": 2995
    },
    {
      "entropy": 5.575883674621582,
      "epoch": 0.2881844380403458,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995826403430942,
      "loss": 5.5194,
      "mean_token_accuracy": 0.18421110808849334,
      "num_tokens": 6887026.0,
      "step": 3000
    },
    {
      "epoch": 0.2881844380403458,
      "eval_entropy": 5.505591410883806,
      "eval_loss": 5.596490859985352,
      "eval_mean_token_accuracy": 0.18258213421383437,
      "eval_num_tokens": 6887026.0,
      "eval_runtime": 26.8269,
      "eval_samples_per_second": 1223.214,
      "eval_steps_per_second": 152.906,
      "step": 3000
    },
    {
      "entropy": 5.759801149368286,
      "epoch": 0.2886647454370797,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995805505395328,
      "loss": 5.6473,
      "mean_token_accuracy": 0.17763538807630538,
      "num_tokens": 6900118.0,
      "step": 3005
    },
    {
      "entropy": 5.654341077804565,
      "epoch": 0.28914505283381364,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004995784555218778,
      "loss": 5.5503,
      "mean_token_accuracy": 0.18039704859256744,
      "num_tokens": 6912135.0,
      "step": 3010
    },
    {
      "entropy": 5.539843368530273,
      "epoch": 0.2896253602305475,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995763552901779,
      "loss": 5.4696,
      "mean_token_accuracy": 0.18159575015306473,
      "num_tokens": 6923781.0,
      "step": 3015
    },
    {
      "entropy": 5.630472469329834,
      "epoch": 0.29010566762728146,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995742498444818,
      "loss": 5.5374,
      "mean_token_accuracy": 0.1715562164783478,
      "num_tokens": 6935804.0,
      "step": 3020
    },
    {
      "entropy": 5.558583688735962,
      "epoch": 0.2905859750240154,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995721391848387,
      "loss": 5.5948,
      "mean_token_accuracy": 0.17960504740476607,
      "num_tokens": 6948289.0,
      "step": 3025
    },
    {
      "entropy": 5.634057235717774,
      "epoch": 0.2910662824207493,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995700233112972,
      "loss": 5.5321,
      "mean_token_accuracy": 0.1770654082298279,
      "num_tokens": 6958719.0,
      "step": 3030
    },
    {
      "entropy": 5.600251197814941,
      "epoch": 0.2915465898174832,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995679022239066,
      "loss": 5.5581,
      "mean_token_accuracy": 0.18355003297328948,
      "num_tokens": 6969041.0,
      "step": 3035
    },
    {
      "entropy": 5.606411504745483,
      "epoch": 0.2920268972142171,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995657759227162,
      "loss": 5.6522,
      "mean_token_accuracy": 0.17753843814134598,
      "num_tokens": 6981206.0,
      "step": 3040
    },
    {
      "entropy": 5.691739225387574,
      "epoch": 0.29250720461095103,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995636444077751,
      "loss": 5.5391,
      "mean_token_accuracy": 0.17880928218364717,
      "num_tokens": 6993169.0,
      "step": 3045
    },
    {
      "entropy": 5.6123960494995115,
      "epoch": 0.2929875120076849,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995615076791333,
      "loss": 5.5922,
      "mean_token_accuracy": 0.17676615267992019,
      "num_tokens": 7005175.0,
      "step": 3050
    },
    {
      "entropy": 5.647955417633057,
      "epoch": 0.29346781940441885,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995593657368399,
      "loss": 5.6063,
      "mean_token_accuracy": 0.17792469263076782,
      "num_tokens": 7016738.0,
      "step": 3055
    },
    {
      "entropy": 5.59646692276001,
      "epoch": 0.29394812680115273,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499557218580945,
      "loss": 5.4979,
      "mean_token_accuracy": 0.1870897263288498,
      "num_tokens": 7028399.0,
      "step": 3060
    },
    {
      "entropy": 5.627456712722778,
      "epoch": 0.29442843419788667,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995550662114981,
      "loss": 5.6468,
      "mean_token_accuracy": 0.17034640163183212,
      "num_tokens": 7039528.0,
      "step": 3065
    },
    {
      "entropy": 5.5611835479736325,
      "epoch": 0.29490874159462055,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995529086285495,
      "loss": 5.434,
      "mean_token_accuracy": 0.19027520716190338,
      "num_tokens": 7051399.0,
      "step": 3070
    },
    {
      "entropy": 5.44168701171875,
      "epoch": 0.2953890489913545,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499550745832149,
      "loss": 5.4353,
      "mean_token_accuracy": 0.18632688075304032,
      "num_tokens": 7062452.0,
      "step": 3075
    },
    {
      "entropy": 5.616183185577393,
      "epoch": 0.29586935638808837,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995485778223471,
      "loss": 5.5695,
      "mean_token_accuracy": 0.18081469535827638,
      "num_tokens": 7074052.0,
      "step": 3080
    },
    {
      "entropy": 5.668457651138306,
      "epoch": 0.2963496637848223,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995464045991939,
      "loss": 5.62,
      "mean_token_accuracy": 0.16842284351587294,
      "num_tokens": 7086692.0,
      "step": 3085
    },
    {
      "entropy": 5.605998182296753,
      "epoch": 0.2968299711815562,
      "grad_norm": 1.125,
      "learning_rate": 0.00049954422616274,
      "loss": 5.5883,
      "mean_token_accuracy": 0.1737152323126793,
      "num_tokens": 7098221.0,
      "step": 3090
    },
    {
      "entropy": 5.5584465026855465,
      "epoch": 0.2973102785782901,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995420425130359,
      "loss": 5.5395,
      "mean_token_accuracy": 0.18584298193454743,
      "num_tokens": 7110129.0,
      "step": 3095
    },
    {
      "entropy": 5.653409051895141,
      "epoch": 0.297790585975024,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995398536501324,
      "loss": 5.6176,
      "mean_token_accuracy": 0.17346330732107162,
      "num_tokens": 7121523.0,
      "step": 3100
    },
    {
      "entropy": 5.588115596771241,
      "epoch": 0.29827089337175794,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995376595740801,
      "loss": 5.5328,
      "mean_token_accuracy": 0.1846982017159462,
      "num_tokens": 7133482.0,
      "step": 3105
    },
    {
      "entropy": 5.538929033279419,
      "epoch": 0.2987512007684918,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995354602849302,
      "loss": 5.518,
      "mean_token_accuracy": 0.17959461808204652,
      "num_tokens": 7144031.0,
      "step": 3110
    },
    {
      "entropy": 5.694006443023682,
      "epoch": 0.29923150816522576,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995332557827337,
      "loss": 5.609,
      "mean_token_accuracy": 0.1750166431069374,
      "num_tokens": 7155469.0,
      "step": 3115
    },
    {
      "entropy": 5.6590465068817135,
      "epoch": 0.29971181556195964,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995310460675416,
      "loss": 5.5302,
      "mean_token_accuracy": 0.17468951493501664,
      "num_tokens": 7167077.0,
      "step": 3120
    },
    {
      "entropy": 5.488217449188232,
      "epoch": 0.3001921229586936,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995288311394053,
      "loss": 5.4798,
      "mean_token_accuracy": 0.18271518647670745,
      "num_tokens": 7178917.0,
      "step": 3125
    },
    {
      "entropy": 5.68494348526001,
      "epoch": 0.30067243035542746,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995266109983764,
      "loss": 5.5797,
      "mean_token_accuracy": 0.1793813332915306,
      "num_tokens": 7190284.0,
      "step": 3130
    },
    {
      "entropy": 5.534503746032715,
      "epoch": 0.3011527377521614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995243856445062,
      "loss": 5.5953,
      "mean_token_accuracy": 0.17944199591875076,
      "num_tokens": 7202391.0,
      "step": 3135
    },
    {
      "entropy": 5.6232444763183596,
      "epoch": 0.3016330451488953,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995221550778466,
      "loss": 5.498,
      "mean_token_accuracy": 0.1835529923439026,
      "num_tokens": 7212882.0,
      "step": 3140
    },
    {
      "entropy": 5.4999878883361815,
      "epoch": 0.3021133525456292,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995199192984491,
      "loss": 5.4883,
      "mean_token_accuracy": 0.18379289358854295,
      "num_tokens": 7224102.0,
      "step": 3145
    },
    {
      "entropy": 5.539435005187988,
      "epoch": 0.3025936599423631,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995176783063657,
      "loss": 5.4357,
      "mean_token_accuracy": 0.18271342515945435,
      "num_tokens": 7234579.0,
      "step": 3150
    },
    {
      "entropy": 5.616196155548096,
      "epoch": 0.30307396733909703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995154321016487,
      "loss": 5.582,
      "mean_token_accuracy": 0.17273370623588563,
      "num_tokens": 7246065.0,
      "step": 3155
    },
    {
      "entropy": 5.621270561218262,
      "epoch": 0.3035542747358309,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995131806843499,
      "loss": 5.5318,
      "mean_token_accuracy": 0.1730843797326088,
      "num_tokens": 7258899.0,
      "step": 3160
    },
    {
      "entropy": 5.612247276306152,
      "epoch": 0.30403458213256485,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995109240545218,
      "loss": 5.621,
      "mean_token_accuracy": 0.16662402749061583,
      "num_tokens": 7271887.0,
      "step": 3165
    },
    {
      "entropy": 5.621712684631348,
      "epoch": 0.3045148895292987,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995086622122167,
      "loss": 5.5654,
      "mean_token_accuracy": 0.17996519356966018,
      "num_tokens": 7282773.0,
      "step": 3170
    },
    {
      "entropy": 5.571623659133911,
      "epoch": 0.30499519692603266,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995063951574871,
      "loss": 5.4093,
      "mean_token_accuracy": 0.1852883964776993,
      "num_tokens": 7294460.0,
      "step": 3175
    },
    {
      "entropy": 5.486154985427857,
      "epoch": 0.30547550432276654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995041228903856,
      "loss": 5.469,
      "mean_token_accuracy": 0.18554151356220244,
      "num_tokens": 7305125.0,
      "step": 3180
    },
    {
      "entropy": 5.455041456222534,
      "epoch": 0.3059558117195005,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499501845410965,
      "loss": 5.4735,
      "mean_token_accuracy": 0.17831799238920212,
      "num_tokens": 7317738.0,
      "step": 3185
    },
    {
      "entropy": 5.621507024765014,
      "epoch": 0.30643611911623436,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994995627192781,
      "loss": 5.489,
      "mean_token_accuracy": 0.18256354928016663,
      "num_tokens": 7329820.0,
      "step": 3190
    },
    {
      "entropy": 5.568184757232666,
      "epoch": 0.3069164265129683,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994972748153781,
      "loss": 5.5336,
      "mean_token_accuracy": 0.1765450730919838,
      "num_tokens": 7341471.0,
      "step": 3195
    },
    {
      "entropy": 5.628265857696533,
      "epoch": 0.30739673390970224,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499494981699318,
      "loss": 5.6641,
      "mean_token_accuracy": 0.1794617623090744,
      "num_tokens": 7352459.0,
      "step": 3200
    },
    {
      "entropy": 5.5684881687164305,
      "epoch": 0.3078770413064361,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499492683371151,
      "loss": 5.4615,
      "mean_token_accuracy": 0.18775735944509506,
      "num_tokens": 7363129.0,
      "step": 3205
    },
    {
      "entropy": 5.567241525650024,
      "epoch": 0.30835734870317005,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994903798309306,
      "loss": 5.5224,
      "mean_token_accuracy": 0.17995841205120086,
      "num_tokens": 7373726.0,
      "step": 3210
    },
    {
      "entropy": 5.6475639820098875,
      "epoch": 0.30883765609990393,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994880710787102,
      "loss": 5.5825,
      "mean_token_accuracy": 0.18032457083463668,
      "num_tokens": 7383847.0,
      "step": 3215
    },
    {
      "entropy": 5.585547542572021,
      "epoch": 0.30931796349663787,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994857571145432,
      "loss": 5.5572,
      "mean_token_accuracy": 0.1745512455701828,
      "num_tokens": 7396425.0,
      "step": 3220
    },
    {
      "entropy": 5.675054025650025,
      "epoch": 0.30979827089337175,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994834379384837,
      "loss": 5.637,
      "mean_token_accuracy": 0.17244965434074402,
      "num_tokens": 7407591.0,
      "step": 3225
    },
    {
      "entropy": 5.5843696117401125,
      "epoch": 0.3102785782901057,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994811135505851,
      "loss": 5.4937,
      "mean_token_accuracy": 0.1801697939634323,
      "num_tokens": 7419668.0,
      "step": 3230
    },
    {
      "entropy": 5.6134058952331545,
      "epoch": 0.31075888568683957,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004994787839509018,
      "loss": 5.597,
      "mean_token_accuracy": 0.17638275176286697,
      "num_tokens": 7431969.0,
      "step": 3235
    },
    {
      "entropy": 5.685126161575317,
      "epoch": 0.3112391930835735,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994764491394876,
      "loss": 5.5912,
      "mean_token_accuracy": 0.17804454267024994,
      "num_tokens": 7443467.0,
      "step": 3240
    },
    {
      "entropy": 5.568789148330689,
      "epoch": 0.3117195004803074,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994741091163969,
      "loss": 5.5749,
      "mean_token_accuracy": 0.1806941583752632,
      "num_tokens": 7455220.0,
      "step": 3245
    },
    {
      "entropy": 5.590549516677856,
      "epoch": 0.3121998078770413,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499471763881684,
      "loss": 5.4928,
      "mean_token_accuracy": 0.18144142478704453,
      "num_tokens": 7466452.0,
      "step": 3250
    },
    {
      "entropy": 5.539456510543824,
      "epoch": 0.3126801152737752,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994694134354031,
      "loss": 5.5385,
      "mean_token_accuracy": 0.18830569088459015,
      "num_tokens": 7478798.0,
      "step": 3255
    },
    {
      "entropy": 5.634785175323486,
      "epoch": 0.31316042267050914,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499467057777609,
      "loss": 5.6215,
      "mean_token_accuracy": 0.17719642966985702,
      "num_tokens": 7491039.0,
      "step": 3260
    },
    {
      "entropy": 5.5484857082366945,
      "epoch": 0.313640730067243,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994646969083565,
      "loss": 5.4851,
      "mean_token_accuracy": 0.1823827013373375,
      "num_tokens": 7501455.0,
      "step": 3265
    },
    {
      "entropy": 5.601222896575928,
      "epoch": 0.31412103746397696,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994623308277002,
      "loss": 5.5459,
      "mean_token_accuracy": 0.17547287046909332,
      "num_tokens": 7513484.0,
      "step": 3270
    },
    {
      "entropy": 5.628010320663452,
      "epoch": 0.31460134486071084,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499459959535695,
      "loss": 5.5376,
      "mean_token_accuracy": 0.1749982163310051,
      "num_tokens": 7524868.0,
      "step": 3275
    },
    {
      "entropy": 5.4064342975616455,
      "epoch": 0.3150816522574448,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994575830323962,
      "loss": 5.4192,
      "mean_token_accuracy": 0.1855955168604851,
      "num_tokens": 7536290.0,
      "step": 3280
    },
    {
      "entropy": 5.506337738037109,
      "epoch": 0.31556195965417866,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994552013178586,
      "loss": 5.5097,
      "mean_token_accuracy": 0.1780158206820488,
      "num_tokens": 7546994.0,
      "step": 3285
    },
    {
      "entropy": 5.58304352760315,
      "epoch": 0.3160422670509126,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499452814392138,
      "loss": 5.4966,
      "mean_token_accuracy": 0.17756061851978303,
      "num_tokens": 7559231.0,
      "step": 3290
    },
    {
      "entropy": 5.535794878005982,
      "epoch": 0.3165225744476465,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994504222552894,
      "loss": 5.4618,
      "mean_token_accuracy": 0.18046220391988754,
      "num_tokens": 7570894.0,
      "step": 3295
    },
    {
      "entropy": 5.609767913818359,
      "epoch": 0.3170028818443804,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004994480249073684,
      "loss": 5.5913,
      "mean_token_accuracy": 0.17232506275177,
      "num_tokens": 7581765.0,
      "step": 3300
    },
    {
      "entropy": 5.565992927551269,
      "epoch": 0.3174831892411143,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994456223484308,
      "loss": 5.4398,
      "mean_token_accuracy": 0.18191327899694443,
      "num_tokens": 7591828.0,
      "step": 3305
    },
    {
      "entropy": 5.515632772445679,
      "epoch": 0.31796349663784823,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994432145785323,
      "loss": 5.5109,
      "mean_token_accuracy": 0.17749277800321578,
      "num_tokens": 7604230.0,
      "step": 3310
    },
    {
      "entropy": 5.533628845214844,
      "epoch": 0.3184438040345821,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994408015977288,
      "loss": 5.4705,
      "mean_token_accuracy": 0.18082427084445954,
      "num_tokens": 7615780.0,
      "step": 3315
    },
    {
      "entropy": 5.5580182552337645,
      "epoch": 0.31892411143131605,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994383834060764,
      "loss": 5.4232,
      "mean_token_accuracy": 0.17906945943832397,
      "num_tokens": 7627633.0,
      "step": 3320
    },
    {
      "entropy": 5.5594038486480715,
      "epoch": 0.31940441882804993,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994359600036311,
      "loss": 5.5512,
      "mean_token_accuracy": 0.1759330376982689,
      "num_tokens": 7639487.0,
      "step": 3325
    },
    {
      "entropy": 5.568274879455567,
      "epoch": 0.31988472622478387,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994335313904493,
      "loss": 5.471,
      "mean_token_accuracy": 0.18615336418151857,
      "num_tokens": 7649976.0,
      "step": 3330
    },
    {
      "entropy": 5.708322763442993,
      "epoch": 0.32036503362151775,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994310975665873,
      "loss": 5.6526,
      "mean_token_accuracy": 0.17140694707632065,
      "num_tokens": 7662555.0,
      "step": 3335
    },
    {
      "entropy": 5.492502927780151,
      "epoch": 0.3208453410182517,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004994286585321017,
      "loss": 5.4349,
      "mean_token_accuracy": 0.18513449132442475,
      "num_tokens": 7673869.0,
      "step": 3340
    },
    {
      "entropy": 5.479371738433838,
      "epoch": 0.32132564841498557,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499426214287049,
      "loss": 5.5361,
      "mean_token_accuracy": 0.17399935126304628,
      "num_tokens": 7687149.0,
      "step": 3345
    },
    {
      "entropy": 5.676080560684204,
      "epoch": 0.3218059558117195,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994237648314862,
      "loss": 5.5787,
      "mean_token_accuracy": 0.17774081528186797,
      "num_tokens": 7698352.0,
      "step": 3350
    },
    {
      "entropy": 5.5003471851348875,
      "epoch": 0.3222862632084534,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994213101654697,
      "loss": 5.4051,
      "mean_token_accuracy": 0.1902243122458458,
      "num_tokens": 7709117.0,
      "step": 3355
    },
    {
      "entropy": 5.504093313217163,
      "epoch": 0.3227665706051873,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499418850289057,
      "loss": 5.4423,
      "mean_token_accuracy": 0.18837745487689972,
      "num_tokens": 7721731.0,
      "step": 3360
    },
    {
      "entropy": 5.468392086029053,
      "epoch": 0.32324687800192126,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004994163852023048,
      "loss": 5.4233,
      "mean_token_accuracy": 0.17894796431064605,
      "num_tokens": 7733010.0,
      "step": 3365
    },
    {
      "entropy": 5.519420766830445,
      "epoch": 0.32372718539865514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004994139149052706,
      "loss": 5.4045,
      "mean_token_accuracy": 0.19315167963504792,
      "num_tokens": 7743905.0,
      "step": 3370
    },
    {
      "entropy": 5.549196529388428,
      "epoch": 0.3242074927953891,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994114393980117,
      "loss": 5.44,
      "mean_token_accuracy": 0.18877158761024476,
      "num_tokens": 7754891.0,
      "step": 3375
    },
    {
      "entropy": 5.571593856811523,
      "epoch": 0.32468780019212296,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994089586805856,
      "loss": 5.5592,
      "mean_token_accuracy": 0.17476010024547578,
      "num_tokens": 7766831.0,
      "step": 3380
    },
    {
      "entropy": 5.5558350563049315,
      "epoch": 0.3251681075888569,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994064727530496,
      "loss": 5.631,
      "mean_token_accuracy": 0.17296807169914247,
      "num_tokens": 7778239.0,
      "step": 3385
    },
    {
      "entropy": 5.601680278778076,
      "epoch": 0.3256484149855908,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994039816154618,
      "loss": 5.4295,
      "mean_token_accuracy": 0.18522366285324096,
      "num_tokens": 7789449.0,
      "step": 3390
    },
    {
      "entropy": 5.6180826187133786,
      "epoch": 0.3261287223823247,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049940148526788,
      "loss": 5.5284,
      "mean_token_accuracy": 0.18058868050575255,
      "num_tokens": 7799507.0,
      "step": 3395
    },
    {
      "entropy": 5.582130336761475,
      "epoch": 0.3266090297790586,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993989837103618,
      "loss": 5.4949,
      "mean_token_accuracy": 0.18623894602060317,
      "num_tokens": 7811730.0,
      "step": 3400
    },
    {
      "entropy": 5.45947756767273,
      "epoch": 0.3270893371757925,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993964769429657,
      "loss": 5.3984,
      "mean_token_accuracy": 0.1905061572790146,
      "num_tokens": 7823108.0,
      "step": 3405
    },
    {
      "entropy": 5.528728246688843,
      "epoch": 0.3275696445725264,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993939649657498,
      "loss": 5.4801,
      "mean_token_accuracy": 0.18378316909074782,
      "num_tokens": 7835384.0,
      "step": 3410
    },
    {
      "entropy": 5.6040123462677,
      "epoch": 0.32804995196926034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993914477787721,
      "loss": 5.5187,
      "mean_token_accuracy": 0.18680190294981003,
      "num_tokens": 7845860.0,
      "step": 3415
    },
    {
      "entropy": 5.592914009094239,
      "epoch": 0.3285302593659942,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993889253820915,
      "loss": 5.5828,
      "mean_token_accuracy": 0.1734114795923233,
      "num_tokens": 7856861.0,
      "step": 3420
    },
    {
      "entropy": 5.585071706771851,
      "epoch": 0.32901056676272816,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004993863977757663,
      "loss": 5.5291,
      "mean_token_accuracy": 0.18304805755615233,
      "num_tokens": 7868902.0,
      "step": 3425
    },
    {
      "entropy": 5.6103928565979,
      "epoch": 0.32949087415946204,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004993838649598552,
      "loss": 5.5504,
      "mean_token_accuracy": 0.17545123770833015,
      "num_tokens": 7880369.0,
      "step": 3430
    },
    {
      "entropy": 5.6020973205566404,
      "epoch": 0.329971181556196,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004993813269344171,
      "loss": 5.5382,
      "mean_token_accuracy": 0.17541403621435164,
      "num_tokens": 7892041.0,
      "step": 3435
    },
    {
      "entropy": 5.504408979415894,
      "epoch": 0.33045148895292986,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993787836995108,
      "loss": 5.4772,
      "mean_token_accuracy": 0.18386240005493165,
      "num_tokens": 7903644.0,
      "step": 3440
    },
    {
      "entropy": 5.571684551239014,
      "epoch": 0.3309317963496638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993762352551954,
      "loss": 5.5414,
      "mean_token_accuracy": 0.17612113058567047,
      "num_tokens": 7915679.0,
      "step": 3445
    },
    {
      "entropy": 5.607144117355347,
      "epoch": 0.3314121037463977,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004993736816015301,
      "loss": 5.5596,
      "mean_token_accuracy": 0.1792970895767212,
      "num_tokens": 7928020.0,
      "step": 3450
    },
    {
      "entropy": 5.629954481124878,
      "epoch": 0.3318924111431316,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993711227385742,
      "loss": 5.6058,
      "mean_token_accuracy": 0.17515352368354797,
      "num_tokens": 7940521.0,
      "step": 3455
    },
    {
      "entropy": 5.631735610961914,
      "epoch": 0.3323727185398655,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993685586663871,
      "loss": 5.5452,
      "mean_token_accuracy": 0.1744692325592041,
      "num_tokens": 7952234.0,
      "step": 3460
    },
    {
      "entropy": 5.602184581756592,
      "epoch": 0.33285302593659943,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993659893850281,
      "loss": 5.6133,
      "mean_token_accuracy": 0.17432404905557633,
      "num_tokens": 7964900.0,
      "step": 3465
    },
    {
      "entropy": 5.642161321640015,
      "epoch": 0.3333333333333333,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993634148945573,
      "loss": 5.6248,
      "mean_token_accuracy": 0.17408506870269774,
      "num_tokens": 7976661.0,
      "step": 3470
    },
    {
      "entropy": 5.5621062278747555,
      "epoch": 0.33381364073006725,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004993608351950341,
      "loss": 5.4427,
      "mean_token_accuracy": 0.18498605787754058,
      "num_tokens": 7987605.0,
      "step": 3475
    },
    {
      "entropy": 5.447353935241699,
      "epoch": 0.33429394812680113,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993582502865185,
      "loss": 5.5135,
      "mean_token_accuracy": 0.18175396621227263,
      "num_tokens": 7998879.0,
      "step": 3480
    },
    {
      "entropy": 5.589016914367676,
      "epoch": 0.33477425552353507,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993556601690706,
      "loss": 5.4948,
      "mean_token_accuracy": 0.18219438046216965,
      "num_tokens": 8010634.0,
      "step": 3485
    },
    {
      "entropy": 5.575048971176147,
      "epoch": 0.33525456292026895,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993530648427505,
      "loss": 5.5089,
      "mean_token_accuracy": 0.17399886548519133,
      "num_tokens": 8022134.0,
      "step": 3490
    },
    {
      "entropy": 5.5292055130004885,
      "epoch": 0.3357348703170029,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993504643076184,
      "loss": 5.4626,
      "mean_token_accuracy": 0.18072339594364167,
      "num_tokens": 8034454.0,
      "step": 3495
    },
    {
      "entropy": 5.571810293197632,
      "epoch": 0.33621517771373677,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004993478585637347,
      "loss": 5.5419,
      "mean_token_accuracy": 0.17695319652557373,
      "num_tokens": 8046175.0,
      "step": 3500
    },
    {
      "entropy": 5.671894598007202,
      "epoch": 0.3366954851104707,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004993452476111599,
      "loss": 5.5915,
      "mean_token_accuracy": 0.17551082521677017,
      "num_tokens": 8057335.0,
      "step": 3505
    },
    {
      "entropy": 5.517408752441407,
      "epoch": 0.3371757925072046,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993426314499546,
      "loss": 5.4127,
      "mean_token_accuracy": 0.18328601419925689,
      "num_tokens": 8068860.0,
      "step": 3510
    },
    {
      "entropy": 5.5245567798614506,
      "epoch": 0.3376560999039385,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004993400100801796,
      "loss": 5.4942,
      "mean_token_accuracy": 0.17597640454769134,
      "num_tokens": 8078957.0,
      "step": 3515
    },
    {
      "entropy": 5.632298755645752,
      "epoch": 0.3381364073006724,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993373835018956,
      "loss": 5.6054,
      "mean_token_accuracy": 0.17132381796836854,
      "num_tokens": 8090143.0,
      "step": 3520
    },
    {
      "entropy": 5.442738676071167,
      "epoch": 0.33861671469740634,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993347517151638,
      "loss": 5.4018,
      "mean_token_accuracy": 0.18353821486234664,
      "num_tokens": 8101961.0,
      "step": 3525
    },
    {
      "entropy": 5.426606559753418,
      "epoch": 0.3390970220941403,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993321147200452,
      "loss": 5.3369,
      "mean_token_accuracy": 0.19012405425310136,
      "num_tokens": 8112635.0,
      "step": 3530
    },
    {
      "entropy": 5.612124490737915,
      "epoch": 0.33957732949087416,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499329472516601,
      "loss": 5.5334,
      "mean_token_accuracy": 0.17269296050071717,
      "num_tokens": 8123730.0,
      "step": 3535
    },
    {
      "entropy": 5.578820610046387,
      "epoch": 0.3400576368876081,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993268251048925,
      "loss": 5.5442,
      "mean_token_accuracy": 0.1789178341627121,
      "num_tokens": 8135258.0,
      "step": 3540
    },
    {
      "entropy": 5.491145992279053,
      "epoch": 0.340537944284342,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993241724849814,
      "loss": 5.4012,
      "mean_token_accuracy": 0.18719657361507416,
      "num_tokens": 8145946.0,
      "step": 3545
    },
    {
      "entropy": 5.56555209159851,
      "epoch": 0.3410182516810759,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499321514656929,
      "loss": 5.6069,
      "mean_token_accuracy": 0.17737515419721603,
      "num_tokens": 8158422.0,
      "step": 3550
    },
    {
      "entropy": 5.589611339569092,
      "epoch": 0.3414985590778098,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004993188516207972,
      "loss": 5.5237,
      "mean_token_accuracy": 0.17171558737754822,
      "num_tokens": 8170849.0,
      "step": 3555
    },
    {
      "entropy": 5.5105218410491945,
      "epoch": 0.34197886647454373,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004993161833766478,
      "loss": 5.4811,
      "mean_token_accuracy": 0.18131685256958008,
      "num_tokens": 8180907.0,
      "step": 3560
    },
    {
      "entropy": 5.629140090942383,
      "epoch": 0.3424591738712776,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004993135099245426,
      "loss": 5.5395,
      "mean_token_accuracy": 0.17697139978408813,
      "num_tokens": 8193418.0,
      "step": 3565
    },
    {
      "entropy": 5.557410335540771,
      "epoch": 0.34293948126801155,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993108312645438,
      "loss": 5.6283,
      "mean_token_accuracy": 0.17542220205068587,
      "num_tokens": 8203873.0,
      "step": 3570
    },
    {
      "entropy": 5.671187067031861,
      "epoch": 0.34341978866474543,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993081473967135,
      "loss": 5.617,
      "mean_token_accuracy": 0.1747312679886818,
      "num_tokens": 8215150.0,
      "step": 3575
    },
    {
      "entropy": 5.629827690124512,
      "epoch": 0.34390009606147937,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993054583211143,
      "loss": 5.5844,
      "mean_token_accuracy": 0.17693494409322738,
      "num_tokens": 8225732.0,
      "step": 3580
    },
    {
      "entropy": 5.582673597335815,
      "epoch": 0.34438040345821325,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993027640378081,
      "loss": 5.5138,
      "mean_token_accuracy": 0.1794430896639824,
      "num_tokens": 8237523.0,
      "step": 3585
    },
    {
      "entropy": 5.557817363739014,
      "epoch": 0.3448607108549472,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499300064546858,
      "loss": 5.5008,
      "mean_token_accuracy": 0.18324483335018157,
      "num_tokens": 8249660.0,
      "step": 3590
    },
    {
      "entropy": 5.491367626190185,
      "epoch": 0.34534101825168106,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004992973598483264,
      "loss": 5.3554,
      "mean_token_accuracy": 0.18711050301790239,
      "num_tokens": 8261224.0,
      "step": 3595
    },
    {
      "entropy": 5.422523736953735,
      "epoch": 0.345821325648415,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499294649942276,
      "loss": 5.5787,
      "mean_token_accuracy": 0.17804848700761794,
      "num_tokens": 8272177.0,
      "step": 3600
    },
    {
      "entropy": 5.582265853881836,
      "epoch": 0.3463016330451489,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992919348287699,
      "loss": 5.4751,
      "mean_token_accuracy": 0.18154810816049577,
      "num_tokens": 8284449.0,
      "step": 3605
    },
    {
      "entropy": 5.555227851867675,
      "epoch": 0.3467819404418828,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992892145078711,
      "loss": 5.5457,
      "mean_token_accuracy": 0.18382474184036254,
      "num_tokens": 8297028.0,
      "step": 3610
    },
    {
      "entropy": 5.5002641677856445,
      "epoch": 0.3472622478386167,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992864889796427,
      "loss": 5.4254,
      "mean_token_accuracy": 0.19357240349054336,
      "num_tokens": 8308348.0,
      "step": 3615
    },
    {
      "entropy": 5.512722015380859,
      "epoch": 0.34774255523535064,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004992837582441481,
      "loss": 5.4302,
      "mean_token_accuracy": 0.19045159369707107,
      "num_tokens": 8318951.0,
      "step": 3620
    },
    {
      "entropy": 5.600104761123657,
      "epoch": 0.3482228626320845,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004992810223014506,
      "loss": 5.5352,
      "mean_token_accuracy": 0.176507867872715,
      "num_tokens": 8329979.0,
      "step": 3625
    },
    {
      "entropy": 5.581473445892334,
      "epoch": 0.34870317002881845,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004992782811516137,
      "loss": 5.4725,
      "mean_token_accuracy": 0.18507088273763656,
      "num_tokens": 8340332.0,
      "step": 3630
    },
    {
      "entropy": 5.521006298065186,
      "epoch": 0.34918347742555234,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992755347947011,
      "loss": 5.4858,
      "mean_token_accuracy": 0.1741225928068161,
      "num_tokens": 8350376.0,
      "step": 3635
    },
    {
      "entropy": 5.531734037399292,
      "epoch": 0.34966378482228627,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992727832307766,
      "loss": 5.4211,
      "mean_token_accuracy": 0.1854601576924324,
      "num_tokens": 8361884.0,
      "step": 3640
    },
    {
      "entropy": 5.414716958999634,
      "epoch": 0.35014409221902015,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992700264599039,
      "loss": 5.354,
      "mean_token_accuracy": 0.18897753357887268,
      "num_tokens": 8372914.0,
      "step": 3645
    },
    {
      "entropy": 5.556543397903442,
      "epoch": 0.3506243996157541,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992672644821473,
      "loss": 5.5339,
      "mean_token_accuracy": 0.17634458988904952,
      "num_tokens": 8385800.0,
      "step": 3650
    },
    {
      "entropy": 5.483111810684204,
      "epoch": 0.35110470701248797,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004992644972975707,
      "loss": 5.4252,
      "mean_token_accuracy": 0.18766560405492783,
      "num_tokens": 8397783.0,
      "step": 3655
    },
    {
      "entropy": 5.507390546798706,
      "epoch": 0.3515850144092219,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992617249062383,
      "loss": 5.4762,
      "mean_token_accuracy": 0.18290266394615173,
      "num_tokens": 8409096.0,
      "step": 3660
    },
    {
      "entropy": 5.480173349380493,
      "epoch": 0.3520653218059558,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992589473082147,
      "loss": 5.3858,
      "mean_token_accuracy": 0.189688578248024,
      "num_tokens": 8420833.0,
      "step": 3665
    },
    {
      "entropy": 5.531547975540161,
      "epoch": 0.3525456292026897,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992561645035641,
      "loss": 5.467,
      "mean_token_accuracy": 0.1827012225985527,
      "num_tokens": 8431848.0,
      "step": 3670
    },
    {
      "entropy": 5.645762300491333,
      "epoch": 0.3530259365994236,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992533764923515,
      "loss": 5.5514,
      "mean_token_accuracy": 0.1750529944896698,
      "num_tokens": 8443620.0,
      "step": 3675
    },
    {
      "entropy": 5.499345874786377,
      "epoch": 0.35350624399615754,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004992505832746412,
      "loss": 5.5536,
      "mean_token_accuracy": 0.18342854380607604,
      "num_tokens": 8455339.0,
      "step": 3680
    },
    {
      "entropy": 5.469972848892212,
      "epoch": 0.3539865513928914,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992477848504983,
      "loss": 5.3863,
      "mean_token_accuracy": 0.19783511906862258,
      "num_tokens": 8466342.0,
      "step": 3685
    },
    {
      "entropy": 5.556191253662109,
      "epoch": 0.35446685878962536,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992449812199877,
      "loss": 5.42,
      "mean_token_accuracy": 0.18476486057043076,
      "num_tokens": 8476955.0,
      "step": 3690
    },
    {
      "entropy": 5.387001466751099,
      "epoch": 0.3549471661863593,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004992421723831745,
      "loss": 5.398,
      "mean_token_accuracy": 0.18051135241985322,
      "num_tokens": 8488148.0,
      "step": 3695
    },
    {
      "entropy": 5.5487226963043215,
      "epoch": 0.3554274735830932,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004992393583401239,
      "loss": 5.416,
      "mean_token_accuracy": 0.18518388420343398,
      "num_tokens": 8499014.0,
      "step": 3700
    },
    {
      "entropy": 5.573843050003052,
      "epoch": 0.3559077809798271,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992365390909014,
      "loss": 5.4837,
      "mean_token_accuracy": 0.17986021488904952,
      "num_tokens": 8510191.0,
      "step": 3705
    },
    {
      "entropy": 5.432552337646484,
      "epoch": 0.356388088376561,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992337146355721,
      "loss": 5.3647,
      "mean_token_accuracy": 0.19069588631391526,
      "num_tokens": 8522368.0,
      "step": 3710
    },
    {
      "entropy": 5.480603408813477,
      "epoch": 0.35686839577329493,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004992308849742019,
      "loss": 5.4021,
      "mean_token_accuracy": 0.19026098847389222,
      "num_tokens": 8533692.0,
      "step": 3715
    },
    {
      "entropy": 5.511455345153808,
      "epoch": 0.3573487031700288,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992280501068563,
      "loss": 5.4858,
      "mean_token_accuracy": 0.18425101935863494,
      "num_tokens": 8545283.0,
      "step": 3720
    },
    {
      "entropy": 5.538014602661133,
      "epoch": 0.35782901056676275,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992252100336012,
      "loss": 5.4613,
      "mean_token_accuracy": 0.18866845965385437,
      "num_tokens": 8555094.0,
      "step": 3725
    },
    {
      "entropy": 5.460282516479492,
      "epoch": 0.35830931796349663,
      "grad_norm": 1.0,
      "learning_rate": 0.0004992223647545027,
      "loss": 5.4866,
      "mean_token_accuracy": 0.17785904705524444,
      "num_tokens": 8565179.0,
      "step": 3730
    },
    {
      "entropy": 5.519764852523804,
      "epoch": 0.35878962536023057,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992195142696266,
      "loss": 5.3856,
      "mean_token_accuracy": 0.1784725680947304,
      "num_tokens": 8575764.0,
      "step": 3735
    },
    {
      "entropy": 5.373103046417237,
      "epoch": 0.35926993275696445,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992166585790391,
      "loss": 5.3547,
      "mean_token_accuracy": 0.19188399910926818,
      "num_tokens": 8586052.0,
      "step": 3740
    },
    {
      "entropy": 5.565570116043091,
      "epoch": 0.3597502401536984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992137976828067,
      "loss": 5.4894,
      "mean_token_accuracy": 0.1816360756754875,
      "num_tokens": 8597832.0,
      "step": 3745
    },
    {
      "entropy": 5.512455606460572,
      "epoch": 0.36023054755043227,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992109315809955,
      "loss": 5.4831,
      "mean_token_accuracy": 0.18780883997678757,
      "num_tokens": 8609917.0,
      "step": 3750
    },
    {
      "entropy": 5.651883316040039,
      "epoch": 0.3607108549471662,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004992080602736725,
      "loss": 5.5063,
      "mean_token_accuracy": 0.1780880644917488,
      "num_tokens": 8620137.0,
      "step": 3755
    },
    {
      "entropy": 5.4667160987854,
      "epoch": 0.3611911623439001,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004992051837609039,
      "loss": 5.5175,
      "mean_token_accuracy": 0.18372666984796523,
      "num_tokens": 8630620.0,
      "step": 3760
    },
    {
      "entropy": 5.473104810714721,
      "epoch": 0.361671469740634,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004992023020427568,
      "loss": 5.3897,
      "mean_token_accuracy": 0.18974411636590957,
      "num_tokens": 8642820.0,
      "step": 3765
    },
    {
      "entropy": 5.489130640029908,
      "epoch": 0.3621517771373679,
      "grad_norm": 1.125,
      "learning_rate": 0.0004991994151192979,
      "loss": 5.3797,
      "mean_token_accuracy": 0.1936124101281166,
      "num_tokens": 8654081.0,
      "step": 3770
    },
    {
      "entropy": 5.4860845565795895,
      "epoch": 0.36263208453410184,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991965229905943,
      "loss": 5.4065,
      "mean_token_accuracy": 0.18499931544065476,
      "num_tokens": 8665097.0,
      "step": 3775
    },
    {
      "entropy": 5.463890933990479,
      "epoch": 0.3631123919308357,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991936256567133,
      "loss": 5.4007,
      "mean_token_accuracy": 0.19077470451593398,
      "num_tokens": 8676576.0,
      "step": 3780
    },
    {
      "entropy": 5.47804822921753,
      "epoch": 0.36359269932756966,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499190723117722,
      "loss": 5.4153,
      "mean_token_accuracy": 0.18302079886198044,
      "num_tokens": 8689217.0,
      "step": 3785
    },
    {
      "entropy": 5.586573553085327,
      "epoch": 0.36407300672430354,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004991878153736877,
      "loss": 5.5437,
      "mean_token_accuracy": 0.17407068759202957,
      "num_tokens": 8701101.0,
      "step": 3790
    },
    {
      "entropy": 5.552250385284424,
      "epoch": 0.3645533141210375,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991849024246781,
      "loss": 5.4694,
      "mean_token_accuracy": 0.18277360647916793,
      "num_tokens": 8712769.0,
      "step": 3795
    },
    {
      "entropy": 5.396585083007812,
      "epoch": 0.36503362151777136,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991819842707608,
      "loss": 5.3912,
      "mean_token_accuracy": 0.1880662336945534,
      "num_tokens": 8724691.0,
      "step": 3800
    },
    {
      "entropy": 5.441560173034668,
      "epoch": 0.3655139289145053,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991790609120035,
      "loss": 5.351,
      "mean_token_accuracy": 0.19300616383552552,
      "num_tokens": 8738008.0,
      "step": 3805
    },
    {
      "entropy": 5.561624002456665,
      "epoch": 0.3659942363112392,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000499176132348474,
      "loss": 5.5646,
      "mean_token_accuracy": 0.17607511281967164,
      "num_tokens": 8750353.0,
      "step": 3810
    },
    {
      "entropy": 5.486063623428345,
      "epoch": 0.3664745437079731,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991731985802405,
      "loss": 5.3771,
      "mean_token_accuracy": 0.18948758393526077,
      "num_tokens": 8761319.0,
      "step": 3815
    },
    {
      "entropy": 5.444846677780151,
      "epoch": 0.366954851104707,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991702596073708,
      "loss": 5.432,
      "mean_token_accuracy": 0.18246179223060607,
      "num_tokens": 8772722.0,
      "step": 3820
    },
    {
      "entropy": 5.616757345199585,
      "epoch": 0.36743515850144093,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991673154299335,
      "loss": 5.5217,
      "mean_token_accuracy": 0.17923828363418579,
      "num_tokens": 8783886.0,
      "step": 3825
    },
    {
      "entropy": 5.505244636535645,
      "epoch": 0.3679154658981748,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991643660479967,
      "loss": 5.4452,
      "mean_token_accuracy": 0.18481753170490264,
      "num_tokens": 8795161.0,
      "step": 3830
    },
    {
      "entropy": 5.41953592300415,
      "epoch": 0.36839577329490875,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004991614114616289,
      "loss": 5.41,
      "mean_token_accuracy": 0.18685935735702514,
      "num_tokens": 8805417.0,
      "step": 3835
    },
    {
      "entropy": 5.451370763778686,
      "epoch": 0.3688760806916426,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991584516708988,
      "loss": 5.4909,
      "mean_token_accuracy": 0.1781032308936119,
      "num_tokens": 8817980.0,
      "step": 3840
    },
    {
      "entropy": 5.429128694534302,
      "epoch": 0.36935638808837656,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991554866758751,
      "loss": 5.2727,
      "mean_token_accuracy": 0.19496323615312577,
      "num_tokens": 8829814.0,
      "step": 3845
    },
    {
      "entropy": 5.4406510353088375,
      "epoch": 0.36983669548511044,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991525164766265,
      "loss": 5.3801,
      "mean_token_accuracy": 0.18468381762504577,
      "num_tokens": 8840220.0,
      "step": 3850
    },
    {
      "entropy": 5.498718690872193,
      "epoch": 0.3703170028818444,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991495410732222,
      "loss": 5.5152,
      "mean_token_accuracy": 0.18211611360311508,
      "num_tokens": 8852156.0,
      "step": 3855
    },
    {
      "entropy": 5.56840672492981,
      "epoch": 0.37079731027857826,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991465604657311,
      "loss": 5.4345,
      "mean_token_accuracy": 0.179707895219326,
      "num_tokens": 8863362.0,
      "step": 3860
    },
    {
      "entropy": 5.438712120056152,
      "epoch": 0.3712776176753122,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004991435746542224,
      "loss": 5.4468,
      "mean_token_accuracy": 0.18730352818965912,
      "num_tokens": 8874957.0,
      "step": 3865
    },
    {
      "entropy": 5.513497734069825,
      "epoch": 0.37175792507204614,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991405836387655,
      "loss": 5.3911,
      "mean_token_accuracy": 0.18872731029987336,
      "num_tokens": 8886171.0,
      "step": 3870
    },
    {
      "entropy": 5.511085510253906,
      "epoch": 0.37223823246878,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991375874194298,
      "loss": 5.5136,
      "mean_token_accuracy": 0.18334945291280746,
      "num_tokens": 8898103.0,
      "step": 3875
    },
    {
      "entropy": 5.533126401901245,
      "epoch": 0.37271853986551395,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499134585996285,
      "loss": 5.4776,
      "mean_token_accuracy": 0.17990458458662034,
      "num_tokens": 8910195.0,
      "step": 3880
    },
    {
      "entropy": 5.476465272903442,
      "epoch": 0.37319884726224783,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991315793694004,
      "loss": 5.4307,
      "mean_token_accuracy": 0.18315743207931517,
      "num_tokens": 8922865.0,
      "step": 3885
    },
    {
      "entropy": 5.469011640548706,
      "epoch": 0.37367915465898177,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991285675388463,
      "loss": 5.4788,
      "mean_token_accuracy": 0.18122199028730393,
      "num_tokens": 8933913.0,
      "step": 3890
    },
    {
      "entropy": 5.487670183181763,
      "epoch": 0.37415946205571565,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991255505046922,
      "loss": 5.3889,
      "mean_token_accuracy": 0.18213097453117372,
      "num_tokens": 8945749.0,
      "step": 3895
    },
    {
      "entropy": 5.5673089027404785,
      "epoch": 0.3746397694524496,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004991225282670083,
      "loss": 5.3971,
      "mean_token_accuracy": 0.18210901021957399,
      "num_tokens": 8957454.0,
      "step": 3900
    },
    {
      "entropy": 5.531436252593994,
      "epoch": 0.37512007684918347,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499119500825865,
      "loss": 5.4453,
      "mean_token_accuracy": 0.18178220391273497,
      "num_tokens": 8968173.0,
      "step": 3905
    },
    {
      "entropy": 5.436434650421143,
      "epoch": 0.3756003842459174,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991164681813323,
      "loss": 5.4747,
      "mean_token_accuracy": 0.1811651349067688,
      "num_tokens": 8979187.0,
      "step": 3910
    },
    {
      "entropy": 5.502574014663696,
      "epoch": 0.3760806916426513,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991134303334807,
      "loss": 5.3675,
      "mean_token_accuracy": 0.19395778924226761,
      "num_tokens": 8990007.0,
      "step": 3915
    },
    {
      "entropy": 5.470153713226319,
      "epoch": 0.3765609990393852,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004991103872823807,
      "loss": 5.4978,
      "mean_token_accuracy": 0.18003395944833755,
      "num_tokens": 9001108.0,
      "step": 3920
    },
    {
      "entropy": 5.592336893081665,
      "epoch": 0.3770413064361191,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000499107339028103,
      "loss": 5.5035,
      "mean_token_accuracy": 0.18068006932735442,
      "num_tokens": 9012704.0,
      "step": 3925
    },
    {
      "entropy": 5.5123508930206295,
      "epoch": 0.37752161383285304,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004991042855707184,
      "loss": 5.4553,
      "mean_token_accuracy": 0.1849782407283783,
      "num_tokens": 9024351.0,
      "step": 3930
    },
    {
      "entropy": 5.420115041732788,
      "epoch": 0.3780019212295869,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004991012269102977,
      "loss": 5.3705,
      "mean_token_accuracy": 0.18681385815143586,
      "num_tokens": 9035099.0,
      "step": 3935
    },
    {
      "entropy": 5.55017318725586,
      "epoch": 0.37848222862632086,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004990981630469119,
      "loss": 5.4126,
      "mean_token_accuracy": 0.1835462212562561,
      "num_tokens": 9045675.0,
      "step": 3940
    },
    {
      "entropy": 5.506099081039428,
      "epoch": 0.37896253602305474,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004990950939806323,
      "loss": 5.3838,
      "mean_token_accuracy": 0.18631053268909453,
      "num_tokens": 9057133.0,
      "step": 3945
    },
    {
      "entropy": 5.420458745956421,
      "epoch": 0.3794428434197887,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049909201971153,
      "loss": 5.3884,
      "mean_token_accuracy": 0.18825268894433975,
      "num_tokens": 9067806.0,
      "step": 3950
    },
    {
      "entropy": 5.383552074432373,
      "epoch": 0.37992315081652256,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990889402396763,
      "loss": 5.3998,
      "mean_token_accuracy": 0.17989789545536042,
      "num_tokens": 9079893.0,
      "step": 3955
    },
    {
      "entropy": 5.504645681381225,
      "epoch": 0.3804034582132565,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990858555651431,
      "loss": 5.4222,
      "mean_token_accuracy": 0.18166604787111282,
      "num_tokens": 9091583.0,
      "step": 3960
    },
    {
      "entropy": 5.475414657592774,
      "epoch": 0.3808837656099904,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990827656880015,
      "loss": 5.4323,
      "mean_token_accuracy": 0.18971748352050782,
      "num_tokens": 9102919.0,
      "step": 3965
    },
    {
      "entropy": 5.508712434768677,
      "epoch": 0.3813640730067243,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990796706083235,
      "loss": 5.4015,
      "mean_token_accuracy": 0.18812174201011658,
      "num_tokens": 9113876.0,
      "step": 3970
    },
    {
      "entropy": 5.560408449172973,
      "epoch": 0.3818443804034582,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004990765703261809,
      "loss": 5.553,
      "mean_token_accuracy": 0.17641301304101945,
      "num_tokens": 9125630.0,
      "step": 3975
    },
    {
      "entropy": 5.512589597702027,
      "epoch": 0.38232468780019213,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004990734648416458,
      "loss": 5.46,
      "mean_token_accuracy": 0.18518878221511842,
      "num_tokens": 9137302.0,
      "step": 3980
    },
    {
      "entropy": 5.500503492355347,
      "epoch": 0.382804995196926,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004990703541547901,
      "loss": 5.4199,
      "mean_token_accuracy": 0.1892622396349907,
      "num_tokens": 9148048.0,
      "step": 3985
    },
    {
      "entropy": 5.572421550750732,
      "epoch": 0.38328530259365995,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004990672382656863,
      "loss": 5.4794,
      "mean_token_accuracy": 0.18170765340328215,
      "num_tokens": 9158656.0,
      "step": 3990
    },
    {
      "entropy": 5.5106470584869385,
      "epoch": 0.38376560999039383,
      "grad_norm": 1.125,
      "learning_rate": 0.0004990641171744064,
      "loss": 5.3745,
      "mean_token_accuracy": 0.1809146761894226,
      "num_tokens": 9169324.0,
      "step": 3995
    },
    {
      "entropy": 5.328002834320069,
      "epoch": 0.38424591738712777,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990609908810231,
      "loss": 5.3403,
      "mean_token_accuracy": 0.18658177107572554,
      "num_tokens": 9180238.0,
      "step": 4000
    },
    {
      "entropy": 5.325616550445557,
      "epoch": 0.38472622478386165,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990578593856089,
      "loss": 5.1987,
      "mean_token_accuracy": 0.198738731443882,
      "num_tokens": 9190766.0,
      "step": 4005
    },
    {
      "entropy": 5.372707653045654,
      "epoch": 0.3852065321805956,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990547226882366,
      "loss": 5.3568,
      "mean_token_accuracy": 0.18937029242515563,
      "num_tokens": 9202445.0,
      "step": 4010
    },
    {
      "entropy": 5.489741611480713,
      "epoch": 0.38568683957732947,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990515807889788,
      "loss": 5.4378,
      "mean_token_accuracy": 0.19018339067697526,
      "num_tokens": 9212383.0,
      "step": 4015
    },
    {
      "entropy": 5.438530158996582,
      "epoch": 0.3861671469740634,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004990484336879087,
      "loss": 5.3276,
      "mean_token_accuracy": 0.19339991062879563,
      "num_tokens": 9223619.0,
      "step": 4020
    },
    {
      "entropy": 5.475633049011231,
      "epoch": 0.3866474543707973,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990452813850992,
      "loss": 5.4781,
      "mean_token_accuracy": 0.18672976344823838,
      "num_tokens": 9234137.0,
      "step": 4025
    },
    {
      "entropy": 5.586398506164551,
      "epoch": 0.3871277617675312,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004990421238806236,
      "loss": 5.4159,
      "mean_token_accuracy": 0.1866623803973198,
      "num_tokens": 9246093.0,
      "step": 4030
    },
    {
      "entropy": 5.503391218185425,
      "epoch": 0.38760806916426516,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004990389611745551,
      "loss": 5.5072,
      "mean_token_accuracy": 0.17473563998937608,
      "num_tokens": 9258021.0,
      "step": 4035
    },
    {
      "entropy": 5.527388000488282,
      "epoch": 0.38808837656099904,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004990357932669672,
      "loss": 5.5463,
      "mean_token_accuracy": 0.18486131727695465,
      "num_tokens": 9267668.0,
      "step": 4040
    },
    {
      "entropy": 5.555417251586914,
      "epoch": 0.388568683957733,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990326201579335,
      "loss": 5.4357,
      "mean_token_accuracy": 0.18652143627405166,
      "num_tokens": 9279381.0,
      "step": 4045
    },
    {
      "entropy": 5.49374942779541,
      "epoch": 0.38904899135446686,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990294418475274,
      "loss": 5.5262,
      "mean_token_accuracy": 0.1776568189263344,
      "num_tokens": 9292578.0,
      "step": 4050
    },
    {
      "entropy": 5.527858591079712,
      "epoch": 0.3895292987512008,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004990262583358231,
      "loss": 5.3043,
      "mean_token_accuracy": 0.18736467957496644,
      "num_tokens": 9304116.0,
      "step": 4055
    },
    {
      "entropy": 5.54514856338501,
      "epoch": 0.3900096061479347,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990230696228943,
      "loss": 5.5257,
      "mean_token_accuracy": 0.18316859900951385,
      "num_tokens": 9315877.0,
      "step": 4060
    },
    {
      "entropy": 5.448623847961426,
      "epoch": 0.3904899135446686,
      "grad_norm": 1.125,
      "learning_rate": 0.0004990198757088149,
      "loss": 5.4296,
      "mean_token_accuracy": 0.18576246052980422,
      "num_tokens": 9326859.0,
      "step": 4065
    },
    {
      "entropy": 5.530019092559814,
      "epoch": 0.3909702209414025,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004990166765936593,
      "loss": 5.4965,
      "mean_token_accuracy": 0.175887767970562,
      "num_tokens": 9338067.0,
      "step": 4070
    },
    {
      "entropy": 5.5362382411956785,
      "epoch": 0.3914505283381364,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990134722775016,
      "loss": 5.5252,
      "mean_token_accuracy": 0.18217928111553192,
      "num_tokens": 9349564.0,
      "step": 4075
    },
    {
      "entropy": 5.4288777828216555,
      "epoch": 0.3919308357348703,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004990102627604162,
      "loss": 5.3357,
      "mean_token_accuracy": 0.1894381582736969,
      "num_tokens": 9359962.0,
      "step": 4080
    },
    {
      "entropy": 5.420823383331299,
      "epoch": 0.39241114313160425,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004990070480424778,
      "loss": 5.3477,
      "mean_token_accuracy": 0.1889730766415596,
      "num_tokens": 9370251.0,
      "step": 4085
    },
    {
      "entropy": 5.543254041671753,
      "epoch": 0.3928914505283381,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990038281237608,
      "loss": 5.477,
      "mean_token_accuracy": 0.17700579166412353,
      "num_tokens": 9381848.0,
      "step": 4090
    },
    {
      "entropy": 5.4344888687133786,
      "epoch": 0.39337175792507206,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990006030043401,
      "loss": 5.389,
      "mean_token_accuracy": 0.18838301450014114,
      "num_tokens": 9392475.0,
      "step": 4095
    },
    {
      "entropy": 5.484930610656738,
      "epoch": 0.39385206532180594,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989973726842906,
      "loss": 5.5131,
      "mean_token_accuracy": 0.1790087714791298,
      "num_tokens": 9404014.0,
      "step": 4100
    },
    {
      "entropy": 5.546372270584106,
      "epoch": 0.3943323727185399,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989941371636872,
      "loss": 5.3699,
      "mean_token_accuracy": 0.17934942245483398,
      "num_tokens": 9415590.0,
      "step": 4105
    },
    {
      "entropy": 5.545758581161499,
      "epoch": 0.39481268011527376,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989908964426051,
      "loss": 5.5446,
      "mean_token_accuracy": 0.1808340519666672,
      "num_tokens": 9427486.0,
      "step": 4110
    },
    {
      "entropy": 5.498834037780762,
      "epoch": 0.3952929875120077,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989876505211194,
      "loss": 5.4192,
      "mean_token_accuracy": 0.18967240750789643,
      "num_tokens": 9438212.0,
      "step": 4115
    },
    {
      "entropy": 5.464620971679688,
      "epoch": 0.3957732949087416,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989843993993056,
      "loss": 5.367,
      "mean_token_accuracy": 0.19194960594177246,
      "num_tokens": 9450922.0,
      "step": 4120
    },
    {
      "entropy": 5.4229480743408205,
      "epoch": 0.3962536023054755,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989811430772392,
      "loss": 5.3588,
      "mean_token_accuracy": 0.18508039712905883,
      "num_tokens": 9461676.0,
      "step": 4125
    },
    {
      "entropy": 5.3516346454620365,
      "epoch": 0.3967339097022094,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989778815549957,
      "loss": 5.3103,
      "mean_token_accuracy": 0.19300473630428314,
      "num_tokens": 9473517.0,
      "step": 4130
    },
    {
      "entropy": 5.531142520904541,
      "epoch": 0.39721421709894333,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989746148326508,
      "loss": 5.4542,
      "mean_token_accuracy": 0.17928824275732042,
      "num_tokens": 9485660.0,
      "step": 4135
    },
    {
      "entropy": 5.516511487960815,
      "epoch": 0.3976945244956772,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004989713429102805,
      "loss": 5.4569,
      "mean_token_accuracy": 0.18583335876464843,
      "num_tokens": 9495983.0,
      "step": 4140
    },
    {
      "entropy": 5.448024129867553,
      "epoch": 0.39817483189241115,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989680657879607,
      "loss": 5.411,
      "mean_token_accuracy": 0.17801857739686966,
      "num_tokens": 9507438.0,
      "step": 4145
    },
    {
      "entropy": 5.389159393310547,
      "epoch": 0.39865513928914503,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989647834657675,
      "loss": 5.3197,
      "mean_token_accuracy": 0.1888060450553894,
      "num_tokens": 9517508.0,
      "step": 4150
    },
    {
      "entropy": 5.444024705886841,
      "epoch": 0.39913544668587897,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498961495943777,
      "loss": 5.3357,
      "mean_token_accuracy": 0.19338419288396835,
      "num_tokens": 9527641.0,
      "step": 4155
    },
    {
      "entropy": 5.49064793586731,
      "epoch": 0.39961575408261285,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989582032220656,
      "loss": 5.4178,
      "mean_token_accuracy": 0.18565522730350495,
      "num_tokens": 9538762.0,
      "step": 4160
    },
    {
      "entropy": 5.476654815673828,
      "epoch": 0.4000960614793468,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004989549053007096,
      "loss": 5.4622,
      "mean_token_accuracy": 0.18227968066930772,
      "num_tokens": 9550246.0,
      "step": 4165
    },
    {
      "entropy": 5.459161758422852,
      "epoch": 0.40057636887608067,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989516021797858,
      "loss": 5.3955,
      "mean_token_accuracy": 0.1861029401421547,
      "num_tokens": 9562341.0,
      "step": 4170
    },
    {
      "entropy": 5.536681127548218,
      "epoch": 0.4010566762728146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000498948293859371,
      "loss": 5.4813,
      "mean_token_accuracy": 0.18325578421354294,
      "num_tokens": 9574484.0,
      "step": 4175
    },
    {
      "entropy": 5.51067886352539,
      "epoch": 0.4015369836695485,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004989449803395415,
      "loss": 5.4233,
      "mean_token_accuracy": 0.18632330000400543,
      "num_tokens": 9585241.0,
      "step": 4180
    },
    {
      "entropy": 5.4815582752227785,
      "epoch": 0.4020172910662824,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989416616203747,
      "loss": 5.428,
      "mean_token_accuracy": 0.18480357229709626,
      "num_tokens": 9598683.0,
      "step": 4185
    },
    {
      "entropy": 5.494958734512329,
      "epoch": 0.4024975984630163,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989383377019476,
      "loss": 5.4552,
      "mean_token_accuracy": 0.1796182617545128,
      "num_tokens": 9610127.0,
      "step": 4190
    },
    {
      "entropy": 5.493792247772217,
      "epoch": 0.40297790585975024,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004989350085843371,
      "loss": 5.3587,
      "mean_token_accuracy": 0.18685469925403594,
      "num_tokens": 9620009.0,
      "step": 4195
    },
    {
      "entropy": 5.425169229507446,
      "epoch": 0.4034582132564842,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004989316742676207,
      "loss": 5.4341,
      "mean_token_accuracy": 0.1924751803278923,
      "num_tokens": 9632479.0,
      "step": 4200
    },
    {
      "entropy": 5.540858888626099,
      "epoch": 0.40393852065321806,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989283347518757,
      "loss": 5.5052,
      "mean_token_accuracy": 0.18447671532630922,
      "num_tokens": 9644447.0,
      "step": 4205
    },
    {
      "entropy": 5.488879728317261,
      "epoch": 0.404418828049952,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989249900371797,
      "loss": 5.4069,
      "mean_token_accuracy": 0.17978627979755402,
      "num_tokens": 9656886.0,
      "step": 4210
    },
    {
      "entropy": 5.470227193832398,
      "epoch": 0.4048991354466859,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989216401236103,
      "loss": 5.4419,
      "mean_token_accuracy": 0.1867706671357155,
      "num_tokens": 9667079.0,
      "step": 4215
    },
    {
      "entropy": 5.5312903881072994,
      "epoch": 0.4053794428434198,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989182850112455,
      "loss": 5.4929,
      "mean_token_accuracy": 0.18182475715875626,
      "num_tokens": 9678237.0,
      "step": 4220
    },
    {
      "entropy": 5.445821523666382,
      "epoch": 0.4058597502401537,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989149247001629,
      "loss": 5.3531,
      "mean_token_accuracy": 0.18974065333604812,
      "num_tokens": 9688809.0,
      "step": 4225
    },
    {
      "entropy": 5.536614418029785,
      "epoch": 0.40634005763688763,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004989115591904407,
      "loss": 5.4582,
      "mean_token_accuracy": 0.19188262522220612,
      "num_tokens": 9700460.0,
      "step": 4230
    },
    {
      "entropy": 5.425814390182495,
      "epoch": 0.4068203650336215,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004989081884821569,
      "loss": 5.3915,
      "mean_token_accuracy": 0.18563754558563234,
      "num_tokens": 9711379.0,
      "step": 4235
    },
    {
      "entropy": 5.427301549911499,
      "epoch": 0.40730067243035545,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989048125753899,
      "loss": 5.334,
      "mean_token_accuracy": 0.18708637952804566,
      "num_tokens": 9722792.0,
      "step": 4240
    },
    {
      "entropy": 5.495201301574707,
      "epoch": 0.40778097982708933,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498901431470218,
      "loss": 5.3744,
      "mean_token_accuracy": 0.19147287607192992,
      "num_tokens": 9733524.0,
      "step": 4245
    },
    {
      "entropy": 5.50749192237854,
      "epoch": 0.40826128722382327,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988980451667198,
      "loss": 5.4474,
      "mean_token_accuracy": 0.1811591386795044,
      "num_tokens": 9745430.0,
      "step": 4250
    },
    {
      "entropy": 5.418610286712647,
      "epoch": 0.40874159462055715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988946536649737,
      "loss": 5.3488,
      "mean_token_accuracy": 0.18888413310050964,
      "num_tokens": 9756927.0,
      "step": 4255
    },
    {
      "entropy": 5.43942403793335,
      "epoch": 0.4092219020172911,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988912569650585,
      "loss": 5.3968,
      "mean_token_accuracy": 0.18817504793405532,
      "num_tokens": 9768350.0,
      "step": 4260
    },
    {
      "entropy": 5.442109155654907,
      "epoch": 0.40970220941402496,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004988878550670533,
      "loss": 5.3322,
      "mean_token_accuracy": 0.19701816290616989,
      "num_tokens": 9779238.0,
      "step": 4265
    },
    {
      "entropy": 5.525984859466552,
      "epoch": 0.4101825168107589,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004988844479710369,
      "loss": 5.5082,
      "mean_token_accuracy": 0.1816016837954521,
      "num_tokens": 9789262.0,
      "step": 4270
    },
    {
      "entropy": 5.533520936965942,
      "epoch": 0.4106628242074928,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004988810356770884,
      "loss": 5.4711,
      "mean_token_accuracy": 0.18100447058677674,
      "num_tokens": 9800909.0,
      "step": 4275
    },
    {
      "entropy": 5.4796051502227785,
      "epoch": 0.4111431316042267,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498877618185287,
      "loss": 5.4232,
      "mean_token_accuracy": 0.1838565245270729,
      "num_tokens": 9812538.0,
      "step": 4280
    },
    {
      "entropy": 5.491272401809693,
      "epoch": 0.4116234390009606,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004988741954957121,
      "loss": 5.4335,
      "mean_token_accuracy": 0.18602810204029083,
      "num_tokens": 9824071.0,
      "step": 4285
    },
    {
      "entropy": 5.599289321899414,
      "epoch": 0.41210374639769454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988707676084432,
      "loss": 5.5216,
      "mean_token_accuracy": 0.17931292653083802,
      "num_tokens": 9836713.0,
      "step": 4290
    },
    {
      "entropy": 5.429421710968017,
      "epoch": 0.4125840537944284,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004988673345235597,
      "loss": 5.3548,
      "mean_token_accuracy": 0.1887418046593666,
      "num_tokens": 9847336.0,
      "step": 4295
    },
    {
      "entropy": 5.373703384399414,
      "epoch": 0.41306436119116235,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988638962411416,
      "loss": 5.3534,
      "mean_token_accuracy": 0.18685731738805772,
      "num_tokens": 9858567.0,
      "step": 4300
    },
    {
      "entropy": 5.560068035125733,
      "epoch": 0.41354466858789624,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988604527612685,
      "loss": 5.4298,
      "mean_token_accuracy": 0.17734273374080659,
      "num_tokens": 9869254.0,
      "step": 4305
    },
    {
      "entropy": 5.451811265945435,
      "epoch": 0.4140249759846302,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988570040840205,
      "loss": 5.368,
      "mean_token_accuracy": 0.19178945571184158,
      "num_tokens": 9879635.0,
      "step": 4310
    },
    {
      "entropy": 5.510119676589966,
      "epoch": 0.41450528338136405,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004988535502094774,
      "loss": 5.4348,
      "mean_token_accuracy": 0.18689756840467453,
      "num_tokens": 9891728.0,
      "step": 4315
    },
    {
      "entropy": 5.416639900207519,
      "epoch": 0.414985590778098,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988500911377198,
      "loss": 5.2427,
      "mean_token_accuracy": 0.194938962161541,
      "num_tokens": 9903140.0,
      "step": 4320
    },
    {
      "entropy": 5.350975036621094,
      "epoch": 0.41546589817483187,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988466268688276,
      "loss": 5.392,
      "mean_token_accuracy": 0.18834515213966369,
      "num_tokens": 9915562.0,
      "step": 4325
    },
    {
      "entropy": 5.46756157875061,
      "epoch": 0.4159462055715658,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004988431574028814,
      "loss": 5.3438,
      "mean_token_accuracy": 0.18281278908252716,
      "num_tokens": 9925720.0,
      "step": 4330
    },
    {
      "entropy": 5.494148445129395,
      "epoch": 0.4164265129682997,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004988396827399618,
      "loss": 5.4082,
      "mean_token_accuracy": 0.18209784477949142,
      "num_tokens": 9936492.0,
      "step": 4335
    },
    {
      "entropy": 5.370842695236206,
      "epoch": 0.4169068203650336,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988362028801495,
      "loss": 5.3411,
      "mean_token_accuracy": 0.18397990614175797,
      "num_tokens": 9948658.0,
      "step": 4340
    },
    {
      "entropy": 5.3537579536437985,
      "epoch": 0.4173871277617675,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988327178235253,
      "loss": 5.2904,
      "mean_token_accuracy": 0.1929224193096161,
      "num_tokens": 9959710.0,
      "step": 4345
    },
    {
      "entropy": 5.423714590072632,
      "epoch": 0.41786743515850144,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988292275701699,
      "loss": 5.2838,
      "mean_token_accuracy": 0.1938964068889618,
      "num_tokens": 9970302.0,
      "step": 4350
    },
    {
      "entropy": 5.444131565093994,
      "epoch": 0.4183477425552353,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988257321201646,
      "loss": 5.4102,
      "mean_token_accuracy": 0.18316021859645842,
      "num_tokens": 9981776.0,
      "step": 4355
    },
    {
      "entropy": 5.462032699584961,
      "epoch": 0.41882804995196926,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988222314735902,
      "loss": 5.3951,
      "mean_token_accuracy": 0.18458667248487473,
      "num_tokens": 9993471.0,
      "step": 4360
    },
    {
      "entropy": 5.5117199420928955,
      "epoch": 0.41930835734870314,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988187256305284,
      "loss": 5.4003,
      "mean_token_accuracy": 0.1906923159956932,
      "num_tokens": 10003818.0,
      "step": 4365
    },
    {
      "entropy": 5.434498977661133,
      "epoch": 0.4197886647454371,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988152145910603,
      "loss": 5.4907,
      "mean_token_accuracy": 0.17460868656635284,
      "num_tokens": 10015507.0,
      "step": 4370
    },
    {
      "entropy": 5.493358182907104,
      "epoch": 0.420268972142171,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004988116983552675,
      "loss": 5.3176,
      "mean_token_accuracy": 0.1922619968652725,
      "num_tokens": 10026994.0,
      "step": 4375
    },
    {
      "entropy": 5.465518522262573,
      "epoch": 0.4207492795389049,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004988081769232317,
      "loss": 5.4581,
      "mean_token_accuracy": 0.1843365788459778,
      "num_tokens": 10039547.0,
      "step": 4380
    },
    {
      "entropy": 5.48856520652771,
      "epoch": 0.42122958693563883,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004988046502950346,
      "loss": 5.4761,
      "mean_token_accuracy": 0.1799684524536133,
      "num_tokens": 10050568.0,
      "step": 4385
    },
    {
      "entropy": 5.524022388458252,
      "epoch": 0.4217098943323727,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000498801118470758,
      "loss": 5.3961,
      "mean_token_accuracy": 0.18780332952737808,
      "num_tokens": 10061780.0,
      "step": 4390
    },
    {
      "entropy": 5.356531238555908,
      "epoch": 0.42219020172910665,
      "grad_norm": 1.0625,
      "learning_rate": 0.000498797581450484,
      "loss": 5.3561,
      "mean_token_accuracy": 0.18687074333429338,
      "num_tokens": 10073101.0,
      "step": 4395
    },
    {
      "entropy": 5.373044490814209,
      "epoch": 0.42267050912584053,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004987940392342948,
      "loss": 5.2842,
      "mean_token_accuracy": 0.20293211042881013,
      "num_tokens": 10084397.0,
      "step": 4400
    },
    {
      "entropy": 5.490303087234497,
      "epoch": 0.42315081652257447,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987904918222726,
      "loss": 5.3625,
      "mean_token_accuracy": 0.19282422363758087,
      "num_tokens": 10095635.0,
      "step": 4405
    },
    {
      "entropy": 5.440840864181519,
      "epoch": 0.42363112391930835,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987869392144996,
      "loss": 5.4444,
      "mean_token_accuracy": 0.1886908993124962,
      "num_tokens": 10107744.0,
      "step": 4410
    },
    {
      "entropy": 5.47255048751831,
      "epoch": 0.4241114313160423,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004987833814110584,
      "loss": 5.3646,
      "mean_token_accuracy": 0.18840182721614837,
      "num_tokens": 10118552.0,
      "step": 4415
    },
    {
      "entropy": 5.531020736694336,
      "epoch": 0.42459173871277617,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987798184120316,
      "loss": 5.4909,
      "mean_token_accuracy": 0.17462566494941711,
      "num_tokens": 10130813.0,
      "step": 4420
    },
    {
      "entropy": 5.504419231414795,
      "epoch": 0.4250720461095101,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004987762502175018,
      "loss": 5.5036,
      "mean_token_accuracy": 0.18759763091802598,
      "num_tokens": 10142196.0,
      "step": 4425
    },
    {
      "entropy": 5.4172688007354735,
      "epoch": 0.425552353506244,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498772676827552,
      "loss": 5.3984,
      "mean_token_accuracy": 0.18967483788728715,
      "num_tokens": 10155012.0,
      "step": 4430
    },
    {
      "entropy": 5.544624853134155,
      "epoch": 0.4260326609029779,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987690982422652,
      "loss": 5.4981,
      "mean_token_accuracy": 0.17872706949710845,
      "num_tokens": 10166160.0,
      "step": 4435
    },
    {
      "entropy": 5.461737442016601,
      "epoch": 0.4265129682997118,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987655144617243,
      "loss": 5.2654,
      "mean_token_accuracy": 0.19589474946260452,
      "num_tokens": 10177663.0,
      "step": 4440
    },
    {
      "entropy": 5.346937131881714,
      "epoch": 0.42699327569644574,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004987619254860126,
      "loss": 5.3591,
      "mean_token_accuracy": 0.1953091338276863,
      "num_tokens": 10189127.0,
      "step": 4445
    },
    {
      "entropy": 5.564468431472778,
      "epoch": 0.4274735830931796,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004987583313152134,
      "loss": 5.5259,
      "mean_token_accuracy": 0.17608540058135985,
      "num_tokens": 10200765.0,
      "step": 4450
    },
    {
      "entropy": 5.487565374374389,
      "epoch": 0.42795389048991356,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004987547319494104,
      "loss": 5.3779,
      "mean_token_accuracy": 0.1892255961894989,
      "num_tokens": 10213047.0,
      "step": 4455
    },
    {
      "entropy": 5.387438154220581,
      "epoch": 0.42843419788664744,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004987511273886867,
      "loss": 5.3464,
      "mean_token_accuracy": 0.18826134651899337,
      "num_tokens": 10223790.0,
      "step": 4460
    },
    {
      "entropy": 5.451563596725464,
      "epoch": 0.4289145052833814,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004987475176331263,
      "loss": 5.3455,
      "mean_token_accuracy": 0.18942897468805314,
      "num_tokens": 10235077.0,
      "step": 4465
    },
    {
      "entropy": 5.471985483169556,
      "epoch": 0.42939481268011526,
      "grad_norm": 1.125,
      "learning_rate": 0.0004987439026828129,
      "loss": 5.4138,
      "mean_token_accuracy": 0.17842211425304413,
      "num_tokens": 10248844.0,
      "step": 4470
    },
    {
      "entropy": 5.426624011993408,
      "epoch": 0.4298751200768492,
      "grad_norm": 1.125,
      "learning_rate": 0.0004987402825378305,
      "loss": 5.3354,
      "mean_token_accuracy": 0.19654394835233688,
      "num_tokens": 10259013.0,
      "step": 4475
    },
    {
      "entropy": 5.380995845794677,
      "epoch": 0.4303554274735831,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987366571982631,
      "loss": 5.3561,
      "mean_token_accuracy": 0.18739008903503418,
      "num_tokens": 10271415.0,
      "step": 4480
    },
    {
      "entropy": 5.38149881362915,
      "epoch": 0.430835734870317,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004987330266641948,
      "loss": 5.3675,
      "mean_token_accuracy": 0.1893654227256775,
      "num_tokens": 10283204.0,
      "step": 4485
    },
    {
      "entropy": 5.551987552642823,
      "epoch": 0.4313160422670509,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004987293909357101,
      "loss": 5.5348,
      "mean_token_accuracy": 0.18340096920728682,
      "num_tokens": 10294111.0,
      "step": 4490
    },
    {
      "entropy": 5.553006505966186,
      "epoch": 0.43179634966378483,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004987257500128933,
      "loss": 5.427,
      "mean_token_accuracy": 0.17463399320840836,
      "num_tokens": 10305417.0,
      "step": 4495
    },
    {
      "entropy": 5.338708400726318,
      "epoch": 0.4322766570605187,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004987221038958288,
      "loss": 5.3095,
      "mean_token_accuracy": 0.18922245055437087,
      "num_tokens": 10316416.0,
      "step": 4500
    },
    {
      "entropy": 5.512272214889526,
      "epoch": 0.43275696445725265,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987184525846015,
      "loss": 5.4741,
      "mean_token_accuracy": 0.18813111782073974,
      "num_tokens": 10328251.0,
      "step": 4505
    },
    {
      "entropy": 5.521883487701416,
      "epoch": 0.4332372718539865,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987147960792958,
      "loss": 5.4424,
      "mean_token_accuracy": 0.1921009048819542,
      "num_tokens": 10340189.0,
      "step": 4510
    },
    {
      "entropy": 5.440450286865234,
      "epoch": 0.43371757925072046,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004987111343799971,
      "loss": 5.3682,
      "mean_token_accuracy": 0.1887647196650505,
      "num_tokens": 10351660.0,
      "step": 4515
    },
    {
      "entropy": 5.4706371307373045,
      "epoch": 0.43419788664745435,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00049870746748679,
      "loss": 5.4136,
      "mean_token_accuracy": 0.1934151217341423,
      "num_tokens": 10362240.0,
      "step": 4520
    },
    {
      "entropy": 5.569837665557861,
      "epoch": 0.4346781940441883,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004987037953997598,
      "loss": 5.4576,
      "mean_token_accuracy": 0.18817198127508164,
      "num_tokens": 10373646.0,
      "step": 4525
    },
    {
      "entropy": 5.466810464859009,
      "epoch": 0.43515850144092216,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987001181189918,
      "loss": 5.4553,
      "mean_token_accuracy": 0.18504131883382796,
      "num_tokens": 10385540.0,
      "step": 4530
    },
    {
      "entropy": 5.428154420852661,
      "epoch": 0.4356388088376561,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986964356445713,
      "loss": 5.4545,
      "mean_token_accuracy": 0.17844501584768296,
      "num_tokens": 10396641.0,
      "step": 4535
    },
    {
      "entropy": 5.524321556091309,
      "epoch": 0.43611911623439004,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986927479765837,
      "loss": 5.4176,
      "mean_token_accuracy": 0.18325759023427962,
      "num_tokens": 10408165.0,
      "step": 4540
    },
    {
      "entropy": 5.473247861862182,
      "epoch": 0.4365994236311239,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004986890551151148,
      "loss": 5.3827,
      "mean_token_accuracy": 0.19138990789651872,
      "num_tokens": 10419646.0,
      "step": 4545
    },
    {
      "entropy": 5.4596000671386715,
      "epoch": 0.43707973102785785,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986853570602503,
      "loss": 5.3892,
      "mean_token_accuracy": 0.1846207231283188,
      "num_tokens": 10431051.0,
      "step": 4550
    },
    {
      "entropy": 5.499724292755127,
      "epoch": 0.43756003842459174,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004986816538120758,
      "loss": 5.4599,
      "mean_token_accuracy": 0.18422451615333557,
      "num_tokens": 10441833.0,
      "step": 4555
    },
    {
      "entropy": 5.472308540344239,
      "epoch": 0.43804034582132567,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004986779453706778,
      "loss": 5.4165,
      "mean_token_accuracy": 0.18615081906318665,
      "num_tokens": 10453544.0,
      "step": 4560
    },
    {
      "entropy": 5.42457594871521,
      "epoch": 0.43852065321805955,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986742317361419,
      "loss": 5.2621,
      "mean_token_accuracy": 0.19069198071956633,
      "num_tokens": 10464140.0,
      "step": 4565
    },
    {
      "entropy": 5.3433092594146725,
      "epoch": 0.4390009606147935,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986705129085546,
      "loss": 5.3205,
      "mean_token_accuracy": 0.19364836066961288,
      "num_tokens": 10475739.0,
      "step": 4570
    },
    {
      "entropy": 5.33997106552124,
      "epoch": 0.43948126801152737,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986667888880021,
      "loss": 5.3782,
      "mean_token_accuracy": 0.19127382785081865,
      "num_tokens": 10487451.0,
      "step": 4575
    },
    {
      "entropy": 5.4929262638092045,
      "epoch": 0.4399615754082613,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986630596745709,
      "loss": 5.4304,
      "mean_token_accuracy": 0.17770482748746871,
      "num_tokens": 10499367.0,
      "step": 4580
    },
    {
      "entropy": 5.458431196212769,
      "epoch": 0.4404418828049952,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004986593252683477,
      "loss": 5.2515,
      "mean_token_accuracy": 0.194043530523777,
      "num_tokens": 10511192.0,
      "step": 4585
    },
    {
      "entropy": 5.283456087112427,
      "epoch": 0.4409221902017291,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986555856694191,
      "loss": 5.2839,
      "mean_token_accuracy": 0.1867209032177925,
      "num_tokens": 10522754.0,
      "step": 4590
    },
    {
      "entropy": 5.477353239059449,
      "epoch": 0.441402497598463,
      "grad_norm": 1.125,
      "learning_rate": 0.0004986518408778718,
      "loss": 5.4302,
      "mean_token_accuracy": 0.18063050508499146,
      "num_tokens": 10533439.0,
      "step": 4595
    },
    {
      "entropy": 5.547759294509888,
      "epoch": 0.44188280499519694,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986480908937929,
      "loss": 5.3776,
      "mean_token_accuracy": 0.18738025575876235,
      "num_tokens": 10543654.0,
      "step": 4600
    },
    {
      "entropy": 5.373477554321289,
      "epoch": 0.4423631123919308,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986443357172695,
      "loss": 5.2886,
      "mean_token_accuracy": 0.1928374096751213,
      "num_tokens": 10554936.0,
      "step": 4605
    },
    {
      "entropy": 5.421936082839966,
      "epoch": 0.44284341978866476,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004986405753483887,
      "loss": 5.3615,
      "mean_token_accuracy": 0.18666858673095704,
      "num_tokens": 10566410.0,
      "step": 4610
    },
    {
      "entropy": 5.434760856628418,
      "epoch": 0.44332372718539864,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004986368097872377,
      "loss": 5.3533,
      "mean_token_accuracy": 0.1969812572002411,
      "num_tokens": 10577258.0,
      "step": 4615
    },
    {
      "entropy": 5.429739856719971,
      "epoch": 0.4438040345821326,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986330390339042,
      "loss": 5.4615,
      "mean_token_accuracy": 0.18065902441740037,
      "num_tokens": 10588783.0,
      "step": 4620
    },
    {
      "entropy": 5.423320722579956,
      "epoch": 0.44428434197886646,
      "grad_norm": 1.125,
      "learning_rate": 0.0004986292630884755,
      "loss": 5.3362,
      "mean_token_accuracy": 0.18651131838560103,
      "num_tokens": 10599778.0,
      "step": 4625
    },
    {
      "entropy": 5.456650638580323,
      "epoch": 0.4447646493756004,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004986254819510393,
      "loss": 5.3593,
      "mean_token_accuracy": 0.194785675406456,
      "num_tokens": 10611211.0,
      "step": 4630
    },
    {
      "entropy": 5.34439058303833,
      "epoch": 0.4452449567723343,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986216956216835,
      "loss": 5.2885,
      "mean_token_accuracy": 0.19258749336004258,
      "num_tokens": 10624535.0,
      "step": 4635
    },
    {
      "entropy": 5.364201498031616,
      "epoch": 0.4457252641690682,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000498617904100496,
      "loss": 5.3311,
      "mean_token_accuracy": 0.19969772100448607,
      "num_tokens": 10635206.0,
      "step": 4640
    },
    {
      "entropy": 5.487021160125733,
      "epoch": 0.4462055715658021,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986141073875646,
      "loss": 5.3739,
      "mean_token_accuracy": 0.19442293643951417,
      "num_tokens": 10646993.0,
      "step": 4645
    },
    {
      "entropy": 5.398804473876953,
      "epoch": 0.44668587896253603,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004986103054829779,
      "loss": 5.3279,
      "mean_token_accuracy": 0.19260976910591127,
      "num_tokens": 10657262.0,
      "step": 4650
    },
    {
      "entropy": 5.412921714782715,
      "epoch": 0.4471661863592699,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004986064983868237,
      "loss": 5.3561,
      "mean_token_accuracy": 0.1881539523601532,
      "num_tokens": 10668962.0,
      "step": 4655
    },
    {
      "entropy": 5.478003215789795,
      "epoch": 0.44764649375600385,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986026860991906,
      "loss": 5.3638,
      "mean_token_accuracy": 0.19651059955358505,
      "num_tokens": 10680183.0,
      "step": 4660
    },
    {
      "entropy": 5.438195371627808,
      "epoch": 0.44812680115273773,
      "grad_norm": 1.125,
      "learning_rate": 0.0004985988686201672,
      "loss": 5.3068,
      "mean_token_accuracy": 0.19921039342880248,
      "num_tokens": 10691221.0,
      "step": 4665
    },
    {
      "entropy": 5.342834854125977,
      "epoch": 0.44860710854947167,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004985950459498419,
      "loss": 5.3352,
      "mean_token_accuracy": 0.1852909281849861,
      "num_tokens": 10703462.0,
      "step": 4670
    },
    {
      "entropy": 5.480234575271607,
      "epoch": 0.44908741594620555,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985912180883037,
      "loss": 5.324,
      "mean_token_accuracy": 0.19051701128482817,
      "num_tokens": 10713774.0,
      "step": 4675
    },
    {
      "entropy": 5.520461845397949,
      "epoch": 0.4495677233429395,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004985873850356411,
      "loss": 5.5299,
      "mean_token_accuracy": 0.1753215327858925,
      "num_tokens": 10725109.0,
      "step": 4680
    },
    {
      "entropy": 5.374638605117798,
      "epoch": 0.45004803073967337,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004985835467919436,
      "loss": 5.3629,
      "mean_token_accuracy": 0.1957563117146492,
      "num_tokens": 10735695.0,
      "step": 4685
    },
    {
      "entropy": 5.44810266494751,
      "epoch": 0.4505283381364073,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004985797033572999,
      "loss": 5.4252,
      "mean_token_accuracy": 0.1872813656926155,
      "num_tokens": 10747554.0,
      "step": 4690
    },
    {
      "entropy": 5.553777599334717,
      "epoch": 0.4510086455331412,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004985758547317994,
      "loss": 5.4286,
      "mean_token_accuracy": 0.18837624043226242,
      "num_tokens": 10759175.0,
      "step": 4695
    },
    {
      "entropy": 5.4048378467559814,
      "epoch": 0.4514889529298751,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004985720009155315,
      "loss": 5.2941,
      "mean_token_accuracy": 0.19457966089248657,
      "num_tokens": 10770365.0,
      "step": 4700
    },
    {
      "entropy": 5.40498456954956,
      "epoch": 0.45196926032660906,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985681419085856,
      "loss": 5.3985,
      "mean_token_accuracy": 0.1860631972551346,
      "num_tokens": 10783096.0,
      "step": 4705
    },
    {
      "entropy": 5.552437353134155,
      "epoch": 0.45244956772334294,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004985642777110513,
      "loss": 5.4805,
      "mean_token_accuracy": 0.18021069020032882,
      "num_tokens": 10795312.0,
      "step": 4710
    },
    {
      "entropy": 5.431002426147461,
      "epoch": 0.4529298751200769,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985604083230183,
      "loss": 5.3158,
      "mean_token_accuracy": 0.1942341759800911,
      "num_tokens": 10806451.0,
      "step": 4715
    },
    {
      "entropy": 5.548686075210571,
      "epoch": 0.45341018251681076,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004985565337445765,
      "loss": 5.4729,
      "mean_token_accuracy": 0.18171751648187637,
      "num_tokens": 10818185.0,
      "step": 4720
    },
    {
      "entropy": 5.41652479171753,
      "epoch": 0.4538904899135447,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004985526539758158,
      "loss": 5.3332,
      "mean_token_accuracy": 0.1988949477672577,
      "num_tokens": 10830203.0,
      "step": 4725
    },
    {
      "entropy": 5.401014375686645,
      "epoch": 0.4543707973102786,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985487690168263,
      "loss": 5.3805,
      "mean_token_accuracy": 0.1784240871667862,
      "num_tokens": 10842953.0,
      "step": 4730
    },
    {
      "entropy": 5.578172969818115,
      "epoch": 0.4548511047070125,
      "grad_norm": 1.15625,
      "learning_rate": 0.000498544878867698,
      "loss": 5.5163,
      "mean_token_accuracy": 0.18030222356319428,
      "num_tokens": 10853322.0,
      "step": 4735
    },
    {
      "entropy": 5.483536863327027,
      "epoch": 0.4553314121037464,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985409835285215,
      "loss": 5.4091,
      "mean_token_accuracy": 0.18444757163524628,
      "num_tokens": 10865480.0,
      "step": 4740
    },
    {
      "entropy": 5.377573156356812,
      "epoch": 0.45581171950048033,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004985370829993873,
      "loss": 5.2728,
      "mean_token_accuracy": 0.1988300397992134,
      "num_tokens": 10875700.0,
      "step": 4745
    },
    {
      "entropy": 5.383729887008667,
      "epoch": 0.4562920268972142,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004985331772803857,
      "loss": 5.3748,
      "mean_token_accuracy": 0.19026633203029633,
      "num_tokens": 10887350.0,
      "step": 4750
    },
    {
      "entropy": 5.479515409469604,
      "epoch": 0.45677233429394815,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004985292663716074,
      "loss": 5.2939,
      "mean_token_accuracy": 0.1913094103336334,
      "num_tokens": 10899355.0,
      "step": 4755
    },
    {
      "entropy": 5.385722637176514,
      "epoch": 0.457252641690682,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004985253502731435,
      "loss": 5.3158,
      "mean_token_accuracy": 0.19433442801237105,
      "num_tokens": 10911521.0,
      "step": 4760
    },
    {
      "entropy": 5.362583923339844,
      "epoch": 0.45773294908741596,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004985214289850845,
      "loss": 5.3815,
      "mean_token_accuracy": 0.18801847100257874,
      "num_tokens": 10923363.0,
      "step": 4765
    },
    {
      "entropy": 5.475541305541992,
      "epoch": 0.45821325648414984,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004985175025075217,
      "loss": 5.3834,
      "mean_token_accuracy": 0.18536795675754547,
      "num_tokens": 10934965.0,
      "step": 4770
    },
    {
      "entropy": 5.416789484024048,
      "epoch": 0.4586935638808838,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004985135708405462,
      "loss": 5.3724,
      "mean_token_accuracy": 0.185246542096138,
      "num_tokens": 10946126.0,
      "step": 4775
    },
    {
      "entropy": 5.453929138183594,
      "epoch": 0.45917387127761766,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004985096339842493,
      "loss": 5.4453,
      "mean_token_accuracy": 0.18530180007219316,
      "num_tokens": 10957667.0,
      "step": 4780
    },
    {
      "entropy": 5.5232569694519045,
      "epoch": 0.4596541786743516,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004985056919387224,
      "loss": 5.5361,
      "mean_token_accuracy": 0.17576603665947915,
      "num_tokens": 10969999.0,
      "step": 4785
    },
    {
      "entropy": 5.412357807159424,
      "epoch": 0.4601344860710855,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985017447040569,
      "loss": 5.2236,
      "mean_token_accuracy": 0.19395162016153336,
      "num_tokens": 10980672.0,
      "step": 4790
    },
    {
      "entropy": 5.412126064300537,
      "epoch": 0.4606147934678194,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004984977922803447,
      "loss": 5.335,
      "mean_token_accuracy": 0.18984059989452362,
      "num_tokens": 10991471.0,
      "step": 4795
    },
    {
      "entropy": 5.407252502441406,
      "epoch": 0.4610951008645533,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004984938346676772,
      "loss": 5.4545,
      "mean_token_accuracy": 0.1849531650543213,
      "num_tokens": 11001960.0,
      "step": 4800
    },
    {
      "entropy": 5.488133907318115,
      "epoch": 0.46157540826128723,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004984898718661468,
      "loss": 5.3017,
      "mean_token_accuracy": 0.19120146483182907,
      "num_tokens": 11013274.0,
      "step": 4805
    },
    {
      "entropy": 5.404259872436524,
      "epoch": 0.4620557156580211,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004984859038758451,
      "loss": 5.3174,
      "mean_token_accuracy": 0.20086986273527146,
      "num_tokens": 11024635.0,
      "step": 4810
    },
    {
      "entropy": 5.320150423049927,
      "epoch": 0.46253602305475505,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984819306968642,
      "loss": 5.3349,
      "mean_token_accuracy": 0.18675200492143632,
      "num_tokens": 11036317.0,
      "step": 4815
    },
    {
      "entropy": 5.420429325103759,
      "epoch": 0.46301633045148893,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004984779523292966,
      "loss": 5.3447,
      "mean_token_accuracy": 0.19013421088457108,
      "num_tokens": 11047776.0,
      "step": 4820
    },
    {
      "entropy": 5.432637882232666,
      "epoch": 0.46349663784822287,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984739687732345,
      "loss": 5.3555,
      "mean_token_accuracy": 0.19237087965011596,
      "num_tokens": 11058746.0,
      "step": 4825
    },
    {
      "entropy": 5.461240530014038,
      "epoch": 0.46397694524495675,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004984699800287705,
      "loss": 5.3972,
      "mean_token_accuracy": 0.19075821787118913,
      "num_tokens": 11068892.0,
      "step": 4830
    },
    {
      "entropy": 5.464425754547119,
      "epoch": 0.4644572526416907,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498465986095997,
      "loss": 5.3943,
      "mean_token_accuracy": 0.1893223136663437,
      "num_tokens": 11080038.0,
      "step": 4835
    },
    {
      "entropy": 5.4291871070861815,
      "epoch": 0.46493756003842457,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004984619869750069,
      "loss": 5.3257,
      "mean_token_accuracy": 0.18744712322950363,
      "num_tokens": 11091261.0,
      "step": 4840
    },
    {
      "entropy": 5.323261499404907,
      "epoch": 0.4654178674351585,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000498457982665893,
      "loss": 5.2868,
      "mean_token_accuracy": 0.19903900176286698,
      "num_tokens": 11103015.0,
      "step": 4845
    },
    {
      "entropy": 5.413032341003418,
      "epoch": 0.4658981748318924,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004984539731687483,
      "loss": 5.2812,
      "mean_token_accuracy": 0.19835793524980544,
      "num_tokens": 11115075.0,
      "step": 4850
    },
    {
      "entropy": 5.392369365692138,
      "epoch": 0.4663784822286263,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984499584836659,
      "loss": 5.2502,
      "mean_token_accuracy": 0.19013367593288422,
      "num_tokens": 11125937.0,
      "step": 4855
    },
    {
      "entropy": 5.366612768173217,
      "epoch": 0.4668587896253602,
      "grad_norm": 1.109375,
      "learning_rate": 0.000498445938610739,
      "loss": 5.3437,
      "mean_token_accuracy": 0.1912291631102562,
      "num_tokens": 11136863.0,
      "step": 4860
    },
    {
      "entropy": 5.492154788970947,
      "epoch": 0.46733909702209414,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984419135500608,
      "loss": 5.4152,
      "mean_token_accuracy": 0.18139687925577164,
      "num_tokens": 11149934.0,
      "step": 4865
    },
    {
      "entropy": 5.430637550354004,
      "epoch": 0.4678194044188281,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004984378833017249,
      "loss": 5.3459,
      "mean_token_accuracy": 0.18646805584430695,
      "num_tokens": 11162208.0,
      "step": 4870
    },
    {
      "entropy": 5.417062759399414,
      "epoch": 0.46829971181556196,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004984338478658248,
      "loss": 5.3868,
      "mean_token_accuracy": 0.19303259253501892,
      "num_tokens": 11172892.0,
      "step": 4875
    },
    {
      "entropy": 5.419336986541748,
      "epoch": 0.4687800192122959,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984298072424542,
      "loss": 5.3334,
      "mean_token_accuracy": 0.19210954904556274,
      "num_tokens": 11184339.0,
      "step": 4880
    },
    {
      "entropy": 5.446358251571655,
      "epoch": 0.4692603266090298,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498425761431707,
      "loss": 5.4112,
      "mean_token_accuracy": 0.1854233607649803,
      "num_tokens": 11195681.0,
      "step": 4885
    },
    {
      "entropy": 5.5017109394073485,
      "epoch": 0.4697406340057637,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498421710433677,
      "loss": 5.3411,
      "mean_token_accuracy": 0.18262007385492324,
      "num_tokens": 11207982.0,
      "step": 4890
    },
    {
      "entropy": 5.409586191177368,
      "epoch": 0.4702209414024976,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004984176542484584,
      "loss": 5.3942,
      "mean_token_accuracy": 0.18591129779815674,
      "num_tokens": 11218901.0,
      "step": 4895
    },
    {
      "entropy": 5.383301591873169,
      "epoch": 0.47070124879923153,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004984135928761452,
      "loss": 5.372,
      "mean_token_accuracy": 0.1926772803068161,
      "num_tokens": 11230764.0,
      "step": 4900
    },
    {
      "entropy": 5.410336017608643,
      "epoch": 0.4711815561959654,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004984095263168317,
      "loss": 5.2801,
      "mean_token_accuracy": 0.19638104289770125,
      "num_tokens": 11241489.0,
      "step": 4905
    },
    {
      "entropy": 5.258735799789429,
      "epoch": 0.47166186359269935,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004984054545706124,
      "loss": 5.2629,
      "mean_token_accuracy": 0.19580549299716948,
      "num_tokens": 11254500.0,
      "step": 4910
    },
    {
      "entropy": 5.497035169601441,
      "epoch": 0.47214217098943323,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498401377637582,
      "loss": 5.3973,
      "mean_token_accuracy": 0.19251825511455536,
      "num_tokens": 11266235.0,
      "step": 4915
    },
    {
      "entropy": 5.469391107559204,
      "epoch": 0.47262247838616717,
      "grad_norm": 1.140625,
      "learning_rate": 0.000498397295517835,
      "loss": 5.3417,
      "mean_token_accuracy": 0.1986463412642479,
      "num_tokens": 11277471.0,
      "step": 4920
    },
    {
      "entropy": 5.250287437438965,
      "epoch": 0.47310278578290105,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983932082114659,
      "loss": 5.2552,
      "mean_token_accuracy": 0.18656531274318694,
      "num_tokens": 11290501.0,
      "step": 4925
    },
    {
      "entropy": 5.340760612487793,
      "epoch": 0.473583093179635,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004983891157185699,
      "loss": 5.3226,
      "mean_token_accuracy": 0.18932406455278397,
      "num_tokens": 11301933.0,
      "step": 4930
    },
    {
      "entropy": 5.490198087692261,
      "epoch": 0.47406340057636887,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004983850180392421,
      "loss": 5.3696,
      "mean_token_accuracy": 0.18665430545806885,
      "num_tokens": 11314539.0,
      "step": 4935
    },
    {
      "entropy": 5.378583288192749,
      "epoch": 0.4745437079731028,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983809151735775,
      "loss": 5.3007,
      "mean_token_accuracy": 0.19274419844150542,
      "num_tokens": 11326641.0,
      "step": 4940
    },
    {
      "entropy": 5.416019296646118,
      "epoch": 0.4750240153698367,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004983768071216713,
      "loss": 5.3924,
      "mean_token_accuracy": 0.18405500501394273,
      "num_tokens": 11337214.0,
      "step": 4945
    },
    {
      "entropy": 5.496784973144531,
      "epoch": 0.4755043227665706,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004983726938836189,
      "loss": 5.3445,
      "mean_token_accuracy": 0.18876688331365585,
      "num_tokens": 11349540.0,
      "step": 4950
    },
    {
      "entropy": 5.4341717720031735,
      "epoch": 0.4759846301633045,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004983685754595159,
      "loss": 5.3355,
      "mean_token_accuracy": 0.18595123589038848,
      "num_tokens": 11359817.0,
      "step": 4955
    },
    {
      "entropy": 5.4430341720581055,
      "epoch": 0.47646493756003844,
      "grad_norm": 1.125,
      "learning_rate": 0.0004983644518494578,
      "loss": 5.3972,
      "mean_token_accuracy": 0.186477367579937,
      "num_tokens": 11371120.0,
      "step": 4960
    },
    {
      "entropy": 5.37560772895813,
      "epoch": 0.4769452449567723,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004983603230535403,
      "loss": 5.3846,
      "mean_token_accuracy": 0.1843823730945587,
      "num_tokens": 11382071.0,
      "step": 4965
    },
    {
      "entropy": 5.486975955963135,
      "epoch": 0.47742555235350626,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004983561890718594,
      "loss": 5.3735,
      "mean_token_accuracy": 0.18451013863086702,
      "num_tokens": 11392938.0,
      "step": 4970
    },
    {
      "entropy": 5.438824224472046,
      "epoch": 0.47790585975024014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498352049904511,
      "loss": 5.383,
      "mean_token_accuracy": 0.19188303351402283,
      "num_tokens": 11405416.0,
      "step": 4975
    },
    {
      "entropy": 5.407766199111938,
      "epoch": 0.4783861671469741,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004983479055515914,
      "loss": 5.3749,
      "mean_token_accuracy": 0.1977550283074379,
      "num_tokens": 11417008.0,
      "step": 4980
    },
    {
      "entropy": 5.512362051010132,
      "epoch": 0.47886647454370795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983437560131964,
      "loss": 5.3685,
      "mean_token_accuracy": 0.19162377417087556,
      "num_tokens": 11428154.0,
      "step": 4985
    },
    {
      "entropy": 5.459709596633911,
      "epoch": 0.4793467819404419,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004983396012894228,
      "loss": 5.3601,
      "mean_token_accuracy": 0.1884807288646698,
      "num_tokens": 11438559.0,
      "step": 4990
    },
    {
      "entropy": 5.487937021255493,
      "epoch": 0.47982708933717577,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004983354413803666,
      "loss": 5.4531,
      "mean_token_accuracy": 0.1842357411980629,
      "num_tokens": 11449501.0,
      "step": 4995
    },
    {
      "entropy": 5.4470375061035154,
      "epoch": 0.4803073967339097,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004983312762861248,
      "loss": 5.4254,
      "mean_token_accuracy": 0.18518142253160477,
      "num_tokens": 11461841.0,
      "step": 5000
    },
    {
      "entropy": 5.418512010574341,
      "epoch": 0.4807877041306436,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004983271060067939,
      "loss": 5.3162,
      "mean_token_accuracy": 0.18898764699697496,
      "num_tokens": 11472876.0,
      "step": 5005
    },
    {
      "entropy": 5.425581312179565,
      "epoch": 0.4812680115273775,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983229305424707,
      "loss": 5.3921,
      "mean_token_accuracy": 0.18778188526630402,
      "num_tokens": 11484541.0,
      "step": 5010
    },
    {
      "entropy": 5.346675109863281,
      "epoch": 0.4817483189241114,
      "grad_norm": 1.0,
      "learning_rate": 0.0004983187498932522,
      "loss": 5.226,
      "mean_token_accuracy": 0.196119824051857,
      "num_tokens": 11496071.0,
      "step": 5015
    },
    {
      "entropy": 5.4374021053314205,
      "epoch": 0.48222862632084534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004983145640592354,
      "loss": 5.3134,
      "mean_token_accuracy": 0.19321988373994828,
      "num_tokens": 11507574.0,
      "step": 5020
    },
    {
      "entropy": 5.359785890579223,
      "epoch": 0.4827089337175792,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983103730405176,
      "loss": 5.3524,
      "mean_token_accuracy": 0.19146872609853743,
      "num_tokens": 11519738.0,
      "step": 5025
    },
    {
      "entropy": 5.446642827987671,
      "epoch": 0.48318924111431316,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498306176837196,
      "loss": 5.3622,
      "mean_token_accuracy": 0.1942235380411148,
      "num_tokens": 11530669.0,
      "step": 5030
    },
    {
      "entropy": 5.344743299484253,
      "epoch": 0.48366954851104704,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004983019754493681,
      "loss": 5.2577,
      "mean_token_accuracy": 0.19626683443784715,
      "num_tokens": 11541680.0,
      "step": 5035
    },
    {
      "entropy": 5.373384809494018,
      "epoch": 0.484149855907781,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004982977688771314,
      "loss": 5.3186,
      "mean_token_accuracy": 0.19459987729787825,
      "num_tokens": 11553204.0,
      "step": 5040
    },
    {
      "entropy": 5.398615169525146,
      "epoch": 0.4846301633045149,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004982935571205835,
      "loss": 5.3498,
      "mean_token_accuracy": 0.18056421130895614,
      "num_tokens": 11564841.0,
      "step": 5045
    },
    {
      "entropy": 5.423081302642823,
      "epoch": 0.4851104707012488,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982893401798223,
      "loss": 5.3459,
      "mean_token_accuracy": 0.18329853266477586,
      "num_tokens": 11576723.0,
      "step": 5050
    },
    {
      "entropy": 5.493621873855591,
      "epoch": 0.48559077809798273,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004982851180549456,
      "loss": 5.3218,
      "mean_token_accuracy": 0.1943645089864731,
      "num_tokens": 11587727.0,
      "step": 5055
    },
    {
      "entropy": 5.324885272979737,
      "epoch": 0.4860710854947166,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004982808907460515,
      "loss": 5.3091,
      "mean_token_accuracy": 0.18765472620725632,
      "num_tokens": 11598508.0,
      "step": 5060
    },
    {
      "entropy": 5.383871364593506,
      "epoch": 0.48655139289145055,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004982766582532382,
      "loss": 5.2494,
      "mean_token_accuracy": 0.2020281434059143,
      "num_tokens": 11609684.0,
      "step": 5065
    },
    {
      "entropy": 5.415210008621216,
      "epoch": 0.48703170028818443,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004982724205766038,
      "loss": 5.3835,
      "mean_token_accuracy": 0.18427674770355223,
      "num_tokens": 11621642.0,
      "step": 5070
    },
    {
      "entropy": 5.398483371734619,
      "epoch": 0.48751200768491837,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004982681777162468,
      "loss": 5.3031,
      "mean_token_accuracy": 0.19016396552324294,
      "num_tokens": 11632778.0,
      "step": 5075
    },
    {
      "entropy": 5.4051717758178714,
      "epoch": 0.48799231508165225,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004982639296722657,
      "loss": 5.3022,
      "mean_token_accuracy": 0.19790378510951995,
      "num_tokens": 11643581.0,
      "step": 5080
    },
    {
      "entropy": 5.526760387420654,
      "epoch": 0.4884726224783862,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004982596764447591,
      "loss": 5.4796,
      "mean_token_accuracy": 0.1833423614501953,
      "num_tokens": 11656126.0,
      "step": 5085
    },
    {
      "entropy": 5.4993753910064695,
      "epoch": 0.48895292987512007,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004982554180338258,
      "loss": 5.3813,
      "mean_token_accuracy": 0.19216351956129074,
      "num_tokens": 11667765.0,
      "step": 5090
    },
    {
      "entropy": 5.422559785842895,
      "epoch": 0.489433237271854,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004982511544395646,
      "loss": 5.3914,
      "mean_token_accuracy": 0.18769994080066682,
      "num_tokens": 11679268.0,
      "step": 5095
    },
    {
      "entropy": 5.398586463928223,
      "epoch": 0.4899135446685879,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982468856620745,
      "loss": 5.3082,
      "mean_token_accuracy": 0.193277807533741,
      "num_tokens": 11690049.0,
      "step": 5100
    },
    {
      "entropy": 5.420659923553467,
      "epoch": 0.4903938520653218,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982426117014545,
      "loss": 5.3036,
      "mean_token_accuracy": 0.19995406419038772,
      "num_tokens": 11701782.0,
      "step": 5105
    },
    {
      "entropy": 5.31513409614563,
      "epoch": 0.4908741594620557,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004982383325578041,
      "loss": 5.3225,
      "mean_token_accuracy": 0.1961298793554306,
      "num_tokens": 11712488.0,
      "step": 5110
    },
    {
      "entropy": 5.434733724594116,
      "epoch": 0.49135446685878964,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982340482312226,
      "loss": 5.3462,
      "mean_token_accuracy": 0.19116350561380385,
      "num_tokens": 11724036.0,
      "step": 5115
    },
    {
      "entropy": 5.3705686092376705,
      "epoch": 0.4918347742555235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004982297587218092,
      "loss": 5.3336,
      "mean_token_accuracy": 0.19269275665283203,
      "num_tokens": 11736461.0,
      "step": 5120
    },
    {
      "entropy": 5.322543287277222,
      "epoch": 0.49231508165225746,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004982254640296637,
      "loss": 5.272,
      "mean_token_accuracy": 0.1959492564201355,
      "num_tokens": 11747549.0,
      "step": 5125
    },
    {
      "entropy": 5.481142234802246,
      "epoch": 0.49279538904899134,
      "grad_norm": 1.125,
      "learning_rate": 0.0004982211641548857,
      "loss": 5.4402,
      "mean_token_accuracy": 0.18007722944021226,
      "num_tokens": 11759012.0,
      "step": 5130
    },
    {
      "entropy": 5.494666719436646,
      "epoch": 0.4932756964457253,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982168590975752,
      "loss": 5.36,
      "mean_token_accuracy": 0.1868395149707794,
      "num_tokens": 11770535.0,
      "step": 5135
    },
    {
      "entropy": 5.391727876663208,
      "epoch": 0.49375600384245916,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004982125488578321,
      "loss": 5.3915,
      "mean_token_accuracy": 0.18696723580360414,
      "num_tokens": 11783708.0,
      "step": 5140
    },
    {
      "entropy": 5.398859977722168,
      "epoch": 0.4942363112391931,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004982082334357563,
      "loss": 5.3384,
      "mean_token_accuracy": 0.18925444781780243,
      "num_tokens": 11794565.0,
      "step": 5145
    },
    {
      "entropy": 5.48035364151001,
      "epoch": 0.494716618635927,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004982039128314481,
      "loss": 5.3589,
      "mean_token_accuracy": 0.19286278039216995,
      "num_tokens": 11805631.0,
      "step": 5150
    },
    {
      "entropy": 5.4589136123657225,
      "epoch": 0.4951969260326609,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004981995870450079,
      "loss": 5.3771,
      "mean_token_accuracy": 0.18388352394104004,
      "num_tokens": 11817041.0,
      "step": 5155
    },
    {
      "entropy": 5.439020109176636,
      "epoch": 0.4956772334293948,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981952560765361,
      "loss": 5.3039,
      "mean_token_accuracy": 0.19377968311309815,
      "num_tokens": 11829090.0,
      "step": 5160
    },
    {
      "entropy": 5.489640522003174,
      "epoch": 0.49615754082612873,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004981909199261331,
      "loss": 5.4491,
      "mean_token_accuracy": 0.18131639063358307,
      "num_tokens": 11839980.0,
      "step": 5165
    },
    {
      "entropy": 5.448116111755371,
      "epoch": 0.4966378482228626,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004981865785938998,
      "loss": 5.3416,
      "mean_token_accuracy": 0.19141905605793,
      "num_tokens": 11851217.0,
      "step": 5170
    },
    {
      "entropy": 5.380829620361328,
      "epoch": 0.49711815561959655,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004981822320799367,
      "loss": 5.3673,
      "mean_token_accuracy": 0.19062818437814713,
      "num_tokens": 11862506.0,
      "step": 5175
    },
    {
      "entropy": 5.465060424804688,
      "epoch": 0.49759846301633043,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004981778803843449,
      "loss": 5.3817,
      "mean_token_accuracy": 0.18010984510183334,
      "num_tokens": 11874844.0,
      "step": 5180
    },
    {
      "entropy": 5.368805027008056,
      "epoch": 0.49807877041306436,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004981735235072256,
      "loss": 5.2962,
      "mean_token_accuracy": 0.19733415246009828,
      "num_tokens": 11885231.0,
      "step": 5185
    },
    {
      "entropy": 5.4106865406036375,
      "epoch": 0.49855907780979825,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004981691614486796,
      "loss": 5.3523,
      "mean_token_accuracy": 0.18457550406455994,
      "num_tokens": 11897222.0,
      "step": 5190
    },
    {
      "entropy": 5.4215858459472654,
      "epoch": 0.4990393852065322,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004981647942088084,
      "loss": 5.3103,
      "mean_token_accuracy": 0.19085955619812012,
      "num_tokens": 11909493.0,
      "step": 5195
    },
    {
      "entropy": 5.5145354747772215,
      "epoch": 0.49951969260326606,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004981604217877135,
      "loss": 5.3002,
      "mean_token_accuracy": 0.18946620672941208,
      "num_tokens": 11920833.0,
      "step": 5200
    },
    {
      "entropy": 5.260312604904175,
      "epoch": 0.5,
      "grad_norm": 1.125,
      "learning_rate": 0.000498156044185496,
      "loss": 5.2053,
      "mean_token_accuracy": 0.19985561072826385,
      "num_tokens": 11932910.0,
      "step": 5205
    },
    {
      "entropy": 5.403596210479736,
      "epoch": 0.5004803073967339,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004981516614022579,
      "loss": 5.3438,
      "mean_token_accuracy": 0.19838465005159378,
      "num_tokens": 11944536.0,
      "step": 5210
    },
    {
      "entropy": 5.425229215621949,
      "epoch": 0.5009606147934679,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004981472734381008,
      "loss": 5.3316,
      "mean_token_accuracy": 0.19838564842939377,
      "num_tokens": 11955724.0,
      "step": 5215
    },
    {
      "entropy": 5.381053066253662,
      "epoch": 0.5014409221902018,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004981428802931267,
      "loss": 5.3053,
      "mean_token_accuracy": 0.19322786033153533,
      "num_tokens": 11965921.0,
      "step": 5220
    },
    {
      "entropy": 5.342360305786133,
      "epoch": 0.5019212295869356,
      "grad_norm": 1.125,
      "learning_rate": 0.0004981384819674375,
      "loss": 5.2759,
      "mean_token_accuracy": 0.1968229666352272,
      "num_tokens": 11976356.0,
      "step": 5225
    },
    {
      "entropy": 5.427505350112915,
      "epoch": 0.5024015369836695,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981340784611354,
      "loss": 5.318,
      "mean_token_accuracy": 0.19113422334194183,
      "num_tokens": 11987959.0,
      "step": 5230
    },
    {
      "entropy": 5.4219732761383055,
      "epoch": 0.5028818443804035,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004981296697743224,
      "loss": 5.3388,
      "mean_token_accuracy": 0.1909692406654358,
      "num_tokens": 12000089.0,
      "step": 5235
    },
    {
      "entropy": 5.436326313018799,
      "epoch": 0.5033621517771374,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004981252559071012,
      "loss": 5.406,
      "mean_token_accuracy": 0.18826683312654496,
      "num_tokens": 12011286.0,
      "step": 5240
    },
    {
      "entropy": 5.405138444900513,
      "epoch": 0.5038424591738713,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004981208368595739,
      "loss": 5.286,
      "mean_token_accuracy": 0.1921167328953743,
      "num_tokens": 12022809.0,
      "step": 5245
    },
    {
      "entropy": 5.369298505783081,
      "epoch": 0.5043227665706052,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004981164126318435,
      "loss": 5.2619,
      "mean_token_accuracy": 0.19603474587202072,
      "num_tokens": 12034311.0,
      "step": 5250
    },
    {
      "entropy": 5.307434558868408,
      "epoch": 0.5048030739673391,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004981119832240124,
      "loss": 5.2772,
      "mean_token_accuracy": 0.19639792144298554,
      "num_tokens": 12046405.0,
      "step": 5255
    },
    {
      "entropy": 5.359806489944458,
      "epoch": 0.505283381364073,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004981075486361837,
      "loss": 5.3371,
      "mean_token_accuracy": 0.18984507620334626,
      "num_tokens": 12059022.0,
      "step": 5260
    },
    {
      "entropy": 5.461632156372071,
      "epoch": 0.5057636887608069,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004981031088684601,
      "loss": 5.3381,
      "mean_token_accuracy": 0.19137645810842513,
      "num_tokens": 12070666.0,
      "step": 5265
    },
    {
      "entropy": 5.34938530921936,
      "epoch": 0.5062439961575408,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004980986639209448,
      "loss": 5.331,
      "mean_token_accuracy": 0.19039150476455688,
      "num_tokens": 12082702.0,
      "step": 5270
    },
    {
      "entropy": 5.382027053833008,
      "epoch": 0.5067243035542748,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498094213793741,
      "loss": 5.3506,
      "mean_token_accuracy": 0.19194735437631608,
      "num_tokens": 12094034.0,
      "step": 5275
    },
    {
      "entropy": 5.4603675365447994,
      "epoch": 0.5072046109510087,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000498089758486952,
      "loss": 5.3328,
      "mean_token_accuracy": 0.18038630187511445,
      "num_tokens": 12104597.0,
      "step": 5280
    },
    {
      "entropy": 5.419860029220581,
      "epoch": 0.5076849183477425,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980852980006812,
      "loss": 5.3261,
      "mean_token_accuracy": 0.18869162499904632,
      "num_tokens": 12116577.0,
      "step": 5285
    },
    {
      "entropy": 5.338322496414184,
      "epoch": 0.5081652257444764,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004980808323350323,
      "loss": 5.2814,
      "mean_token_accuracy": 0.190215539932251,
      "num_tokens": 12128005.0,
      "step": 5290
    },
    {
      "entropy": 5.425300359725952,
      "epoch": 0.5086455331412104,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980763614901089,
      "loss": 5.3881,
      "mean_token_accuracy": 0.18722544312477113,
      "num_tokens": 12138591.0,
      "step": 5295
    },
    {
      "entropy": 5.391377878189087,
      "epoch": 0.5091258405379443,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004980718854660146,
      "loss": 5.3657,
      "mean_token_accuracy": 0.18914763033390045,
      "num_tokens": 12150010.0,
      "step": 5300
    },
    {
      "entropy": 5.437503385543823,
      "epoch": 0.5096061479346782,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004980674042628537,
      "loss": 5.3307,
      "mean_token_accuracy": 0.18339596837759017,
      "num_tokens": 12161268.0,
      "step": 5305
    },
    {
      "entropy": 5.360389041900635,
      "epoch": 0.5100864553314121,
      "grad_norm": 1.1875,
      "learning_rate": 0.00049806291788073,
      "loss": 5.2831,
      "mean_token_accuracy": 0.19422477781772612,
      "num_tokens": 12172647.0,
      "step": 5310
    },
    {
      "entropy": 5.3478131771087645,
      "epoch": 0.510566762728146,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004980584263197477,
      "loss": 5.377,
      "mean_token_accuracy": 0.1917024940252304,
      "num_tokens": 12183521.0,
      "step": 5315
    },
    {
      "entropy": 5.576809692382812,
      "epoch": 0.5110470701248799,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004980539295800111,
      "loss": 5.4053,
      "mean_token_accuracy": 0.18107457906007768,
      "num_tokens": 12193616.0,
      "step": 5320
    },
    {
      "entropy": 5.356401634216309,
      "epoch": 0.5115273775216138,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004980494276616246,
      "loss": 5.3489,
      "mean_token_accuracy": 0.18917052149772645,
      "num_tokens": 12204817.0,
      "step": 5325
    },
    {
      "entropy": 5.45305643081665,
      "epoch": 0.5120076849183477,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004980449205646926,
      "loss": 5.4062,
      "mean_token_accuracy": 0.18327575474977492,
      "num_tokens": 12217281.0,
      "step": 5330
    },
    {
      "entropy": 5.4598816394805905,
      "epoch": 0.5124879923150817,
      "grad_norm": 1.140625,
      "learning_rate": 0.00049804040828932,
      "loss": 5.3044,
      "mean_token_accuracy": 0.18908292204141616,
      "num_tokens": 12228705.0,
      "step": 5335
    },
    {
      "entropy": 5.419664621353149,
      "epoch": 0.5129682997118156,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004980358908356113,
      "loss": 5.4013,
      "mean_token_accuracy": 0.18576061427593232,
      "num_tokens": 12240010.0,
      "step": 5340
    },
    {
      "entropy": 5.485500288009644,
      "epoch": 0.5134486071085494,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004980313682036717,
      "loss": 5.4381,
      "mean_token_accuracy": 0.18056011199951172,
      "num_tokens": 12251384.0,
      "step": 5345
    },
    {
      "entropy": 5.409132957458496,
      "epoch": 0.5139289145052833,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004980268403936058,
      "loss": 5.2898,
      "mean_token_accuracy": 0.18928261548280717,
      "num_tokens": 12262548.0,
      "step": 5350
    },
    {
      "entropy": 5.3435169696807865,
      "epoch": 0.5144092219020173,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004980223074055189,
      "loss": 5.2304,
      "mean_token_accuracy": 0.20114988833665848,
      "num_tokens": 12274022.0,
      "step": 5355
    },
    {
      "entropy": 5.342687797546387,
      "epoch": 0.5148895292987512,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004980177692395164,
      "loss": 5.2322,
      "mean_token_accuracy": 0.1975323513150215,
      "num_tokens": 12285000.0,
      "step": 5360
    },
    {
      "entropy": 5.349481010437012,
      "epoch": 0.5153698366954851,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004980132258957035,
      "loss": 5.3018,
      "mean_token_accuracy": 0.19179490953683853,
      "num_tokens": 12297472.0,
      "step": 5365
    },
    {
      "entropy": 5.4828574657440186,
      "epoch": 0.515850144092219,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004980086773741856,
      "loss": 5.3512,
      "mean_token_accuracy": 0.19311970025300979,
      "num_tokens": 12309310.0,
      "step": 5370
    },
    {
      "entropy": 5.473441410064697,
      "epoch": 0.516330451488953,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004980041236750685,
      "loss": 5.4145,
      "mean_token_accuracy": 0.18272912204265596,
      "num_tokens": 12321897.0,
      "step": 5375
    },
    {
      "entropy": 5.334092903137207,
      "epoch": 0.5168107588856868,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004979995647984577,
      "loss": 5.2589,
      "mean_token_accuracy": 0.20549208521842957,
      "num_tokens": 12333484.0,
      "step": 5380
    },
    {
      "entropy": 5.388194036483765,
      "epoch": 0.5172910662824207,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004979950007444593,
      "loss": 5.3015,
      "mean_token_accuracy": 0.1908046305179596,
      "num_tokens": 12345154.0,
      "step": 5385
    },
    {
      "entropy": 5.483976411819458,
      "epoch": 0.5177713736791547,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004979904315131792,
      "loss": 5.3425,
      "mean_token_accuracy": 0.19354143887758254,
      "num_tokens": 12354934.0,
      "step": 5390
    },
    {
      "entropy": 5.4896973133087155,
      "epoch": 0.5182516810758886,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979858571047233,
      "loss": 5.5147,
      "mean_token_accuracy": 0.17918144166469574,
      "num_tokens": 12367729.0,
      "step": 5395
    },
    {
      "entropy": 5.4514892578125,
      "epoch": 0.5187319884726225,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004979812775191979,
      "loss": 5.3031,
      "mean_token_accuracy": 0.19406510442495345,
      "num_tokens": 12380366.0,
      "step": 5400
    },
    {
      "entropy": 5.379918050765991,
      "epoch": 0.5192122958693564,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004979766927567094,
      "loss": 5.3495,
      "mean_token_accuracy": 0.1916754201054573,
      "num_tokens": 12392338.0,
      "step": 5405
    },
    {
      "entropy": 5.45466833114624,
      "epoch": 0.5196926032660903,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004979721028173643,
      "loss": 5.2961,
      "mean_token_accuracy": 0.19356513321399688,
      "num_tokens": 12403280.0,
      "step": 5410
    },
    {
      "entropy": 5.399883699417114,
      "epoch": 0.5201729106628242,
      "grad_norm": 1.109375,
      "learning_rate": 0.000497967507701269,
      "loss": 5.2851,
      "mean_token_accuracy": 0.20053905695676805,
      "num_tokens": 12414279.0,
      "step": 5415
    },
    {
      "entropy": 5.331335401535034,
      "epoch": 0.5206532180595581,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004979629074085303,
      "loss": 5.2121,
      "mean_token_accuracy": 0.19684697985649108,
      "num_tokens": 12425486.0,
      "step": 5420
    },
    {
      "entropy": 5.340576839447022,
      "epoch": 0.521133525456292,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004979583019392548,
      "loss": 5.3285,
      "mean_token_accuracy": 0.18615050464868546,
      "num_tokens": 12437295.0,
      "step": 5425
    },
    {
      "entropy": 5.423246288299561,
      "epoch": 0.521613832853026,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004979536912935497,
      "loss": 5.2884,
      "mean_token_accuracy": 0.19045187085866927,
      "num_tokens": 12448684.0,
      "step": 5430
    },
    {
      "entropy": 5.315219831466675,
      "epoch": 0.5220941402497599,
      "grad_norm": 1.15625,
      "learning_rate": 0.000497949075471522,
      "loss": 5.2373,
      "mean_token_accuracy": 0.2019451752305031,
      "num_tokens": 12461209.0,
      "step": 5435
    },
    {
      "entropy": 5.387474918365479,
      "epoch": 0.5225744476464937,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004979444544732786,
      "loss": 5.3127,
      "mean_token_accuracy": 0.19375288784503936,
      "num_tokens": 12472863.0,
      "step": 5440
    },
    {
      "entropy": 5.460505723953247,
      "epoch": 0.5230547550432276,
      "grad_norm": 1.0625,
      "learning_rate": 0.000497939828298927,
      "loss": 5.3404,
      "mean_token_accuracy": 0.1916773334145546,
      "num_tokens": 12483510.0,
      "step": 5445
    },
    {
      "entropy": 5.424359321594238,
      "epoch": 0.5235350624399616,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004979351969485747,
      "loss": 5.3636,
      "mean_token_accuracy": 0.1898370161652565,
      "num_tokens": 12495532.0,
      "step": 5450
    },
    {
      "entropy": 5.264410161972046,
      "epoch": 0.5240153698366955,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979305604223291,
      "loss": 5.2657,
      "mean_token_accuracy": 0.1976442515850067,
      "num_tokens": 12508413.0,
      "step": 5455
    },
    {
      "entropy": 5.4739661693573,
      "epoch": 0.5244956772334294,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004979259187202978,
      "loss": 5.3693,
      "mean_token_accuracy": 0.1878237634897232,
      "num_tokens": 12520705.0,
      "step": 5460
    },
    {
      "entropy": 5.385482740402222,
      "epoch": 0.5249759846301633,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004979212718425887,
      "loss": 5.3163,
      "mean_token_accuracy": 0.19141486287117004,
      "num_tokens": 12531560.0,
      "step": 5465
    },
    {
      "entropy": 5.37366361618042,
      "epoch": 0.5254562920268973,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004979166197893096,
      "loss": 5.3446,
      "mean_token_accuracy": 0.1845066502690315,
      "num_tokens": 12542539.0,
      "step": 5470
    },
    {
      "entropy": 5.47672176361084,
      "epoch": 0.5259365994236311,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004979119625605683,
      "loss": 5.3676,
      "mean_token_accuracy": 0.18837804198265076,
      "num_tokens": 12555228.0,
      "step": 5475
    },
    {
      "entropy": 5.413539218902588,
      "epoch": 0.526416906820365,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004979073001564734,
      "loss": 5.2946,
      "mean_token_accuracy": 0.19506368041038513,
      "num_tokens": 12565715.0,
      "step": 5480
    },
    {
      "entropy": 5.36220235824585,
      "epoch": 0.5268972142170989,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004979026325771328,
      "loss": 5.2431,
      "mean_token_accuracy": 0.1950734779238701,
      "num_tokens": 12576896.0,
      "step": 5485
    },
    {
      "entropy": 5.430029487609863,
      "epoch": 0.5273775216138329,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004978979598226549,
      "loss": 5.298,
      "mean_token_accuracy": 0.19185535460710526,
      "num_tokens": 12588117.0,
      "step": 5490
    },
    {
      "entropy": 5.3773425102233885,
      "epoch": 0.5278578290105668,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004978932818931483,
      "loss": 5.3014,
      "mean_token_accuracy": 0.1949531376361847,
      "num_tokens": 12599987.0,
      "step": 5495
    },
    {
      "entropy": 5.376591968536377,
      "epoch": 0.5283381364073007,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004978885987887216,
      "loss": 5.3954,
      "mean_token_accuracy": 0.18627863228321076,
      "num_tokens": 12611453.0,
      "step": 5500
    },
    {
      "entropy": 5.488531637191772,
      "epoch": 0.5288184438040345,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004978839105094833,
      "loss": 5.3385,
      "mean_token_accuracy": 0.18823210150003433,
      "num_tokens": 12623068.0,
      "step": 5505
    },
    {
      "entropy": 5.422285699844361,
      "epoch": 0.5292987512007685,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004978792170555426,
      "loss": 5.3642,
      "mean_token_accuracy": 0.19297946691513063,
      "num_tokens": 12634952.0,
      "step": 5510
    },
    {
      "entropy": 5.376954555511475,
      "epoch": 0.5297790585975024,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978745184270083,
      "loss": 5.3206,
      "mean_token_accuracy": 0.18644773960113525,
      "num_tokens": 12647027.0,
      "step": 5515
    },
    {
      "entropy": 5.3268852710723875,
      "epoch": 0.5302593659942363,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004978698146239893,
      "loss": 5.2083,
      "mean_token_accuracy": 0.19966192543506622,
      "num_tokens": 12657249.0,
      "step": 5520
    },
    {
      "entropy": 5.368176126480103,
      "epoch": 0.5307396733909702,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004978651056465952,
      "loss": 5.2175,
      "mean_token_accuracy": 0.20413198322057724,
      "num_tokens": 12666714.0,
      "step": 5525
    },
    {
      "entropy": 5.348070764541626,
      "epoch": 0.5312199807877042,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000497860391494935,
      "loss": 5.3037,
      "mean_token_accuracy": 0.1925904095172882,
      "num_tokens": 12678627.0,
      "step": 5530
    },
    {
      "entropy": 5.399757194519043,
      "epoch": 0.531700288184438,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004978556721691183,
      "loss": 5.3398,
      "mean_token_accuracy": 0.18929341584444045,
      "num_tokens": 12691009.0,
      "step": 5535
    },
    {
      "entropy": 5.447514438629151,
      "epoch": 0.5321805955811719,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004978509476692547,
      "loss": 5.3182,
      "mean_token_accuracy": 0.18941847383975982,
      "num_tokens": 12702503.0,
      "step": 5540
    },
    {
      "entropy": 5.424411153793335,
      "epoch": 0.5326609029779059,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004978462179954538,
      "loss": 5.3488,
      "mean_token_accuracy": 0.18748276233673095,
      "num_tokens": 12714018.0,
      "step": 5545
    },
    {
      "entropy": 5.4570845603942875,
      "epoch": 0.5331412103746398,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004978414831478253,
      "loss": 5.4458,
      "mean_token_accuracy": 0.19093400090932847,
      "num_tokens": 12726726.0,
      "step": 5550
    },
    {
      "entropy": 5.397979068756103,
      "epoch": 0.5336215177713737,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004978367431264794,
      "loss": 5.3064,
      "mean_token_accuracy": 0.19297455698251725,
      "num_tokens": 12737790.0,
      "step": 5555
    },
    {
      "entropy": 5.341822910308838,
      "epoch": 0.5341018251681076,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004978319979315261,
      "loss": 5.1932,
      "mean_token_accuracy": 0.19440237581729888,
      "num_tokens": 12749473.0,
      "step": 5560
    },
    {
      "entropy": 5.31666841506958,
      "epoch": 0.5345821325648416,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004978272475630752,
      "loss": 5.3324,
      "mean_token_accuracy": 0.1933175280690193,
      "num_tokens": 12762220.0,
      "step": 5565
    },
    {
      "entropy": 5.233556509017944,
      "epoch": 0.5350624399615754,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004978224920212374,
      "loss": 5.1367,
      "mean_token_accuracy": 0.20168466866016388,
      "num_tokens": 12772788.0,
      "step": 5570
    },
    {
      "entropy": 5.348819255828857,
      "epoch": 0.5355427473583093,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004978177313061232,
      "loss": 5.2401,
      "mean_token_accuracy": 0.19780543446540833,
      "num_tokens": 12784552.0,
      "step": 5575
    },
    {
      "entropy": 5.375419950485229,
      "epoch": 0.5360230547550432,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004978129654178426,
      "loss": 5.2973,
      "mean_token_accuracy": 0.19902891665697098,
      "num_tokens": 12797618.0,
      "step": 5580
    },
    {
      "entropy": 5.334380292892456,
      "epoch": 0.5365033621517772,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004978081943565067,
      "loss": 5.1996,
      "mean_token_accuracy": 0.19755010455846786,
      "num_tokens": 12807789.0,
      "step": 5585
    },
    {
      "entropy": 5.376867866516113,
      "epoch": 0.5369836695485111,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004978034181222261,
      "loss": 5.3639,
      "mean_token_accuracy": 0.19576447308063508,
      "num_tokens": 12819020.0,
      "step": 5590
    },
    {
      "entropy": 5.401323890686035,
      "epoch": 0.537463976945245,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004977986367151119,
      "loss": 5.323,
      "mean_token_accuracy": 0.18985025882720946,
      "num_tokens": 12829527.0,
      "step": 5595
    },
    {
      "entropy": 5.446440935134888,
      "epoch": 0.5379442843419788,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004977938501352747,
      "loss": 5.3045,
      "mean_token_accuracy": 0.1857060894370079,
      "num_tokens": 12840821.0,
      "step": 5600
    },
    {
      "entropy": 5.387655591964721,
      "epoch": 0.5384245917387128,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004977890583828259,
      "loss": 5.2981,
      "mean_token_accuracy": 0.19810779690742492,
      "num_tokens": 12853314.0,
      "step": 5605
    },
    {
      "entropy": 5.364499950408936,
      "epoch": 0.5389048991354467,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004977842614578768,
      "loss": 5.3054,
      "mean_token_accuracy": 0.19122883379459382,
      "num_tokens": 12863507.0,
      "step": 5610
    },
    {
      "entropy": 5.418261241912842,
      "epoch": 0.5393852065321806,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004977794593605386,
      "loss": 5.3245,
      "mean_token_accuracy": 0.19565323889255523,
      "num_tokens": 12874142.0,
      "step": 5615
    },
    {
      "entropy": 5.4371335983276365,
      "epoch": 0.5398655139289145,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497774652090923,
      "loss": 5.3303,
      "mean_token_accuracy": 0.1994064599275589,
      "num_tokens": 12884548.0,
      "step": 5620
    },
    {
      "entropy": 5.4369251251220705,
      "epoch": 0.5403458213256485,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004977698396491414,
      "loss": 5.321,
      "mean_token_accuracy": 0.1944878250360489,
      "num_tokens": 12895562.0,
      "step": 5625
    },
    {
      "entropy": 5.317808818817139,
      "epoch": 0.5408261287223823,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004977650220353055,
      "loss": 5.2973,
      "mean_token_accuracy": 0.20060104578733445,
      "num_tokens": 12905308.0,
      "step": 5630
    },
    {
      "entropy": 5.379275178909301,
      "epoch": 0.5413064361191162,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004977601992495274,
      "loss": 5.332,
      "mean_token_accuracy": 0.18896620124578475,
      "num_tokens": 12916731.0,
      "step": 5635
    },
    {
      "entropy": 5.508238649368286,
      "epoch": 0.5417867435158501,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004977553712919189,
      "loss": 5.4035,
      "mean_token_accuracy": 0.18990837335586547,
      "num_tokens": 12927609.0,
      "step": 5640
    },
    {
      "entropy": 5.4735791206359865,
      "epoch": 0.5422670509125841,
      "grad_norm": 1.0,
      "learning_rate": 0.0004977505381625921,
      "loss": 5.3185,
      "mean_token_accuracy": 0.19049646854400634,
      "num_tokens": 12939445.0,
      "step": 5645
    },
    {
      "entropy": 5.34160418510437,
      "epoch": 0.542747358309318,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004977456998616593,
      "loss": 5.3106,
      "mean_token_accuracy": 0.19311475306749343,
      "num_tokens": 12951638.0,
      "step": 5650
    },
    {
      "entropy": 5.457923841476441,
      "epoch": 0.5432276657060519,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004977408563892327,
      "loss": 5.285,
      "mean_token_accuracy": 0.1956470564007759,
      "num_tokens": 12963157.0,
      "step": 5655
    },
    {
      "entropy": 5.365251064300537,
      "epoch": 0.5437079731027857,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004977360077454249,
      "loss": 5.2703,
      "mean_token_accuracy": 0.18992200046777724,
      "num_tokens": 12974187.0,
      "step": 5660
    },
    {
      "entropy": 5.396293973922729,
      "epoch": 0.5441882804995197,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004977311539303483,
      "loss": 5.3596,
      "mean_token_accuracy": 0.19024358987808226,
      "num_tokens": 12986612.0,
      "step": 5665
    },
    {
      "entropy": 5.451742506027221,
      "epoch": 0.5446685878962536,
      "grad_norm": 1.125,
      "learning_rate": 0.0004977262949441158,
      "loss": 5.3537,
      "mean_token_accuracy": 0.18752253502607347,
      "num_tokens": 12998343.0,
      "step": 5670
    },
    {
      "entropy": 5.346511554718018,
      "epoch": 0.5451488952929875,
      "grad_norm": 1.125,
      "learning_rate": 0.0004977214307868399,
      "loss": 5.3423,
      "mean_token_accuracy": 0.18622989505529403,
      "num_tokens": 13009272.0,
      "step": 5675
    },
    {
      "entropy": 5.4496993064880375,
      "epoch": 0.5456292026897214,
      "grad_norm": 1.140625,
      "learning_rate": 0.000497716561458634,
      "loss": 5.309,
      "mean_token_accuracy": 0.1973856955766678,
      "num_tokens": 13019961.0,
      "step": 5680
    },
    {
      "entropy": 5.332365226745606,
      "epoch": 0.5461095100864554,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004977116869596107,
      "loss": 5.263,
      "mean_token_accuracy": 0.2018553301692009,
      "num_tokens": 13030827.0,
      "step": 5685
    },
    {
      "entropy": 5.522433185577393,
      "epoch": 0.5465898174831892,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004977068072898834,
      "loss": 5.4959,
      "mean_token_accuracy": 0.18335341066122054,
      "num_tokens": 13043702.0,
      "step": 5690
    },
    {
      "entropy": 5.330508804321289,
      "epoch": 0.5470701248799231,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004977019224495652,
      "loss": 5.1429,
      "mean_token_accuracy": 0.20818858295679094,
      "num_tokens": 13054474.0,
      "step": 5695
    },
    {
      "entropy": 5.365789175033569,
      "epoch": 0.547550432276657,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004976970324387698,
      "loss": 5.3109,
      "mean_token_accuracy": 0.1911236196756363,
      "num_tokens": 13064920.0,
      "step": 5700
    },
    {
      "entropy": 5.4297404289245605,
      "epoch": 0.548030739673391,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004976921372576104,
      "loss": 5.4015,
      "mean_token_accuracy": 0.1894885390996933,
      "num_tokens": 13077214.0,
      "step": 5705
    },
    {
      "entropy": 5.426016092300415,
      "epoch": 0.5485110470701249,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004976872369062011,
      "loss": 5.3051,
      "mean_token_accuracy": 0.18989336490631104,
      "num_tokens": 13089472.0,
      "step": 5710
    },
    {
      "entropy": 5.3725920677185055,
      "epoch": 0.5489913544668588,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004976823313846552,
      "loss": 5.3788,
      "mean_token_accuracy": 0.1942233443260193,
      "num_tokens": 13100369.0,
      "step": 5715
    },
    {
      "entropy": 5.428851795196533,
      "epoch": 0.5494716618635928,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004976774206930869,
      "loss": 5.3343,
      "mean_token_accuracy": 0.18011238425970078,
      "num_tokens": 13110813.0,
      "step": 5720
    },
    {
      "entropy": 5.406992053985595,
      "epoch": 0.5499519692603266,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004976725048316101,
      "loss": 5.3225,
      "mean_token_accuracy": 0.1953043147921562,
      "num_tokens": 13121946.0,
      "step": 5725
    },
    {
      "entropy": 5.350114202499389,
      "epoch": 0.5504322766570605,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004976675838003388,
      "loss": 5.2526,
      "mean_token_accuracy": 0.19921243637800218,
      "num_tokens": 13133286.0,
      "step": 5730
    },
    {
      "entropy": 5.2883875370025635,
      "epoch": 0.5509125840537944,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004976626575993877,
      "loss": 5.2488,
      "mean_token_accuracy": 0.19969585686922073,
      "num_tokens": 13145023.0,
      "step": 5735
    },
    {
      "entropy": 5.378187274932861,
      "epoch": 0.5513928914505284,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004976577262288705,
      "loss": 5.2073,
      "mean_token_accuracy": 0.19632900804281234,
      "num_tokens": 13156379.0,
      "step": 5740
    },
    {
      "entropy": 5.279415082931519,
      "epoch": 0.5518731988472623,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004976527896889023,
      "loss": 5.2516,
      "mean_token_accuracy": 0.20334208607673646,
      "num_tokens": 13166313.0,
      "step": 5745
    },
    {
      "entropy": 5.396868944168091,
      "epoch": 0.5523535062439962,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004976478479795974,
      "loss": 5.3345,
      "mean_token_accuracy": 0.1866173580288887,
      "num_tokens": 13177968.0,
      "step": 5750
    },
    {
      "entropy": 5.4609315395355225,
      "epoch": 0.55283381364073,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004976429011010706,
      "loss": 5.3396,
      "mean_token_accuracy": 0.19105368852615356,
      "num_tokens": 13189464.0,
      "step": 5755
    },
    {
      "entropy": 5.379194164276123,
      "epoch": 0.553314121037464,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004976379490534366,
      "loss": 5.347,
      "mean_token_accuracy": 0.185107459127903,
      "num_tokens": 13201716.0,
      "step": 5760
    },
    {
      "entropy": 5.316136741638184,
      "epoch": 0.5537944284341979,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004976329918368107,
      "loss": 5.1813,
      "mean_token_accuracy": 0.19924976974725722,
      "num_tokens": 13212976.0,
      "step": 5765
    },
    {
      "entropy": 5.28137092590332,
      "epoch": 0.5542747358309318,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004976280294513079,
      "loss": 5.2152,
      "mean_token_accuracy": 0.2058737576007843,
      "num_tokens": 13225265.0,
      "step": 5770
    },
    {
      "entropy": 5.542584228515625,
      "epoch": 0.5547550432276657,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004976230618970431,
      "loss": 5.4889,
      "mean_token_accuracy": 0.17671897858381272,
      "num_tokens": 13236443.0,
      "step": 5775
    },
    {
      "entropy": 5.398683643341064,
      "epoch": 0.5552353506243997,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000497618089174132,
      "loss": 5.3588,
      "mean_token_accuracy": 0.19621264338493347,
      "num_tokens": 13248681.0,
      "step": 5780
    },
    {
      "entropy": 5.354566526412964,
      "epoch": 0.5557156580211335,
      "grad_norm": 1.125,
      "learning_rate": 0.0004976131112826898,
      "loss": 5.3284,
      "mean_token_accuracy": 0.19065317511558533,
      "num_tokens": 13258719.0,
      "step": 5785
    },
    {
      "entropy": 5.409298324584961,
      "epoch": 0.5561959654178674,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004976081282228323,
      "loss": 5.2901,
      "mean_token_accuracy": 0.18838194459676744,
      "num_tokens": 13270544.0,
      "step": 5790
    },
    {
      "entropy": 5.464127492904663,
      "epoch": 0.5566762728146013,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000497603139994675,
      "loss": 5.2391,
      "mean_token_accuracy": 0.1958411604166031,
      "num_tokens": 13280759.0,
      "step": 5795
    },
    {
      "entropy": 5.303540468215942,
      "epoch": 0.5571565802113353,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004975981465983338,
      "loss": 5.3385,
      "mean_token_accuracy": 0.19360581785440445,
      "num_tokens": 13292496.0,
      "step": 5800
    },
    {
      "entropy": 5.35555362701416,
      "epoch": 0.5576368876080692,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004975931480339246,
      "loss": 5.3508,
      "mean_token_accuracy": 0.18778592944145203,
      "num_tokens": 13303003.0,
      "step": 5805
    },
    {
      "entropy": 5.443258905410767,
      "epoch": 0.5581171950048031,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004975881443015635,
      "loss": 5.3204,
      "mean_token_accuracy": 0.1869622528553009,
      "num_tokens": 13315036.0,
      "step": 5810
    },
    {
      "entropy": 5.386992502212524,
      "epoch": 0.5585975024015369,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004975831354013667,
      "loss": 5.2981,
      "mean_token_accuracy": 0.19159692227840425,
      "num_tokens": 13327529.0,
      "step": 5815
    },
    {
      "entropy": 5.356541442871094,
      "epoch": 0.5590778097982709,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004975781213334503,
      "loss": 5.2877,
      "mean_token_accuracy": 0.19408222883939744,
      "num_tokens": 13339494.0,
      "step": 5820
    },
    {
      "entropy": 5.377878713607788,
      "epoch": 0.5595581171950048,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004975731020979309,
      "loss": 5.3153,
      "mean_token_accuracy": 0.18473985493183137,
      "num_tokens": 13351184.0,
      "step": 5825
    },
    {
      "entropy": 5.344663667678833,
      "epoch": 0.5600384245917387,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004975680776949249,
      "loss": 5.2138,
      "mean_token_accuracy": 0.1968445360660553,
      "num_tokens": 13362402.0,
      "step": 5830
    },
    {
      "entropy": 5.34788408279419,
      "epoch": 0.5605187319884726,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004975630481245492,
      "loss": 5.3829,
      "mean_token_accuracy": 0.18990558981895447,
      "num_tokens": 13375257.0,
      "step": 5835
    },
    {
      "entropy": 5.456510972976685,
      "epoch": 0.5609990393852066,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004975580133869202,
      "loss": 5.3346,
      "mean_token_accuracy": 0.1925740733742714,
      "num_tokens": 13385917.0,
      "step": 5840
    },
    {
      "entropy": 5.459752130508423,
      "epoch": 0.5614793467819404,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004975529734821552,
      "loss": 5.3348,
      "mean_token_accuracy": 0.18876757323741913,
      "num_tokens": 13397962.0,
      "step": 5845
    },
    {
      "entropy": 5.325477743148804,
      "epoch": 0.5619596541786743,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004975479284103708,
      "loss": 5.2102,
      "mean_token_accuracy": 0.20007061213254929,
      "num_tokens": 13408654.0,
      "step": 5850
    },
    {
      "entropy": 5.398826169967651,
      "epoch": 0.5624399615754082,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004975428781716845,
      "loss": 5.3348,
      "mean_token_accuracy": 0.19162471294403077,
      "num_tokens": 13419790.0,
      "step": 5855
    },
    {
      "entropy": 5.36678409576416,
      "epoch": 0.5629202689721422,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004975378227662134,
      "loss": 5.254,
      "mean_token_accuracy": 0.19414303749799727,
      "num_tokens": 13430305.0,
      "step": 5860
    },
    {
      "entropy": 5.374216604232788,
      "epoch": 0.5634005763688761,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004975327621940746,
      "loss": 5.3505,
      "mean_token_accuracy": 0.18956543356180192,
      "num_tokens": 13441730.0,
      "step": 5865
    },
    {
      "entropy": 5.518111515045166,
      "epoch": 0.56388088376561,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004975276964553861,
      "loss": 5.3513,
      "mean_token_accuracy": 0.19055971801280974,
      "num_tokens": 13452003.0,
      "step": 5870
    },
    {
      "entropy": 5.39121823310852,
      "epoch": 0.5643611911623438,
      "grad_norm": 1.125,
      "learning_rate": 0.0004975226255502651,
      "loss": 5.2249,
      "mean_token_accuracy": 0.19949804097414017,
      "num_tokens": 13463953.0,
      "step": 5875
    },
    {
      "entropy": 5.382539319992065,
      "epoch": 0.5648414985590778,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004975175494788297,
      "loss": 5.3672,
      "mean_token_accuracy": 0.18837829530239106,
      "num_tokens": 13473820.0,
      "step": 5880
    },
    {
      "entropy": 5.444622421264649,
      "epoch": 0.5653218059558117,
      "grad_norm": 1.125,
      "learning_rate": 0.0004975124682411974,
      "loss": 5.3837,
      "mean_token_accuracy": 0.18933936804533005,
      "num_tokens": 13484181.0,
      "step": 5885
    },
    {
      "entropy": 5.426729583740235,
      "epoch": 0.5658021133525456,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004975073818374863,
      "loss": 5.3139,
      "mean_token_accuracy": 0.18750295639038086,
      "num_tokens": 13494480.0,
      "step": 5890
    },
    {
      "entropy": 5.305769681930542,
      "epoch": 0.5662824207492796,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004975022902678145,
      "loss": 5.2678,
      "mean_token_accuracy": 0.19684486091136932,
      "num_tokens": 13506026.0,
      "step": 5895
    },
    {
      "entropy": 5.416043663024903,
      "epoch": 0.5667627281460135,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004974971935323003,
      "loss": 5.37,
      "mean_token_accuracy": 0.18816518485546113,
      "num_tokens": 13516849.0,
      "step": 5900
    },
    {
      "entropy": 5.42561354637146,
      "epoch": 0.5672430355427474,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004974920916310619,
      "loss": 5.267,
      "mean_token_accuracy": 0.19366928786039353,
      "num_tokens": 13527386.0,
      "step": 5905
    },
    {
      "entropy": 5.365526628494263,
      "epoch": 0.5677233429394812,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004974869845642178,
      "loss": 5.3841,
      "mean_token_accuracy": 0.1874728262424469,
      "num_tokens": 13540238.0,
      "step": 5910
    },
    {
      "entropy": 5.29762454032898,
      "epoch": 0.5682036503362152,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004974818723318866,
      "loss": 5.2464,
      "mean_token_accuracy": 0.20166910886764527,
      "num_tokens": 13551808.0,
      "step": 5915
    },
    {
      "entropy": 5.331619691848755,
      "epoch": 0.5686839577329491,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004974767549341868,
      "loss": 5.2827,
      "mean_token_accuracy": 0.1904905468225479,
      "num_tokens": 13563795.0,
      "step": 5920
    },
    {
      "entropy": 5.444057130813599,
      "epoch": 0.569164265129683,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004974716323712376,
      "loss": 5.2752,
      "mean_token_accuracy": 0.19699456691741943,
      "num_tokens": 13575168.0,
      "step": 5925
    },
    {
      "entropy": 5.402334833145142,
      "epoch": 0.5696445725264169,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004974665046431576,
      "loss": 5.3738,
      "mean_token_accuracy": 0.19210525304079057,
      "num_tokens": 13587013.0,
      "step": 5930
    },
    {
      "entropy": 5.38467845916748,
      "epoch": 0.5701248799231509,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004974613717500659,
      "loss": 5.2216,
      "mean_token_accuracy": 0.19611396342515947,
      "num_tokens": 13598138.0,
      "step": 5935
    },
    {
      "entropy": 5.4071392059326175,
      "epoch": 0.5706051873198847,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004974562336920818,
      "loss": 5.3223,
      "mean_token_accuracy": 0.18657563924789428,
      "num_tokens": 13609174.0,
      "step": 5940
    },
    {
      "entropy": 5.276412105560302,
      "epoch": 0.5710854947166186,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004974510904693245,
      "loss": 5.235,
      "mean_token_accuracy": 0.19945335686206817,
      "num_tokens": 13619724.0,
      "step": 5945
    },
    {
      "entropy": 5.370401430130005,
      "epoch": 0.5715658021133525,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004974459420819134,
      "loss": 5.2966,
      "mean_token_accuracy": 0.19626145213842391,
      "num_tokens": 13631114.0,
      "step": 5950
    },
    {
      "entropy": 5.352457571029663,
      "epoch": 0.5720461095100865,
      "grad_norm": 1.125,
      "learning_rate": 0.000497440788529968,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19957654625177385,
      "num_tokens": 13643035.0,
      "step": 5955
    },
    {
      "entropy": 5.406207418441772,
      "epoch": 0.5725264169068204,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004974356298136081,
      "loss": 5.3002,
      "mean_token_accuracy": 0.19240035265684127,
      "num_tokens": 13656603.0,
      "step": 5960
    },
    {
      "entropy": 5.3621937274932865,
      "epoch": 0.5730067243035543,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004974304659329533,
      "loss": 5.3349,
      "mean_token_accuracy": 0.19261169880628587,
      "num_tokens": 13668420.0,
      "step": 5965
    },
    {
      "entropy": 5.428332567214966,
      "epoch": 0.5734870317002881,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004974252968881236,
      "loss": 5.3357,
      "mean_token_accuracy": 0.18741582483053207,
      "num_tokens": 13680521.0,
      "step": 5970
    },
    {
      "entropy": 5.332199621200561,
      "epoch": 0.5739673390970221,
      "grad_norm": 1.171875,
      "learning_rate": 0.000497420122679239,
      "loss": 5.2217,
      "mean_token_accuracy": 0.19596464633941652,
      "num_tokens": 13692503.0,
      "step": 5975
    },
    {
      "entropy": 5.274319314956665,
      "epoch": 0.574447646493756,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004974149433064196,
      "loss": 5.1162,
      "mean_token_accuracy": 0.2053505852818489,
      "num_tokens": 13703318.0,
      "step": 5980
    },
    {
      "entropy": 5.401137018203736,
      "epoch": 0.5749279538904899,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004974097587697856,
      "loss": 5.2912,
      "mean_token_accuracy": 0.19216251373291016,
      "num_tokens": 13713957.0,
      "step": 5985
    },
    {
      "entropy": 5.422296524047852,
      "epoch": 0.5754082612872238,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004974045690694575,
      "loss": 5.3962,
      "mean_token_accuracy": 0.18840026259422302,
      "num_tokens": 13725300.0,
      "step": 5990
    },
    {
      "entropy": 5.405480527877808,
      "epoch": 0.5758885686839578,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004973993742055557,
      "loss": 5.3731,
      "mean_token_accuracy": 0.1879885569214821,
      "num_tokens": 13735463.0,
      "step": 5995
    },
    {
      "entropy": 5.463148498535157,
      "epoch": 0.5763688760806917,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004973941741782007,
      "loss": 5.4366,
      "mean_token_accuracy": 0.1818365767598152,
      "num_tokens": 13749062.0,
      "step": 6000
    },
    {
      "epoch": 0.5763688760806917,
      "eval_entropy": 5.28873029364312,
      "eval_loss": 5.337101936340332,
      "eval_mean_token_accuracy": 0.19782284849928625,
      "eval_num_tokens": 13749062.0,
      "eval_runtime": 26.6574,
      "eval_samples_per_second": 1230.992,
      "eval_steps_per_second": 153.879,
      "step": 6000
    },
    {
      "entropy": 5.432906007766723,
      "epoch": 0.5768491834774255,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004973889689875135,
      "loss": 5.3874,
      "mean_token_accuracy": 0.18486691117286683,
      "num_tokens": 13761766.0,
      "step": 6005
    },
    {
      "entropy": 5.379178285598755,
      "epoch": 0.5773294908741594,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004973837586336147,
      "loss": 5.3504,
      "mean_token_accuracy": 0.19981864988803863,
      "num_tokens": 13774702.0,
      "step": 6010
    },
    {
      "entropy": 5.373263025283814,
      "epoch": 0.5778097982708934,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004973785431166254,
      "loss": 5.2723,
      "mean_token_accuracy": 0.19134992957115174,
      "num_tokens": 13786215.0,
      "step": 6015
    },
    {
      "entropy": 5.299070358276367,
      "epoch": 0.5782901056676273,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004973733224366666,
      "loss": 5.2924,
      "mean_token_accuracy": 0.19437947422266005,
      "num_tokens": 13797697.0,
      "step": 6020
    },
    {
      "entropy": 5.42590913772583,
      "epoch": 0.5787704130643612,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004973680965938597,
      "loss": 5.3631,
      "mean_token_accuracy": 0.1909841701388359,
      "num_tokens": 13810085.0,
      "step": 6025
    },
    {
      "entropy": 5.435785818099975,
      "epoch": 0.579250720461095,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004973628655883258,
      "loss": 5.2611,
      "mean_token_accuracy": 0.198626346886158,
      "num_tokens": 13822054.0,
      "step": 6030
    },
    {
      "entropy": 5.315575313568115,
      "epoch": 0.579731027857829,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004973576294201865,
      "loss": 5.2304,
      "mean_token_accuracy": 0.19967078119516374,
      "num_tokens": 13834342.0,
      "step": 6035
    },
    {
      "entropy": 5.370406818389893,
      "epoch": 0.5802113352545629,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004973523880895633,
      "loss": 5.3191,
      "mean_token_accuracy": 0.18895659744739532,
      "num_tokens": 13846598.0,
      "step": 6040
    },
    {
      "entropy": 5.372568225860595,
      "epoch": 0.5806916426512968,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004973471415965779,
      "loss": 5.2484,
      "mean_token_accuracy": 0.19362230747938156,
      "num_tokens": 13857515.0,
      "step": 6045
    },
    {
      "entropy": 5.272423505783081,
      "epoch": 0.5811719500480308,
      "grad_norm": 1.109375,
      "learning_rate": 0.000497341889941352,
      "loss": 5.1887,
      "mean_token_accuracy": 0.20049561262130738,
      "num_tokens": 13868333.0,
      "step": 6050
    },
    {
      "entropy": 5.373748350143432,
      "epoch": 0.5816522574447647,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004973366331240078,
      "loss": 5.3312,
      "mean_token_accuracy": 0.1898027777671814,
      "num_tokens": 13879048.0,
      "step": 6055
    },
    {
      "entropy": 5.438288068771362,
      "epoch": 0.5821325648414986,
      "grad_norm": 1.125,
      "learning_rate": 0.000497331371144667,
      "loss": 5.4014,
      "mean_token_accuracy": 0.18395356237888336,
      "num_tokens": 13891358.0,
      "step": 6060
    },
    {
      "entropy": 5.545960187911987,
      "epoch": 0.5826128722382324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004973261040034521,
      "loss": 5.4286,
      "mean_token_accuracy": 0.18392524719238282,
      "num_tokens": 13902680.0,
      "step": 6065
    },
    {
      "entropy": 5.435374736785889,
      "epoch": 0.5830931796349664,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004973208317004852,
      "loss": 5.3248,
      "mean_token_accuracy": 0.18987623304128648,
      "num_tokens": 13913863.0,
      "step": 6070
    },
    {
      "entropy": 5.304476976394653,
      "epoch": 0.5835734870317003,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004973155542358889,
      "loss": 5.2903,
      "mean_token_accuracy": 0.19581366330385208,
      "num_tokens": 13924917.0,
      "step": 6075
    },
    {
      "entropy": 5.233635711669922,
      "epoch": 0.5840537944284342,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004973102716097853,
      "loss": 5.0835,
      "mean_token_accuracy": 0.2015526384115219,
      "num_tokens": 13936022.0,
      "step": 6080
    },
    {
      "entropy": 5.385754346847534,
      "epoch": 0.5845341018251681,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004973049838222973,
      "loss": 5.3265,
      "mean_token_accuracy": 0.19295449256896974,
      "num_tokens": 13947594.0,
      "step": 6085
    },
    {
      "entropy": 5.434674978256226,
      "epoch": 0.5850144092219021,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004972996908735479,
      "loss": 5.3065,
      "mean_token_accuracy": 0.19354501515626907,
      "num_tokens": 13960698.0,
      "step": 6090
    },
    {
      "entropy": 5.399975061416626,
      "epoch": 0.585494716618636,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004972943927636597,
      "loss": 5.3143,
      "mean_token_accuracy": 0.1904390811920166,
      "num_tokens": 13971004.0,
      "step": 6095
    },
    {
      "entropy": 5.305690145492553,
      "epoch": 0.5859750240153698,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004972890894927558,
      "loss": 5.2939,
      "mean_token_accuracy": 0.1975315123796463,
      "num_tokens": 13982500.0,
      "step": 6100
    },
    {
      "entropy": 5.517695808410645,
      "epoch": 0.5864553314121037,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004972837810609592,
      "loss": 5.455,
      "mean_token_accuracy": 0.18700396716594697,
      "num_tokens": 13994129.0,
      "step": 6105
    },
    {
      "entropy": 5.393400573730469,
      "epoch": 0.5869356388088377,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004972784674683933,
      "loss": 5.2204,
      "mean_token_accuracy": 0.1971786603331566,
      "num_tokens": 14005581.0,
      "step": 6110
    },
    {
      "entropy": 5.357818222045898,
      "epoch": 0.5874159462055716,
      "grad_norm": 1.125,
      "learning_rate": 0.0004972731487151815,
      "loss": 5.3146,
      "mean_token_accuracy": 0.18786782622337342,
      "num_tokens": 14015866.0,
      "step": 6115
    },
    {
      "entropy": 5.357286787033081,
      "epoch": 0.5878962536023055,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004972678248014471,
      "loss": 5.3024,
      "mean_token_accuracy": 0.1992377445101738,
      "num_tokens": 14027002.0,
      "step": 6120
    },
    {
      "entropy": 5.35838565826416,
      "epoch": 0.5883765609990393,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004972624957273139,
      "loss": 5.3162,
      "mean_token_accuracy": 0.1934517115354538,
      "num_tokens": 14038962.0,
      "step": 6125
    },
    {
      "entropy": 5.412038087844849,
      "epoch": 0.5888568683957733,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004972571614929055,
      "loss": 5.3318,
      "mean_token_accuracy": 0.19270756095647812,
      "num_tokens": 14049956.0,
      "step": 6130
    },
    {
      "entropy": 5.385321140289307,
      "epoch": 0.5893371757925072,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004972518220983457,
      "loss": 5.3284,
      "mean_token_accuracy": 0.18838195651769638,
      "num_tokens": 14062413.0,
      "step": 6135
    },
    {
      "entropy": 5.426869869232178,
      "epoch": 0.5898174831892411,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004972464775437586,
      "loss": 5.308,
      "mean_token_accuracy": 0.19127641767263412,
      "num_tokens": 14075166.0,
      "step": 6140
    },
    {
      "entropy": 5.3534454822540285,
      "epoch": 0.590297790585975,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004972411278292683,
      "loss": 5.2269,
      "mean_token_accuracy": 0.19504850208759308,
      "num_tokens": 14085208.0,
      "step": 6145
    },
    {
      "entropy": 5.4879467487335205,
      "epoch": 0.590778097982709,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004972357729549988,
      "loss": 5.3832,
      "mean_token_accuracy": 0.18505072146654128,
      "num_tokens": 14096688.0,
      "step": 6150
    },
    {
      "entropy": 5.2428711414337155,
      "epoch": 0.5912584053794429,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004972304129210746,
      "loss": 5.2011,
      "mean_token_accuracy": 0.20279280245304107,
      "num_tokens": 14108737.0,
      "step": 6155
    },
    {
      "entropy": 5.351060056686402,
      "epoch": 0.5917387127761767,
      "grad_norm": 1.125,
      "learning_rate": 0.0004972250477276202,
      "loss": 5.2656,
      "mean_token_accuracy": 0.2006557896733284,
      "num_tokens": 14120121.0,
      "step": 6160
    },
    {
      "entropy": 5.339959716796875,
      "epoch": 0.5922190201729106,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004972196773747599,
      "loss": 5.3246,
      "mean_token_accuracy": 0.19507967084646224,
      "num_tokens": 14132157.0,
      "step": 6165
    },
    {
      "entropy": 5.4116638660430905,
      "epoch": 0.5926993275696446,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004972143018626186,
      "loss": 5.2529,
      "mean_token_accuracy": 0.19307645857334138,
      "num_tokens": 14143092.0,
      "step": 6170
    },
    {
      "entropy": 5.3788917541503904,
      "epoch": 0.5931796349663785,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004972089211913211,
      "loss": 5.2573,
      "mean_token_accuracy": 0.19232773035764694,
      "num_tokens": 14153973.0,
      "step": 6175
    },
    {
      "entropy": 5.458052015304565,
      "epoch": 0.5936599423631124,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004972035353609923,
      "loss": 5.3343,
      "mean_token_accuracy": 0.18674774318933487,
      "num_tokens": 14164919.0,
      "step": 6180
    },
    {
      "entropy": 5.364097261428833,
      "epoch": 0.5941402497598463,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004971981443717572,
      "loss": 5.3158,
      "mean_token_accuracy": 0.19318037629127502,
      "num_tokens": 14176803.0,
      "step": 6185
    },
    {
      "entropy": 5.363823699951172,
      "epoch": 0.5946205571565802,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004971927482237409,
      "loss": 5.218,
      "mean_token_accuracy": 0.194744873046875,
      "num_tokens": 14190087.0,
      "step": 6190
    },
    {
      "entropy": 5.3881336688995365,
      "epoch": 0.5951008645533141,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004971873469170689,
      "loss": 5.4667,
      "mean_token_accuracy": 0.1798494800925255,
      "num_tokens": 14202242.0,
      "step": 6195
    },
    {
      "entropy": 5.399596929550171,
      "epoch": 0.595581171950048,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004971819404518664,
      "loss": 5.3532,
      "mean_token_accuracy": 0.18845622539520263,
      "num_tokens": 14214020.0,
      "step": 6200
    },
    {
      "entropy": 5.379790782928467,
      "epoch": 0.5960614793467819,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000497176528828259,
      "loss": 5.2514,
      "mean_token_accuracy": 0.20230887979269027,
      "num_tokens": 14225455.0,
      "step": 6205
    },
    {
      "entropy": 5.354551553726196,
      "epoch": 0.5965417867435159,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004971711120463722,
      "loss": 5.2398,
      "mean_token_accuracy": 0.19716615378856658,
      "num_tokens": 14235955.0,
      "step": 6210
    },
    {
      "entropy": 5.328782463073731,
      "epoch": 0.5970220941402498,
      "grad_norm": 1.203125,
      "learning_rate": 0.000497165690106332,
      "loss": 5.3163,
      "mean_token_accuracy": 0.1913655862212181,
      "num_tokens": 14246107.0,
      "step": 6215
    },
    {
      "entropy": 5.40400710105896,
      "epoch": 0.5975024015369836,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004971602630082642,
      "loss": 5.265,
      "mean_token_accuracy": 0.19884584099054337,
      "num_tokens": 14257042.0,
      "step": 6220
    },
    {
      "entropy": 5.449358034133911,
      "epoch": 0.5979827089337176,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004971548307522947,
      "loss": 5.4455,
      "mean_token_accuracy": 0.18984502404928208,
      "num_tokens": 14268670.0,
      "step": 6225
    },
    {
      "entropy": 5.356748294830322,
      "epoch": 0.5984630163304515,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004971493933385498,
      "loss": 5.2143,
      "mean_token_accuracy": 0.19889178425073623,
      "num_tokens": 14279538.0,
      "step": 6230
    },
    {
      "entropy": 5.384539842605591,
      "epoch": 0.5989433237271854,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004971439507671556,
      "loss": 5.3323,
      "mean_token_accuracy": 0.19492403864860536,
      "num_tokens": 14291928.0,
      "step": 6235
    },
    {
      "entropy": 5.352553987503052,
      "epoch": 0.5994236311239193,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004971385030382384,
      "loss": 5.2556,
      "mean_token_accuracy": 0.1985200271010399,
      "num_tokens": 14302958.0,
      "step": 6240
    },
    {
      "entropy": 5.374592351913452,
      "epoch": 0.5999039385206533,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004971330501519248,
      "loss": 5.2992,
      "mean_token_accuracy": 0.1964127853512764,
      "num_tokens": 14315585.0,
      "step": 6245
    },
    {
      "entropy": 5.396578741073609,
      "epoch": 0.6003842459173871,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004971275921083414,
      "loss": 5.3562,
      "mean_token_accuracy": 0.1905662462115288,
      "num_tokens": 14327228.0,
      "step": 6250
    },
    {
      "entropy": 5.469001436233521,
      "epoch": 0.600864553314121,
      "grad_norm": 1.109375,
      "learning_rate": 0.000497122128907615,
      "loss": 5.3837,
      "mean_token_accuracy": 0.18429897576570511,
      "num_tokens": 14338079.0,
      "step": 6255
    },
    {
      "entropy": 5.410619163513184,
      "epoch": 0.6013448607108549,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004971166605498722,
      "loss": 5.2525,
      "mean_token_accuracy": 0.19853390008211136,
      "num_tokens": 14350220.0,
      "step": 6260
    },
    {
      "entropy": 5.3692981719970705,
      "epoch": 0.6018251681075889,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004971111870352401,
      "loss": 5.3265,
      "mean_token_accuracy": 0.18665691316127778,
      "num_tokens": 14360761.0,
      "step": 6265
    },
    {
      "entropy": 5.387164545059204,
      "epoch": 0.6023054755043228,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004971057083638458,
      "loss": 5.314,
      "mean_token_accuracy": 0.1931974396109581,
      "num_tokens": 14372056.0,
      "step": 6270
    },
    {
      "entropy": 5.450940132141113,
      "epoch": 0.6027857829010567,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004971002245358166,
      "loss": 5.3214,
      "mean_token_accuracy": 0.19571699649095536,
      "num_tokens": 14382601.0,
      "step": 6275
    },
    {
      "entropy": 5.531852293014526,
      "epoch": 0.6032660902977905,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004970947355512795,
      "loss": 5.4301,
      "mean_token_accuracy": 0.18212406933307648,
      "num_tokens": 14393749.0,
      "step": 6280
    },
    {
      "entropy": 5.4366333961486815,
      "epoch": 0.6037463976945245,
      "grad_norm": 1.125,
      "learning_rate": 0.0004970892414103622,
      "loss": 5.3051,
      "mean_token_accuracy": 0.18854477107524872,
      "num_tokens": 14405001.0,
      "step": 6285
    },
    {
      "entropy": 5.278852748870849,
      "epoch": 0.6042267050912584,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004970837421131921,
      "loss": 5.1717,
      "mean_token_accuracy": 0.1954176038503647,
      "num_tokens": 14415875.0,
      "step": 6290
    },
    {
      "entropy": 5.333696174621582,
      "epoch": 0.6047070124879923,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004970782376598972,
      "loss": 5.2192,
      "mean_token_accuracy": 0.1950991630554199,
      "num_tokens": 14426896.0,
      "step": 6295
    },
    {
      "entropy": 5.343789434432983,
      "epoch": 0.6051873198847262,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004970727280506048,
      "loss": 5.2868,
      "mean_token_accuracy": 0.18748259097337722,
      "num_tokens": 14438712.0,
      "step": 6300
    },
    {
      "entropy": 5.367167329788208,
      "epoch": 0.6056676272814602,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004970672132854431,
      "loss": 5.2373,
      "mean_token_accuracy": 0.1957060143351555,
      "num_tokens": 14450952.0,
      "step": 6305
    },
    {
      "entropy": 5.389392614364624,
      "epoch": 0.6061479346781941,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004970616933645403,
      "loss": 5.2686,
      "mean_token_accuracy": 0.19645968675613404,
      "num_tokens": 14462139.0,
      "step": 6310
    },
    {
      "entropy": 5.307354831695557,
      "epoch": 0.6066282420749279,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004970561682880242,
      "loss": 5.3487,
      "mean_token_accuracy": 0.19032770842313768,
      "num_tokens": 14472817.0,
      "step": 6315
    },
    {
      "entropy": 5.448327732086182,
      "epoch": 0.6071085494716618,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004970506380560233,
      "loss": 5.3616,
      "mean_token_accuracy": 0.18767159134149553,
      "num_tokens": 14484241.0,
      "step": 6320
    },
    {
      "entropy": 5.326076173782349,
      "epoch": 0.6075888568683958,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004970451026686659,
      "loss": 5.2438,
      "mean_token_accuracy": 0.19500181525945665,
      "num_tokens": 14496327.0,
      "step": 6325
    },
    {
      "entropy": 5.418546390533447,
      "epoch": 0.6080691642651297,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004970395621260806,
      "loss": 5.2774,
      "mean_token_accuracy": 0.19401371479034424,
      "num_tokens": 14508041.0,
      "step": 6330
    },
    {
      "entropy": 5.399137830734253,
      "epoch": 0.6085494716618636,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000497034016428396,
      "loss": 5.2845,
      "mean_token_accuracy": 0.20141435861587526,
      "num_tokens": 14519023.0,
      "step": 6335
    },
    {
      "entropy": 5.31208758354187,
      "epoch": 0.6090297790585975,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004970284655757409,
      "loss": 5.275,
      "mean_token_accuracy": 0.1909711867570877,
      "num_tokens": 14530783.0,
      "step": 6340
    },
    {
      "entropy": 5.3341593742370605,
      "epoch": 0.6095100864553314,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004970229095682439,
      "loss": 5.2339,
      "mean_token_accuracy": 0.20182230621576308,
      "num_tokens": 14541680.0,
      "step": 6345
    },
    {
      "entropy": 5.430689191818237,
      "epoch": 0.6099903938520653,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004970173484060344,
      "loss": 5.2624,
      "mean_token_accuracy": 0.18791322112083436,
      "num_tokens": 14553792.0,
      "step": 6350
    },
    {
      "entropy": 5.3776082515716555,
      "epoch": 0.6104707012487992,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004970117820892414,
      "loss": 5.3363,
      "mean_token_accuracy": 0.18898752927780152,
      "num_tokens": 14565410.0,
      "step": 6355
    },
    {
      "entropy": 5.33832197189331,
      "epoch": 0.6109510086455331,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004970062106179939,
      "loss": 5.1854,
      "mean_token_accuracy": 0.1978567749261856,
      "num_tokens": 14576722.0,
      "step": 6360
    },
    {
      "entropy": 5.275474786758423,
      "epoch": 0.6114313160422671,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004970006339924214,
      "loss": 5.1999,
      "mean_token_accuracy": 0.1971910908818245,
      "num_tokens": 14589443.0,
      "step": 6365
    },
    {
      "entropy": 5.449461269378662,
      "epoch": 0.611911623439001,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004969950522126534,
      "loss": 5.3611,
      "mean_token_accuracy": 0.18972034901380538,
      "num_tokens": 14600731.0,
      "step": 6370
    },
    {
      "entropy": 5.35175347328186,
      "epoch": 0.6123919308357348,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004969894652788196,
      "loss": 5.2135,
      "mean_token_accuracy": 0.193570838868618,
      "num_tokens": 14611152.0,
      "step": 6375
    },
    {
      "entropy": 5.3688249588012695,
      "epoch": 0.6128722382324687,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004969838731910494,
      "loss": 5.3429,
      "mean_token_accuracy": 0.18554913401603698,
      "num_tokens": 14624526.0,
      "step": 6380
    },
    {
      "entropy": 5.410442590713501,
      "epoch": 0.6133525456292027,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004969782759494729,
      "loss": 5.2783,
      "mean_token_accuracy": 0.19764875620603561,
      "num_tokens": 14635527.0,
      "step": 6385
    },
    {
      "entropy": 5.432242918014526,
      "epoch": 0.6138328530259366,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00049697267355422,
      "loss": 5.3173,
      "mean_token_accuracy": 0.19151121228933335,
      "num_tokens": 14647372.0,
      "step": 6390
    },
    {
      "entropy": 5.435008716583252,
      "epoch": 0.6143131604226705,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004969670660054208,
      "loss": 5.3429,
      "mean_token_accuracy": 0.1899779334664345,
      "num_tokens": 14659216.0,
      "step": 6395
    },
    {
      "entropy": 5.3866535186767575,
      "epoch": 0.6147934678194045,
      "grad_norm": 1.375,
      "learning_rate": 0.0004969614533032054,
      "loss": 5.2899,
      "mean_token_accuracy": 0.1927552193403244,
      "num_tokens": 14670336.0,
      "step": 6400
    },
    {
      "entropy": 5.30097541809082,
      "epoch": 0.6152737752161384,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004969558354477041,
      "loss": 5.1445,
      "mean_token_accuracy": 0.20126848816871643,
      "num_tokens": 14680668.0,
      "step": 6405
    },
    {
      "entropy": 5.355036401748658,
      "epoch": 0.6157540826128722,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004969502124390474,
      "loss": 5.3718,
      "mean_token_accuracy": 0.19198231548070907,
      "num_tokens": 14691521.0,
      "step": 6410
    },
    {
      "entropy": 5.371433544158935,
      "epoch": 0.6162343900096061,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004969445842773658,
      "loss": 5.2355,
      "mean_token_accuracy": 0.18961513191461563,
      "num_tokens": 14702473.0,
      "step": 6415
    },
    {
      "entropy": 5.285498094558716,
      "epoch": 0.6167146974063401,
      "grad_norm": 1.171875,
      "learning_rate": 0.00049693895096279,
      "loss": 5.1942,
      "mean_token_accuracy": 0.1993377149105072,
      "num_tokens": 14713057.0,
      "step": 6420
    },
    {
      "entropy": 5.454256153106689,
      "epoch": 0.617195004803074,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004969333124954508,
      "loss": 5.3115,
      "mean_token_accuracy": 0.18584847450256348,
      "num_tokens": 14725203.0,
      "step": 6425
    },
    {
      "entropy": 5.364003372192383,
      "epoch": 0.6176753121998079,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004969276688754791,
      "loss": 5.2688,
      "mean_token_accuracy": 0.196619975566864,
      "num_tokens": 14735851.0,
      "step": 6430
    },
    {
      "entropy": 5.448974227905273,
      "epoch": 0.6181556195965417,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004969220201030059,
      "loss": 5.3531,
      "mean_token_accuracy": 0.1924108862876892,
      "num_tokens": 14747876.0,
      "step": 6435
    },
    {
      "entropy": 5.36148099899292,
      "epoch": 0.6186359269932757,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004969163661781624,
      "loss": 5.2723,
      "mean_token_accuracy": 0.19365184605121613,
      "num_tokens": 14758673.0,
      "step": 6440
    },
    {
      "entropy": 5.327347469329834,
      "epoch": 0.6191162343900096,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004969107071010798,
      "loss": 5.2161,
      "mean_token_accuracy": 0.1983474686741829,
      "num_tokens": 14770505.0,
      "step": 6445
    },
    {
      "entropy": 5.382239484786988,
      "epoch": 0.6195965417867435,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004969050428718895,
      "loss": 5.2877,
      "mean_token_accuracy": 0.19152792096138,
      "num_tokens": 14782518.0,
      "step": 6450
    },
    {
      "entropy": 5.311021184921264,
      "epoch": 0.6200768491834774,
      "grad_norm": 1.21875,
      "learning_rate": 0.000496899373490723,
      "loss": 5.2288,
      "mean_token_accuracy": 0.19708690643310547,
      "num_tokens": 14794608.0,
      "step": 6455
    },
    {
      "entropy": 5.280953884124756,
      "epoch": 0.6205571565802114,
      "grad_norm": 1.1875,
      "learning_rate": 0.000496893698957712,
      "loss": 5.1606,
      "mean_token_accuracy": 0.20110323578119277,
      "num_tokens": 14804725.0,
      "step": 6460
    },
    {
      "entropy": 5.385286521911621,
      "epoch": 0.6210374639769453,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004968880192729882,
      "loss": 5.2925,
      "mean_token_accuracy": 0.19567234069108963,
      "num_tokens": 14816178.0,
      "step": 6465
    },
    {
      "entropy": 5.387650489807129,
      "epoch": 0.6215177713736791,
      "grad_norm": 1.125,
      "learning_rate": 0.0004968823344366835,
      "loss": 5.2533,
      "mean_token_accuracy": 0.18687826544046401,
      "num_tokens": 14827808.0,
      "step": 6470
    },
    {
      "entropy": 5.298763418197632,
      "epoch": 0.621998078770413,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004968766444489298,
      "loss": 5.2177,
      "mean_token_accuracy": 0.1932251736521721,
      "num_tokens": 14839423.0,
      "step": 6475
    },
    {
      "entropy": 5.432989549636841,
      "epoch": 0.622478386167147,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004968709493098593,
      "loss": 5.36,
      "mean_token_accuracy": 0.1934700608253479,
      "num_tokens": 14849678.0,
      "step": 6480
    },
    {
      "entropy": 5.334136581420898,
      "epoch": 0.6229586935638809,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004968652490196041,
      "loss": 5.2476,
      "mean_token_accuracy": 0.19755996763706207,
      "num_tokens": 14860912.0,
      "step": 6485
    },
    {
      "entropy": 5.304739952087402,
      "epoch": 0.6234390009606148,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004968595435782967,
      "loss": 5.2405,
      "mean_token_accuracy": 0.19799728244543074,
      "num_tokens": 14873028.0,
      "step": 6490
    },
    {
      "entropy": 5.415500736236572,
      "epoch": 0.6239193083573487,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004968538329860695,
      "loss": 5.237,
      "mean_token_accuracy": 0.18860090374946595,
      "num_tokens": 14883985.0,
      "step": 6495
    },
    {
      "entropy": 5.354244995117187,
      "epoch": 0.6243996157540826,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004968481172430549,
      "loss": 5.2898,
      "mean_token_accuracy": 0.1904371902346611,
      "num_tokens": 14896569.0,
      "step": 6500
    },
    {
      "entropy": 5.336266136169433,
      "epoch": 0.6248799231508165,
      "grad_norm": 1.125,
      "learning_rate": 0.000496842396349386,
      "loss": 5.2838,
      "mean_token_accuracy": 0.20230574309825897,
      "num_tokens": 14907874.0,
      "step": 6505
    },
    {
      "entropy": 5.395018672943115,
      "epoch": 0.6253602305475504,
      "grad_norm": 1.125,
      "learning_rate": 0.0004968366703051952,
      "loss": 5.2934,
      "mean_token_accuracy": 0.19793783128261566,
      "num_tokens": 14919501.0,
      "step": 6510
    },
    {
      "entropy": 5.359877443313598,
      "epoch": 0.6258405379442843,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004968309391106157,
      "loss": 5.2604,
      "mean_token_accuracy": 0.19375524967908858,
      "num_tokens": 14929876.0,
      "step": 6515
    },
    {
      "entropy": 5.326406478881836,
      "epoch": 0.6263208453410183,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004968252027657806,
      "loss": 5.3383,
      "mean_token_accuracy": 0.18928777873516084,
      "num_tokens": 14941317.0,
      "step": 6520
    },
    {
      "entropy": 5.4428253173828125,
      "epoch": 0.6268011527377522,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004968194612708229,
      "loss": 5.3023,
      "mean_token_accuracy": 0.1928161546587944,
      "num_tokens": 14953038.0,
      "step": 6525
    },
    {
      "entropy": 5.294983148574829,
      "epoch": 0.627281460134486,
      "grad_norm": 1.25,
      "learning_rate": 0.0004968137146258759,
      "loss": 5.2343,
      "mean_token_accuracy": 0.1946828156709671,
      "num_tokens": 14964664.0,
      "step": 6530
    },
    {
      "entropy": 5.3447856426239015,
      "epoch": 0.6277617675312199,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004968079628310732,
      "loss": 5.2245,
      "mean_token_accuracy": 0.1996602311730385,
      "num_tokens": 14975761.0,
      "step": 6535
    },
    {
      "entropy": 5.375889730453491,
      "epoch": 0.6282420749279539,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004968022058865482,
      "loss": 5.1679,
      "mean_token_accuracy": 0.19762679636478425,
      "num_tokens": 14987670.0,
      "step": 6540
    },
    {
      "entropy": 5.2437666893005375,
      "epoch": 0.6287223823246878,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004967964437924346,
      "loss": 5.1296,
      "mean_token_accuracy": 0.20173428654670716,
      "num_tokens": 14999390.0,
      "step": 6545
    },
    {
      "entropy": 5.284714603424073,
      "epoch": 0.6292026897214217,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004967906765488662,
      "loss": 5.2594,
      "mean_token_accuracy": 0.19322776645421982,
      "num_tokens": 15011691.0,
      "step": 6550
    },
    {
      "entropy": 5.467965459823608,
      "epoch": 0.6296829971181557,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004967849041559769,
      "loss": 5.388,
      "mean_token_accuracy": 0.189520226418972,
      "num_tokens": 15023539.0,
      "step": 6555
    },
    {
      "entropy": 5.409853982925415,
      "epoch": 0.6301633045148896,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004967791266139006,
      "loss": 5.3433,
      "mean_token_accuracy": 0.19641104936599732,
      "num_tokens": 15035529.0,
      "step": 6560
    },
    {
      "entropy": 5.38854022026062,
      "epoch": 0.6306436119116234,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004967733439227716,
      "loss": 5.2778,
      "mean_token_accuracy": 0.1941842630505562,
      "num_tokens": 15047072.0,
      "step": 6565
    },
    {
      "entropy": 5.458043432235717,
      "epoch": 0.6311239193083573,
      "grad_norm": 1.203125,
      "learning_rate": 0.000496767556082724,
      "loss": 5.3389,
      "mean_token_accuracy": 0.1900546908378601,
      "num_tokens": 15058951.0,
      "step": 6570
    },
    {
      "entropy": 5.343481874465942,
      "epoch": 0.6316042267050913,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004967617630938924,
      "loss": 5.3234,
      "mean_token_accuracy": 0.1954561561346054,
      "num_tokens": 15070647.0,
      "step": 6575
    },
    {
      "entropy": 5.331634378433227,
      "epoch": 0.6320845341018252,
      "grad_norm": 1.203125,
      "learning_rate": 0.000496755964956411,
      "loss": 5.2437,
      "mean_token_accuracy": 0.20260208249092101,
      "num_tokens": 15082324.0,
      "step": 6580
    },
    {
      "entropy": 5.4098306655883786,
      "epoch": 0.6325648414985591,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004967501616704147,
      "loss": 5.3996,
      "mean_token_accuracy": 0.18187952488660813,
      "num_tokens": 15094641.0,
      "step": 6585
    },
    {
      "entropy": 5.495353412628174,
      "epoch": 0.633045148895293,
      "grad_norm": 1.09375,
      "learning_rate": 0.000496744353236038,
      "loss": 5.3082,
      "mean_token_accuracy": 0.1930881470441818,
      "num_tokens": 15106610.0,
      "step": 6590
    },
    {
      "entropy": 5.279156351089478,
      "epoch": 0.633525456292027,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000496738539653416,
      "loss": 5.2624,
      "mean_token_accuracy": 0.19616489559412004,
      "num_tokens": 15118463.0,
      "step": 6595
    },
    {
      "entropy": 5.33190884590149,
      "epoch": 0.6340057636887608,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004967327209226835,
      "loss": 5.2618,
      "mean_token_accuracy": 0.1938682034611702,
      "num_tokens": 15130154.0,
      "step": 6600
    },
    {
      "entropy": 5.350794792175293,
      "epoch": 0.6344860710854947,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004967268970439755,
      "loss": 5.2286,
      "mean_token_accuracy": 0.2024156391620636,
      "num_tokens": 15140854.0,
      "step": 6605
    },
    {
      "entropy": 5.323513507843018,
      "epoch": 0.6349663784822286,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004967210680174274,
      "loss": 5.2456,
      "mean_token_accuracy": 0.20193173736333847,
      "num_tokens": 15151075.0,
      "step": 6610
    },
    {
      "entropy": 5.316777849197388,
      "epoch": 0.6354466858789626,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004967152338431745,
      "loss": 5.2009,
      "mean_token_accuracy": 0.2029179885983467,
      "num_tokens": 15162091.0,
      "step": 6615
    },
    {
      "entropy": 5.4659806251525875,
      "epoch": 0.6359269932756965,
      "grad_norm": 1.0,
      "learning_rate": 0.0004967093945213523,
      "loss": 5.3548,
      "mean_token_accuracy": 0.1876670867204666,
      "num_tokens": 15175321.0,
      "step": 6620
    },
    {
      "entropy": 5.319769954681396,
      "epoch": 0.6364073006724303,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004967035500520962,
      "loss": 5.2784,
      "mean_token_accuracy": 0.1991502821445465,
      "num_tokens": 15187058.0,
      "step": 6625
    },
    {
      "entropy": 5.327629947662354,
      "epoch": 0.6368876080691642,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004966977004355421,
      "loss": 5.2387,
      "mean_token_accuracy": 0.19580726474523544,
      "num_tokens": 15198886.0,
      "step": 6630
    },
    {
      "entropy": 5.41781702041626,
      "epoch": 0.6373679154658982,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004966918456718256,
      "loss": 5.3124,
      "mean_token_accuracy": 0.19024166464805603,
      "num_tokens": 15211026.0,
      "step": 6635
    },
    {
      "entropy": 5.2639116764068605,
      "epoch": 0.6378482228626321,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004966859857610828,
      "loss": 5.2149,
      "mean_token_accuracy": 0.21087726056575776,
      "num_tokens": 15222094.0,
      "step": 6640
    },
    {
      "entropy": 5.3419280529022215,
      "epoch": 0.638328530259366,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004966801207034497,
      "loss": 5.3074,
      "mean_token_accuracy": 0.1960823729634285,
      "num_tokens": 15232464.0,
      "step": 6645
    },
    {
      "entropy": 5.402856349945068,
      "epoch": 0.6388088376560999,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004966742504990624,
      "loss": 5.2153,
      "mean_token_accuracy": 0.1991950288414955,
      "num_tokens": 15244564.0,
      "step": 6650
    },
    {
      "entropy": 5.392894601821899,
      "epoch": 0.6392891450528339,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004966683751480572,
      "loss": 5.3525,
      "mean_token_accuracy": 0.19364752173423766,
      "num_tokens": 15255822.0,
      "step": 6655
    },
    {
      "entropy": 5.35927906036377,
      "epoch": 0.6397694524495677,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004966624946505706,
      "loss": 5.2595,
      "mean_token_accuracy": 0.19099513441324234,
      "num_tokens": 15267336.0,
      "step": 6660
    },
    {
      "entropy": 5.339018154144287,
      "epoch": 0.6402497598463016,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004966566090067391,
      "loss": 5.168,
      "mean_token_accuracy": 0.20301739871501923,
      "num_tokens": 15278073.0,
      "step": 6665
    },
    {
      "entropy": 5.3801665782928465,
      "epoch": 0.6407300672430355,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004966507182166993,
      "loss": 5.3671,
      "mean_token_accuracy": 0.18923889547586442,
      "num_tokens": 15290119.0,
      "step": 6670
    },
    {
      "entropy": 5.2976888656616214,
      "epoch": 0.6412103746397695,
      "grad_norm": 1.03125,
      "learning_rate": 0.000496644822280588,
      "loss": 5.149,
      "mean_token_accuracy": 0.20653512626886367,
      "num_tokens": 15302235.0,
      "step": 6675
    },
    {
      "entropy": 5.390843057632447,
      "epoch": 0.6416906820365034,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000496638921198542,
      "loss": 5.3622,
      "mean_token_accuracy": 0.18563737869262695,
      "num_tokens": 15312924.0,
      "step": 6680
    },
    {
      "entropy": 5.259305429458618,
      "epoch": 0.6421709894332372,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004966330149706984,
      "loss": 5.1602,
      "mean_token_accuracy": 0.20792033523321152,
      "num_tokens": 15323453.0,
      "step": 6685
    },
    {
      "entropy": 5.326416826248169,
      "epoch": 0.6426512968299711,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004966271035971944,
      "loss": 5.2034,
      "mean_token_accuracy": 0.2021821066737175,
      "num_tokens": 15334907.0,
      "step": 6690
    },
    {
      "entropy": 5.330428266525269,
      "epoch": 0.6431316042267051,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004966211870781671,
      "loss": 5.2001,
      "mean_token_accuracy": 0.20158731639385224,
      "num_tokens": 15347211.0,
      "step": 6695
    },
    {
      "entropy": 5.393175458908081,
      "epoch": 0.643611911623439,
      "grad_norm": 1.140625,
      "learning_rate": 0.000496615265413754,
      "loss": 5.3765,
      "mean_token_accuracy": 0.18601736277341843,
      "num_tokens": 15360171.0,
      "step": 6700
    },
    {
      "entropy": 5.509953927993775,
      "epoch": 0.6440922190201729,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004966093386040923,
      "loss": 5.4442,
      "mean_token_accuracy": 0.18850017637014388,
      "num_tokens": 15372209.0,
      "step": 6705
    },
    {
      "entropy": 5.478467750549316,
      "epoch": 0.6445725264169068,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00049660340664932,
      "loss": 5.34,
      "mean_token_accuracy": 0.18766234666109086,
      "num_tokens": 15383753.0,
      "step": 6710
    },
    {
      "entropy": 5.406390905380249,
      "epoch": 0.6450528338136408,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004965974695495746,
      "loss": 5.2796,
      "mean_token_accuracy": 0.1964322566986084,
      "num_tokens": 15395202.0,
      "step": 6715
    },
    {
      "entropy": 5.378592443466187,
      "epoch": 0.6455331412103746,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004965915273049938,
      "loss": 5.2808,
      "mean_token_accuracy": 0.1904027968645096,
      "num_tokens": 15407121.0,
      "step": 6720
    },
    {
      "entropy": 5.294044017791748,
      "epoch": 0.6460134486071085,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004965855799157158,
      "loss": 5.2421,
      "mean_token_accuracy": 0.19476212114095687,
      "num_tokens": 15418063.0,
      "step": 6725
    },
    {
      "entropy": 5.343578624725342,
      "epoch": 0.6464937560038425,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004965796273818787,
      "loss": 5.2633,
      "mean_token_accuracy": 0.19227101057767867,
      "num_tokens": 15428830.0,
      "step": 6730
    },
    {
      "entropy": 5.308950996398925,
      "epoch": 0.6469740634005764,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004965736697036206,
      "loss": 5.195,
      "mean_token_accuracy": 0.19371612668037413,
      "num_tokens": 15441009.0,
      "step": 6735
    },
    {
      "entropy": 5.3475734233856205,
      "epoch": 0.6474543707973103,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004965677068810798,
      "loss": 5.2611,
      "mean_token_accuracy": 0.19994836747646333,
      "num_tokens": 15453008.0,
      "step": 6740
    },
    {
      "entropy": 5.4246358394622805,
      "epoch": 0.6479346781940442,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004965617389143947,
      "loss": 5.282,
      "mean_token_accuracy": 0.1944934219121933,
      "num_tokens": 15463907.0,
      "step": 6745
    },
    {
      "entropy": 5.37232518196106,
      "epoch": 0.6484149855907781,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000496555765803704,
      "loss": 5.1994,
      "mean_token_accuracy": 0.193635892868042,
      "num_tokens": 15474706.0,
      "step": 6750
    },
    {
      "entropy": 5.34351658821106,
      "epoch": 0.648895292987512,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004965497875491462,
      "loss": 5.2753,
      "mean_token_accuracy": 0.19635723829269408,
      "num_tokens": 15486355.0,
      "step": 6755
    },
    {
      "entropy": 5.387782621383667,
      "epoch": 0.6493756003842459,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004965438041508604,
      "loss": 5.2812,
      "mean_token_accuracy": 0.19190554916858674,
      "num_tokens": 15498184.0,
      "step": 6760
    },
    {
      "entropy": 5.4388158321380615,
      "epoch": 0.6498559077809798,
      "grad_norm": 1.25,
      "learning_rate": 0.0004965378156089851,
      "loss": 5.3281,
      "mean_token_accuracy": 0.19422216564416886,
      "num_tokens": 15509461.0,
      "step": 6765
    },
    {
      "entropy": 5.317912912368774,
      "epoch": 0.6503362151777138,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004965318219236597,
      "loss": 5.2599,
      "mean_token_accuracy": 0.19328572750091552,
      "num_tokens": 15520169.0,
      "step": 6770
    },
    {
      "entropy": 5.382796430587769,
      "epoch": 0.6508165225744477,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004965258230950232,
      "loss": 5.342,
      "mean_token_accuracy": 0.19026800096035004,
      "num_tokens": 15532783.0,
      "step": 6775
    },
    {
      "entropy": 5.408096408843994,
      "epoch": 0.6512968299711815,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004965198191232148,
      "loss": 5.253,
      "mean_token_accuracy": 0.18796288073062897,
      "num_tokens": 15543639.0,
      "step": 6780
    },
    {
      "entropy": 5.297939920425415,
      "epoch": 0.6517771373679154,
      "grad_norm": 1.140625,
      "learning_rate": 0.000496513810008374,
      "loss": 5.2321,
      "mean_token_accuracy": 0.2001722812652588,
      "num_tokens": 15555149.0,
      "step": 6785
    },
    {
      "entropy": 5.416798400878906,
      "epoch": 0.6522574447646494,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004965077957506403,
      "loss": 5.2907,
      "mean_token_accuracy": 0.1960344597697258,
      "num_tokens": 15566922.0,
      "step": 6790
    },
    {
      "entropy": 5.447213792800904,
      "epoch": 0.6527377521613833,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004965017763501533,
      "loss": 5.3465,
      "mean_token_accuracy": 0.19247231036424636,
      "num_tokens": 15578047.0,
      "step": 6795
    },
    {
      "entropy": 5.411528253555298,
      "epoch": 0.6532180595581172,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004964957518070529,
      "loss": 5.3574,
      "mean_token_accuracy": 0.19036312103271485,
      "num_tokens": 15590378.0,
      "step": 6800
    },
    {
      "entropy": 5.408063554763794,
      "epoch": 0.6536983669548511,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004964897221214788,
      "loss": 5.3044,
      "mean_token_accuracy": 0.19512173235416413,
      "num_tokens": 15600866.0,
      "step": 6805
    },
    {
      "entropy": 5.3664919376373295,
      "epoch": 0.654178674351585,
      "grad_norm": 1.1875,
      "learning_rate": 0.000496483687293571,
      "loss": 5.1507,
      "mean_token_accuracy": 0.19889784157276152,
      "num_tokens": 15612617.0,
      "step": 6810
    },
    {
      "entropy": 5.328865575790405,
      "epoch": 0.6546589817483189,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004964776473234696,
      "loss": 5.2509,
      "mean_token_accuracy": 0.20222360342741014,
      "num_tokens": 15624446.0,
      "step": 6815
    },
    {
      "entropy": 5.322605085372925,
      "epoch": 0.6551392891450528,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000496471602211315,
      "loss": 5.2365,
      "mean_token_accuracy": 0.1987505465745926,
      "num_tokens": 15635300.0,
      "step": 6820
    },
    {
      "entropy": 5.267195844650269,
      "epoch": 0.6556195965417867,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004964655519572475,
      "loss": 5.1461,
      "mean_token_accuracy": 0.19999029636383056,
      "num_tokens": 15646452.0,
      "step": 6825
    },
    {
      "entropy": 5.353394269943237,
      "epoch": 0.6560999039385207,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004964594965614072,
      "loss": 5.3331,
      "mean_token_accuracy": 0.18658129572868348,
      "num_tokens": 15658514.0,
      "step": 6830
    },
    {
      "entropy": 5.433322191238403,
      "epoch": 0.6565802113352546,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004964534360239353,
      "loss": 5.2464,
      "mean_token_accuracy": 0.19038560837507248,
      "num_tokens": 15670222.0,
      "step": 6835
    },
    {
      "entropy": 5.336945056915283,
      "epoch": 0.6570605187319885,
      "grad_norm": 1.359375,
      "learning_rate": 0.000496447370344972,
      "loss": 5.28,
      "mean_token_accuracy": 0.19730362743139268,
      "num_tokens": 15681627.0,
      "step": 6840
    },
    {
      "entropy": 5.369326829910278,
      "epoch": 0.6575408261287223,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004964412995246584,
      "loss": 5.2347,
      "mean_token_accuracy": 0.19277772605419158,
      "num_tokens": 15692632.0,
      "step": 6845
    },
    {
      "entropy": 5.311785697937012,
      "epoch": 0.6580211335254563,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004964352235631354,
      "loss": 5.2285,
      "mean_token_accuracy": 0.19828074127435685,
      "num_tokens": 15704736.0,
      "step": 6850
    },
    {
      "entropy": 5.362556028366089,
      "epoch": 0.6585014409221902,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000496429142460544,
      "loss": 5.3624,
      "mean_token_accuracy": 0.19175844937562941,
      "num_tokens": 15717057.0,
      "step": 6855
    },
    {
      "entropy": 5.483156824111939,
      "epoch": 0.6589817483189241,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004964230562170254,
      "loss": 5.407,
      "mean_token_accuracy": 0.185353022813797,
      "num_tokens": 15728577.0,
      "step": 6860
    },
    {
      "entropy": 5.43200044631958,
      "epoch": 0.659462055715658,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004964169648327209,
      "loss": 5.2441,
      "mean_token_accuracy": 0.1919241353869438,
      "num_tokens": 15739714.0,
      "step": 6865
    },
    {
      "entropy": 5.32646222114563,
      "epoch": 0.659942363112392,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000496410868307772,
      "loss": 5.1554,
      "mean_token_accuracy": 0.2041032612323761,
      "num_tokens": 15751339.0,
      "step": 6870
    },
    {
      "entropy": 5.442872047424316,
      "epoch": 0.6604226705091258,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004964047666423203,
      "loss": 5.4414,
      "mean_token_accuracy": 0.18352966606616974,
      "num_tokens": 15763177.0,
      "step": 6875
    },
    {
      "entropy": 5.410501432418823,
      "epoch": 0.6609029779058597,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004963986598365072,
      "loss": 5.3132,
      "mean_token_accuracy": 0.19398682415485383,
      "num_tokens": 15774691.0,
      "step": 6880
    },
    {
      "entropy": 5.423373603820801,
      "epoch": 0.6613832853025937,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004963925478904746,
      "loss": 5.3238,
      "mean_token_accuracy": 0.19534589797258378,
      "num_tokens": 15785561.0,
      "step": 6885
    },
    {
      "entropy": 5.427777433395386,
      "epoch": 0.6618635926993276,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004963864308043645,
      "loss": 5.3473,
      "mean_token_accuracy": 0.19008113145828248,
      "num_tokens": 15795687.0,
      "step": 6890
    },
    {
      "entropy": 5.322005414962769,
      "epoch": 0.6623439000960615,
      "grad_norm": 1.125,
      "learning_rate": 0.0004963803085783189,
      "loss": 5.2968,
      "mean_token_accuracy": 0.19727164804935454,
      "num_tokens": 15807191.0,
      "step": 6895
    },
    {
      "entropy": 5.386359739303589,
      "epoch": 0.6628242074927954,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004963741812124799,
      "loss": 5.2711,
      "mean_token_accuracy": 0.1963063135743141,
      "num_tokens": 15819585.0,
      "step": 6900
    },
    {
      "entropy": 5.359811210632325,
      "epoch": 0.6633045148895294,
      "grad_norm": 1.125,
      "learning_rate": 0.0004963680487069898,
      "loss": 5.3277,
      "mean_token_accuracy": 0.19686192870140076,
      "num_tokens": 15831580.0,
      "step": 6905
    },
    {
      "entropy": 5.333573579788208,
      "epoch": 0.6637848222862632,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004963619110619908,
      "loss": 5.2747,
      "mean_token_accuracy": 0.19445305466651916,
      "num_tokens": 15843848.0,
      "step": 6910
    },
    {
      "entropy": 5.479167890548706,
      "epoch": 0.6642651296829971,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004963557682776256,
      "loss": 5.3684,
      "mean_token_accuracy": 0.18923174887895583,
      "num_tokens": 15855091.0,
      "step": 6915
    },
    {
      "entropy": 5.4101043224334715,
      "epoch": 0.664745437079731,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004963496203540368,
      "loss": 5.2985,
      "mean_token_accuracy": 0.19317937940359114,
      "num_tokens": 15866676.0,
      "step": 6920
    },
    {
      "entropy": 5.293942308425903,
      "epoch": 0.665225744476465,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004963434672913671,
      "loss": 5.147,
      "mean_token_accuracy": 0.20060354620218276,
      "num_tokens": 15877901.0,
      "step": 6925
    },
    {
      "entropy": 5.347819805145264,
      "epoch": 0.6657060518731989,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004963373090897592,
      "loss": 5.2403,
      "mean_token_accuracy": 0.1864198699593544,
      "num_tokens": 15889284.0,
      "step": 6930
    },
    {
      "entropy": 5.327149057388306,
      "epoch": 0.6661863592699327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004963311457493563,
      "loss": 5.1984,
      "mean_token_accuracy": 0.20588819086551666,
      "num_tokens": 15901218.0,
      "step": 6935
    },
    {
      "entropy": 5.2322813987731935,
      "epoch": 0.6666666666666666,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004963249772703015,
      "loss": 5.1419,
      "mean_token_accuracy": 0.2050909638404846,
      "num_tokens": 15912452.0,
      "step": 6940
    },
    {
      "entropy": 5.295687818527222,
      "epoch": 0.6671469740634006,
      "grad_norm": 1.125,
      "learning_rate": 0.0004963188036527378,
      "loss": 5.2594,
      "mean_token_accuracy": 0.19967744946479798,
      "num_tokens": 15924353.0,
      "step": 6945
    },
    {
      "entropy": 5.396337509155273,
      "epoch": 0.6676272814601345,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004963126248968087,
      "loss": 5.2725,
      "mean_token_accuracy": 0.19790602028369902,
      "num_tokens": 15936295.0,
      "step": 6950
    },
    {
      "entropy": 5.379715824127198,
      "epoch": 0.6681075888568684,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004963064410026577,
      "loss": 5.2659,
      "mean_token_accuracy": 0.18990197032690048,
      "num_tokens": 15948407.0,
      "step": 6955
    },
    {
      "entropy": 5.314703035354614,
      "epoch": 0.6685878962536023,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004963002519704281,
      "loss": 5.2353,
      "mean_token_accuracy": 0.19275419861078263,
      "num_tokens": 15959975.0,
      "step": 6960
    },
    {
      "entropy": 5.391917753219604,
      "epoch": 0.6690682036503363,
      "grad_norm": 1.03125,
      "learning_rate": 0.000496294057800264,
      "loss": 5.2139,
      "mean_token_accuracy": 0.20132070779800415,
      "num_tokens": 15973084.0,
      "step": 6965
    },
    {
      "entropy": 5.308930826187134,
      "epoch": 0.6695485110470701,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004962878584923089,
      "loss": 5.1491,
      "mean_token_accuracy": 0.20476654171943665,
      "num_tokens": 15984444.0,
      "step": 6970
    },
    {
      "entropy": 5.301266193389893,
      "epoch": 0.670028818443804,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004962816540467068,
      "loss": 5.2644,
      "mean_token_accuracy": 0.19526363015174866,
      "num_tokens": 15996611.0,
      "step": 6975
    },
    {
      "entropy": 5.338674736022949,
      "epoch": 0.6705091258405379,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004962754444636017,
      "loss": 5.2394,
      "mean_token_accuracy": 0.19197880029678344,
      "num_tokens": 16007144.0,
      "step": 6980
    },
    {
      "entropy": 5.348417282104492,
      "epoch": 0.6709894332372719,
      "grad_norm": 1.1875,
      "learning_rate": 0.000496269229743138,
      "loss": 5.2473,
      "mean_token_accuracy": 0.203453728556633,
      "num_tokens": 16018878.0,
      "step": 6985
    },
    {
      "entropy": 5.393905067443848,
      "epoch": 0.6714697406340058,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004962630098854597,
      "loss": 5.2694,
      "mean_token_accuracy": 0.19071220159530639,
      "num_tokens": 16031273.0,
      "step": 6990
    },
    {
      "entropy": 5.379865980148315,
      "epoch": 0.6719500480307397,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004962567848907113,
      "loss": 5.3279,
      "mean_token_accuracy": 0.18957776874303817,
      "num_tokens": 16043101.0,
      "step": 6995
    },
    {
      "entropy": 5.292806005477905,
      "epoch": 0.6724303554274735,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004962505547590374,
      "loss": 5.1986,
      "mean_token_accuracy": 0.1949713036417961,
      "num_tokens": 16054922.0,
      "step": 7000
    },
    {
      "entropy": 5.397235536575318,
      "epoch": 0.6729106628242075,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004962443194905826,
      "loss": 5.2773,
      "mean_token_accuracy": 0.18930272608995438,
      "num_tokens": 16065461.0,
      "step": 7005
    },
    {
      "entropy": 5.362657260894776,
      "epoch": 0.6733909702209414,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004962380790854916,
      "loss": 5.3708,
      "mean_token_accuracy": 0.18719131946563722,
      "num_tokens": 16076715.0,
      "step": 7010
    },
    {
      "entropy": 5.357847881317139,
      "epoch": 0.6738712776176753,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004962318335439094,
      "loss": 5.2656,
      "mean_token_accuracy": 0.19243426769971847,
      "num_tokens": 16088937.0,
      "step": 7015
    },
    {
      "entropy": 5.359144830703736,
      "epoch": 0.6743515850144092,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004962255828659809,
      "loss": 5.2121,
      "mean_token_accuracy": 0.19907149225473403,
      "num_tokens": 16100004.0,
      "step": 7020
    },
    {
      "entropy": 5.350142574310302,
      "epoch": 0.6748318924111432,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004962193270518513,
      "loss": 5.2747,
      "mean_token_accuracy": 0.19897734969854355,
      "num_tokens": 16112435.0,
      "step": 7025
    },
    {
      "entropy": 5.408117961883545,
      "epoch": 0.675312199807877,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004962130661016659,
      "loss": 5.4017,
      "mean_token_accuracy": 0.19031336307525634,
      "num_tokens": 16124246.0,
      "step": 7030
    },
    {
      "entropy": 5.48540472984314,
      "epoch": 0.6757925072046109,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004962068000155699,
      "loss": 5.3686,
      "mean_token_accuracy": 0.1813717767596245,
      "num_tokens": 16136291.0,
      "step": 7035
    },
    {
      "entropy": 5.315890216827393,
      "epoch": 0.6762728146013448,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004962005287937088,
      "loss": 5.1985,
      "mean_token_accuracy": 0.19860848486423494,
      "num_tokens": 16148641.0,
      "step": 7040
    },
    {
      "entropy": 5.2963779926300045,
      "epoch": 0.6767531219980788,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004961942524362283,
      "loss": 5.182,
      "mean_token_accuracy": 0.2081290066242218,
      "num_tokens": 16159152.0,
      "step": 7045
    },
    {
      "entropy": 5.398535346984863,
      "epoch": 0.6772334293948127,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004961879709432741,
      "loss": 5.2921,
      "mean_token_accuracy": 0.18922872543334962,
      "num_tokens": 16170019.0,
      "step": 7050
    },
    {
      "entropy": 5.392918443679809,
      "epoch": 0.6777137367915466,
      "grad_norm": 1.0625,
      "learning_rate": 0.000496181684314992,
      "loss": 5.3019,
      "mean_token_accuracy": 0.19140758663415908,
      "num_tokens": 16181714.0,
      "step": 7055
    },
    {
      "entropy": 5.388112592697143,
      "epoch": 0.6781940441882806,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004961753925515279,
      "loss": 5.337,
      "mean_token_accuracy": 0.19100531041622162,
      "num_tokens": 16192822.0,
      "step": 7060
    },
    {
      "entropy": 5.272073221206665,
      "epoch": 0.6786743515850144,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000496169095653028,
      "loss": 5.1562,
      "mean_token_accuracy": 0.19710003435611725,
      "num_tokens": 16203660.0,
      "step": 7065
    },
    {
      "entropy": 5.312820339202881,
      "epoch": 0.6791546589817483,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961627936196384,
      "loss": 5.2391,
      "mean_token_accuracy": 0.1993875563144684,
      "num_tokens": 16215495.0,
      "step": 7070
    },
    {
      "entropy": 5.330344867706299,
      "epoch": 0.6796349663784822,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004961564864515055,
      "loss": 5.1942,
      "mean_token_accuracy": 0.20265206396579744,
      "num_tokens": 16227532.0,
      "step": 7075
    },
    {
      "entropy": 5.186900949478149,
      "epoch": 0.6801152737752162,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004961501741487757,
      "loss": 5.0614,
      "mean_token_accuracy": 0.211708764731884,
      "num_tokens": 16238979.0,
      "step": 7080
    },
    {
      "entropy": 5.220660448074341,
      "epoch": 0.6805955811719501,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004961438567115955,
      "loss": 5.1144,
      "mean_token_accuracy": 0.2064817115664482,
      "num_tokens": 16249146.0,
      "step": 7085
    },
    {
      "entropy": 5.327303171157837,
      "epoch": 0.681075888568684,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004961375341401116,
      "loss": 5.2266,
      "mean_token_accuracy": 0.19437379091978074,
      "num_tokens": 16262269.0,
      "step": 7090
    },
    {
      "entropy": 5.292921733856201,
      "epoch": 0.6815561959654178,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004961312064344708,
      "loss": 5.201,
      "mean_token_accuracy": 0.20271383374929428,
      "num_tokens": 16273877.0,
      "step": 7095
    },
    {
      "entropy": 5.432090902328492,
      "epoch": 0.6820365033621518,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00049612487359482,
      "loss": 5.4507,
      "mean_token_accuracy": 0.19284398406744002,
      "num_tokens": 16285623.0,
      "step": 7100
    },
    {
      "entropy": 5.316836929321289,
      "epoch": 0.6825168107588857,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004961185356213062,
      "loss": 5.167,
      "mean_token_accuracy": 0.20173908919095992,
      "num_tokens": 16296242.0,
      "step": 7105
    },
    {
      "entropy": 5.411134338378906,
      "epoch": 0.6829971181556196,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004961121925140767,
      "loss": 5.3382,
      "mean_token_accuracy": 0.19266834557056428,
      "num_tokens": 16308400.0,
      "step": 7110
    },
    {
      "entropy": 5.402437973022461,
      "epoch": 0.6834774255523535,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004961058442732786,
      "loss": 5.2395,
      "mean_token_accuracy": 0.19521282911300658,
      "num_tokens": 16319201.0,
      "step": 7115
    },
    {
      "entropy": 5.3299806118011475,
      "epoch": 0.6839577329490875,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004960994908990594,
      "loss": 5.2823,
      "mean_token_accuracy": 0.1932195782661438,
      "num_tokens": 16331462.0,
      "step": 7120
    },
    {
      "entropy": 5.354860544204712,
      "epoch": 0.6844380403458213,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004960931323915665,
      "loss": 5.2824,
      "mean_token_accuracy": 0.19656197130680084,
      "num_tokens": 16343428.0,
      "step": 7125
    },
    {
      "entropy": 5.348445177078247,
      "epoch": 0.6849183477425552,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004960867687509475,
      "loss": 5.1814,
      "mean_token_accuracy": 0.19947498589754104,
      "num_tokens": 16356270.0,
      "step": 7130
    },
    {
      "entropy": 5.283010482788086,
      "epoch": 0.6853986551392891,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004960803999773504,
      "loss": 5.2257,
      "mean_token_accuracy": 0.19890724122524261,
      "num_tokens": 16367039.0,
      "step": 7135
    },
    {
      "entropy": 5.259706640243531,
      "epoch": 0.6858789625360231,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004960740260709228,
      "loss": 5.2877,
      "mean_token_accuracy": 0.1950128495693207,
      "num_tokens": 16378601.0,
      "step": 7140
    },
    {
      "entropy": 5.429637050628662,
      "epoch": 0.686359269932757,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004960676470318128,
      "loss": 5.3105,
      "mean_token_accuracy": 0.1963237076997757,
      "num_tokens": 16391096.0,
      "step": 7145
    },
    {
      "entropy": 5.357524394989014,
      "epoch": 0.6868395773294909,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004960612628601683,
      "loss": 5.1478,
      "mean_token_accuracy": 0.2030516669154167,
      "num_tokens": 16401527.0,
      "step": 7150
    },
    {
      "entropy": 5.340597486495971,
      "epoch": 0.6873198847262247,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004960548735561379,
      "loss": 5.3001,
      "mean_token_accuracy": 0.18366778790950775,
      "num_tokens": 16413948.0,
      "step": 7155
    },
    {
      "entropy": 5.371850776672363,
      "epoch": 0.6878001921229587,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004960484791198697,
      "loss": 5.1962,
      "mean_token_accuracy": 0.20123261511325835,
      "num_tokens": 16425718.0,
      "step": 7160
    },
    {
      "entropy": 5.292998313903809,
      "epoch": 0.6882804995196926,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004960420795515121,
      "loss": 5.2326,
      "mean_token_accuracy": 0.1958795204758644,
      "num_tokens": 16437606.0,
      "step": 7165
    },
    {
      "entropy": 5.364447832107544,
      "epoch": 0.6887608069164265,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004960356748512138,
      "loss": 5.2633,
      "mean_token_accuracy": 0.19923620373010636,
      "num_tokens": 16448964.0,
      "step": 7170
    },
    {
      "entropy": 5.326768207550049,
      "epoch": 0.6892411143131604,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004960292650191236,
      "loss": 5.26,
      "mean_token_accuracy": 0.19667254239320756,
      "num_tokens": 16459974.0,
      "step": 7175
    },
    {
      "entropy": 5.3668365478515625,
      "epoch": 0.6897214217098944,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004960228500553899,
      "loss": 5.2112,
      "mean_token_accuracy": 0.19811732769012452,
      "num_tokens": 16471143.0,
      "step": 7180
    },
    {
      "entropy": 5.33847975730896,
      "epoch": 0.6902017291066282,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004960164299601623,
      "loss": 5.224,
      "mean_token_accuracy": 0.2031083270907402,
      "num_tokens": 16482090.0,
      "step": 7185
    },
    {
      "entropy": 5.2586156845092775,
      "epoch": 0.6906820365033621,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004960100047335892,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19742683172225953,
      "num_tokens": 16493695.0,
      "step": 7190
    },
    {
      "entropy": 5.3455345153808596,
      "epoch": 0.691162343900096,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004960035743758202,
      "loss": 5.3085,
      "mean_token_accuracy": 0.19393242746591569,
      "num_tokens": 16505317.0,
      "step": 7195
    },
    {
      "entropy": 5.342055892944336,
      "epoch": 0.69164265129683,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004959971388870044,
      "loss": 5.2247,
      "mean_token_accuracy": 0.19544664472341539,
      "num_tokens": 16516719.0,
      "step": 7200
    },
    {
      "entropy": 5.327449655532837,
      "epoch": 0.6921229586935639,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004959906982672912,
      "loss": 5.1725,
      "mean_token_accuracy": 0.20105513781309128,
      "num_tokens": 16528302.0,
      "step": 7205
    },
    {
      "entropy": 5.3697254180908205,
      "epoch": 0.6926032660902978,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004959842525168302,
      "loss": 5.2074,
      "mean_token_accuracy": 0.19357937276363374,
      "num_tokens": 16538471.0,
      "step": 7210
    },
    {
      "entropy": 5.371806049346924,
      "epoch": 0.6930835734870316,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004959778016357712,
      "loss": 5.2173,
      "mean_token_accuracy": 0.2014424204826355,
      "num_tokens": 16549283.0,
      "step": 7215
    },
    {
      "entropy": 5.266766023635864,
      "epoch": 0.6935638808837656,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004959713456242637,
      "loss": 5.1941,
      "mean_token_accuracy": 0.1935098111629486,
      "num_tokens": 16560895.0,
      "step": 7220
    },
    {
      "entropy": 5.34767713546753,
      "epoch": 0.6940441882804995,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004959648844824576,
      "loss": 5.2859,
      "mean_token_accuracy": 0.19510417133569719,
      "num_tokens": 16571306.0,
      "step": 7225
    },
    {
      "entropy": 5.353601598739624,
      "epoch": 0.6945244956772334,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004959584182105032,
      "loss": 5.2121,
      "mean_token_accuracy": 0.1987803727388382,
      "num_tokens": 16582399.0,
      "step": 7230
    },
    {
      "entropy": 5.284029293060303,
      "epoch": 0.6950048030739674,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004959519468085504,
      "loss": 5.178,
      "mean_token_accuracy": 0.19841667264699936,
      "num_tokens": 16593617.0,
      "step": 7235
    },
    {
      "entropy": 5.262728929519653,
      "epoch": 0.6954851104707013,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004959454702767494,
      "loss": 5.2439,
      "mean_token_accuracy": 0.19724708050489426,
      "num_tokens": 16604109.0,
      "step": 7240
    },
    {
      "entropy": 5.418563985824585,
      "epoch": 0.6959654178674352,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004959389886152507,
      "loss": 5.3341,
      "mean_token_accuracy": 0.19159986227750778,
      "num_tokens": 16616007.0,
      "step": 7245
    },
    {
      "entropy": 5.3621745109558105,
      "epoch": 0.696445725264169,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004959325018242048,
      "loss": 5.1938,
      "mean_token_accuracy": 0.19810741096735002,
      "num_tokens": 16627609.0,
      "step": 7250
    },
    {
      "entropy": 5.341673946380615,
      "epoch": 0.696926032660903,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004959260099037622,
      "loss": 5.1821,
      "mean_token_accuracy": 0.20069833397865294,
      "num_tokens": 16639842.0,
      "step": 7255
    },
    {
      "entropy": 5.351086235046386,
      "epoch": 0.6974063400576369,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004959195128540737,
      "loss": 5.3314,
      "mean_token_accuracy": 0.18700163513422013,
      "num_tokens": 16651620.0,
      "step": 7260
    },
    {
      "entropy": 5.36456618309021,
      "epoch": 0.6978866474543708,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00049591301067529,
      "loss": 5.2616,
      "mean_token_accuracy": 0.19532680958509446,
      "num_tokens": 16663637.0,
      "step": 7265
    },
    {
      "entropy": 5.399697017669678,
      "epoch": 0.6983669548511047,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004959065033675623,
      "loss": 5.3128,
      "mean_token_accuracy": 0.18797222077846526,
      "num_tokens": 16675685.0,
      "step": 7270
    },
    {
      "entropy": 5.361882829666138,
      "epoch": 0.6988472622478387,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004958999909310414,
      "loss": 5.2287,
      "mean_token_accuracy": 0.18990648239851,
      "num_tokens": 16686268.0,
      "step": 7275
    },
    {
      "entropy": 5.26370759010315,
      "epoch": 0.6993275696445725,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004958934733658788,
      "loss": 5.2589,
      "mean_token_accuracy": 0.19714254140853882,
      "num_tokens": 16698382.0,
      "step": 7280
    },
    {
      "entropy": 5.461414051055908,
      "epoch": 0.6998078770413064,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004958869506722256,
      "loss": 5.3292,
      "mean_token_accuracy": 0.18784838318824768,
      "num_tokens": 16710421.0,
      "step": 7285
    },
    {
      "entropy": 5.341016960144043,
      "epoch": 0.7002881844380403,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004958804228502332,
      "loss": 5.1729,
      "mean_token_accuracy": 0.19566827416419982,
      "num_tokens": 16721658.0,
      "step": 7290
    },
    {
      "entropy": 5.273524188995362,
      "epoch": 0.7007684918347743,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004958738899000534,
      "loss": 5.1663,
      "mean_token_accuracy": 0.20114559680223465,
      "num_tokens": 16734100.0,
      "step": 7295
    },
    {
      "entropy": 5.314267253875732,
      "epoch": 0.7012487992315082,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004958673518218377,
      "loss": 5.2191,
      "mean_token_accuracy": 0.19652467370033264,
      "num_tokens": 16746386.0,
      "step": 7300
    },
    {
      "entropy": 5.409583663940429,
      "epoch": 0.7017291066282421,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004958608086157379,
      "loss": 5.2783,
      "mean_token_accuracy": 0.19342856258153915,
      "num_tokens": 16756967.0,
      "step": 7305
    },
    {
      "entropy": 5.250471496582032,
      "epoch": 0.7022094140249759,
      "grad_norm": 1.234375,
      "learning_rate": 0.000495854260281906,
      "loss": 5.1656,
      "mean_token_accuracy": 0.1986488074064255,
      "num_tokens": 16767454.0,
      "step": 7310
    },
    {
      "entropy": 5.270944738388062,
      "epoch": 0.7026897214217099,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004958477068204941,
      "loss": 5.2215,
      "mean_token_accuracy": 0.20177811831235887,
      "num_tokens": 16779533.0,
      "step": 7315
    },
    {
      "entropy": 5.3583348274230955,
      "epoch": 0.7031700288184438,
      "grad_norm": 1.34375,
      "learning_rate": 0.000495841148231654,
      "loss": 5.2559,
      "mean_token_accuracy": 0.1946343943476677,
      "num_tokens": 16792019.0,
      "step": 7320
    },
    {
      "entropy": 5.283969402313232,
      "epoch": 0.7036503362151777,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004958345845155383,
      "loss": 5.1938,
      "mean_token_accuracy": 0.19848118722438812,
      "num_tokens": 16803089.0,
      "step": 7325
    },
    {
      "entropy": 5.314704513549804,
      "epoch": 0.7041306436119116,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004958280156722992,
      "loss": 5.1562,
      "mean_token_accuracy": 0.19987927973270417,
      "num_tokens": 16813651.0,
      "step": 7330
    },
    {
      "entropy": 5.3096870422363285,
      "epoch": 0.7046109510086456,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004958214417020894,
      "loss": 5.1868,
      "mean_token_accuracy": 0.20127215087413788,
      "num_tokens": 16825460.0,
      "step": 7335
    },
    {
      "entropy": 5.279446840286255,
      "epoch": 0.7050912584053795,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004958148626050614,
      "loss": 5.1809,
      "mean_token_accuracy": 0.19999829083681106,
      "num_tokens": 16836344.0,
      "step": 7340
    },
    {
      "entropy": 5.347747135162353,
      "epoch": 0.7055715658021133,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000495808278381368,
      "loss": 5.2215,
      "mean_token_accuracy": 0.19129775315523148,
      "num_tokens": 16848274.0,
      "step": 7345
    },
    {
      "entropy": 5.343861198425293,
      "epoch": 0.7060518731988472,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000495801689031162,
      "loss": 5.2443,
      "mean_token_accuracy": 0.19473153352737427,
      "num_tokens": 16860363.0,
      "step": 7350
    },
    {
      "entropy": 5.339507293701172,
      "epoch": 0.7065321805955812,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004957950945545965,
      "loss": 5.2011,
      "mean_token_accuracy": 0.19773450940847398,
      "num_tokens": 16872037.0,
      "step": 7355
    },
    {
      "entropy": 5.40102596282959,
      "epoch": 0.7070124879923151,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004957884949518246,
      "loss": 5.3039,
      "mean_token_accuracy": 0.18888331800699235,
      "num_tokens": 16883845.0,
      "step": 7360
    },
    {
      "entropy": 5.403661298751831,
      "epoch": 0.707492795389049,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004957818902229992,
      "loss": 5.268,
      "mean_token_accuracy": 0.1890197679400444,
      "num_tokens": 16894965.0,
      "step": 7365
    },
    {
      "entropy": 5.327176189422607,
      "epoch": 0.7079731027857828,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004957752803682741,
      "loss": 5.2615,
      "mean_token_accuracy": 0.19063124507665635,
      "num_tokens": 16907475.0,
      "step": 7370
    },
    {
      "entropy": 5.305124950408936,
      "epoch": 0.7084534101825168,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004957686653878024,
      "loss": 5.147,
      "mean_token_accuracy": 0.20095228403806686,
      "num_tokens": 16918825.0,
      "step": 7375
    },
    {
      "entropy": 5.37538423538208,
      "epoch": 0.7089337175792507,
      "grad_norm": 1.203125,
      "learning_rate": 0.000495762045281738,
      "loss": 5.1997,
      "mean_token_accuracy": 0.1956855207681656,
      "num_tokens": 16930663.0,
      "step": 7380
    },
    {
      "entropy": 5.342241430282593,
      "epoch": 0.7094140249759846,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004957554200502344,
      "loss": 5.232,
      "mean_token_accuracy": 0.19982221871614456,
      "num_tokens": 16942089.0,
      "step": 7385
    },
    {
      "entropy": 5.255606842041016,
      "epoch": 0.7098943323727186,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004957487896934454,
      "loss": 5.2691,
      "mean_token_accuracy": 0.1964074969291687,
      "num_tokens": 16954203.0,
      "step": 7390
    },
    {
      "entropy": 5.346683406829834,
      "epoch": 0.7103746397694525,
      "grad_norm": 1.21875,
      "learning_rate": 0.000495742154211525,
      "loss": 5.2147,
      "mean_token_accuracy": 0.19777590781450272,
      "num_tokens": 16966047.0,
      "step": 7395
    },
    {
      "entropy": 5.278224611282349,
      "epoch": 0.7108549471661864,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004957355136046272,
      "loss": 5.1218,
      "mean_token_accuracy": 0.20880722999572754,
      "num_tokens": 16977726.0,
      "step": 7400
    },
    {
      "entropy": 5.296135282516479,
      "epoch": 0.7113352545629202,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004957288678729064,
      "loss": 5.195,
      "mean_token_accuracy": 0.20421857982873917,
      "num_tokens": 16990917.0,
      "step": 7405
    },
    {
      "entropy": 5.297642230987549,
      "epoch": 0.7118155619596542,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004957222170165166,
      "loss": 5.219,
      "mean_token_accuracy": 0.20040077418088914,
      "num_tokens": 17001703.0,
      "step": 7410
    },
    {
      "entropy": 5.316972875595093,
      "epoch": 0.7122958693563881,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004957155610356124,
      "loss": 5.1098,
      "mean_token_accuracy": 0.20611813366413118,
      "num_tokens": 17011918.0,
      "step": 7415
    },
    {
      "entropy": 5.408256387710571,
      "epoch": 0.712776176753122,
      "grad_norm": 1.375,
      "learning_rate": 0.0004957088999303481,
      "loss": 5.2644,
      "mean_token_accuracy": 0.19444481879472733,
      "num_tokens": 17022653.0,
      "step": 7420
    },
    {
      "entropy": 5.269344520568848,
      "epoch": 0.7132564841498559,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004957022337008787,
      "loss": 5.1293,
      "mean_token_accuracy": 0.20770153105258943,
      "num_tokens": 17034775.0,
      "step": 7425
    },
    {
      "entropy": 5.292084741592407,
      "epoch": 0.7137367915465899,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004956955623473587,
      "loss": 5.183,
      "mean_token_accuracy": 0.19573891758918763,
      "num_tokens": 17046835.0,
      "step": 7430
    },
    {
      "entropy": 5.417061853408813,
      "epoch": 0.7142170989433237,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000495688885869943,
      "loss": 5.2836,
      "mean_token_accuracy": 0.19740686118602752,
      "num_tokens": 17059163.0,
      "step": 7435
    },
    {
      "entropy": 5.388308763504028,
      "epoch": 0.7146974063400576,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004956822042687868,
      "loss": 5.2888,
      "mean_token_accuracy": 0.19233759194612504,
      "num_tokens": 17070777.0,
      "step": 7440
    },
    {
      "entropy": 5.3594683647155765,
      "epoch": 0.7151777137367915,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004956755175440451,
      "loss": 5.2487,
      "mean_token_accuracy": 0.19894787967205046,
      "num_tokens": 17083770.0,
      "step": 7445
    },
    {
      "entropy": 5.305829000473023,
      "epoch": 0.7156580211335255,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004956688256958731,
      "loss": 5.2559,
      "mean_token_accuracy": 0.19449299275875093,
      "num_tokens": 17096049.0,
      "step": 7450
    },
    {
      "entropy": 5.240206718444824,
      "epoch": 0.7161383285302594,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004956621287244262,
      "loss": 5.1125,
      "mean_token_accuracy": 0.20599310249090194,
      "num_tokens": 17107030.0,
      "step": 7455
    },
    {
      "entropy": 5.26452603340149,
      "epoch": 0.7166186359269933,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004956554266298599,
      "loss": 5.1767,
      "mean_token_accuracy": 0.19949893951416015,
      "num_tokens": 17118638.0,
      "step": 7460
    },
    {
      "entropy": 5.383656024932861,
      "epoch": 0.7170989433237271,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004956487194123298,
      "loss": 5.2607,
      "mean_token_accuracy": 0.20003731399774552,
      "num_tokens": 17129783.0,
      "step": 7465
    },
    {
      "entropy": 5.402425765991211,
      "epoch": 0.7175792507204611,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004956420070719918,
      "loss": 5.2545,
      "mean_token_accuracy": 0.19425942599773408,
      "num_tokens": 17141939.0,
      "step": 7470
    },
    {
      "entropy": 5.311451816558838,
      "epoch": 0.718059558117195,
      "grad_norm": 1.125,
      "learning_rate": 0.0004956352896090014,
      "loss": 5.2055,
      "mean_token_accuracy": 0.19686287343502046,
      "num_tokens": 17153471.0,
      "step": 7475
    },
    {
      "entropy": 5.391935014724732,
      "epoch": 0.7185398655139289,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004956285670235147,
      "loss": 5.2813,
      "mean_token_accuracy": 0.18988336622714996,
      "num_tokens": 17164659.0,
      "step": 7480
    },
    {
      "entropy": 5.360458946228027,
      "epoch": 0.7190201729106628,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004956218393156879,
      "loss": 5.2016,
      "mean_token_accuracy": 0.20056852102279663,
      "num_tokens": 17175749.0,
      "step": 7485
    },
    {
      "entropy": 5.311680746078491,
      "epoch": 0.7195004803073968,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004956151064856772,
      "loss": 5.2763,
      "mean_token_accuracy": 0.19060041904449462,
      "num_tokens": 17187482.0,
      "step": 7490
    },
    {
      "entropy": 5.332025766372681,
      "epoch": 0.7199807877041307,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004956083685336386,
      "loss": 5.2233,
      "mean_token_accuracy": 0.1963557183742523,
      "num_tokens": 17199904.0,
      "step": 7495
    },
    {
      "entropy": 5.342145347595215,
      "epoch": 0.7204610951008645,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004956016254597289,
      "loss": 5.1696,
      "mean_token_accuracy": 0.19562121331691742,
      "num_tokens": 17210649.0,
      "step": 7500
    },
    {
      "entropy": 5.367743062973022,
      "epoch": 0.7209414024975984,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004955948772641044,
      "loss": 5.2475,
      "mean_token_accuracy": 0.19091322869062424,
      "num_tokens": 17223069.0,
      "step": 7505
    },
    {
      "entropy": 5.329480791091919,
      "epoch": 0.7214217098943324,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000495588123946922,
      "loss": 5.188,
      "mean_token_accuracy": 0.2017177537083626,
      "num_tokens": 17235144.0,
      "step": 7510
    },
    {
      "entropy": 5.268783140182495,
      "epoch": 0.7219020172910663,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004955813655083384,
      "loss": 5.1576,
      "mean_token_accuracy": 0.19690847843885423,
      "num_tokens": 17246816.0,
      "step": 7515
    },
    {
      "entropy": 5.267178010940552,
      "epoch": 0.7223823246878002,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004955746019485104,
      "loss": 5.2267,
      "mean_token_accuracy": 0.20049119889736175,
      "num_tokens": 17259056.0,
      "step": 7520
    },
    {
      "entropy": 5.429098844528198,
      "epoch": 0.722862632084534,
      "grad_norm": 1.140625,
      "learning_rate": 0.000495567833267595,
      "loss": 5.2558,
      "mean_token_accuracy": 0.19504401981830596,
      "num_tokens": 17270391.0,
      "step": 7525
    },
    {
      "entropy": 5.385547494888305,
      "epoch": 0.723342939481268,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004955610594657496,
      "loss": 5.174,
      "mean_token_accuracy": 0.19319403022527695,
      "num_tokens": 17281608.0,
      "step": 7530
    },
    {
      "entropy": 5.3089500904083256,
      "epoch": 0.7238232468780019,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004955542805431313,
      "loss": 5.295,
      "mean_token_accuracy": 0.19207192063331605,
      "num_tokens": 17293649.0,
      "step": 7535
    },
    {
      "entropy": 5.340746021270752,
      "epoch": 0.7243035542747358,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004955474964998976,
      "loss": 5.254,
      "mean_token_accuracy": 0.19785054177045822,
      "num_tokens": 17305548.0,
      "step": 7540
    },
    {
      "entropy": 5.40021243095398,
      "epoch": 0.7247838616714697,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004955407073362058,
      "loss": 5.2161,
      "mean_token_accuracy": 0.19872122257947922,
      "num_tokens": 17317380.0,
      "step": 7545
    },
    {
      "entropy": 5.271648931503296,
      "epoch": 0.7252641690682037,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004955339130522136,
      "loss": 5.1946,
      "mean_token_accuracy": 0.19641194343566895,
      "num_tokens": 17329263.0,
      "step": 7550
    },
    {
      "entropy": 5.372245836257934,
      "epoch": 0.7257444764649376,
      "grad_norm": 1.203125,
      "learning_rate": 0.000495527113648079,
      "loss": 5.3551,
      "mean_token_accuracy": 0.18608965426683427,
      "num_tokens": 17343268.0,
      "step": 7555
    },
    {
      "entropy": 5.358549547195435,
      "epoch": 0.7262247838616714,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004955203091239596,
      "loss": 5.1811,
      "mean_token_accuracy": 0.20911150574684143,
      "num_tokens": 17353979.0,
      "step": 7560
    },
    {
      "entropy": 5.325914001464843,
      "epoch": 0.7267050912584054,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004955134994800134,
      "loss": 5.2656,
      "mean_token_accuracy": 0.19534502774477006,
      "num_tokens": 17367587.0,
      "step": 7565
    },
    {
      "entropy": 5.348534297943115,
      "epoch": 0.7271853986551393,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004955066847163984,
      "loss": 5.2593,
      "mean_token_accuracy": 0.19680333137512207,
      "num_tokens": 17379093.0,
      "step": 7570
    },
    {
      "entropy": 5.3767317771911625,
      "epoch": 0.7276657060518732,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004954998648332731,
      "loss": 5.2326,
      "mean_token_accuracy": 0.19793325662612915,
      "num_tokens": 17391254.0,
      "step": 7575
    },
    {
      "entropy": 5.268054533004761,
      "epoch": 0.7281460134486071,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004954930398307956,
      "loss": 5.205,
      "mean_token_accuracy": 0.19531721919775008,
      "num_tokens": 17402532.0,
      "step": 7580
    },
    {
      "entropy": 5.361209297180176,
      "epoch": 0.7286263208453411,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004954862097091245,
      "loss": 5.2548,
      "mean_token_accuracy": 0.2014056921005249,
      "num_tokens": 17413849.0,
      "step": 7585
    },
    {
      "entropy": 5.248220777511596,
      "epoch": 0.729106628242075,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004954793744684184,
      "loss": 5.145,
      "mean_token_accuracy": 0.20903123468160628,
      "num_tokens": 17424982.0,
      "step": 7590
    },
    {
      "entropy": 5.41027979850769,
      "epoch": 0.7295869356388088,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004954725341088358,
      "loss": 5.3738,
      "mean_token_accuracy": 0.18464324921369552,
      "num_tokens": 17437774.0,
      "step": 7595
    },
    {
      "entropy": 5.345804929733276,
      "epoch": 0.7300672430355427,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004954656886305356,
      "loss": 5.237,
      "mean_token_accuracy": 0.20515668243169785,
      "num_tokens": 17448676.0,
      "step": 7600
    },
    {
      "entropy": 5.398682117462158,
      "epoch": 0.7305475504322767,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004954588380336768,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19920091927051545,
      "num_tokens": 17459576.0,
      "step": 7605
    },
    {
      "entropy": 5.255887222290039,
      "epoch": 0.7310278578290106,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004954519823184184,
      "loss": 5.1351,
      "mean_token_accuracy": 0.20705800354480744,
      "num_tokens": 17471220.0,
      "step": 7610
    },
    {
      "entropy": 5.313342332839966,
      "epoch": 0.7315081652257445,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004954451214849196,
      "loss": 5.2786,
      "mean_token_accuracy": 0.18875118643045424,
      "num_tokens": 17483939.0,
      "step": 7615
    },
    {
      "entropy": 5.34489803314209,
      "epoch": 0.7319884726224783,
      "grad_norm": 1.125,
      "learning_rate": 0.0004954382555333397,
      "loss": 5.2374,
      "mean_token_accuracy": 0.2002933979034424,
      "num_tokens": 17496058.0,
      "step": 7620
    },
    {
      "entropy": 5.342188978195191,
      "epoch": 0.7324687800192123,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000495431384463838,
      "loss": 5.2773,
      "mean_token_accuracy": 0.19607143700122834,
      "num_tokens": 17507750.0,
      "step": 7625
    },
    {
      "entropy": 5.411129331588745,
      "epoch": 0.7329490874159462,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004954245082765741,
      "loss": 5.2076,
      "mean_token_accuracy": 0.20007972717285155,
      "num_tokens": 17520208.0,
      "step": 7630
    },
    {
      "entropy": 5.262160348892212,
      "epoch": 0.7334293948126801,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004954176269717077,
      "loss": 5.1287,
      "mean_token_accuracy": 0.20108612030744552,
      "num_tokens": 17531655.0,
      "step": 7635
    },
    {
      "entropy": 5.235440397262574,
      "epoch": 0.733909702209414,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004954107405493984,
      "loss": 5.1932,
      "mean_token_accuracy": 0.19531163871288298,
      "num_tokens": 17543766.0,
      "step": 7640
    },
    {
      "entropy": 5.413145065307617,
      "epoch": 0.734390009606148,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004954038490098064,
      "loss": 5.2705,
      "mean_token_accuracy": 0.19128247052431108,
      "num_tokens": 17555710.0,
      "step": 7645
    },
    {
      "entropy": 5.288508081436158,
      "epoch": 0.7348703170028819,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004953969523530914,
      "loss": 5.135,
      "mean_token_accuracy": 0.20539151281118392,
      "num_tokens": 17566674.0,
      "step": 7650
    },
    {
      "entropy": 5.34483232498169,
      "epoch": 0.7353506243996157,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004953900505794136,
      "loss": 5.2402,
      "mean_token_accuracy": 0.19354115873575212,
      "num_tokens": 17579320.0,
      "step": 7655
    },
    {
      "entropy": 5.276857471466064,
      "epoch": 0.7358309317963496,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004953831436889332,
      "loss": 5.1941,
      "mean_token_accuracy": 0.1925780177116394,
      "num_tokens": 17590679.0,
      "step": 7660
    },
    {
      "entropy": 5.370760297775268,
      "epoch": 0.7363112391930836,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004953762316818106,
      "loss": 5.167,
      "mean_token_accuracy": 0.20880268663167953,
      "num_tokens": 17600928.0,
      "step": 7665
    },
    {
      "entropy": 5.18484320640564,
      "epoch": 0.7367915465898175,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004953693145582064,
      "loss": 5.1384,
      "mean_token_accuracy": 0.20616303533315658,
      "num_tokens": 17611906.0,
      "step": 7670
    },
    {
      "entropy": 5.25790548324585,
      "epoch": 0.7372718539865514,
      "grad_norm": 1.09375,
      "learning_rate": 0.000495362392318281,
      "loss": 5.3002,
      "mean_token_accuracy": 0.19541302174329758,
      "num_tokens": 17623165.0,
      "step": 7675
    },
    {
      "entropy": 5.427854824066162,
      "epoch": 0.7377521613832853,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004953554649621951,
      "loss": 5.1839,
      "mean_token_accuracy": 0.20174630135297775,
      "num_tokens": 17633632.0,
      "step": 7680
    },
    {
      "entropy": 5.282119560241699,
      "epoch": 0.7382324687800192,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004953485324901098,
      "loss": 5.2085,
      "mean_token_accuracy": 0.20252581536769867,
      "num_tokens": 17644801.0,
      "step": 7685
    },
    {
      "entropy": 5.267820739746094,
      "epoch": 0.7387127761767531,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004953415949021857,
      "loss": 5.2161,
      "mean_token_accuracy": 0.19662141650915146,
      "num_tokens": 17656852.0,
      "step": 7690
    },
    {
      "entropy": 5.416881322860718,
      "epoch": 0.739193083573487,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004953346521985843,
      "loss": 5.2221,
      "mean_token_accuracy": 0.1946612551808357,
      "num_tokens": 17667538.0,
      "step": 7695
    },
    {
      "entropy": 5.318049335479737,
      "epoch": 0.7396733909702209,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004953277043794663,
      "loss": 5.1425,
      "mean_token_accuracy": 0.20722710937261582,
      "num_tokens": 17680049.0,
      "step": 7700
    },
    {
      "entropy": 5.194083118438721,
      "epoch": 0.7401536983669549,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004953207514449933,
      "loss": 5.2015,
      "mean_token_accuracy": 0.19716615080833436,
      "num_tokens": 17692611.0,
      "step": 7705
    },
    {
      "entropy": 5.2996772766113285,
      "epoch": 0.7406340057636888,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004953137933953267,
      "loss": 5.1177,
      "mean_token_accuracy": 0.19896381348371506,
      "num_tokens": 17704145.0,
      "step": 7710
    },
    {
      "entropy": 5.351549863815308,
      "epoch": 0.7411143131604226,
      "grad_norm": 1.234375,
      "learning_rate": 0.000495306830230628,
      "loss": 5.2198,
      "mean_token_accuracy": 0.1914952352643013,
      "num_tokens": 17715467.0,
      "step": 7715
    },
    {
      "entropy": 5.332698917388916,
      "epoch": 0.7415946205571565,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004952998619510589,
      "loss": 5.2145,
      "mean_token_accuracy": 0.19876985400915145,
      "num_tokens": 17726709.0,
      "step": 7720
    },
    {
      "entropy": 5.281695175170898,
      "epoch": 0.7420749279538905,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004952928885567811,
      "loss": 5.147,
      "mean_token_accuracy": 0.2050749897956848,
      "num_tokens": 17736896.0,
      "step": 7725
    },
    {
      "entropy": 5.339973306655883,
      "epoch": 0.7425552353506244,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004952859100479566,
      "loss": 5.2185,
      "mean_token_accuracy": 0.19460239559412001,
      "num_tokens": 17747658.0,
      "step": 7730
    },
    {
      "entropy": 5.313399171829223,
      "epoch": 0.7430355427473583,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004952789264247474,
      "loss": 5.1909,
      "mean_token_accuracy": 0.2034624010324478,
      "num_tokens": 17757833.0,
      "step": 7735
    },
    {
      "entropy": 5.2951913356781,
      "epoch": 0.7435158501440923,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004952719376873156,
      "loss": 5.163,
      "mean_token_accuracy": 0.1979545459151268,
      "num_tokens": 17769144.0,
      "step": 7740
    },
    {
      "entropy": 5.3401695728302006,
      "epoch": 0.7439961575408262,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004952649438358234,
      "loss": 5.2641,
      "mean_token_accuracy": 0.19062443673610688,
      "num_tokens": 17781357.0,
      "step": 7745
    },
    {
      "entropy": 5.365611982345581,
      "epoch": 0.74447646493756,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004952579448704334,
      "loss": 5.2222,
      "mean_token_accuracy": 0.19461191743612288,
      "num_tokens": 17793640.0,
      "step": 7750
    },
    {
      "entropy": 5.306923484802246,
      "epoch": 0.7449567723342939,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000495250940791308,
      "loss": 5.1776,
      "mean_token_accuracy": 0.1966538682579994,
      "num_tokens": 17804228.0,
      "step": 7755
    },
    {
      "entropy": 5.306346750259399,
      "epoch": 0.7454370797310279,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004952439315986096,
      "loss": 5.2114,
      "mean_token_accuracy": 0.20089666694402694,
      "num_tokens": 17816153.0,
      "step": 7760
    },
    {
      "entropy": 5.319047594070435,
      "epoch": 0.7459173871277618,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004952369172925012,
      "loss": 5.2401,
      "mean_token_accuracy": 0.19817813932895662,
      "num_tokens": 17828464.0,
      "step": 7765
    },
    {
      "entropy": 5.23803014755249,
      "epoch": 0.7463976945244957,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004952298978731454,
      "loss": 5.1098,
      "mean_token_accuracy": 0.19798599630594255,
      "num_tokens": 17840126.0,
      "step": 7770
    },
    {
      "entropy": 5.2483419418334964,
      "epoch": 0.7468780019212296,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004952228733407055,
      "loss": 5.1722,
      "mean_token_accuracy": 0.2030227154493332,
      "num_tokens": 17851735.0,
      "step": 7775
    },
    {
      "entropy": 5.397327709197998,
      "epoch": 0.7473583093179635,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004952158436953444,
      "loss": 5.1442,
      "mean_token_accuracy": 0.20113309919834138,
      "num_tokens": 17862897.0,
      "step": 7780
    },
    {
      "entropy": 5.351913595199585,
      "epoch": 0.7478386167146974,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004952088089372252,
      "loss": 5.2262,
      "mean_token_accuracy": 0.2005309656262398,
      "num_tokens": 17874235.0,
      "step": 7785
    },
    {
      "entropy": 5.283905076980591,
      "epoch": 0.7483189241114313,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004952017690665114,
      "loss": 5.2272,
      "mean_token_accuracy": 0.1962098091840744,
      "num_tokens": 17885677.0,
      "step": 7790
    },
    {
      "entropy": 5.298684692382812,
      "epoch": 0.7487992315081652,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004951947240833664,
      "loss": 5.1413,
      "mean_token_accuracy": 0.20765664130449296,
      "num_tokens": 17896204.0,
      "step": 7795
    },
    {
      "entropy": 5.394489669799805,
      "epoch": 0.7492795389048992,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951876739879537,
      "loss": 5.2382,
      "mean_token_accuracy": 0.19582002013921737,
      "num_tokens": 17908514.0,
      "step": 7800
    },
    {
      "entropy": 5.326735401153565,
      "epoch": 0.7497598463016331,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004951806187804371,
      "loss": 5.1381,
      "mean_token_accuracy": 0.20370566248893737,
      "num_tokens": 17918493.0,
      "step": 7805
    },
    {
      "entropy": 5.279997825622559,
      "epoch": 0.7502401536983669,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004951735584609804,
      "loss": 5.1516,
      "mean_token_accuracy": 0.20330339819192886,
      "num_tokens": 17929016.0,
      "step": 7810
    },
    {
      "entropy": 5.217226362228393,
      "epoch": 0.7507204610951008,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004951664930297474,
      "loss": 5.1957,
      "mean_token_accuracy": 0.20084228962659836,
      "num_tokens": 17940674.0,
      "step": 7815
    },
    {
      "entropy": 5.383158826828003,
      "epoch": 0.7512007684918348,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000495159422486902,
      "loss": 5.2386,
      "mean_token_accuracy": 0.19554046392440796,
      "num_tokens": 17951431.0,
      "step": 7820
    },
    {
      "entropy": 5.443437767028809,
      "epoch": 0.7516810758885687,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004951523468326088,
      "loss": 5.2688,
      "mean_token_accuracy": 0.1874414101243019,
      "num_tokens": 17961434.0,
      "step": 7825
    },
    {
      "entropy": 5.179300117492676,
      "epoch": 0.7521613832853026,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004951452660670317,
      "loss": 5.104,
      "mean_token_accuracy": 0.20570982694625856,
      "num_tokens": 17971917.0,
      "step": 7830
    },
    {
      "entropy": 5.339613819122315,
      "epoch": 0.7526416906820365,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004951381801903352,
      "loss": 5.1772,
      "mean_token_accuracy": 0.2078887030482292,
      "num_tokens": 17983754.0,
      "step": 7835
    },
    {
      "entropy": 5.254813480377197,
      "epoch": 0.7531219980787704,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004951310892026839,
      "loss": 5.1694,
      "mean_token_accuracy": 0.199359530210495,
      "num_tokens": 17994088.0,
      "step": 7840
    },
    {
      "entropy": 5.292300176620484,
      "epoch": 0.7536023054755043,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004951239931042424,
      "loss": 5.1374,
      "mean_token_accuracy": 0.20227789878845215,
      "num_tokens": 18004847.0,
      "step": 7845
    },
    {
      "entropy": 5.276197671890259,
      "epoch": 0.7540826128722382,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004951168918951753,
      "loss": 5.1805,
      "mean_token_accuracy": 0.21156322956085205,
      "num_tokens": 18015756.0,
      "step": 7850
    },
    {
      "entropy": 5.294304466247558,
      "epoch": 0.7545629202689721,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004951097855756476,
      "loss": 5.2781,
      "mean_token_accuracy": 0.19650911390781403,
      "num_tokens": 18026417.0,
      "step": 7855
    },
    {
      "entropy": 5.306974935531616,
      "epoch": 0.7550432276657061,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004951026741458243,
      "loss": 5.1955,
      "mean_token_accuracy": 0.19948124140501022,
      "num_tokens": 18038205.0,
      "step": 7860
    },
    {
      "entropy": 5.3613261699676515,
      "epoch": 0.75552353506244,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004950955576058705,
      "loss": 5.2377,
      "mean_token_accuracy": 0.19970234781503676,
      "num_tokens": 18049223.0,
      "step": 7865
    },
    {
      "entropy": 5.318355083465576,
      "epoch": 0.7560038424591738,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004950884359559513,
      "loss": 5.1988,
      "mean_token_accuracy": 0.199899822473526,
      "num_tokens": 18061855.0,
      "step": 7870
    },
    {
      "entropy": 5.312740421295166,
      "epoch": 0.7564841498559077,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004950813091962324,
      "loss": 5.1879,
      "mean_token_accuracy": 0.20412675887346268,
      "num_tokens": 18073969.0,
      "step": 7875
    },
    {
      "entropy": 5.285601329803467,
      "epoch": 0.7569644572526417,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004950741773268788,
      "loss": 5.1304,
      "mean_token_accuracy": 0.20552899390459062,
      "num_tokens": 18085786.0,
      "step": 7880
    },
    {
      "entropy": 5.2859865665435795,
      "epoch": 0.7574447646493756,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004950670403480562,
      "loss": 5.1648,
      "mean_token_accuracy": 0.20141587555408477,
      "num_tokens": 18096263.0,
      "step": 7885
    },
    {
      "entropy": 5.319224119186401,
      "epoch": 0.7579250720461095,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004950598982599306,
      "loss": 5.2758,
      "mean_token_accuracy": 0.1892411604523659,
      "num_tokens": 18108769.0,
      "step": 7890
    },
    {
      "entropy": 5.3544079780578615,
      "epoch": 0.7584053794428435,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004950527510626675,
      "loss": 5.2489,
      "mean_token_accuracy": 0.1985387086868286,
      "num_tokens": 18120255.0,
      "step": 7895
    },
    {
      "entropy": 5.327633810043335,
      "epoch": 0.7588856868395774,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004950455987564329,
      "loss": 5.2386,
      "mean_token_accuracy": 0.200568225979805,
      "num_tokens": 18132016.0,
      "step": 7900
    },
    {
      "entropy": 5.270994043350219,
      "epoch": 0.7593659942363112,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004950384413413931,
      "loss": 5.1369,
      "mean_token_accuracy": 0.19926393777132034,
      "num_tokens": 18144101.0,
      "step": 7905
    },
    {
      "entropy": 5.3310620307922365,
      "epoch": 0.7598463016330451,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004950312788177139,
      "loss": 5.1401,
      "mean_token_accuracy": 0.20418404936790466,
      "num_tokens": 18155086.0,
      "step": 7910
    },
    {
      "entropy": 5.268496751785278,
      "epoch": 0.7603266090297791,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004950241111855618,
      "loss": 5.201,
      "mean_token_accuracy": 0.20112455636262894,
      "num_tokens": 18166902.0,
      "step": 7915
    },
    {
      "entropy": 5.27679409980774,
      "epoch": 0.760806916426513,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004950169384451031,
      "loss": 5.1624,
      "mean_token_accuracy": 0.2017876386642456,
      "num_tokens": 18178258.0,
      "step": 7920
    },
    {
      "entropy": 5.306259822845459,
      "epoch": 0.7612872238232469,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004950097605965045,
      "loss": 5.2094,
      "mean_token_accuracy": 0.19705730825662612,
      "num_tokens": 18190580.0,
      "step": 7925
    },
    {
      "entropy": 5.3387962818145756,
      "epoch": 0.7617675312199808,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004950025776399326,
      "loss": 5.2761,
      "mean_token_accuracy": 0.19709905683994294,
      "num_tokens": 18204912.0,
      "step": 7930
    },
    {
      "entropy": 5.3384746551513675,
      "epoch": 0.7622478386167147,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000494995389575554,
      "loss": 5.1698,
      "mean_token_accuracy": 0.1957654118537903,
      "num_tokens": 18215129.0,
      "step": 7935
    },
    {
      "entropy": 5.320358848571777,
      "epoch": 0.7627281460134486,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004949881964035357,
      "loss": 5.2923,
      "mean_token_accuracy": 0.19477613121271134,
      "num_tokens": 18225640.0,
      "step": 7940
    },
    {
      "entropy": 5.299406623840332,
      "epoch": 0.7632084534101825,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004949809981240448,
      "loss": 5.1613,
      "mean_token_accuracy": 0.20336330831050872,
      "num_tokens": 18236641.0,
      "step": 7945
    },
    {
      "entropy": 5.237579679489135,
      "epoch": 0.7636887608069164,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004949737947372483,
      "loss": 5.1593,
      "mean_token_accuracy": 0.20016419440507888,
      "num_tokens": 18249244.0,
      "step": 7950
    },
    {
      "entropy": 5.313110446929931,
      "epoch": 0.7641690682036504,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004949665862433134,
      "loss": 5.2275,
      "mean_token_accuracy": 0.19830468595027922,
      "num_tokens": 18261595.0,
      "step": 7955
    },
    {
      "entropy": 5.338136386871338,
      "epoch": 0.7646493756003843,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004949593726424077,
      "loss": 5.2093,
      "mean_token_accuracy": 0.20016187578439712,
      "num_tokens": 18272609.0,
      "step": 7960
    },
    {
      "entropy": 5.358658504486084,
      "epoch": 0.7651296829971181,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004949521539346985,
      "loss": 5.3439,
      "mean_token_accuracy": 0.19711480289697647,
      "num_tokens": 18283886.0,
      "step": 7965
    },
    {
      "entropy": 5.308894777297974,
      "epoch": 0.765609990393852,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004949449301203533,
      "loss": 5.2631,
      "mean_token_accuracy": 0.19835685640573503,
      "num_tokens": 18294794.0,
      "step": 7970
    },
    {
      "entropy": 5.411302852630615,
      "epoch": 0.766090297790586,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004949377011995399,
      "loss": 5.2466,
      "mean_token_accuracy": 0.19544450491666793,
      "num_tokens": 18306250.0,
      "step": 7975
    },
    {
      "entropy": 5.350947046279908,
      "epoch": 0.7665706051873199,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004949304671724263,
      "loss": 5.222,
      "mean_token_accuracy": 0.20425568670034408,
      "num_tokens": 18317743.0,
      "step": 7980
    },
    {
      "entropy": 5.171504354476928,
      "epoch": 0.7670509125840538,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004949232280391802,
      "loss": 5.0487,
      "mean_token_accuracy": 0.2057706281542778,
      "num_tokens": 18328287.0,
      "step": 7985
    },
    {
      "entropy": 5.311772012710572,
      "epoch": 0.7675312199807877,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004949159837999698,
      "loss": 5.2706,
      "mean_token_accuracy": 0.19622322618961335,
      "num_tokens": 18341354.0,
      "step": 7990
    },
    {
      "entropy": 5.365178871154785,
      "epoch": 0.7680115273775217,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004949087344549633,
      "loss": 5.2492,
      "mean_token_accuracy": 0.20166341662406922,
      "num_tokens": 18353822.0,
      "step": 7995
    },
    {
      "entropy": 5.389340400695801,
      "epoch": 0.7684918347742555,
      "grad_norm": 1.234375,
      "learning_rate": 0.000494901480004329,
      "loss": 5.1981,
      "mean_token_accuracy": 0.19579378217458726,
      "num_tokens": 18365491.0,
      "step": 8000
    },
    {
      "entropy": 5.231784915924072,
      "epoch": 0.7689721421709894,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004948942204482351,
      "loss": 5.1543,
      "mean_token_accuracy": 0.20263582468032837,
      "num_tokens": 18376329.0,
      "step": 8005
    },
    {
      "entropy": 5.3295105457305905,
      "epoch": 0.7694524495677233,
      "grad_norm": 1.375,
      "learning_rate": 0.0004948869557868506,
      "loss": 5.2452,
      "mean_token_accuracy": 0.19202990233898162,
      "num_tokens": 18387105.0,
      "step": 8010
    },
    {
      "entropy": 5.2895866394042965,
      "epoch": 0.7699327569644573,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004948796860203439,
      "loss": 5.197,
      "mean_token_accuracy": 0.19792436212301254,
      "num_tokens": 18398113.0,
      "step": 8015
    },
    {
      "entropy": 5.3291889190673825,
      "epoch": 0.7704130643611912,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004948724111488838,
      "loss": 5.1696,
      "mean_token_accuracy": 0.20074525475502014,
      "num_tokens": 18409315.0,
      "step": 8020
    },
    {
      "entropy": 5.211002588272095,
      "epoch": 0.770893371757925,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004948651311726391,
      "loss": 5.1972,
      "mean_token_accuracy": 0.19440497905015947,
      "num_tokens": 18421110.0,
      "step": 8025
    },
    {
      "entropy": 5.354279613494873,
      "epoch": 0.7713736791546589,
      "grad_norm": 1.25,
      "learning_rate": 0.0004948578460917789,
      "loss": 5.2705,
      "mean_token_accuracy": 0.19524814039468766,
      "num_tokens": 18433145.0,
      "step": 8030
    },
    {
      "entropy": 5.40636043548584,
      "epoch": 0.7718539865513929,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004948505559064725,
      "loss": 5.2084,
      "mean_token_accuracy": 0.1970051944255829,
      "num_tokens": 18444689.0,
      "step": 8035
    },
    {
      "entropy": 5.249115419387818,
      "epoch": 0.7723342939481268,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004948432606168889,
      "loss": 5.1649,
      "mean_token_accuracy": 0.20348201990127562,
      "num_tokens": 18455943.0,
      "step": 8040
    },
    {
      "entropy": 5.230320262908935,
      "epoch": 0.7728146013448607,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004948359602231976,
      "loss": 5.1166,
      "mean_token_accuracy": 0.2080174058675766,
      "num_tokens": 18466763.0,
      "step": 8045
    },
    {
      "entropy": 5.314593410491943,
      "epoch": 0.7732949087415946,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004948286547255681,
      "loss": 5.2639,
      "mean_token_accuracy": 0.19204829633235931,
      "num_tokens": 18477647.0,
      "step": 8050
    },
    {
      "entropy": 5.277676153182983,
      "epoch": 0.7737752161383286,
      "grad_norm": 1.140625,
      "learning_rate": 0.00049482134412417,
      "loss": 5.1434,
      "mean_token_accuracy": 0.20424923151731492,
      "num_tokens": 18488465.0,
      "step": 8055
    },
    {
      "entropy": 5.372780132293701,
      "epoch": 0.7742555235350624,
      "grad_norm": 1.234375,
      "learning_rate": 0.000494814028419173,
      "loss": 5.301,
      "mean_token_accuracy": 0.19462932646274567,
      "num_tokens": 18500705.0,
      "step": 8060
    },
    {
      "entropy": 5.299214839935303,
      "epoch": 0.7747358309317963,
      "grad_norm": 1.34375,
      "learning_rate": 0.000494806707610747,
      "loss": 5.1763,
      "mean_token_accuracy": 0.19690685123205184,
      "num_tokens": 18512253.0,
      "step": 8065
    },
    {
      "entropy": 5.223462295532227,
      "epoch": 0.7752161383285303,
      "grad_norm": 1.0625,
      "learning_rate": 0.000494799381699062,
      "loss": 5.1694,
      "mean_token_accuracy": 0.20119058191776276,
      "num_tokens": 18524231.0,
      "step": 8070
    },
    {
      "entropy": 5.341293525695801,
      "epoch": 0.7756964457252642,
      "grad_norm": 1.375,
      "learning_rate": 0.0004947920506842879,
      "loss": 5.2358,
      "mean_token_accuracy": 0.20328855961561204,
      "num_tokens": 18535824.0,
      "step": 8075
    },
    {
      "entropy": 5.4111504554748535,
      "epoch": 0.7761767531219981,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004947847145665951,
      "loss": 5.3011,
      "mean_token_accuracy": 0.20030797272920609,
      "num_tokens": 18547957.0,
      "step": 8080
    },
    {
      "entropy": 5.265676784515381,
      "epoch": 0.776657060518732,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004947773733461539,
      "loss": 5.2152,
      "mean_token_accuracy": 0.19701437056064605,
      "num_tokens": 18559297.0,
      "step": 8085
    },
    {
      "entropy": 5.332735204696656,
      "epoch": 0.777137367915466,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004947700270231347,
      "loss": 5.28,
      "mean_token_accuracy": 0.19301839768886567,
      "num_tokens": 18570264.0,
      "step": 8090
    },
    {
      "entropy": 5.3696177959442135,
      "epoch": 0.7776176753121998,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004947626755977079,
      "loss": 5.1884,
      "mean_token_accuracy": 0.20004567205905915,
      "num_tokens": 18582417.0,
      "step": 8095
    },
    {
      "entropy": 5.315948390960694,
      "epoch": 0.7780979827089337,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004947553190700444,
      "loss": 5.1986,
      "mean_token_accuracy": 0.19527707397937774,
      "num_tokens": 18593703.0,
      "step": 8100
    },
    {
      "entropy": 5.237201690673828,
      "epoch": 0.7785782901056676,
      "grad_norm": 1.328125,
      "learning_rate": 0.000494747957440315,
      "loss": 5.1748,
      "mean_token_accuracy": 0.20488553494215012,
      "num_tokens": 18605050.0,
      "step": 8105
    },
    {
      "entropy": 5.3131767272949215,
      "epoch": 0.7790585975024016,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004947405907086905,
      "loss": 5.2388,
      "mean_token_accuracy": 0.1944789320230484,
      "num_tokens": 18616962.0,
      "step": 8110
    },
    {
      "entropy": 5.384182596206665,
      "epoch": 0.7795389048991355,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004947332188753419,
      "loss": 5.1984,
      "mean_token_accuracy": 0.1946897104382515,
      "num_tokens": 18628640.0,
      "step": 8115
    },
    {
      "entropy": 5.31669397354126,
      "epoch": 0.7800192122958693,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004947258419404405,
      "loss": 5.2179,
      "mean_token_accuracy": 0.2010749787092209,
      "num_tokens": 18640488.0,
      "step": 8120
    },
    {
      "entropy": 5.276207971572876,
      "epoch": 0.7804995196926032,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004947184599041576,
      "loss": 5.1843,
      "mean_token_accuracy": 0.2055606335401535,
      "num_tokens": 18652095.0,
      "step": 8125
    },
    {
      "entropy": 5.349359655380249,
      "epoch": 0.7809798270893372,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004947110727666644,
      "loss": 5.1875,
      "mean_token_accuracy": 0.19877787679433823,
      "num_tokens": 18662767.0,
      "step": 8130
    },
    {
      "entropy": 5.293517398834228,
      "epoch": 0.7814601344860711,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004947036805281325,
      "loss": 5.2406,
      "mean_token_accuracy": 0.19749340564012527,
      "num_tokens": 18673568.0,
      "step": 8135
    },
    {
      "entropy": 5.345055627822876,
      "epoch": 0.781940441882805,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004946962831887336,
      "loss": 5.1836,
      "mean_token_accuracy": 0.20542929619550704,
      "num_tokens": 18684456.0,
      "step": 8140
    },
    {
      "entropy": 5.317824125289917,
      "epoch": 0.7824207492795389,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004946888807486393,
      "loss": 5.2323,
      "mean_token_accuracy": 0.1958246186375618,
      "num_tokens": 18695942.0,
      "step": 8145
    },
    {
      "entropy": 5.408920955657959,
      "epoch": 0.7829010566762729,
      "grad_norm": 1.125,
      "learning_rate": 0.0004946814732080214,
      "loss": 5.2717,
      "mean_token_accuracy": 0.19965537786483764,
      "num_tokens": 18708110.0,
      "step": 8150
    },
    {
      "entropy": 5.418804025650024,
      "epoch": 0.7833813640730067,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004946740605670523,
      "loss": 5.2616,
      "mean_token_accuracy": 0.19457316547632217,
      "num_tokens": 18718864.0,
      "step": 8155
    },
    {
      "entropy": 5.282051515579224,
      "epoch": 0.7838616714697406,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004946666428259037,
      "loss": 5.1372,
      "mean_token_accuracy": 0.19738233238458633,
      "num_tokens": 18730338.0,
      "step": 8160
    },
    {
      "entropy": 5.305791234970092,
      "epoch": 0.7843419788664745,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004946592199847478,
      "loss": 5.3095,
      "mean_token_accuracy": 0.18705110996961594,
      "num_tokens": 18743433.0,
      "step": 8165
    },
    {
      "entropy": 5.370831775665283,
      "epoch": 0.7848222862632085,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004946517920437571,
      "loss": 5.1913,
      "mean_token_accuracy": 0.196962109208107,
      "num_tokens": 18754631.0,
      "step": 8170
    },
    {
      "entropy": 5.31728196144104,
      "epoch": 0.7853025936599424,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004946443590031041,
      "loss": 5.1996,
      "mean_token_accuracy": 0.19670164436101914,
      "num_tokens": 18766118.0,
      "step": 8175
    },
    {
      "entropy": 5.3038763999938965,
      "epoch": 0.7857829010566763,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004946369208629613,
      "loss": 5.21,
      "mean_token_accuracy": 0.19857015162706376,
      "num_tokens": 18777743.0,
      "step": 8180
    },
    {
      "entropy": 5.239198875427246,
      "epoch": 0.7862632084534101,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004946294776235013,
      "loss": 5.172,
      "mean_token_accuracy": 0.20766065567731856,
      "num_tokens": 18789684.0,
      "step": 8185
    },
    {
      "entropy": 5.266590356826782,
      "epoch": 0.7867435158501441,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004946220292848971,
      "loss": 5.1356,
      "mean_token_accuracy": 0.20451817512512208,
      "num_tokens": 18801088.0,
      "step": 8190
    },
    {
      "entropy": 5.357203960418701,
      "epoch": 0.787223823246878,
      "grad_norm": 1.25,
      "learning_rate": 0.0004946145758473214,
      "loss": 5.2337,
      "mean_token_accuracy": 0.19293261617422103,
      "num_tokens": 18812478.0,
      "step": 8195
    },
    {
      "entropy": 5.405508279800415,
      "epoch": 0.7877041306436119,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004946071173109475,
      "loss": 5.2451,
      "mean_token_accuracy": 0.198164926469326,
      "num_tokens": 18823573.0,
      "step": 8200
    },
    {
      "entropy": 5.263828086853027,
      "epoch": 0.7881844380403458,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004945996536759484,
      "loss": 5.1502,
      "mean_token_accuracy": 0.20268950760364532,
      "num_tokens": 18833880.0,
      "step": 8205
    },
    {
      "entropy": 5.271388101577759,
      "epoch": 0.7886647454370798,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004945921849424974,
      "loss": 5.1162,
      "mean_token_accuracy": 0.2040919005870819,
      "num_tokens": 18844367.0,
      "step": 8210
    },
    {
      "entropy": 5.307074642181396,
      "epoch": 0.7891450528338136,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004945847111107679,
      "loss": 5.1871,
      "mean_token_accuracy": 0.19504126757383347,
      "num_tokens": 18856309.0,
      "step": 8215
    },
    {
      "entropy": 5.287567758560181,
      "epoch": 0.7896253602305475,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004945772321809334,
      "loss": 5.1378,
      "mean_token_accuracy": 0.20097096860408784,
      "num_tokens": 18868873.0,
      "step": 8220
    },
    {
      "entropy": 5.244015455245972,
      "epoch": 0.7901056676272814,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004945697481531677,
      "loss": 5.1245,
      "mean_token_accuracy": 0.2060120776295662,
      "num_tokens": 18879953.0,
      "step": 8225
    },
    {
      "entropy": 5.337798738479615,
      "epoch": 0.7905859750240154,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004945622590276443,
      "loss": 5.2697,
      "mean_token_accuracy": 0.19190877228975295,
      "num_tokens": 18892220.0,
      "step": 8230
    },
    {
      "entropy": 5.315177011489868,
      "epoch": 0.7910662824207493,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004945547648045373,
      "loss": 5.1825,
      "mean_token_accuracy": 0.19964509755373,
      "num_tokens": 18903261.0,
      "step": 8235
    },
    {
      "entropy": 5.275335741043091,
      "epoch": 0.7915465898174832,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004945472654840206,
      "loss": 5.2145,
      "mean_token_accuracy": 0.2043684035539627,
      "num_tokens": 18913702.0,
      "step": 8240
    },
    {
      "entropy": 5.3623268604278564,
      "epoch": 0.7920268972142172,
      "grad_norm": 1.875,
      "learning_rate": 0.0004945397610662683,
      "loss": 5.2666,
      "mean_token_accuracy": 0.19344153106212617,
      "num_tokens": 18925024.0,
      "step": 8245
    },
    {
      "entropy": 5.332998275756836,
      "epoch": 0.792507204610951,
      "grad_norm": 1.125,
      "learning_rate": 0.0004945322515514547,
      "loss": 5.1709,
      "mean_token_accuracy": 0.2051153913140297,
      "num_tokens": 18935939.0,
      "step": 8250
    },
    {
      "entropy": 5.280615758895874,
      "epoch": 0.7929875120076849,
      "grad_norm": 1.15625,
      "learning_rate": 0.000494524736939754,
      "loss": 5.1427,
      "mean_token_accuracy": 0.19560697674751282,
      "num_tokens": 18946091.0,
      "step": 8255
    },
    {
      "entropy": 5.375056886672974,
      "epoch": 0.7934678194044188,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004945172172313408,
      "loss": 5.2202,
      "mean_token_accuracy": 0.19347982108592987,
      "num_tokens": 18957471.0,
      "step": 8260
    },
    {
      "entropy": 5.422613477706909,
      "epoch": 0.7939481268011528,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004945096924263896,
      "loss": 5.3593,
      "mean_token_accuracy": 0.19303269386291505,
      "num_tokens": 18969462.0,
      "step": 8265
    },
    {
      "entropy": 5.37666425704956,
      "epoch": 0.7944284341978867,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004945021625250753,
      "loss": 5.2036,
      "mean_token_accuracy": 0.19561776220798494,
      "num_tokens": 18980228.0,
      "step": 8270
    },
    {
      "entropy": 5.423059034347534,
      "epoch": 0.7949087415946205,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004944946275275724,
      "loss": 5.3322,
      "mean_token_accuracy": 0.19468947052955626,
      "num_tokens": 18991661.0,
      "step": 8275
    },
    {
      "entropy": 5.338334417343139,
      "epoch": 0.7953890489913544,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004944870874340561,
      "loss": 5.2226,
      "mean_token_accuracy": 0.2015485867857933,
      "num_tokens": 19003082.0,
      "step": 8280
    },
    {
      "entropy": 5.3358570575714115,
      "epoch": 0.7958693563880884,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004944795422447013,
      "loss": 5.1076,
      "mean_token_accuracy": 0.20028071105480194,
      "num_tokens": 19015408.0,
      "step": 8285
    },
    {
      "entropy": 5.319946527481079,
      "epoch": 0.7963496637848223,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004944719919596835,
      "loss": 5.1963,
      "mean_token_accuracy": 0.1946763277053833,
      "num_tokens": 19027246.0,
      "step": 8290
    },
    {
      "entropy": 5.275589942932129,
      "epoch": 0.7968299711815562,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004944644365791776,
      "loss": 5.1891,
      "mean_token_accuracy": 0.2022460073232651,
      "num_tokens": 19038126.0,
      "step": 8295
    },
    {
      "entropy": 5.254959535598755,
      "epoch": 0.7973102785782901,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000494456876103359,
      "loss": 5.1288,
      "mean_token_accuracy": 0.2032008394598961,
      "num_tokens": 19047998.0,
      "step": 8300
    },
    {
      "entropy": 5.254001712799072,
      "epoch": 0.7977905859750241,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004944493105324035,
      "loss": 5.2265,
      "mean_token_accuracy": 0.19854558408260345,
      "num_tokens": 19058969.0,
      "step": 8305
    },
    {
      "entropy": 5.335261631011963,
      "epoch": 0.7982708933717579,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004944417398664866,
      "loss": 5.1423,
      "mean_token_accuracy": 0.20597300827503204,
      "num_tokens": 19070314.0,
      "step": 8310
    },
    {
      "entropy": 5.366521310806275,
      "epoch": 0.7987512007684918,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004944341641057843,
      "loss": 5.1987,
      "mean_token_accuracy": 0.20353574752807618,
      "num_tokens": 19082575.0,
      "step": 8315
    },
    {
      "entropy": 5.254666090011597,
      "epoch": 0.7992315081652257,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004944265832504721,
      "loss": 5.158,
      "mean_token_accuracy": 0.19918670803308486,
      "num_tokens": 19095719.0,
      "step": 8320
    },
    {
      "entropy": 5.31092848777771,
      "epoch": 0.7997118155619597,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004944189973007262,
      "loss": 5.1515,
      "mean_token_accuracy": 0.20244342982769012,
      "num_tokens": 19106894.0,
      "step": 8325
    },
    {
      "entropy": 5.263830757141113,
      "epoch": 0.8001921229586936,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004944114062567229,
      "loss": 5.1598,
      "mean_token_accuracy": 0.19858704507350922,
      "num_tokens": 19118206.0,
      "step": 8330
    },
    {
      "entropy": 5.3236809253692625,
      "epoch": 0.8006724303554275,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004944038101186381,
      "loss": 5.2252,
      "mean_token_accuracy": 0.19298866987228394,
      "num_tokens": 19129613.0,
      "step": 8335
    },
    {
      "entropy": 5.224341487884521,
      "epoch": 0.8011527377521613,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004943962088866483,
      "loss": 5.155,
      "mean_token_accuracy": 0.20398945957422257,
      "num_tokens": 19142214.0,
      "step": 8340
    },
    {
      "entropy": 5.309964752197265,
      "epoch": 0.8016330451488953,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004943886025609301,
      "loss": 5.2506,
      "mean_token_accuracy": 0.19829992353916168,
      "num_tokens": 19153906.0,
      "step": 8345
    },
    {
      "entropy": 5.336621046066284,
      "epoch": 0.8021133525456292,
      "grad_norm": 1.1875,
      "learning_rate": 0.00049438099114166,
      "loss": 5.1461,
      "mean_token_accuracy": 0.19538349360227586,
      "num_tokens": 19163677.0,
      "step": 8350
    },
    {
      "entropy": 5.351234769821167,
      "epoch": 0.8025936599423631,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004943733746290147,
      "loss": 5.2189,
      "mean_token_accuracy": 0.19923721253871918,
      "num_tokens": 19175240.0,
      "step": 8355
    },
    {
      "entropy": 5.24532732963562,
      "epoch": 0.803073967339097,
      "grad_norm": 1.328125,
      "learning_rate": 0.000494365753023171,
      "loss": 5.0519,
      "mean_token_accuracy": 0.205495922267437,
      "num_tokens": 19185529.0,
      "step": 8360
    },
    {
      "entropy": 5.2331788539886475,
      "epoch": 0.803554274735831,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004943581263243059,
      "loss": 5.1454,
      "mean_token_accuracy": 0.2023431807756424,
      "num_tokens": 19197153.0,
      "step": 8365
    },
    {
      "entropy": 5.347947883605957,
      "epoch": 0.8040345821325648,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004943504945325965,
      "loss": 5.3048,
      "mean_token_accuracy": 0.19195201694965364,
      "num_tokens": 19208439.0,
      "step": 8370
    },
    {
      "entropy": 5.271076202392578,
      "epoch": 0.8045148895292987,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004943428576482198,
      "loss": 5.1347,
      "mean_token_accuracy": 0.2040252387523651,
      "num_tokens": 19218570.0,
      "step": 8375
    },
    {
      "entropy": 5.253508377075195,
      "epoch": 0.8049951969260326,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004943352156713535,
      "loss": 5.139,
      "mean_token_accuracy": 0.20503223240375518,
      "num_tokens": 19229080.0,
      "step": 8380
    },
    {
      "entropy": 5.304885292053223,
      "epoch": 0.8054755043227666,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004943275686021747,
      "loss": 5.2285,
      "mean_token_accuracy": 0.1969524398446083,
      "num_tokens": 19239299.0,
      "step": 8385
    },
    {
      "entropy": 5.336613368988037,
      "epoch": 0.8059558117195005,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000494319916440861,
      "loss": 5.1717,
      "mean_token_accuracy": 0.20321761667728425,
      "num_tokens": 19250691.0,
      "step": 8390
    },
    {
      "entropy": 5.305912017822266,
      "epoch": 0.8064361191162344,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004943122591875901,
      "loss": 5.1622,
      "mean_token_accuracy": 0.20421040058135986,
      "num_tokens": 19260868.0,
      "step": 8395
    },
    {
      "entropy": 5.381595182418823,
      "epoch": 0.8069164265129684,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004943045968425398,
      "loss": 5.2641,
      "mean_token_accuracy": 0.19213174879550934,
      "num_tokens": 19271924.0,
      "step": 8400
    },
    {
      "entropy": 5.307283496856689,
      "epoch": 0.8073967339097022,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004942969294058878,
      "loss": 5.2121,
      "mean_token_accuracy": 0.20062707513570785,
      "num_tokens": 19283340.0,
      "step": 8405
    },
    {
      "entropy": 5.329241847991943,
      "epoch": 0.8078770413064361,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004942892568778125,
      "loss": 5.2659,
      "mean_token_accuracy": 0.19736168384552003,
      "num_tokens": 19294903.0,
      "step": 8410
    },
    {
      "entropy": 5.340913486480713,
      "epoch": 0.80835734870317,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004942815792584917,
      "loss": 5.1534,
      "mean_token_accuracy": 0.19918999075889587,
      "num_tokens": 19305670.0,
      "step": 8415
    },
    {
      "entropy": 5.351573133468628,
      "epoch": 0.808837656099904,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004942738965481038,
      "loss": 5.1627,
      "mean_token_accuracy": 0.19951719492673875,
      "num_tokens": 19317043.0,
      "step": 8420
    },
    {
      "entropy": 5.303330707550049,
      "epoch": 0.8093179634966379,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004942662087468272,
      "loss": 5.2397,
      "mean_token_accuracy": 0.19928821325302123,
      "num_tokens": 19329161.0,
      "step": 8425
    },
    {
      "entropy": 5.315001773834228,
      "epoch": 0.8097982708933718,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004942585158548402,
      "loss": 5.2125,
      "mean_token_accuracy": 0.19904156476259233,
      "num_tokens": 19340706.0,
      "step": 8430
    },
    {
      "entropy": 5.309930896759033,
      "epoch": 0.8102785782901056,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004942508178723214,
      "loss": 5.1982,
      "mean_token_accuracy": 0.1963711068034172,
      "num_tokens": 19352422.0,
      "step": 8435
    },
    {
      "entropy": 5.291172170639038,
      "epoch": 0.8107588856868396,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004942431147994499,
      "loss": 5.2405,
      "mean_token_accuracy": 0.1970633402466774,
      "num_tokens": 19364109.0,
      "step": 8440
    },
    {
      "entropy": 5.343476152420044,
      "epoch": 0.8112391930835735,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004942354066364042,
      "loss": 5.228,
      "mean_token_accuracy": 0.193661467730999,
      "num_tokens": 19377386.0,
      "step": 8445
    },
    {
      "entropy": 5.357939863204956,
      "epoch": 0.8117195004803074,
      "grad_norm": 4.90625,
      "learning_rate": 0.0004942276933833634,
      "loss": 5.1496,
      "mean_token_accuracy": 0.20619347244501113,
      "num_tokens": 19388453.0,
      "step": 8450
    },
    {
      "entropy": 5.2919844627380375,
      "epoch": 0.8121998078770413,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004942199750405064,
      "loss": 5.2856,
      "mean_token_accuracy": 0.1960239440202713,
      "num_tokens": 19402240.0,
      "step": 8455
    },
    {
      "entropy": 5.257682275772095,
      "epoch": 0.8126801152737753,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004942122516080127,
      "loss": 5.1917,
      "mean_token_accuracy": 0.19957529455423356,
      "num_tokens": 19414260.0,
      "step": 8460
    },
    {
      "entropy": 5.275717449188233,
      "epoch": 0.8131604226705091,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004942045230860614,
      "loss": 5.123,
      "mean_token_accuracy": 0.19857604056596756,
      "num_tokens": 19426356.0,
      "step": 8465
    },
    {
      "entropy": 5.300590562820434,
      "epoch": 0.813640730067243,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004941967894748319,
      "loss": 5.2107,
      "mean_token_accuracy": 0.19834148436784743,
      "num_tokens": 19438881.0,
      "step": 8470
    },
    {
      "entropy": 5.3006284713745115,
      "epoch": 0.8141210374639769,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004941890507745039,
      "loss": 5.1564,
      "mean_token_accuracy": 0.1988014116883278,
      "num_tokens": 19450109.0,
      "step": 8475
    },
    {
      "entropy": 5.321816539764404,
      "epoch": 0.8146013448607109,
      "grad_norm": 1.25,
      "learning_rate": 0.0004941813069852569,
      "loss": 5.1508,
      "mean_token_accuracy": 0.204894956946373,
      "num_tokens": 19461155.0,
      "step": 8480
    },
    {
      "entropy": 5.375993394851685,
      "epoch": 0.8150816522574448,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004941735581072708,
      "loss": 5.3499,
      "mean_token_accuracy": 0.19433972388505935,
      "num_tokens": 19474364.0,
      "step": 8485
    },
    {
      "entropy": 5.33035306930542,
      "epoch": 0.8155619596541787,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004941658041407255,
      "loss": 5.1427,
      "mean_token_accuracy": 0.2113051801919937,
      "num_tokens": 19485271.0,
      "step": 8490
    },
    {
      "entropy": 5.386933040618897,
      "epoch": 0.8160422670509125,
      "grad_norm": 1.140625,
      "learning_rate": 0.000494158045085801,
      "loss": 5.3273,
      "mean_token_accuracy": 0.18803139477968217,
      "num_tokens": 19497117.0,
      "step": 8495
    },
    {
      "entropy": 5.335579490661621,
      "epoch": 0.8165225744476465,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004941502809426776,
      "loss": 5.1072,
      "mean_token_accuracy": 0.20256997644901276,
      "num_tokens": 19507035.0,
      "step": 8500
    },
    {
      "entropy": 5.241343879699707,
      "epoch": 0.8170028818443804,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004941425117115354,
      "loss": 5.1069,
      "mean_token_accuracy": 0.20615998059511184,
      "num_tokens": 19517791.0,
      "step": 8505
    },
    {
      "entropy": 5.2908729076385494,
      "epoch": 0.8174831892411143,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004941347373925547,
      "loss": 5.1852,
      "mean_token_accuracy": 0.2008901059627533,
      "num_tokens": 19529577.0,
      "step": 8510
    },
    {
      "entropy": 5.396166372299194,
      "epoch": 0.8179634966378482,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004941269579859161,
      "loss": 5.2814,
      "mean_token_accuracy": 0.19961613565683364,
      "num_tokens": 19542355.0,
      "step": 8515
    },
    {
      "entropy": 5.363604354858398,
      "epoch": 0.8184438040345822,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004941191734918002,
      "loss": 5.256,
      "mean_token_accuracy": 0.1920148551464081,
      "num_tokens": 19553819.0,
      "step": 8520
    },
    {
      "entropy": 5.360704326629639,
      "epoch": 0.818924111431316,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000494111383910388,
      "loss": 5.2644,
      "mean_token_accuracy": 0.1924416497349739,
      "num_tokens": 19565816.0,
      "step": 8525
    },
    {
      "entropy": 5.307268524169922,
      "epoch": 0.8194044188280499,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004941035892418597,
      "loss": 5.1702,
      "mean_token_accuracy": 0.20374006032943726,
      "num_tokens": 19577991.0,
      "step": 8530
    },
    {
      "entropy": 5.254099655151367,
      "epoch": 0.8198847262247838,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004940957894863968,
      "loss": 5.1601,
      "mean_token_accuracy": 0.19929326176643372,
      "num_tokens": 19590225.0,
      "step": 8535
    },
    {
      "entropy": 5.22421612739563,
      "epoch": 0.8203650336215178,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004940879846441804,
      "loss": 5.1926,
      "mean_token_accuracy": 0.20000286549329757,
      "num_tokens": 19602966.0,
      "step": 8540
    },
    {
      "entropy": 5.424680519104004,
      "epoch": 0.8208453410182517,
      "grad_norm": 1.125,
      "learning_rate": 0.0004940801747153914,
      "loss": 5.1902,
      "mean_token_accuracy": 0.1992236241698265,
      "num_tokens": 19613148.0,
      "step": 8545
    },
    {
      "entropy": 5.250468778610229,
      "epoch": 0.8213256484149856,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004940723597002113,
      "loss": 5.0497,
      "mean_token_accuracy": 0.21086280941963195,
      "num_tokens": 19624361.0,
      "step": 8550
    },
    {
      "entropy": 5.195836544036865,
      "epoch": 0.8218059558117194,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004940645395988216,
      "loss": 5.1631,
      "mean_token_accuracy": 0.20194362848997116,
      "num_tokens": 19635915.0,
      "step": 8555
    },
    {
      "entropy": 5.3118428707122805,
      "epoch": 0.8222862632084534,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004940567144114036,
      "loss": 5.1912,
      "mean_token_accuracy": 0.19464145302772523,
      "num_tokens": 19646701.0,
      "step": 8560
    },
    {
      "entropy": 5.403972721099853,
      "epoch": 0.8227665706051873,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004940488841381393,
      "loss": 5.2803,
      "mean_token_accuracy": 0.19274735003709792,
      "num_tokens": 19657804.0,
      "step": 8565
    },
    {
      "entropy": 5.36010856628418,
      "epoch": 0.8232468780019212,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004940410487792103,
      "loss": 5.2756,
      "mean_token_accuracy": 0.19777508080005646,
      "num_tokens": 19670202.0,
      "step": 8570
    },
    {
      "entropy": 5.171558856964111,
      "epoch": 0.8237271853986552,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004940332083347986,
      "loss": 5.0638,
      "mean_token_accuracy": 0.21246689707040786,
      "num_tokens": 19681362.0,
      "step": 8575
    },
    {
      "entropy": 5.269045829772949,
      "epoch": 0.8242074927953891,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004940253628050861,
      "loss": 5.2149,
      "mean_token_accuracy": 0.19713031500577927,
      "num_tokens": 19693429.0,
      "step": 8580
    },
    {
      "entropy": 5.367968606948852,
      "epoch": 0.824687800192123,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004940175121902552,
      "loss": 5.2653,
      "mean_token_accuracy": 0.20200771540403367,
      "num_tokens": 19704667.0,
      "step": 8585
    },
    {
      "entropy": 5.353277063369751,
      "epoch": 0.8251681075888568,
      "grad_norm": 1.25,
      "learning_rate": 0.000494009656490488,
      "loss": 5.192,
      "mean_token_accuracy": 0.19973867982625962,
      "num_tokens": 19716002.0,
      "step": 8590
    },
    {
      "entropy": 5.294351577758789,
      "epoch": 0.8256484149855908,
      "grad_norm": 1.25,
      "learning_rate": 0.0004940017957059668,
      "loss": 5.1946,
      "mean_token_accuracy": 0.2015809714794159,
      "num_tokens": 19726626.0,
      "step": 8595
    },
    {
      "entropy": 5.335646724700927,
      "epoch": 0.8261287223823247,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004939939298368742,
      "loss": 5.1802,
      "mean_token_accuracy": 0.2016742706298828,
      "num_tokens": 19738250.0,
      "step": 8600
    },
    {
      "entropy": 5.406242227554321,
      "epoch": 0.8266090297790586,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004939860588833929,
      "loss": 5.2617,
      "mean_token_accuracy": 0.20098914653062822,
      "num_tokens": 19750119.0,
      "step": 8605
    },
    {
      "entropy": 5.326722812652588,
      "epoch": 0.8270893371757925,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004939781828457055,
      "loss": 5.2468,
      "mean_token_accuracy": 0.19814164340496063,
      "num_tokens": 19761709.0,
      "step": 8610
    },
    {
      "entropy": 5.348292446136474,
      "epoch": 0.8275696445725265,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000493970301723995,
      "loss": 5.2136,
      "mean_token_accuracy": 0.19424867182970046,
      "num_tokens": 19773562.0,
      "step": 8615
    },
    {
      "entropy": 5.430823898315429,
      "epoch": 0.8280499519692603,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004939624155184443,
      "loss": 5.2302,
      "mean_token_accuracy": 0.19674144834280013,
      "num_tokens": 19785327.0,
      "step": 8620
    },
    {
      "entropy": 5.332286167144775,
      "epoch": 0.8285302593659942,
      "grad_norm": 1.375,
      "learning_rate": 0.0004939545242292365,
      "loss": 5.2609,
      "mean_token_accuracy": 0.1936324030160904,
      "num_tokens": 19796760.0,
      "step": 8625
    },
    {
      "entropy": 5.314222717285157,
      "epoch": 0.8290105667627281,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004939466278565547,
      "loss": 5.1711,
      "mean_token_accuracy": 0.19922126531600953,
      "num_tokens": 19807271.0,
      "step": 8630
    },
    {
      "entropy": 5.329996919631958,
      "epoch": 0.8294908741594621,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004939387264005825,
      "loss": 5.1991,
      "mean_token_accuracy": 0.20293248444795609,
      "num_tokens": 19818098.0,
      "step": 8635
    },
    {
      "entropy": 5.303319644927979,
      "epoch": 0.829971181556196,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004939308198615031,
      "loss": 5.1948,
      "mean_token_accuracy": 0.20245611816644668,
      "num_tokens": 19829482.0,
      "step": 8640
    },
    {
      "entropy": 5.323172235488892,
      "epoch": 0.8304514889529299,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004939229082395001,
      "loss": 5.2351,
      "mean_token_accuracy": 0.195654059946537,
      "num_tokens": 19840883.0,
      "step": 8645
    },
    {
      "entropy": 5.278830862045288,
      "epoch": 0.8309317963496637,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004939149915347573,
      "loss": 5.1209,
      "mean_token_accuracy": 0.2011566698551178,
      "num_tokens": 19852619.0,
      "step": 8650
    },
    {
      "entropy": 5.236908102035523,
      "epoch": 0.8314121037463977,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004939070697474585,
      "loss": 5.0943,
      "mean_token_accuracy": 0.1989907518029213,
      "num_tokens": 19863316.0,
      "step": 8655
    },
    {
      "entropy": 5.334206867218017,
      "epoch": 0.8318924111431316,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004938991428777875,
      "loss": 5.2732,
      "mean_token_accuracy": 0.1999387785792351,
      "num_tokens": 19873965.0,
      "step": 8660
    },
    {
      "entropy": 5.345083570480346,
      "epoch": 0.8323727185398655,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004938912109259284,
      "loss": 5.1772,
      "mean_token_accuracy": 0.20287925451993943,
      "num_tokens": 19884685.0,
      "step": 8665
    },
    {
      "entropy": 5.285607051849365,
      "epoch": 0.8328530259365994,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004938832738920654,
      "loss": 5.1363,
      "mean_token_accuracy": 0.20559962391853331,
      "num_tokens": 19897594.0,
      "step": 8670
    },
    {
      "entropy": 5.3275954246521,
      "epoch": 0.8333333333333334,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004938753317763826,
      "loss": 5.2397,
      "mean_token_accuracy": 0.19611236155033113,
      "num_tokens": 19908093.0,
      "step": 8675
    },
    {
      "entropy": 5.387573289871216,
      "epoch": 0.8338136407300673,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004938673845790646,
      "loss": 5.2737,
      "mean_token_accuracy": 0.19050368070602416,
      "num_tokens": 19920502.0,
      "step": 8680
    },
    {
      "entropy": 5.256779432296753,
      "epoch": 0.8342939481268011,
      "grad_norm": 1.25,
      "learning_rate": 0.0004938594323002957,
      "loss": 5.1014,
      "mean_token_accuracy": 0.20578562170267106,
      "num_tokens": 19931437.0,
      "step": 8685
    },
    {
      "entropy": 5.266842889785766,
      "epoch": 0.834774255523535,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004938514749402609,
      "loss": 5.1802,
      "mean_token_accuracy": 0.20510144978761674,
      "num_tokens": 19942592.0,
      "step": 8690
    },
    {
      "entropy": 5.3655835628509525,
      "epoch": 0.835254562920269,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004938435124991447,
      "loss": 5.3449,
      "mean_token_accuracy": 0.18808136880397797,
      "num_tokens": 19953802.0,
      "step": 8695
    },
    {
      "entropy": 5.329191064834594,
      "epoch": 0.8357348703170029,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004938355449771318,
      "loss": 5.1842,
      "mean_token_accuracy": 0.19948012083768846,
      "num_tokens": 19964686.0,
      "step": 8700
    },
    {
      "entropy": 5.297907495498658,
      "epoch": 0.8362151777137368,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004938275723744075,
      "loss": 5.2294,
      "mean_token_accuracy": 0.19733165800571442,
      "num_tokens": 19975603.0,
      "step": 8705
    },
    {
      "entropy": 5.366217851638794,
      "epoch": 0.8366954851104706,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004938195946911567,
      "loss": 5.2722,
      "mean_token_accuracy": 0.1943481370806694,
      "num_tokens": 19986205.0,
      "step": 8710
    },
    {
      "entropy": 5.346731376647949,
      "epoch": 0.8371757925072046,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004938116119275645,
      "loss": 5.1152,
      "mean_token_accuracy": 0.20564762502908707,
      "num_tokens": 19997323.0,
      "step": 8715
    },
    {
      "entropy": 5.2681962013244625,
      "epoch": 0.8376560999039385,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004938036240838166,
      "loss": 5.1644,
      "mean_token_accuracy": 0.21143766790628432,
      "num_tokens": 20007334.0,
      "step": 8720
    },
    {
      "entropy": 5.358319568634033,
      "epoch": 0.8381364073006724,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004937956311600983,
      "loss": 5.1685,
      "mean_token_accuracy": 0.19725634753704072,
      "num_tokens": 20017476.0,
      "step": 8725
    },
    {
      "entropy": 5.228802299499511,
      "epoch": 0.8386167146974063,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004937876331565951,
      "loss": 5.1285,
      "mean_token_accuracy": 0.1960748866200447,
      "num_tokens": 20029323.0,
      "step": 8730
    },
    {
      "entropy": 5.3003987789154055,
      "epoch": 0.8390970220941403,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004937796300734926,
      "loss": 5.1427,
      "mean_token_accuracy": 0.19682454466819763,
      "num_tokens": 20040317.0,
      "step": 8735
    },
    {
      "entropy": 5.307217311859131,
      "epoch": 0.8395773294908742,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004937716219109769,
      "loss": 5.2105,
      "mean_token_accuracy": 0.20332604944705962,
      "num_tokens": 20050495.0,
      "step": 8740
    },
    {
      "entropy": 5.357418155670166,
      "epoch": 0.840057636887608,
      "grad_norm": 1.375,
      "learning_rate": 0.0004937636086692336,
      "loss": 5.2422,
      "mean_token_accuracy": 0.19442333728075029,
      "num_tokens": 20061185.0,
      "step": 8745
    },
    {
      "entropy": 5.3504157066345215,
      "epoch": 0.840537944284342,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000493755590348449,
      "loss": 5.2041,
      "mean_token_accuracy": 0.19920917898416518,
      "num_tokens": 20073746.0,
      "step": 8750
    },
    {
      "entropy": 5.354205751419068,
      "epoch": 0.8410182516810759,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004937475669488091,
      "loss": 5.2324,
      "mean_token_accuracy": 0.19785770773887634,
      "num_tokens": 20085667.0,
      "step": 8755
    },
    {
      "entropy": 5.286267900466919,
      "epoch": 0.8414985590778098,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004937395384705004,
      "loss": 5.118,
      "mean_token_accuracy": 0.20490512996912003,
      "num_tokens": 20096017.0,
      "step": 8760
    },
    {
      "entropy": 5.284736633300781,
      "epoch": 0.8419788664745437,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004937315049137089,
      "loss": 5.2013,
      "mean_token_accuracy": 0.19973714053630828,
      "num_tokens": 20106928.0,
      "step": 8765
    },
    {
      "entropy": 5.300041961669922,
      "epoch": 0.8424591738712777,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004937234662786216,
      "loss": 5.1754,
      "mean_token_accuracy": 0.20246230363845824,
      "num_tokens": 20118768.0,
      "step": 8770
    },
    {
      "entropy": 5.454545783996582,
      "epoch": 0.8429394812680115,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004937154225654246,
      "loss": 5.3209,
      "mean_token_accuracy": 0.1886596366763115,
      "num_tokens": 20129009.0,
      "step": 8775
    },
    {
      "entropy": 5.332577228546143,
      "epoch": 0.8434197886647454,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004937073737743051,
      "loss": 5.1872,
      "mean_token_accuracy": 0.1994388997554779,
      "num_tokens": 20141577.0,
      "step": 8780
    },
    {
      "entropy": 5.377079772949219,
      "epoch": 0.8439000960614793,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004936993199054499,
      "loss": 5.2623,
      "mean_token_accuracy": 0.1997489422559738,
      "num_tokens": 20153636.0,
      "step": 8785
    },
    {
      "entropy": 5.266515445709229,
      "epoch": 0.8443804034582133,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004936912609590458,
      "loss": 5.1527,
      "mean_token_accuracy": 0.20007888674736024,
      "num_tokens": 20165719.0,
      "step": 8790
    },
    {
      "entropy": 5.3406284809112545,
      "epoch": 0.8448607108549472,
      "grad_norm": 1.203125,
      "learning_rate": 0.00049368319693528,
      "loss": 5.2225,
      "mean_token_accuracy": 0.19011954963207245,
      "num_tokens": 20178260.0,
      "step": 8795
    },
    {
      "entropy": 5.293787240982056,
      "epoch": 0.8453410182516811,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004936751278343397,
      "loss": 5.2225,
      "mean_token_accuracy": 0.20058311223983766,
      "num_tokens": 20189667.0,
      "step": 8800
    },
    {
      "entropy": 5.2720568656921385,
      "epoch": 0.845821325648415,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004936670536564123,
      "loss": 5.1872,
      "mean_token_accuracy": 0.1949663758277893,
      "num_tokens": 20201451.0,
      "step": 8805
    },
    {
      "entropy": 5.347260856628418,
      "epoch": 0.8463016330451489,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004936589744016853,
      "loss": 5.2565,
      "mean_token_accuracy": 0.19954086244106292,
      "num_tokens": 20213257.0,
      "step": 8810
    },
    {
      "entropy": 5.352904462814331,
      "epoch": 0.8467819404418828,
      "grad_norm": 1.09375,
      "learning_rate": 0.000493650890070346,
      "loss": 5.2073,
      "mean_token_accuracy": 0.20174627602100373,
      "num_tokens": 20225301.0,
      "step": 8815
    },
    {
      "entropy": 5.3138245105743405,
      "epoch": 0.8472622478386167,
      "grad_norm": 1.125,
      "learning_rate": 0.0004936428006625824,
      "loss": 5.1677,
      "mean_token_accuracy": 0.2056563079357147,
      "num_tokens": 20236183.0,
      "step": 8820
    },
    {
      "entropy": 5.3882440567016605,
      "epoch": 0.8477425552353506,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004936347061785823,
      "loss": 5.2818,
      "mean_token_accuracy": 0.19195735156536103,
      "num_tokens": 20248663.0,
      "step": 8825
    },
    {
      "entropy": 5.279906368255615,
      "epoch": 0.8482228626320846,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004936266066185334,
      "loss": 5.1495,
      "mean_token_accuracy": 0.20857516825199127,
      "num_tokens": 20259293.0,
      "step": 8830
    },
    {
      "entropy": 5.3200846195220945,
      "epoch": 0.8487031700288185,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004936185019826239,
      "loss": 5.2928,
      "mean_token_accuracy": 0.19034613519906998,
      "num_tokens": 20270601.0,
      "step": 8835
    },
    {
      "entropy": 5.354975318908691,
      "epoch": 0.8491834774255523,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004936103922710419,
      "loss": 5.2663,
      "mean_token_accuracy": 0.2018500819802284,
      "num_tokens": 20281471.0,
      "step": 8840
    },
    {
      "entropy": 5.322579717636108,
      "epoch": 0.8496637848222862,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004936022774839759,
      "loss": 5.1065,
      "mean_token_accuracy": 0.20395581126213075,
      "num_tokens": 20291941.0,
      "step": 8845
    },
    {
      "entropy": 5.353834390640259,
      "epoch": 0.8501440922190202,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004935941576216141,
      "loss": 5.2362,
      "mean_token_accuracy": 0.1947695016860962,
      "num_tokens": 20302631.0,
      "step": 8850
    },
    {
      "entropy": 5.345116043090821,
      "epoch": 0.8506243996157541,
      "grad_norm": 1.0625,
      "learning_rate": 0.000493586032684145,
      "loss": 5.3011,
      "mean_token_accuracy": 0.1986623674631119,
      "num_tokens": 20314955.0,
      "step": 8855
    },
    {
      "entropy": 5.36974401473999,
      "epoch": 0.851104707012488,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004935779026717573,
      "loss": 5.2356,
      "mean_token_accuracy": 0.20304753482341767,
      "num_tokens": 20326791.0,
      "step": 8860
    },
    {
      "entropy": 5.298195457458496,
      "epoch": 0.8515850144092219,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004935697675846396,
      "loss": 5.2176,
      "mean_token_accuracy": 0.2029862090945244,
      "num_tokens": 20338912.0,
      "step": 8865
    },
    {
      "entropy": 5.295038652420044,
      "epoch": 0.8520653218059558,
      "grad_norm": 1.25,
      "learning_rate": 0.0004935616274229811,
      "loss": 5.1811,
      "mean_token_accuracy": 0.2046012669801712,
      "num_tokens": 20349414.0,
      "step": 8870
    },
    {
      "entropy": 5.387618350982666,
      "epoch": 0.8525456292026897,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004935534821869705,
      "loss": 5.1979,
      "mean_token_accuracy": 0.1997691035270691,
      "num_tokens": 20360346.0,
      "step": 8875
    },
    {
      "entropy": 5.30146279335022,
      "epoch": 0.8530259365994236,
      "grad_norm": 1.125,
      "learning_rate": 0.0004935453318767971,
      "loss": 5.1782,
      "mean_token_accuracy": 0.20337827503681183,
      "num_tokens": 20371965.0,
      "step": 8880
    },
    {
      "entropy": 5.290291452407837,
      "epoch": 0.8535062439961575,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00049353717649265,
      "loss": 5.2057,
      "mean_token_accuracy": 0.2005314975976944,
      "num_tokens": 20382632.0,
      "step": 8885
    },
    {
      "entropy": 5.321771955490112,
      "epoch": 0.8539865513928915,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004935290160347185,
      "loss": 5.1873,
      "mean_token_accuracy": 0.20742321908473968,
      "num_tokens": 20395796.0,
      "step": 8890
    },
    {
      "entropy": 5.308526229858399,
      "epoch": 0.8544668587896254,
      "grad_norm": 2.234375,
      "learning_rate": 0.0004935208505031922,
      "loss": 5.2273,
      "mean_token_accuracy": 0.2014682650566101,
      "num_tokens": 20407113.0,
      "step": 8895
    },
    {
      "entropy": 5.374697494506836,
      "epoch": 0.8549471661863592,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004935126798982606,
      "loss": 5.2491,
      "mean_token_accuracy": 0.19916778951883315,
      "num_tokens": 20419012.0,
      "step": 8900
    },
    {
      "entropy": 5.3496192455291744,
      "epoch": 0.8554274735830932,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004935045042201135,
      "loss": 5.1823,
      "mean_token_accuracy": 0.19541468918323518,
      "num_tokens": 20430337.0,
      "step": 8905
    },
    {
      "entropy": 5.253158187866211,
      "epoch": 0.8559077809798271,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004934963234689407,
      "loss": 5.2082,
      "mean_token_accuracy": 0.1972067877650261,
      "num_tokens": 20441614.0,
      "step": 8910
    },
    {
      "entropy": 5.273896884918213,
      "epoch": 0.856388088376561,
      "grad_norm": 1.109375,
      "learning_rate": 0.000493488137644932,
      "loss": 5.0723,
      "mean_token_accuracy": 0.21466317027807236,
      "num_tokens": 20453170.0,
      "step": 8915
    },
    {
      "entropy": 5.201868486404419,
      "epoch": 0.8568683957732949,
      "grad_norm": 1.375,
      "learning_rate": 0.0004934799467482774,
      "loss": 5.047,
      "mean_token_accuracy": 0.20770143568515778,
      "num_tokens": 20464585.0,
      "step": 8920
    },
    {
      "entropy": 5.244882345199585,
      "epoch": 0.8573487031700289,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004934717507791673,
      "loss": 5.0398,
      "mean_token_accuracy": 0.21078211218118667,
      "num_tokens": 20473903.0,
      "step": 8925
    },
    {
      "entropy": 5.275483655929565,
      "epoch": 0.8578290105667628,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004934635497377919,
      "loss": 5.1983,
      "mean_token_accuracy": 0.2006663292646408,
      "num_tokens": 20484778.0,
      "step": 8930
    },
    {
      "entropy": 5.293740320205688,
      "epoch": 0.8583093179634966,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004934553436243415,
      "loss": 5.234,
      "mean_token_accuracy": 0.19553713947534562,
      "num_tokens": 20495534.0,
      "step": 8935
    },
    {
      "entropy": 5.452532386779785,
      "epoch": 0.8587896253602305,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004934471324390067,
      "loss": 5.2796,
      "mean_token_accuracy": 0.19752825796604156,
      "num_tokens": 20506853.0,
      "step": 8940
    },
    {
      "entropy": 5.314362573623657,
      "epoch": 0.8592699327569645,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004934389161819783,
      "loss": 5.214,
      "mean_token_accuracy": 0.2019460216164589,
      "num_tokens": 20519112.0,
      "step": 8945
    },
    {
      "entropy": 5.333446455001831,
      "epoch": 0.8597502401536984,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004934306948534467,
      "loss": 5.2155,
      "mean_token_accuracy": 0.19869790822267533,
      "num_tokens": 20530060.0,
      "step": 8950
    },
    {
      "entropy": 5.273174095153808,
      "epoch": 0.8602305475504323,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004934224684536031,
      "loss": 5.1508,
      "mean_token_accuracy": 0.20007785707712172,
      "num_tokens": 20541802.0,
      "step": 8955
    },
    {
      "entropy": 5.2137669086456295,
      "epoch": 0.8607108549471661,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004934142369826382,
      "loss": 5.071,
      "mean_token_accuracy": 0.2069406807422638,
      "num_tokens": 20552494.0,
      "step": 8960
    },
    {
      "entropy": 5.287360286712646,
      "epoch": 0.8611911623439001,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004934060004407434,
      "loss": 5.2092,
      "mean_token_accuracy": 0.20390230864286424,
      "num_tokens": 20564467.0,
      "step": 8965
    },
    {
      "entropy": 5.318108749389649,
      "epoch": 0.861671469740634,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004933977588281099,
      "loss": 5.2399,
      "mean_token_accuracy": 0.1985464796423912,
      "num_tokens": 20575399.0,
      "step": 8970
    },
    {
      "entropy": 5.354484796524048,
      "epoch": 0.8621517771373679,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004933895121449288,
      "loss": 5.2279,
      "mean_token_accuracy": 0.20146728157997132,
      "num_tokens": 20586212.0,
      "step": 8975
    },
    {
      "entropy": 5.365288782119751,
      "epoch": 0.8626320845341018,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004933812603913917,
      "loss": 5.2356,
      "mean_token_accuracy": 0.19713350385427475,
      "num_tokens": 20597020.0,
      "step": 8980
    },
    {
      "entropy": 5.282711505889893,
      "epoch": 0.8631123919308358,
      "grad_norm": 1.25,
      "learning_rate": 0.0004933730035676903,
      "loss": 5.1785,
      "mean_token_accuracy": 0.2024289458990097,
      "num_tokens": 20609447.0,
      "step": 8985
    },
    {
      "entropy": 5.301419448852539,
      "epoch": 0.8635926993275697,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004933647416740161,
      "loss": 5.149,
      "mean_token_accuracy": 0.2009505644440651,
      "num_tokens": 20621780.0,
      "step": 8990
    },
    {
      "entropy": 5.215321397781372,
      "epoch": 0.8640730067243035,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000493356474710561,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20762356817722322,
      "num_tokens": 20632331.0,
      "step": 8995
    },
    {
      "entropy": 5.232652950286865,
      "epoch": 0.8645533141210374,
      "grad_norm": 1.15625,
      "learning_rate": 0.000493348202677517,
      "loss": 5.0668,
      "mean_token_accuracy": 0.2011037364602089,
      "num_tokens": 20643357.0,
      "step": 9000
    },
    {
      "epoch": 0.8645533141210374,
      "eval_entropy": 5.161021471837275,
      "eval_loss": 5.215470314025879,
      "eval_mean_token_accuracy": 0.2061440540901746,
      "eval_num_tokens": 20643357.0,
      "eval_runtime": 26.8016,
      "eval_samples_per_second": 1224.369,
      "eval_steps_per_second": 153.051,
      "step": 9000
    },
    {
      "entropy": 5.247832918167115,
      "epoch": 0.8650336215177714,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004933399255750761,
      "loss": 5.1296,
      "mean_token_accuracy": 0.20930825173854828,
      "num_tokens": 20655212.0,
      "step": 9005
    },
    {
      "entropy": 5.211752796173096,
      "epoch": 0.8655139289145053,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004933316434034304,
      "loss": 5.0077,
      "mean_token_accuracy": 0.21536419838666915,
      "num_tokens": 20666467.0,
      "step": 9010
    },
    {
      "entropy": 5.188072967529297,
      "epoch": 0.8659942363112392,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004933233561627723,
      "loss": 5.1396,
      "mean_token_accuracy": 0.20552490204572677,
      "num_tokens": 20677709.0,
      "step": 9015
    },
    {
      "entropy": 5.285771274566651,
      "epoch": 0.866474543707973,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004933150638532942,
      "loss": 5.1591,
      "mean_token_accuracy": 0.19521578550338745,
      "num_tokens": 20688880.0,
      "step": 9020
    },
    {
      "entropy": 5.344256114959717,
      "epoch": 0.866954851104707,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004933067664751885,
      "loss": 5.31,
      "mean_token_accuracy": 0.18637982457876207,
      "num_tokens": 20701384.0,
      "step": 9025
    },
    {
      "entropy": 5.401659202575684,
      "epoch": 0.8674351585014409,
      "grad_norm": 1.140625,
      "learning_rate": 0.000493298464028648,
      "loss": 5.2538,
      "mean_token_accuracy": 0.19299253970384597,
      "num_tokens": 20713736.0,
      "step": 9030
    },
    {
      "entropy": 5.196385478973388,
      "epoch": 0.8679154658981748,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004932901565138653,
      "loss": 5.0865,
      "mean_token_accuracy": 0.20637259781360626,
      "num_tokens": 20725424.0,
      "step": 9035
    },
    {
      "entropy": 5.26168475151062,
      "epoch": 0.8683957732949087,
      "grad_norm": 1.125,
      "learning_rate": 0.0004932818439310334,
      "loss": 5.0981,
      "mean_token_accuracy": 0.20577118694782257,
      "num_tokens": 20737454.0,
      "step": 9040
    },
    {
      "entropy": 5.390743923187256,
      "epoch": 0.8688760806916427,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004932735262803452,
      "loss": 5.2413,
      "mean_token_accuracy": 0.1979524627327919,
      "num_tokens": 20748882.0,
      "step": 9045
    },
    {
      "entropy": 5.251908588409424,
      "epoch": 0.8693563880883766,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004932652035619939,
      "loss": 5.1169,
      "mean_token_accuracy": 0.20614991784095765,
      "num_tokens": 20759622.0,
      "step": 9050
    },
    {
      "entropy": 5.200184154510498,
      "epoch": 0.8698366954851104,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004932568757761727,
      "loss": 5.1233,
      "mean_token_accuracy": 0.2038133293390274,
      "num_tokens": 20770658.0,
      "step": 9055
    },
    {
      "entropy": 5.346509742736816,
      "epoch": 0.8703170028818443,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004932485429230748,
      "loss": 5.1816,
      "mean_token_accuracy": 0.2016171008348465,
      "num_tokens": 20782176.0,
      "step": 9060
    },
    {
      "entropy": 5.403298854827881,
      "epoch": 0.8707973102785783,
      "grad_norm": 1.203125,
      "learning_rate": 0.000493240205002894,
      "loss": 5.2444,
      "mean_token_accuracy": 0.18923848420381545,
      "num_tokens": 20793592.0,
      "step": 9065
    },
    {
      "entropy": 5.337335729598999,
      "epoch": 0.8712776176753122,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004932318620158235,
      "loss": 5.1389,
      "mean_token_accuracy": 0.19778282940387726,
      "num_tokens": 20804226.0,
      "step": 9070
    },
    {
      "entropy": 5.29606819152832,
      "epoch": 0.8717579250720461,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004932235139620574,
      "loss": 5.2846,
      "mean_token_accuracy": 0.19290715903043748,
      "num_tokens": 20815336.0,
      "step": 9075
    },
    {
      "entropy": 5.301112842559815,
      "epoch": 0.8722382324687801,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004932151608417892,
      "loss": 5.1715,
      "mean_token_accuracy": 0.21021737158298492,
      "num_tokens": 20826936.0,
      "step": 9080
    },
    {
      "entropy": 5.301452398300171,
      "epoch": 0.872718539865514,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004932068026552127,
      "loss": 5.2246,
      "mean_token_accuracy": 0.2009325474500656,
      "num_tokens": 20838159.0,
      "step": 9085
    },
    {
      "entropy": 5.277756261825561,
      "epoch": 0.8731988472622478,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004931984394025224,
      "loss": 5.1192,
      "mean_token_accuracy": 0.1982308253645897,
      "num_tokens": 20848605.0,
      "step": 9090
    },
    {
      "entropy": 5.335677909851074,
      "epoch": 0.8736791546589817,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004931900710839123,
      "loss": 5.2281,
      "mean_token_accuracy": 0.19745373576879502,
      "num_tokens": 20860494.0,
      "step": 9095
    },
    {
      "entropy": 5.340137243270874,
      "epoch": 0.8741594620557157,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004931816976995766,
      "loss": 5.2347,
      "mean_token_accuracy": 0.20225998312234877,
      "num_tokens": 20871327.0,
      "step": 9100
    },
    {
      "entropy": 5.312560939788819,
      "epoch": 0.8746397694524496,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004931733192497097,
      "loss": 5.183,
      "mean_token_accuracy": 0.20286207944154738,
      "num_tokens": 20883032.0,
      "step": 9105
    },
    {
      "entropy": 5.31759204864502,
      "epoch": 0.8751200768491835,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004931649357345062,
      "loss": 5.2189,
      "mean_token_accuracy": 0.19399236440658568,
      "num_tokens": 20894327.0,
      "step": 9110
    },
    {
      "entropy": 5.224873304367065,
      "epoch": 0.8756003842459174,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004931565471541606,
      "loss": 5.0844,
      "mean_token_accuracy": 0.2122832104563713,
      "num_tokens": 20905489.0,
      "step": 9115
    },
    {
      "entropy": 5.271890544891358,
      "epoch": 0.8760806916426513,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004931481535088679,
      "loss": 5.1585,
      "mean_token_accuracy": 0.19868822395801544,
      "num_tokens": 20917044.0,
      "step": 9120
    },
    {
      "entropy": 5.328049230575561,
      "epoch": 0.8765609990393852,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004931397547988229,
      "loss": 5.1521,
      "mean_token_accuracy": 0.20387671887874603,
      "num_tokens": 20928065.0,
      "step": 9125
    },
    {
      "entropy": 5.309719371795654,
      "epoch": 0.8770413064361191,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004931313510242204,
      "loss": 5.1707,
      "mean_token_accuracy": 0.20396066904067994,
      "num_tokens": 20939396.0,
      "step": 9130
    },
    {
      "entropy": 5.367967128753662,
      "epoch": 0.877521613832853,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004931229421852557,
      "loss": 5.2561,
      "mean_token_accuracy": 0.20187842845916748,
      "num_tokens": 20950844.0,
      "step": 9135
    },
    {
      "entropy": 5.272248411178589,
      "epoch": 0.878001921229587,
      "grad_norm": 1.375,
      "learning_rate": 0.000493114528282124,
      "loss": 5.1566,
      "mean_token_accuracy": 0.20478422194719315,
      "num_tokens": 20961567.0,
      "step": 9140
    },
    {
      "entropy": 5.383096599578858,
      "epoch": 0.8784822286263209,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004931061093150206,
      "loss": 5.3168,
      "mean_token_accuracy": 0.19615557342767714,
      "num_tokens": 20972762.0,
      "step": 9145
    },
    {
      "entropy": 5.293380308151245,
      "epoch": 0.8789625360230547,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004930976852841409,
      "loss": 5.1481,
      "mean_token_accuracy": 0.20180506259202957,
      "num_tokens": 20983848.0,
      "step": 9150
    },
    {
      "entropy": 5.283677577972412,
      "epoch": 0.8794428434197886,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004930892561896806,
      "loss": 5.2099,
      "mean_token_accuracy": 0.2011642649769783,
      "num_tokens": 20996418.0,
      "step": 9155
    },
    {
      "entropy": 5.3179511547088625,
      "epoch": 0.8799231508165226,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004930808220318354,
      "loss": 5.1844,
      "mean_token_accuracy": 0.20691795498132706,
      "num_tokens": 21008714.0,
      "step": 9160
    },
    {
      "entropy": 5.3237926959991455,
      "epoch": 0.8804034582132565,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004930723828108012,
      "loss": 5.1381,
      "mean_token_accuracy": 0.20677352994680404,
      "num_tokens": 21020453.0,
      "step": 9165
    },
    {
      "entropy": 5.1607448101043705,
      "epoch": 0.8808837656099904,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004930639385267736,
      "loss": 5.1064,
      "mean_token_accuracy": 0.2048717975616455,
      "num_tokens": 21032211.0,
      "step": 9170
    },
    {
      "entropy": 5.3238568782806395,
      "epoch": 0.8813640730067243,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000493055489179949,
      "loss": 5.2069,
      "mean_token_accuracy": 0.19962812811136246,
      "num_tokens": 21042995.0,
      "step": 9175
    },
    {
      "entropy": 5.244577169418335,
      "epoch": 0.8818443804034583,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004930470347705234,
      "loss": 5.0588,
      "mean_token_accuracy": 0.20873888432979584,
      "num_tokens": 21054306.0,
      "step": 9180
    },
    {
      "entropy": 5.284455680847168,
      "epoch": 0.8823246878001921,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000493038575298693,
      "loss": 5.2305,
      "mean_token_accuracy": 0.19198696613311766,
      "num_tokens": 21067144.0,
      "step": 9185
    },
    {
      "entropy": 5.429376602172852,
      "epoch": 0.882804995196926,
      "grad_norm": 1.125,
      "learning_rate": 0.0004930301107646545,
      "loss": 5.2138,
      "mean_token_accuracy": 0.19381897151470184,
      "num_tokens": 21078885.0,
      "step": 9190
    },
    {
      "entropy": 5.383817958831787,
      "epoch": 0.8832853025936599,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004930216411686042,
      "loss": 5.2733,
      "mean_token_accuracy": 0.19005136936903,
      "num_tokens": 21089482.0,
      "step": 9195
    },
    {
      "entropy": 5.341294479370117,
      "epoch": 0.8837656099903939,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004930131665107387,
      "loss": 5.2547,
      "mean_token_accuracy": 0.19461563378572463,
      "num_tokens": 21102046.0,
      "step": 9200
    },
    {
      "entropy": 5.287715673446655,
      "epoch": 0.8842459173871278,
      "grad_norm": 1.5,
      "learning_rate": 0.000493004686791255,
      "loss": 5.103,
      "mean_token_accuracy": 0.20509903132915497,
      "num_tokens": 21112827.0,
      "step": 9205
    },
    {
      "entropy": 5.312151002883911,
      "epoch": 0.8847262247838616,
      "grad_norm": 1.25,
      "learning_rate": 0.0004929962020103496,
      "loss": 5.2062,
      "mean_token_accuracy": 0.19648258984088898,
      "num_tokens": 21123108.0,
      "step": 9210
    },
    {
      "entropy": 5.313097047805786,
      "epoch": 0.8852065321805955,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004929877121682198,
      "loss": 5.2283,
      "mean_token_accuracy": 0.20041085183620452,
      "num_tokens": 21134955.0,
      "step": 9215
    },
    {
      "entropy": 5.317350006103515,
      "epoch": 0.8856868395773295,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004929792172650627,
      "loss": 5.121,
      "mean_token_accuracy": 0.20926922261714936,
      "num_tokens": 21146415.0,
      "step": 9220
    },
    {
      "entropy": 5.280223178863525,
      "epoch": 0.8861671469740634,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004929707173010753,
      "loss": 5.1489,
      "mean_token_accuracy": 0.20307850390672683,
      "num_tokens": 21158691.0,
      "step": 9225
    },
    {
      "entropy": 5.284905004501343,
      "epoch": 0.8866474543707973,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004929622122764552,
      "loss": 5.1709,
      "mean_token_accuracy": 0.199029840528965,
      "num_tokens": 21170063.0,
      "step": 9230
    },
    {
      "entropy": 5.363232851028442,
      "epoch": 0.8871277617675313,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004929537021913997,
      "loss": 5.2552,
      "mean_token_accuracy": 0.1968018516898155,
      "num_tokens": 21182427.0,
      "step": 9235
    },
    {
      "entropy": 5.299725961685181,
      "epoch": 0.8876080691642652,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004929451870461064,
      "loss": 5.1084,
      "mean_token_accuracy": 0.20508427172899246,
      "num_tokens": 21195011.0,
      "step": 9240
    },
    {
      "entropy": 5.292886400222779,
      "epoch": 0.888088376560999,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004929366668407731,
      "loss": 5.1663,
      "mean_token_accuracy": 0.20353161841630935,
      "num_tokens": 21207886.0,
      "step": 9245
    },
    {
      "entropy": 5.260407400131226,
      "epoch": 0.8885686839577329,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004929281415755974,
      "loss": 5.1334,
      "mean_token_accuracy": 0.20615132749080659,
      "num_tokens": 21219285.0,
      "step": 9250
    },
    {
      "entropy": 5.353756475448608,
      "epoch": 0.8890489913544669,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004929196112507775,
      "loss": 5.1926,
      "mean_token_accuracy": 0.19552750885486603,
      "num_tokens": 21231208.0,
      "step": 9255
    },
    {
      "entropy": 5.249666500091553,
      "epoch": 0.8895292987512008,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004929110758665112,
      "loss": 5.0776,
      "mean_token_accuracy": 0.20356733351945877,
      "num_tokens": 21242573.0,
      "step": 9260
    },
    {
      "entropy": 5.14683256149292,
      "epoch": 0.8900096061479347,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004929025354229969,
      "loss": 5.0631,
      "mean_token_accuracy": 0.2149868592619896,
      "num_tokens": 21254966.0,
      "step": 9265
    },
    {
      "entropy": 5.292044019699096,
      "epoch": 0.8904899135446686,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004928939899204326,
      "loss": 5.1755,
      "mean_token_accuracy": 0.202527816593647,
      "num_tokens": 21266167.0,
      "step": 9270
    },
    {
      "entropy": 5.302504539489746,
      "epoch": 0.8909702209414025,
      "grad_norm": 1.265625,
      "learning_rate": 0.000492885439359017,
      "loss": 5.2135,
      "mean_token_accuracy": 0.19785204976797105,
      "num_tokens": 21277036.0,
      "step": 9275
    },
    {
      "entropy": 5.231221389770508,
      "epoch": 0.8914505283381364,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004928768837389485,
      "loss": 5.1452,
      "mean_token_accuracy": 0.2035413846373558,
      "num_tokens": 21288887.0,
      "step": 9280
    },
    {
      "entropy": 5.355093669891358,
      "epoch": 0.8919308357348703,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004928683230604257,
      "loss": 5.2589,
      "mean_token_accuracy": 0.19448313117027283,
      "num_tokens": 21300664.0,
      "step": 9285
    },
    {
      "entropy": 5.356330108642578,
      "epoch": 0.8924111431316042,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004928597573236474,
      "loss": 5.1198,
      "mean_token_accuracy": 0.20275168716907502,
      "num_tokens": 21311535.0,
      "step": 9290
    },
    {
      "entropy": 5.1858726978302006,
      "epoch": 0.8928914505283382,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004928511865288123,
      "loss": 5.084,
      "mean_token_accuracy": 0.20858484655618667,
      "num_tokens": 21323248.0,
      "step": 9295
    },
    {
      "entropy": 5.246766138076782,
      "epoch": 0.8933717579250721,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004928426106761197,
      "loss": 5.1408,
      "mean_token_accuracy": 0.20377192199230193,
      "num_tokens": 21334901.0,
      "step": 9300
    },
    {
      "entropy": 5.383527708053589,
      "epoch": 0.8938520653218059,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004928340297657685,
      "loss": 5.2368,
      "mean_token_accuracy": 0.20129102617502212,
      "num_tokens": 21345902.0,
      "step": 9305
    },
    {
      "entropy": 5.4028785705566404,
      "epoch": 0.8943323727185398,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004928254437979578,
      "loss": 5.2762,
      "mean_token_accuracy": 0.19476929157972336,
      "num_tokens": 21358240.0,
      "step": 9310
    },
    {
      "entropy": 5.192928838729858,
      "epoch": 0.8948126801152738,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004928168527728873,
      "loss": 5.0287,
      "mean_token_accuracy": 0.2050730064511299,
      "num_tokens": 21368209.0,
      "step": 9315
    },
    {
      "entropy": 5.194242143630982,
      "epoch": 0.8952929875120077,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004928082566907562,
      "loss": 5.1142,
      "mean_token_accuracy": 0.20026154667139054,
      "num_tokens": 21379484.0,
      "step": 9320
    },
    {
      "entropy": 5.282604742050171,
      "epoch": 0.8957732949087416,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004927996555517642,
      "loss": 5.1528,
      "mean_token_accuracy": 0.2018384501338005,
      "num_tokens": 21391750.0,
      "step": 9325
    },
    {
      "entropy": 5.350960302352905,
      "epoch": 0.8962536023054755,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004927910493561109,
      "loss": 5.2259,
      "mean_token_accuracy": 0.19743057787418367,
      "num_tokens": 21402583.0,
      "step": 9330
    },
    {
      "entropy": 5.276013040542603,
      "epoch": 0.8967339097022095,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000492782438103996,
      "loss": 5.1212,
      "mean_token_accuracy": 0.19557443112134934,
      "num_tokens": 21412892.0,
      "step": 9335
    },
    {
      "entropy": 5.305454874038697,
      "epoch": 0.8972142170989433,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004927738217956197,
      "loss": 5.1469,
      "mean_token_accuracy": 0.19979104548692703,
      "num_tokens": 21422828.0,
      "step": 9340
    },
    {
      "entropy": 5.2201439380645756,
      "epoch": 0.8976945244956772,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004927652004311819,
      "loss": 5.1485,
      "mean_token_accuracy": 0.20431360751390457,
      "num_tokens": 21434441.0,
      "step": 9345
    },
    {
      "entropy": 5.338518142700195,
      "epoch": 0.8981748318924111,
      "grad_norm": 1.25,
      "learning_rate": 0.0004927565740108828,
      "loss": 5.2968,
      "mean_token_accuracy": 0.19366048872470856,
      "num_tokens": 21445945.0,
      "step": 9350
    },
    {
      "entropy": 5.408589506149292,
      "epoch": 0.8986551392891451,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004927479425349226,
      "loss": 5.2202,
      "mean_token_accuracy": 0.20081375688314437,
      "num_tokens": 21457265.0,
      "step": 9355
    },
    {
      "entropy": 5.28257417678833,
      "epoch": 0.899135446685879,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004927393060035018,
      "loss": 5.1473,
      "mean_token_accuracy": 0.20172258019447326,
      "num_tokens": 21469090.0,
      "step": 9360
    },
    {
      "entropy": 5.307695293426514,
      "epoch": 0.8996157540826129,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004927306644168207,
      "loss": 5.2034,
      "mean_token_accuracy": 0.19522192627191542,
      "num_tokens": 21480280.0,
      "step": 9365
    },
    {
      "entropy": 5.263973903656006,
      "epoch": 0.9000960614793467,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004927220177750803,
      "loss": 5.143,
      "mean_token_accuracy": 0.21245052367448808,
      "num_tokens": 21491856.0,
      "step": 9370
    },
    {
      "entropy": 5.2963625431060795,
      "epoch": 0.9005763688760807,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004927133660784811,
      "loss": 5.1785,
      "mean_token_accuracy": 0.20074619799852372,
      "num_tokens": 21501814.0,
      "step": 9375
    },
    {
      "entropy": 5.318115234375,
      "epoch": 0.9010566762728146,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004927047093272241,
      "loss": 5.2025,
      "mean_token_accuracy": 0.20094681680202484,
      "num_tokens": 21514725.0,
      "step": 9380
    },
    {
      "entropy": 5.261816883087159,
      "epoch": 0.9015369836695485,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00049269604752151,
      "loss": 5.1414,
      "mean_token_accuracy": 0.205718532204628,
      "num_tokens": 21526523.0,
      "step": 9385
    },
    {
      "entropy": 5.236191749572754,
      "epoch": 0.9020172910662824,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004926873806615403,
      "loss": 5.1102,
      "mean_token_accuracy": 0.20615485012531282,
      "num_tokens": 21537947.0,
      "step": 9390
    },
    {
      "entropy": 5.366571617126465,
      "epoch": 0.9024975984630164,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004926787087475158,
      "loss": 5.2187,
      "mean_token_accuracy": 0.19643158465623856,
      "num_tokens": 21550238.0,
      "step": 9395
    },
    {
      "entropy": 5.2883669376373295,
      "epoch": 0.9029779058597502,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004926700317796382,
      "loss": 5.2131,
      "mean_token_accuracy": 0.19550104886293412,
      "num_tokens": 21561546.0,
      "step": 9400
    },
    {
      "entropy": 5.335643863677978,
      "epoch": 0.9034582132564841,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004926613497581088,
      "loss": 5.1997,
      "mean_token_accuracy": 0.2027388483285904,
      "num_tokens": 21572034.0,
      "step": 9405
    },
    {
      "entropy": 5.274737358093262,
      "epoch": 0.9039385206532181,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004926526626831292,
      "loss": 5.16,
      "mean_token_accuracy": 0.20008518397808076,
      "num_tokens": 21583427.0,
      "step": 9410
    },
    {
      "entropy": 5.296346712112427,
      "epoch": 0.904418828049952,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004926439705549011,
      "loss": 5.1641,
      "mean_token_accuracy": 0.20773767530918122,
      "num_tokens": 21595071.0,
      "step": 9415
    },
    {
      "entropy": 5.360065698623657,
      "epoch": 0.9048991354466859,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004926352733736262,
      "loss": 5.2153,
      "mean_token_accuracy": 0.20213809609413147,
      "num_tokens": 21606007.0,
      "step": 9420
    },
    {
      "entropy": 5.293471193313598,
      "epoch": 0.9053794428434198,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004926265711395065,
      "loss": 5.1611,
      "mean_token_accuracy": 0.20753625482320787,
      "num_tokens": 21617988.0,
      "step": 9425
    },
    {
      "entropy": 5.292473602294922,
      "epoch": 0.9058597502401537,
      "grad_norm": 1.1875,
      "learning_rate": 0.000492617863852744,
      "loss": 5.1621,
      "mean_token_accuracy": 0.2052145004272461,
      "num_tokens": 21629832.0,
      "step": 9430
    },
    {
      "entropy": 5.306193113327026,
      "epoch": 0.9063400576368876,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004926091515135409,
      "loss": 5.1667,
      "mean_token_accuracy": 0.19883570820093155,
      "num_tokens": 21641295.0,
      "step": 9435
    },
    {
      "entropy": 5.181912994384765,
      "epoch": 0.9068203650336215,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004926004341220995,
      "loss": 5.0544,
      "mean_token_accuracy": 0.2099272534251213,
      "num_tokens": 21652829.0,
      "step": 9440
    },
    {
      "entropy": 5.373570823669434,
      "epoch": 0.9073006724303554,
      "grad_norm": 1.75,
      "learning_rate": 0.0004925917116786222,
      "loss": 5.2463,
      "mean_token_accuracy": 0.19399139136075974,
      "num_tokens": 21665032.0,
      "step": 9445
    },
    {
      "entropy": 5.167655181884766,
      "epoch": 0.9077809798270894,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004925829841833114,
      "loss": 5.0818,
      "mean_token_accuracy": 0.2102190375328064,
      "num_tokens": 21675814.0,
      "step": 9450
    },
    {
      "entropy": 5.266166400909424,
      "epoch": 0.9082612872238233,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004925742516363699,
      "loss": 5.1556,
      "mean_token_accuracy": 0.20075963884592057,
      "num_tokens": 21687581.0,
      "step": 9455
    },
    {
      "entropy": 5.350229644775391,
      "epoch": 0.9087415946205571,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004925655140380002,
      "loss": 5.172,
      "mean_token_accuracy": 0.19996838122606278,
      "num_tokens": 21700167.0,
      "step": 9460
    },
    {
      "entropy": 5.29394268989563,
      "epoch": 0.909221902017291,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004925567713884054,
      "loss": 5.2204,
      "mean_token_accuracy": 0.20227693915367126,
      "num_tokens": 21712309.0,
      "step": 9465
    },
    {
      "entropy": 5.287266731262207,
      "epoch": 0.909702209414025,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004925480236877884,
      "loss": 5.1425,
      "mean_token_accuracy": 0.20785951018333435,
      "num_tokens": 21723391.0,
      "step": 9470
    },
    {
      "entropy": 5.266707849502564,
      "epoch": 0.9101825168107589,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004925392709363522,
      "loss": 5.0952,
      "mean_token_accuracy": 0.202304208278656,
      "num_tokens": 21735052.0,
      "step": 9475
    },
    {
      "entropy": 5.286109828948975,
      "epoch": 0.9106628242074928,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004925305131343001,
      "loss": 5.2215,
      "mean_token_accuracy": 0.20100320130586624,
      "num_tokens": 21746424.0,
      "step": 9480
    },
    {
      "entropy": 5.30694842338562,
      "epoch": 0.9111431316042267,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004925217502818355,
      "loss": 5.203,
      "mean_token_accuracy": 0.19759567975997924,
      "num_tokens": 21756960.0,
      "step": 9485
    },
    {
      "entropy": 5.333887338638306,
      "epoch": 0.9116234390009607,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004925129823791616,
      "loss": 5.1542,
      "mean_token_accuracy": 0.20748519748449326,
      "num_tokens": 21767621.0,
      "step": 9490
    },
    {
      "entropy": 5.289246940612793,
      "epoch": 0.9121037463976945,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004925042094264822,
      "loss": 5.197,
      "mean_token_accuracy": 0.1969016820192337,
      "num_tokens": 21779403.0,
      "step": 9495
    },
    {
      "entropy": 5.397540426254272,
      "epoch": 0.9125840537944284,
      "grad_norm": 1.203125,
      "learning_rate": 0.000492495431424001,
      "loss": 5.2978,
      "mean_token_accuracy": 0.1891071692109108,
      "num_tokens": 21791039.0,
      "step": 9500
    },
    {
      "entropy": 5.427066373825073,
      "epoch": 0.9130643611911623,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004924866483719216,
      "loss": 5.2836,
      "mean_token_accuracy": 0.19289563000202178,
      "num_tokens": 21803731.0,
      "step": 9505
    },
    {
      "entropy": 5.333277368545533,
      "epoch": 0.9135446685878963,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004924778602704481,
      "loss": 5.1677,
      "mean_token_accuracy": 0.1957067146897316,
      "num_tokens": 21816139.0,
      "step": 9510
    },
    {
      "entropy": 5.140634536743164,
      "epoch": 0.9140249759846302,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004924690671197845,
      "loss": 5.0477,
      "mean_token_accuracy": 0.21424670368433,
      "num_tokens": 21826861.0,
      "step": 9515
    },
    {
      "entropy": 5.280607748031616,
      "epoch": 0.914505283381364,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004924602689201348,
      "loss": 5.1832,
      "mean_token_accuracy": 0.20522396564483641,
      "num_tokens": 21838083.0,
      "step": 9520
    },
    {
      "entropy": 5.335158777236939,
      "epoch": 0.9149855907780979,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004924514656717034,
      "loss": 5.191,
      "mean_token_accuracy": 0.20277170836925507,
      "num_tokens": 21849232.0,
      "step": 9525
    },
    {
      "entropy": 5.338367891311646,
      "epoch": 0.9154658981748319,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004924426573746948,
      "loss": 5.23,
      "mean_token_accuracy": 0.19856053441762925,
      "num_tokens": 21861650.0,
      "step": 9530
    },
    {
      "entropy": 5.423274517059326,
      "epoch": 0.9159462055715658,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004924338440293131,
      "loss": 5.2783,
      "mean_token_accuracy": 0.1954554572701454,
      "num_tokens": 21873369.0,
      "step": 9535
    },
    {
      "entropy": 5.333283948898315,
      "epoch": 0.9164265129682997,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004924250256357635,
      "loss": 5.1724,
      "mean_token_accuracy": 0.20098866075277327,
      "num_tokens": 21884563.0,
      "step": 9540
    },
    {
      "entropy": 5.277515935897827,
      "epoch": 0.9169068203650336,
      "grad_norm": 1.375,
      "learning_rate": 0.0004924162021942502,
      "loss": 5.1874,
      "mean_token_accuracy": 0.20464817434549332,
      "num_tokens": 21896114.0,
      "step": 9545
    },
    {
      "entropy": 5.335600996017456,
      "epoch": 0.9173871277617676,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004924073737049784,
      "loss": 5.1987,
      "mean_token_accuracy": 0.20126154869794846,
      "num_tokens": 21907386.0,
      "step": 9550
    },
    {
      "entropy": 5.342973852157593,
      "epoch": 0.9178674351585014,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004923985401681528,
      "loss": 5.2488,
      "mean_token_accuracy": 0.19889401346445085,
      "num_tokens": 21918312.0,
      "step": 9555
    },
    {
      "entropy": 5.287489891052246,
      "epoch": 0.9183477425552353,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004923897015839788,
      "loss": 5.1358,
      "mean_token_accuracy": 0.20642819702625276,
      "num_tokens": 21930101.0,
      "step": 9560
    },
    {
      "entropy": 5.31551456451416,
      "epoch": 0.9188280499519692,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004923808579526613,
      "loss": 5.2576,
      "mean_token_accuracy": 0.20030594915151595,
      "num_tokens": 21941739.0,
      "step": 9565
    },
    {
      "entropy": 5.329789638519287,
      "epoch": 0.9193083573487032,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004923720092744059,
      "loss": 5.2323,
      "mean_token_accuracy": 0.18889658749103547,
      "num_tokens": 21953342.0,
      "step": 9570
    },
    {
      "entropy": 5.3859357833862305,
      "epoch": 0.9197886647454371,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004923631555494179,
      "loss": 5.2329,
      "mean_token_accuracy": 0.20384127497673035,
      "num_tokens": 21963506.0,
      "step": 9575
    },
    {
      "entropy": 5.376025152206421,
      "epoch": 0.920268972142171,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004923542967779028,
      "loss": 5.2758,
      "mean_token_accuracy": 0.19540261775255202,
      "num_tokens": 21975078.0,
      "step": 9580
    },
    {
      "entropy": 5.2846917629241945,
      "epoch": 0.920749279538905,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004923454329600664,
      "loss": 5.1249,
      "mean_token_accuracy": 0.20521679222583772,
      "num_tokens": 21987128.0,
      "step": 9585
    },
    {
      "entropy": 5.283337163925171,
      "epoch": 0.9212295869356388,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004923365640961143,
      "loss": 5.2331,
      "mean_token_accuracy": 0.2014299139380455,
      "num_tokens": 22000106.0,
      "step": 9590
    },
    {
      "entropy": 5.2900158882141115,
      "epoch": 0.9217098943323727,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004923276901862526,
      "loss": 5.1304,
      "mean_token_accuracy": 0.20454052090644836,
      "num_tokens": 22010571.0,
      "step": 9595
    },
    {
      "entropy": 5.285225486755371,
      "epoch": 0.9221902017291066,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004923188112306874,
      "loss": 5.0869,
      "mean_token_accuracy": 0.21403514891862868,
      "num_tokens": 22021399.0,
      "step": 9600
    },
    {
      "entropy": 5.250886297225952,
      "epoch": 0.9226705091258406,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004923099272296246,
      "loss": 5.0942,
      "mean_token_accuracy": 0.2096669614315033,
      "num_tokens": 22032302.0,
      "step": 9605
    },
    {
      "entropy": 5.223394346237183,
      "epoch": 0.9231508165225745,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004923010381832706,
      "loss": 5.1207,
      "mean_token_accuracy": 0.2084650442004204,
      "num_tokens": 22043762.0,
      "step": 9610
    },
    {
      "entropy": 5.323691415786743,
      "epoch": 0.9236311239193083,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004922921440918318,
      "loss": 5.1569,
      "mean_token_accuracy": 0.205050228536129,
      "num_tokens": 22054885.0,
      "step": 9615
    },
    {
      "entropy": 5.373482990264892,
      "epoch": 0.9241114313160422,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004922832449555144,
      "loss": 5.2458,
      "mean_token_accuracy": 0.19463663548231125,
      "num_tokens": 22066627.0,
      "step": 9620
    },
    {
      "entropy": 5.186635971069336,
      "epoch": 0.9245917387127762,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004922743407745255,
      "loss": 5.0757,
      "mean_token_accuracy": 0.20978489518165588,
      "num_tokens": 22078062.0,
      "step": 9625
    },
    {
      "entropy": 5.275094032287598,
      "epoch": 0.9250720461095101,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004922654315490714,
      "loss": 5.0956,
      "mean_token_accuracy": 0.20137395858764648,
      "num_tokens": 22089465.0,
      "step": 9630
    },
    {
      "entropy": 5.352103042602539,
      "epoch": 0.925552353506244,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004922565172793593,
      "loss": 5.2231,
      "mean_token_accuracy": 0.2037585958838463,
      "num_tokens": 22101688.0,
      "step": 9635
    },
    {
      "entropy": 5.258252811431885,
      "epoch": 0.9260326609029779,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004922475979655958,
      "loss": 5.2052,
      "mean_token_accuracy": 0.20143339931964874,
      "num_tokens": 22112912.0,
      "step": 9640
    },
    {
      "entropy": 5.349901485443115,
      "epoch": 0.9265129682997119,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004922386736079883,
      "loss": 5.3294,
      "mean_token_accuracy": 0.1897306814789772,
      "num_tokens": 22125079.0,
      "step": 9645
    },
    {
      "entropy": 5.391475772857666,
      "epoch": 0.9269932756964457,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004922297442067438,
      "loss": 5.1895,
      "mean_token_accuracy": 0.20708800256252288,
      "num_tokens": 22137146.0,
      "step": 9650
    },
    {
      "entropy": 5.267781352996826,
      "epoch": 0.9274735830931796,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004922208097620697,
      "loss": 5.1425,
      "mean_token_accuracy": 0.19954589307308196,
      "num_tokens": 22148330.0,
      "step": 9655
    },
    {
      "entropy": 5.225883674621582,
      "epoch": 0.9279538904899135,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004922118702741735,
      "loss": 5.1448,
      "mean_token_accuracy": 0.1999584674835205,
      "num_tokens": 22160511.0,
      "step": 9660
    },
    {
      "entropy": 5.26427435874939,
      "epoch": 0.9284341978866475,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004922029257432625,
      "loss": 5.1271,
      "mean_token_accuracy": 0.2007584884762764,
      "num_tokens": 22171298.0,
      "step": 9665
    },
    {
      "entropy": 5.294016265869141,
      "epoch": 0.9289145052833814,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004921939761695446,
      "loss": 5.1792,
      "mean_token_accuracy": 0.20192683935165406,
      "num_tokens": 22181897.0,
      "step": 9670
    },
    {
      "entropy": 5.356138849258423,
      "epoch": 0.9293948126801153,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004921850215532275,
      "loss": 5.2624,
      "mean_token_accuracy": 0.19709415286779403,
      "num_tokens": 22194630.0,
      "step": 9675
    },
    {
      "entropy": 5.315850687026978,
      "epoch": 0.9298751200768491,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004921760618945192,
      "loss": 5.1989,
      "mean_token_accuracy": 0.20254762321710587,
      "num_tokens": 22206041.0,
      "step": 9680
    },
    {
      "entropy": 5.27511830329895,
      "epoch": 0.9303554274735831,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004921670971936276,
      "loss": 5.1871,
      "mean_token_accuracy": 0.1983708769083023,
      "num_tokens": 22218434.0,
      "step": 9685
    },
    {
      "entropy": 5.3494175434112545,
      "epoch": 0.930835734870317,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004921581274507607,
      "loss": 5.1835,
      "mean_token_accuracy": 0.19647131562232972,
      "num_tokens": 22229182.0,
      "step": 9690
    },
    {
      "entropy": 5.370598602294922,
      "epoch": 0.9313160422670509,
      "grad_norm": 1.203125,
      "learning_rate": 0.000492149152666127,
      "loss": 5.2706,
      "mean_token_accuracy": 0.19563601315021514,
      "num_tokens": 22241142.0,
      "step": 9695
    },
    {
      "entropy": 5.354789066314697,
      "epoch": 0.9317963496637848,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004921401728399348,
      "loss": 5.1755,
      "mean_token_accuracy": 0.20050271451473237,
      "num_tokens": 22252695.0,
      "step": 9700
    },
    {
      "entropy": 5.249253463745117,
      "epoch": 0.9322766570605188,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004921311879723926,
      "loss": 5.1104,
      "mean_token_accuracy": 0.20982398241758346,
      "num_tokens": 22264159.0,
      "step": 9705
    },
    {
      "entropy": 5.3121129989624025,
      "epoch": 0.9327569644572526,
      "grad_norm": 1.125,
      "learning_rate": 0.0004921221980637088,
      "loss": 5.1901,
      "mean_token_accuracy": 0.20158101320266725,
      "num_tokens": 22276154.0,
      "step": 9710
    },
    {
      "entropy": 5.248436164855957,
      "epoch": 0.9332372718539865,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004921132031140925,
      "loss": 5.0658,
      "mean_token_accuracy": 0.2020740792155266,
      "num_tokens": 22288328.0,
      "step": 9715
    },
    {
      "entropy": 5.251942539215088,
      "epoch": 0.9337175792507204,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004921042031237521,
      "loss": 5.166,
      "mean_token_accuracy": 0.19952569305896758,
      "num_tokens": 22300154.0,
      "step": 9720
    },
    {
      "entropy": 5.143451261520386,
      "epoch": 0.9341978866474544,
      "grad_norm": 1.25,
      "learning_rate": 0.0004920951980928969,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21996416300535201,
      "num_tokens": 22311176.0,
      "step": 9725
    },
    {
      "entropy": 5.3122765064239506,
      "epoch": 0.9346781940441883,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004920861880217359,
      "loss": 5.1927,
      "mean_token_accuracy": 0.19527371376752853,
      "num_tokens": 22321953.0,
      "step": 9730
    },
    {
      "entropy": 5.335900115966797,
      "epoch": 0.9351585014409222,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004920771729104781,
      "loss": 5.1623,
      "mean_token_accuracy": 0.2025658443570137,
      "num_tokens": 22333152.0,
      "step": 9735
    },
    {
      "entropy": 5.333852386474609,
      "epoch": 0.9356388088376562,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004920681527593329,
      "loss": 5.1738,
      "mean_token_accuracy": 0.20439306348562242,
      "num_tokens": 22344848.0,
      "step": 9740
    },
    {
      "entropy": 5.286120939254761,
      "epoch": 0.93611911623439,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004920591275685098,
      "loss": 5.1159,
      "mean_token_accuracy": 0.2034482702612877,
      "num_tokens": 22355100.0,
      "step": 9745
    },
    {
      "entropy": 5.253074836730957,
      "epoch": 0.9365994236311239,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004920500973382184,
      "loss": 5.141,
      "mean_token_accuracy": 0.1988856852054596,
      "num_tokens": 22366045.0,
      "step": 9750
    },
    {
      "entropy": 5.284864091873169,
      "epoch": 0.9370797310278578,
      "grad_norm": 1.125,
      "learning_rate": 0.0004920410620686682,
      "loss": 5.1314,
      "mean_token_accuracy": 0.1963018298149109,
      "num_tokens": 22378781.0,
      "step": 9755
    },
    {
      "entropy": 5.344362211227417,
      "epoch": 0.9375600384245918,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004920320217600689,
      "loss": 5.2278,
      "mean_token_accuracy": 0.19963358491659164,
      "num_tokens": 22390121.0,
      "step": 9760
    },
    {
      "entropy": 5.250550842285156,
      "epoch": 0.9380403458213257,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004920229764126306,
      "loss": 5.115,
      "mean_token_accuracy": 0.20942254960536957,
      "num_tokens": 22402261.0,
      "step": 9765
    },
    {
      "entropy": 5.276533889770508,
      "epoch": 0.9385206532180596,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004920139260265632,
      "loss": 5.1912,
      "mean_token_accuracy": 0.2040989562869072,
      "num_tokens": 22413939.0,
      "step": 9770
    },
    {
      "entropy": 5.311789655685425,
      "epoch": 0.9390009606147934,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004920048706020769,
      "loss": 5.1048,
      "mean_token_accuracy": 0.20209636241197587,
      "num_tokens": 22425439.0,
      "step": 9775
    },
    {
      "entropy": 5.33395938873291,
      "epoch": 0.9394812680115274,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004919958101393817,
      "loss": 5.2107,
      "mean_token_accuracy": 0.19799000322818755,
      "num_tokens": 22436989.0,
      "step": 9780
    },
    {
      "entropy": 5.283024454116822,
      "epoch": 0.9399615754082613,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004919867446386883,
      "loss": 5.2202,
      "mean_token_accuracy": 0.1953797608613968,
      "num_tokens": 22449181.0,
      "step": 9785
    },
    {
      "entropy": 5.198832035064697,
      "epoch": 0.9404418828049952,
      "grad_norm": 1.171875,
      "learning_rate": 0.000491977674100207,
      "loss": 5.0667,
      "mean_token_accuracy": 0.20685746520757675,
      "num_tokens": 22460592.0,
      "step": 9790
    },
    {
      "entropy": 5.187889099121094,
      "epoch": 0.9409221902017291,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004919685985241483,
      "loss": 5.1104,
      "mean_token_accuracy": 0.2070457324385643,
      "num_tokens": 22471718.0,
      "step": 9795
    },
    {
      "entropy": 5.318658971786499,
      "epoch": 0.9414024975984631,
      "grad_norm": 1.265625,
      "learning_rate": 0.000491959517910723,
      "loss": 5.1479,
      "mean_token_accuracy": 0.21063447296619414,
      "num_tokens": 22482413.0,
      "step": 9800
    },
    {
      "entropy": 5.236535739898682,
      "epoch": 0.9418828049951969,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004919504322601421,
      "loss": 5.1082,
      "mean_token_accuracy": 0.20763958543539046,
      "num_tokens": 22494016.0,
      "step": 9805
    },
    {
      "entropy": 5.26602611541748,
      "epoch": 0.9423631123919308,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004919413415726162,
      "loss": 5.1832,
      "mean_token_accuracy": 0.19661634266376496,
      "num_tokens": 22505015.0,
      "step": 9810
    },
    {
      "entropy": 5.315104293823242,
      "epoch": 0.9428434197886647,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004919322458483566,
      "loss": 5.2113,
      "mean_token_accuracy": 0.2064582511782646,
      "num_tokens": 22516691.0,
      "step": 9815
    },
    {
      "entropy": 5.280817222595215,
      "epoch": 0.9433237271853987,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004919231450875745,
      "loss": 5.1548,
      "mean_token_accuracy": 0.2063862308859825,
      "num_tokens": 22527740.0,
      "step": 9820
    },
    {
      "entropy": 5.2642027854919435,
      "epoch": 0.9438040345821326,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004919140392904809,
      "loss": 5.1419,
      "mean_token_accuracy": 0.20387705117464067,
      "num_tokens": 22537685.0,
      "step": 9825
    },
    {
      "entropy": 5.26283540725708,
      "epoch": 0.9442843419788665,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004919049284572875,
      "loss": 5.0685,
      "mean_token_accuracy": 0.21276348382234572,
      "num_tokens": 22548282.0,
      "step": 9830
    },
    {
      "entropy": 5.157790803909302,
      "epoch": 0.9447646493756003,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004918958125882058,
      "loss": 5.0137,
      "mean_token_accuracy": 0.20991793572902678,
      "num_tokens": 22559277.0,
      "step": 9835
    },
    {
      "entropy": 5.2666724681854244,
      "epoch": 0.9452449567723343,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004918866916834474,
      "loss": 5.1585,
      "mean_token_accuracy": 0.2033451661467552,
      "num_tokens": 22570664.0,
      "step": 9840
    },
    {
      "entropy": 5.257972288131714,
      "epoch": 0.9457252641690682,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004918775657432239,
      "loss": 5.0912,
      "mean_token_accuracy": 0.21164929270744323,
      "num_tokens": 22580410.0,
      "step": 9845
    },
    {
      "entropy": 5.232131433486939,
      "epoch": 0.9462055715658021,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004918684347677474,
      "loss": 5.1585,
      "mean_token_accuracy": 0.2049712911248207,
      "num_tokens": 22591630.0,
      "step": 9850
    },
    {
      "entropy": 5.264332008361817,
      "epoch": 0.946685878962536,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004918592987572298,
      "loss": 5.156,
      "mean_token_accuracy": 0.20031468272209169,
      "num_tokens": 22603658.0,
      "step": 9855
    },
    {
      "entropy": 5.316908931732177,
      "epoch": 0.94716618635927,
      "grad_norm": 1.25,
      "learning_rate": 0.0004918501577118832,
      "loss": 5.189,
      "mean_token_accuracy": 0.1919010505080223,
      "num_tokens": 22614812.0,
      "step": 9860
    },
    {
      "entropy": 5.346418571472168,
      "epoch": 0.9476464937560038,
      "grad_norm": 1.3125,
      "learning_rate": 0.00049184101163192,
      "loss": 5.1288,
      "mean_token_accuracy": 0.2017146721482277,
      "num_tokens": 22626924.0,
      "step": 9865
    },
    {
      "entropy": 5.305496597290039,
      "epoch": 0.9481268011527377,
      "grad_norm": 1.25,
      "learning_rate": 0.0004918318605175522,
      "loss": 5.1725,
      "mean_token_accuracy": 0.19936136901378632,
      "num_tokens": 22638715.0,
      "step": 9870
    },
    {
      "entropy": 5.364515972137451,
      "epoch": 0.9486071085494716,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004918227043689924,
      "loss": 5.2976,
      "mean_token_accuracy": 0.18901295512914656,
      "num_tokens": 22650257.0,
      "step": 9875
    },
    {
      "entropy": 5.273727941513061,
      "epoch": 0.9490874159462056,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004918135431864534,
      "loss": 5.1591,
      "mean_token_accuracy": 0.1968395531177521,
      "num_tokens": 22662480.0,
      "step": 9880
    },
    {
      "entropy": 5.2737961769104,
      "epoch": 0.9495677233429395,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004918043769701478,
      "loss": 5.1612,
      "mean_token_accuracy": 0.1973642513155937,
      "num_tokens": 22673927.0,
      "step": 9885
    },
    {
      "entropy": 5.418401765823364,
      "epoch": 0.9500480307396734,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004917952057202882,
      "loss": 5.2134,
      "mean_token_accuracy": 0.2050120547413826,
      "num_tokens": 22684670.0,
      "step": 9890
    },
    {
      "entropy": 5.315881538391113,
      "epoch": 0.9505283381364072,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004917860294370877,
      "loss": 5.2735,
      "mean_token_accuracy": 0.20009334683418273,
      "num_tokens": 22696571.0,
      "step": 9895
    },
    {
      "entropy": 5.2041833877563475,
      "epoch": 0.9510086455331412,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004917768481207593,
      "loss": 5.1166,
      "mean_token_accuracy": 0.208723047375679,
      "num_tokens": 22708987.0,
      "step": 9900
    },
    {
      "entropy": 5.307413864135742,
      "epoch": 0.9514889529298751,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004917676617715162,
      "loss": 5.1962,
      "mean_token_accuracy": 0.20884546637535095,
      "num_tokens": 22720769.0,
      "step": 9905
    },
    {
      "entropy": 5.3030843257904055,
      "epoch": 0.951969260326609,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004917584703895717,
      "loss": 5.1164,
      "mean_token_accuracy": 0.2086371064186096,
      "num_tokens": 22732592.0,
      "step": 9910
    },
    {
      "entropy": 5.306847810745239,
      "epoch": 0.952449567723343,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004917492739751391,
      "loss": 5.1529,
      "mean_token_accuracy": 0.19614050388336182,
      "num_tokens": 22744440.0,
      "step": 9915
    },
    {
      "entropy": 5.233251237869263,
      "epoch": 0.9529298751200769,
      "grad_norm": 1.203125,
      "learning_rate": 0.000491740072528432,
      "loss": 5.1429,
      "mean_token_accuracy": 0.20407471507787706,
      "num_tokens": 22756336.0,
      "step": 9920
    },
    {
      "entropy": 5.257513236999512,
      "epoch": 0.9534101825168108,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000491730866049664,
      "loss": 5.163,
      "mean_token_accuracy": 0.20242875814437866,
      "num_tokens": 22768256.0,
      "step": 9925
    },
    {
      "entropy": 5.3625304222106935,
      "epoch": 0.9538904899135446,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004917216545390489,
      "loss": 5.1934,
      "mean_token_accuracy": 0.19828447699546814,
      "num_tokens": 22779693.0,
      "step": 9930
    },
    {
      "entropy": 5.181369018554688,
      "epoch": 0.9543707973102786,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004917124379968004,
      "loss": 5.0266,
      "mean_token_accuracy": 0.20504971891641616,
      "num_tokens": 22791417.0,
      "step": 9935
    },
    {
      "entropy": 5.292864990234375,
      "epoch": 0.9548511047070125,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004917032164231327,
      "loss": 5.13,
      "mean_token_accuracy": 0.20741121768951415,
      "num_tokens": 22803541.0,
      "step": 9940
    },
    {
      "entropy": 5.2767088413238525,
      "epoch": 0.9553314121037464,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004916939898182598,
      "loss": 5.1651,
      "mean_token_accuracy": 0.20157254338264466,
      "num_tokens": 22814648.0,
      "step": 9945
    },
    {
      "entropy": 5.27077088356018,
      "epoch": 0.9558117195004803,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004916847581823958,
      "loss": 5.182,
      "mean_token_accuracy": 0.20345498323440553,
      "num_tokens": 22825672.0,
      "step": 9950
    },
    {
      "entropy": 5.278829050064087,
      "epoch": 0.9562920268972143,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004916755215157552,
      "loss": 5.1281,
      "mean_token_accuracy": 0.19974546879529953,
      "num_tokens": 22837663.0,
      "step": 9955
    },
    {
      "entropy": 5.180472326278687,
      "epoch": 0.9567723342939481,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004916662798185524,
      "loss": 5.0379,
      "mean_token_accuracy": 0.20684915035963058,
      "num_tokens": 22849531.0,
      "step": 9960
    },
    {
      "entropy": 5.29015531539917,
      "epoch": 0.957252641690682,
      "grad_norm": 1.125,
      "learning_rate": 0.0004916570330910019,
      "loss": 5.1899,
      "mean_token_accuracy": 0.19755279123783112,
      "num_tokens": 22861784.0,
      "step": 9965
    },
    {
      "entropy": 5.377617979049683,
      "epoch": 0.9577329490874159,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004916477813333185,
      "loss": 5.199,
      "mean_token_accuracy": 0.19778556376695633,
      "num_tokens": 22873482.0,
      "step": 9970
    },
    {
      "entropy": 5.2757673263549805,
      "epoch": 0.9582132564841499,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004916385245457168,
      "loss": 5.1126,
      "mean_token_accuracy": 0.2054372861981392,
      "num_tokens": 22883917.0,
      "step": 9975
    },
    {
      "entropy": 5.265840911865235,
      "epoch": 0.9586935638808838,
      "grad_norm": 1.140625,
      "learning_rate": 0.000491629262728412,
      "loss": 5.1588,
      "mean_token_accuracy": 0.20182005763053895,
      "num_tokens": 22896167.0,
      "step": 9980
    },
    {
      "entropy": 5.3045097351074215,
      "epoch": 0.9591738712776177,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004916199958816188,
      "loss": 5.1866,
      "mean_token_accuracy": 0.20387232601642608,
      "num_tokens": 22908350.0,
      "step": 9985
    },
    {
      "entropy": 5.433508348464966,
      "epoch": 0.9596541786743515,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004916107240055527,
      "loss": 5.2677,
      "mean_token_accuracy": 0.20209979861974717,
      "num_tokens": 22918913.0,
      "step": 9990
    },
    {
      "entropy": 5.154075622558594,
      "epoch": 0.9601344860710855,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004916014471004287,
      "loss": 5.1004,
      "mean_token_accuracy": 0.20652660876512527,
      "num_tokens": 22929336.0,
      "step": 9995
    },
    {
      "entropy": 5.233493375778198,
      "epoch": 0.9606147934678194,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004915921651664622,
      "loss": 5.1134,
      "mean_token_accuracy": 0.20148668438196182,
      "num_tokens": 22940406.0,
      "step": 10000
    },
    {
      "entropy": 5.255504751205445,
      "epoch": 0.9610951008645533,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000491582878203869,
      "loss": 5.0985,
      "mean_token_accuracy": 0.2031227871775627,
      "num_tokens": 22951518.0,
      "step": 10005
    },
    {
      "entropy": 5.169245910644531,
      "epoch": 0.9615754082612872,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004915735862128643,
      "loss": 5.0833,
      "mean_token_accuracy": 0.20790803879499437,
      "num_tokens": 22962309.0,
      "step": 10010
    },
    {
      "entropy": 5.273991632461548,
      "epoch": 0.9620557156580212,
      "grad_norm": 1.375,
      "learning_rate": 0.0004915642891936641,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20749742090702056,
      "num_tokens": 22973996.0,
      "step": 10015
    },
    {
      "entropy": 5.34439697265625,
      "epoch": 0.962536023054755,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004915549871464841,
      "loss": 5.2362,
      "mean_token_accuracy": 0.2021578460931778,
      "num_tokens": 22984964.0,
      "step": 10020
    },
    {
      "entropy": 5.3410313606262205,
      "epoch": 0.9630163304514889,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004915456800715403,
      "loss": 5.1982,
      "mean_token_accuracy": 0.20068050026893616,
      "num_tokens": 22996058.0,
      "step": 10025
    },
    {
      "entropy": 5.2164530754089355,
      "epoch": 0.9634966378482228,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000491536367969049,
      "loss": 5.0356,
      "mean_token_accuracy": 0.20966239273548126,
      "num_tokens": 23006140.0,
      "step": 10030
    },
    {
      "entropy": 5.271325731277466,
      "epoch": 0.9639769452449568,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004915270508392261,
      "loss": 5.2396,
      "mean_token_accuracy": 0.19378947019577025,
      "num_tokens": 23018129.0,
      "step": 10035
    },
    {
      "entropy": 5.314516592025757,
      "epoch": 0.9644572526416907,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000491517728682288,
      "loss": 5.0614,
      "mean_token_accuracy": 0.2116171583533287,
      "num_tokens": 23028673.0,
      "step": 10040
    },
    {
      "entropy": 5.215683507919311,
      "epoch": 0.9649375600384246,
      "grad_norm": 1.125,
      "learning_rate": 0.0004915084014984512,
      "loss": 5.1048,
      "mean_token_accuracy": 0.20646367222070694,
      "num_tokens": 23040778.0,
      "step": 10045
    },
    {
      "entropy": 5.2242467403411865,
      "epoch": 0.9654178674351584,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004914990692879322,
      "loss": 5.1351,
      "mean_token_accuracy": 0.21372740268707274,
      "num_tokens": 23052170.0,
      "step": 10050
    },
    {
      "entropy": 5.261769151687622,
      "epoch": 0.9658981748318924,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004914897320509478,
      "loss": 5.1258,
      "mean_token_accuracy": 0.20331106930971146,
      "num_tokens": 23063455.0,
      "step": 10055
    },
    {
      "entropy": 5.187658786773682,
      "epoch": 0.9663784822286263,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004914803897877146,
      "loss": 5.0283,
      "mean_token_accuracy": 0.2141657903790474,
      "num_tokens": 23075009.0,
      "step": 10060
    },
    {
      "entropy": 5.2808253288269045,
      "epoch": 0.9668587896253602,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004914710424984495,
      "loss": 5.15,
      "mean_token_accuracy": 0.20173825323581696,
      "num_tokens": 23085921.0,
      "step": 10065
    },
    {
      "entropy": 5.379551076889038,
      "epoch": 0.9673390970220941,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004914616901833696,
      "loss": 5.2769,
      "mean_token_accuracy": 0.1893296718597412,
      "num_tokens": 23098181.0,
      "step": 10070
    },
    {
      "entropy": 5.210932683944702,
      "epoch": 0.9678194044188281,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000491452332842692,
      "loss": 5.1168,
      "mean_token_accuracy": 0.2042910113930702,
      "num_tokens": 23109904.0,
      "step": 10075
    },
    {
      "entropy": 5.282880735397339,
      "epoch": 0.968299711815562,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000491442970476634,
      "loss": 5.2102,
      "mean_token_accuracy": 0.19747001975774764,
      "num_tokens": 23122356.0,
      "step": 10080
    },
    {
      "entropy": 5.342620754241944,
      "epoch": 0.9687800192122958,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004914336030854129,
      "loss": 5.1118,
      "mean_token_accuracy": 0.20807668417692185,
      "num_tokens": 23134055.0,
      "step": 10085
    },
    {
      "entropy": 5.220643091201782,
      "epoch": 0.9692603266090298,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004914242306692461,
      "loss": 5.0546,
      "mean_token_accuracy": 0.21157983988523482,
      "num_tokens": 23145388.0,
      "step": 10090
    },
    {
      "entropy": 5.2309699058532715,
      "epoch": 0.9697406340057637,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004914148532283516,
      "loss": 5.1854,
      "mean_token_accuracy": 0.20171017050743104,
      "num_tokens": 23157083.0,
      "step": 10095
    },
    {
      "entropy": 5.276914501190186,
      "epoch": 0.9702209414024976,
      "grad_norm": 1.25,
      "learning_rate": 0.0004914054707629466,
      "loss": 5.121,
      "mean_token_accuracy": 0.2065420225262642,
      "num_tokens": 23169154.0,
      "step": 10100
    },
    {
      "entropy": 5.22411732673645,
      "epoch": 0.9707012487992315,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004913960832732493,
      "loss": 5.0352,
      "mean_token_accuracy": 0.20972169637680055,
      "num_tokens": 23179879.0,
      "step": 10105
    },
    {
      "entropy": 5.269572401046753,
      "epoch": 0.9711815561959655,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004913866907594774,
      "loss": 5.1381,
      "mean_token_accuracy": 0.21003148853778839,
      "num_tokens": 23190632.0,
      "step": 10110
    },
    {
      "entropy": 5.3404319286346436,
      "epoch": 0.9716618635926993,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004913772932218491,
      "loss": 5.1942,
      "mean_token_accuracy": 0.19655626863241196,
      "num_tokens": 23202060.0,
      "step": 10115
    },
    {
      "entropy": 5.270214700698853,
      "epoch": 0.9721421709894332,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004913678906605825,
      "loss": 5.1722,
      "mean_token_accuracy": 0.1959027722477913,
      "num_tokens": 23214103.0,
      "step": 10120
    },
    {
      "entropy": 5.2885565757751465,
      "epoch": 0.9726224783861671,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004913584830758961,
      "loss": 5.1314,
      "mean_token_accuracy": 0.20674732476472854,
      "num_tokens": 23224676.0,
      "step": 10125
    },
    {
      "entropy": 5.283637619018554,
      "epoch": 0.9731027857829011,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004913490704680081,
      "loss": 5.1484,
      "mean_token_accuracy": 0.2089652106165886,
      "num_tokens": 23235478.0,
      "step": 10130
    },
    {
      "entropy": 5.29101529121399,
      "epoch": 0.973583093179635,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004913396528371371,
      "loss": 5.1968,
      "mean_token_accuracy": 0.2033202901482582,
      "num_tokens": 23246913.0,
      "step": 10135
    },
    {
      "entropy": 5.299480628967285,
      "epoch": 0.9740634005763689,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004913302301835018,
      "loss": 5.1681,
      "mean_token_accuracy": 0.2054528683423996,
      "num_tokens": 23258472.0,
      "step": 10140
    },
    {
      "entropy": 5.350821924209595,
      "epoch": 0.9745437079731027,
      "grad_norm": 1.75,
      "learning_rate": 0.000491320802507321,
      "loss": 5.2531,
      "mean_token_accuracy": 0.19590283781290055,
      "num_tokens": 23270443.0,
      "step": 10145
    },
    {
      "entropy": 5.296839094161987,
      "epoch": 0.9750240153698367,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004913113698088133,
      "loss": 5.1179,
      "mean_token_accuracy": 0.20833492577075957,
      "num_tokens": 23281199.0,
      "step": 10150
    },
    {
      "entropy": 5.223940229415893,
      "epoch": 0.9755043227665706,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000491301932088198,
      "loss": 5.1042,
      "mean_token_accuracy": 0.20876264274120332,
      "num_tokens": 23291980.0,
      "step": 10155
    },
    {
      "entropy": 5.361666345596314,
      "epoch": 0.9759846301633045,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004912924893456942,
      "loss": 5.2188,
      "mean_token_accuracy": 0.20530154407024384,
      "num_tokens": 23303673.0,
      "step": 10160
    },
    {
      "entropy": 5.414979934692383,
      "epoch": 0.9764649375600384,
      "grad_norm": 1.375,
      "learning_rate": 0.000491283041581521,
      "loss": 5.1786,
      "mean_token_accuracy": 0.20272355526685715,
      "num_tokens": 23315830.0,
      "step": 10165
    },
    {
      "entropy": 5.300431537628174,
      "epoch": 0.9769452449567724,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004912735887958978,
      "loss": 5.2114,
      "mean_token_accuracy": 0.1945984348654747,
      "num_tokens": 23329860.0,
      "step": 10170
    },
    {
      "entropy": 5.2571783542633055,
      "epoch": 0.9774255523535063,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004912641309890441,
      "loss": 5.1203,
      "mean_token_accuracy": 0.21061473935842515,
      "num_tokens": 23340610.0,
      "step": 10175
    },
    {
      "entropy": 5.281990909576416,
      "epoch": 0.9779058597502401,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004912546681611794,
      "loss": 5.1319,
      "mean_token_accuracy": 0.2027701184153557,
      "num_tokens": 23352153.0,
      "step": 10180
    },
    {
      "entropy": 5.2378288269042965,
      "epoch": 0.978386167146974,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004912452003125234,
      "loss": 5.0898,
      "mean_token_accuracy": 0.20687327831983565,
      "num_tokens": 23363630.0,
      "step": 10185
    },
    {
      "entropy": 5.282939529418945,
      "epoch": 0.978866474543708,
      "grad_norm": 1.15625,
      "learning_rate": 0.000491235727443296,
      "loss": 5.103,
      "mean_token_accuracy": 0.20540405958890914,
      "num_tokens": 23375922.0,
      "step": 10190
    },
    {
      "entropy": 5.231866836547852,
      "epoch": 0.9793467819404419,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004912262495537171,
      "loss": 5.0696,
      "mean_token_accuracy": 0.2041434481739998,
      "num_tokens": 23386446.0,
      "step": 10195
    },
    {
      "entropy": 5.196145248413086,
      "epoch": 0.9798270893371758,
      "grad_norm": 1.25,
      "learning_rate": 0.0004912167666440068,
      "loss": 5.029,
      "mean_token_accuracy": 0.2104156345129013,
      "num_tokens": 23398458.0,
      "step": 10200
    },
    {
      "entropy": 5.221741008758545,
      "epoch": 0.9803073967339097,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004912072787143852,
      "loss": 5.1202,
      "mean_token_accuracy": 0.20207154750823975,
      "num_tokens": 23409725.0,
      "step": 10205
    },
    {
      "entropy": 5.2421143531799315,
      "epoch": 0.9807877041306436,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004911977857650725,
      "loss": 5.169,
      "mean_token_accuracy": 0.20189805775880815,
      "num_tokens": 23420468.0,
      "step": 10210
    },
    {
      "entropy": 5.280211114883423,
      "epoch": 0.9812680115273775,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004911882877962893,
      "loss": 5.1396,
      "mean_token_accuracy": 0.20175901502370835,
      "num_tokens": 23431386.0,
      "step": 10215
    },
    {
      "entropy": 5.342608022689819,
      "epoch": 0.9817483189241114,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004911787848082559,
      "loss": 5.1661,
      "mean_token_accuracy": 0.19996436685323715,
      "num_tokens": 23442759.0,
      "step": 10220
    },
    {
      "entropy": 5.261898469924927,
      "epoch": 0.9822286263208453,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004911692768011931,
      "loss": 5.1088,
      "mean_token_accuracy": 0.20783044248819352,
      "num_tokens": 23454372.0,
      "step": 10225
    },
    {
      "entropy": 5.319669723510742,
      "epoch": 0.9827089337175793,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004911597637753217,
      "loss": 5.1792,
      "mean_token_accuracy": 0.20771663784980773,
      "num_tokens": 23465557.0,
      "step": 10230
    },
    {
      "entropy": 5.1974317073822025,
      "epoch": 0.9831892411143132,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004911502457308623,
      "loss": 5.0038,
      "mean_token_accuracy": 0.21003925949335098,
      "num_tokens": 23475850.0,
      "step": 10235
    },
    {
      "entropy": 5.185853767395019,
      "epoch": 0.983669548511047,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000491140722668036,
      "loss": 5.0469,
      "mean_token_accuracy": 0.2061360627412796,
      "num_tokens": 23487145.0,
      "step": 10240
    },
    {
      "entropy": 5.232642316818238,
      "epoch": 0.984149855907781,
      "grad_norm": 1.15625,
      "learning_rate": 0.000491131194587064,
      "loss": 5.1513,
      "mean_token_accuracy": 0.19824368953704835,
      "num_tokens": 23498112.0,
      "step": 10245
    },
    {
      "entropy": 5.239406776428223,
      "epoch": 0.9846301633045149,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004911216614881675,
      "loss": 5.0576,
      "mean_token_accuracy": 0.2057785287499428,
      "num_tokens": 23509865.0,
      "step": 10250
    },
    {
      "entropy": 5.339295864105225,
      "epoch": 0.9851104707012488,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004911121233715677,
      "loss": 5.2424,
      "mean_token_accuracy": 0.19286640882492065,
      "num_tokens": 23521153.0,
      "step": 10255
    },
    {
      "entropy": 5.3234961986541744,
      "epoch": 0.9855907780979827,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004911025802374861,
      "loss": 5.2708,
      "mean_token_accuracy": 0.20124376714229583,
      "num_tokens": 23532372.0,
      "step": 10260
    },
    {
      "entropy": 5.323987150192261,
      "epoch": 0.9860710854947167,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004910930320861442,
      "loss": 5.2221,
      "mean_token_accuracy": 0.1999245971441269,
      "num_tokens": 23543494.0,
      "step": 10265
    },
    {
      "entropy": 5.30865707397461,
      "epoch": 0.9865513928914506,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004910834789177639,
      "loss": 5.137,
      "mean_token_accuracy": 0.2011390894651413,
      "num_tokens": 23554020.0,
      "step": 10270
    },
    {
      "entropy": 5.366797685623169,
      "epoch": 0.9870317002881844,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004910739207325668,
      "loss": 5.2929,
      "mean_token_accuracy": 0.19569133520126342,
      "num_tokens": 23565032.0,
      "step": 10275
    },
    {
      "entropy": 5.244882345199585,
      "epoch": 0.9875120076849183,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004910643575307749,
      "loss": 5.1561,
      "mean_token_accuracy": 0.20377912372350693,
      "num_tokens": 23576243.0,
      "step": 10280
    },
    {
      "entropy": 5.356347990036011,
      "epoch": 0.9879923150816523,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004910547893126102,
      "loss": 5.1994,
      "mean_token_accuracy": 0.20131476521492003,
      "num_tokens": 23586111.0,
      "step": 10285
    },
    {
      "entropy": 5.3763329029083256,
      "epoch": 0.9884726224783862,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004910452160782948,
      "loss": 5.2275,
      "mean_token_accuracy": 0.19981012493371964,
      "num_tokens": 23597718.0,
      "step": 10290
    },
    {
      "entropy": 5.327854061126709,
      "epoch": 0.9889529298751201,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000491035637828051,
      "loss": 5.2005,
      "mean_token_accuracy": 0.19184060543775558,
      "num_tokens": 23608241.0,
      "step": 10295
    },
    {
      "entropy": 5.347801780700683,
      "epoch": 0.989433237271854,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004910260545621012,
      "loss": 5.1912,
      "mean_token_accuracy": 0.1934354394674301,
      "num_tokens": 23619120.0,
      "step": 10300
    },
    {
      "entropy": 5.2663328647613525,
      "epoch": 0.9899135446685879,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004910164662806679,
      "loss": 5.0446,
      "mean_token_accuracy": 0.2109780102968216,
      "num_tokens": 23630642.0,
      "step": 10305
    },
    {
      "entropy": 5.242478609085083,
      "epoch": 0.9903938520653218,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004910068729839736,
      "loss": 5.0974,
      "mean_token_accuracy": 0.20741959512233735,
      "num_tokens": 23641987.0,
      "step": 10310
    },
    {
      "entropy": 5.3318517208099365,
      "epoch": 0.9908741594620557,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004909972746722413,
      "loss": 5.2619,
      "mean_token_accuracy": 0.1974133312702179,
      "num_tokens": 23653144.0,
      "step": 10315
    },
    {
      "entropy": 5.320223331451416,
      "epoch": 0.9913544668587896,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004909876713456935,
      "loss": 5.1657,
      "mean_token_accuracy": 0.20280012339353562,
      "num_tokens": 23663807.0,
      "step": 10320
    },
    {
      "entropy": 5.278014993667602,
      "epoch": 0.9918347742555236,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004909780630045534,
      "loss": 5.0933,
      "mean_token_accuracy": 0.20209735482931138,
      "num_tokens": 23674341.0,
      "step": 10325
    },
    {
      "entropy": 5.267170715332031,
      "epoch": 0.9923150816522575,
      "grad_norm": 1.1875,
      "learning_rate": 0.000490968449649044,
      "loss": 5.207,
      "mean_token_accuracy": 0.2007220134139061,
      "num_tokens": 23685415.0,
      "step": 10330
    },
    {
      "entropy": 5.332669401168824,
      "epoch": 0.9927953890489913,
      "grad_norm": 2.15625,
      "learning_rate": 0.0004909588312793884,
      "loss": 5.1863,
      "mean_token_accuracy": 0.20597185343503951,
      "num_tokens": 23696652.0,
      "step": 10335
    },
    {
      "entropy": 5.315309810638428,
      "epoch": 0.9932756964457252,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004909492078958101,
      "loss": 5.1147,
      "mean_token_accuracy": 0.20756062269210815,
      "num_tokens": 23708939.0,
      "step": 10340
    },
    {
      "entropy": 5.277033090591431,
      "epoch": 0.9937560038424592,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004909395794985324,
      "loss": 5.1219,
      "mean_token_accuracy": 0.20166240334510804,
      "num_tokens": 23720587.0,
      "step": 10345
    },
    {
      "entropy": 5.18465633392334,
      "epoch": 0.9942363112391931,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004909299460877788,
      "loss": 5.0915,
      "mean_token_accuracy": 0.20672692507505416,
      "num_tokens": 23732345.0,
      "step": 10350
    },
    {
      "entropy": 5.289073896408081,
      "epoch": 0.994716618635927,
      "grad_norm": 1.25,
      "learning_rate": 0.0004909203076637732,
      "loss": 5.1381,
      "mean_token_accuracy": 0.19841207563877106,
      "num_tokens": 23743149.0,
      "step": 10355
    },
    {
      "entropy": 5.378394365310669,
      "epoch": 0.9951969260326609,
      "grad_norm": 1.375,
      "learning_rate": 0.0004909106642267392,
      "loss": 5.1794,
      "mean_token_accuracy": 0.20339563637971877,
      "num_tokens": 23755508.0,
      "step": 10360
    },
    {
      "entropy": 5.263547658920288,
      "epoch": 0.9956772334293948,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004909010157769006,
      "loss": 5.0412,
      "mean_token_accuracy": 0.1998227596282959,
      "num_tokens": 23766313.0,
      "step": 10365
    },
    {
      "entropy": 5.181618118286133,
      "epoch": 0.9961575408261287,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004908913623144814,
      "loss": 5.0349,
      "mean_token_accuracy": 0.21776023209095002,
      "num_tokens": 23777776.0,
      "step": 10370
    },
    {
      "entropy": 5.289537763595581,
      "epoch": 0.9966378482228626,
      "grad_norm": 1.21875,
      "learning_rate": 0.000490881703839706,
      "loss": 5.2086,
      "mean_token_accuracy": 0.2042216256260872,
      "num_tokens": 23791143.0,
      "step": 10375
    },
    {
      "entropy": 5.3995143413543705,
      "epoch": 0.9971181556195965,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004908720403527984,
      "loss": 5.1692,
      "mean_token_accuracy": 0.20047396123409272,
      "num_tokens": 23802446.0,
      "step": 10380
    },
    {
      "entropy": 5.136845684051513,
      "epoch": 0.9975984630163305,
      "grad_norm": 1.3125,
      "learning_rate": 0.000490862371853983,
      "loss": 5.0426,
      "mean_token_accuracy": 0.21330038458108902,
      "num_tokens": 23813757.0,
      "step": 10385
    },
    {
      "entropy": 5.232293367385864,
      "epoch": 0.9980787704130644,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004908526983434844,
      "loss": 5.1313,
      "mean_token_accuracy": 0.20587840676307678,
      "num_tokens": 23824934.0,
      "step": 10390
    },
    {
      "entropy": 5.416840505599976,
      "epoch": 0.9985590778097982,
      "grad_norm": 1.3125,
      "learning_rate": 0.000490843019821527,
      "loss": 5.2461,
      "mean_token_accuracy": 0.19968908429145812,
      "num_tokens": 23836872.0,
      "step": 10395
    },
    {
      "entropy": 5.360136032104492,
      "epoch": 0.9990393852065321,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004908333362883358,
      "loss": 5.1446,
      "mean_token_accuracy": 0.1978767365217209,
      "num_tokens": 23847853.0,
      "step": 10400
    },
    {
      "entropy": 5.2001087188720705,
      "epoch": 0.9995196926032661,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004908236477441353,
      "loss": 5.0656,
      "mean_token_accuracy": 0.21127649694681166,
      "num_tokens": 23859208.0,
      "step": 10405
    },
    {
      "entropy": 5.192113780975342,
      "epoch": 1.0,
      "grad_norm": 1.9453125,
      "learning_rate": 0.0004908139541891505,
      "loss": 5.0423,
      "mean_token_accuracy": 0.20258355736732483,
      "num_tokens": 23868536.0,
      "step": 10410
    },
    {
      "entropy": 5.268091773986816,
      "epoch": 1.0004803073967339,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004908042556236066,
      "loss": 5.0212,
      "mean_token_accuracy": 0.2069282665848732,
      "num_tokens": 23881335.0,
      "step": 10415
    },
    {
      "entropy": 5.189616250991821,
      "epoch": 1.0009606147934678,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004907945520477286,
      "loss": 4.9664,
      "mean_token_accuracy": 0.21255423575639726,
      "num_tokens": 23891694.0,
      "step": 10420
    },
    {
      "entropy": 5.185774040222168,
      "epoch": 1.0014409221902016,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004907848434617419,
      "loss": 5.0318,
      "mean_token_accuracy": 0.20579083263874054,
      "num_tokens": 23903560.0,
      "step": 10425
    },
    {
      "entropy": 5.414279794692993,
      "epoch": 1.0019212295869357,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000490775129865872,
      "loss": 5.236,
      "mean_token_accuracy": 0.19321727156639099,
      "num_tokens": 23914525.0,
      "step": 10430
    },
    {
      "entropy": 5.323701810836792,
      "epoch": 1.0024015369836696,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004907654112603442,
      "loss": 5.0911,
      "mean_token_accuracy": 0.21004375964403152,
      "num_tokens": 23926492.0,
      "step": 10435
    },
    {
      "entropy": 5.312917995452881,
      "epoch": 1.0028818443804035,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004907556876453843,
      "loss": 5.0633,
      "mean_token_accuracy": 0.2073927566409111,
      "num_tokens": 23936326.0,
      "step": 10440
    },
    {
      "entropy": 5.085734605789185,
      "epoch": 1.0033621517771374,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000490745959021218,
      "loss": 4.8691,
      "mean_token_accuracy": 0.22059814333915712,
      "num_tokens": 23947173.0,
      "step": 10445
    },
    {
      "entropy": 5.240332174301147,
      "epoch": 1.0038424591738713,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004907362253880711,
      "loss": 5.0436,
      "mean_token_accuracy": 0.20943649113178253,
      "num_tokens": 23959771.0,
      "step": 10450
    },
    {
      "entropy": 5.2468465805053714,
      "epoch": 1.0043227665706052,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004907264867461697,
      "loss": 5.1239,
      "mean_token_accuracy": 0.2045529752969742,
      "num_tokens": 23970908.0,
      "step": 10455
    },
    {
      "entropy": 5.2241455078125,
      "epoch": 1.004803073967339,
      "grad_norm": 2.375,
      "learning_rate": 0.0004907167430957399,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20875274240970612,
      "num_tokens": 23981640.0,
      "step": 10460
    },
    {
      "entropy": 5.295232725143433,
      "epoch": 1.005283381364073,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004907069944370077,
      "loss": 5.0705,
      "mean_token_accuracy": 0.2090965822339058,
      "num_tokens": 23992913.0,
      "step": 10465
    },
    {
      "entropy": 5.253322887420654,
      "epoch": 1.005763688760807,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004906972407701998,
      "loss": 5.0396,
      "mean_token_accuracy": 0.20532276779413222,
      "num_tokens": 24003819.0,
      "step": 10470
    },
    {
      "entropy": 5.297573661804199,
      "epoch": 1.006243996157541,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004906874820955423,
      "loss": 5.0672,
      "mean_token_accuracy": 0.20593006312847137,
      "num_tokens": 24013997.0,
      "step": 10475
    },
    {
      "entropy": 5.244946527481079,
      "epoch": 1.0067243035542748,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004906777184132621,
      "loss": 5.037,
      "mean_token_accuracy": 0.20284036844968795,
      "num_tokens": 24025244.0,
      "step": 10480
    },
    {
      "entropy": 5.166749763488769,
      "epoch": 1.0072046109510087,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004906679497235856,
      "loss": 5.0149,
      "mean_token_accuracy": 0.20969751179218293,
      "num_tokens": 24037701.0,
      "step": 10485
    },
    {
      "entropy": 5.330682754516602,
      "epoch": 1.0076849183477425,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004906581760267397,
      "loss": 5.0983,
      "mean_token_accuracy": 0.20062616169452668,
      "num_tokens": 24049369.0,
      "step": 10490
    },
    {
      "entropy": 5.25918288230896,
      "epoch": 1.0081652257444764,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004906483973229513,
      "loss": 5.0572,
      "mean_token_accuracy": 0.205008926987648,
      "num_tokens": 24060625.0,
      "step": 10495
    },
    {
      "entropy": 5.348217487335205,
      "epoch": 1.0086455331412103,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004906386136124476,
      "loss": 5.0781,
      "mean_token_accuracy": 0.20820992141962053,
      "num_tokens": 24072527.0,
      "step": 10500
    },
    {
      "entropy": 5.274533462524414,
      "epoch": 1.0091258405379442,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004906288248954554,
      "loss": 5.1097,
      "mean_token_accuracy": 0.19823974668979644,
      "num_tokens": 24083979.0,
      "step": 10505
    },
    {
      "entropy": 5.186685800552368,
      "epoch": 1.0096061479346783,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004906190311722023,
      "loss": 5.0832,
      "mean_token_accuracy": 0.2083374246954918,
      "num_tokens": 24095926.0,
      "step": 10510
    },
    {
      "entropy": 5.264166927337646,
      "epoch": 1.0100864553314122,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004906092324429155,
      "loss": 4.9807,
      "mean_token_accuracy": 0.21339434385299683,
      "num_tokens": 24107898.0,
      "step": 10515
    },
    {
      "entropy": 5.202907371520996,
      "epoch": 1.010566762728146,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004905994287078227,
      "loss": 5.0107,
      "mean_token_accuracy": 0.2139447033405304,
      "num_tokens": 24118713.0,
      "step": 10520
    },
    {
      "entropy": 5.250958251953125,
      "epoch": 1.01104707012488,
      "grad_norm": 1.25,
      "learning_rate": 0.0004905896199671512,
      "loss": 5.0617,
      "mean_token_accuracy": 0.2110368326306343,
      "num_tokens": 24130410.0,
      "step": 10525
    },
    {
      "entropy": 5.276316595077515,
      "epoch": 1.0115273775216138,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000490579806221129,
      "loss": 5.0628,
      "mean_token_accuracy": 0.2041147232055664,
      "num_tokens": 24140961.0,
      "step": 10530
    },
    {
      "entropy": 5.320595455169678,
      "epoch": 1.0120076849183477,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004905699874699838,
      "loss": 5.1307,
      "mean_token_accuracy": 0.20101449638605118,
      "num_tokens": 24153467.0,
      "step": 10535
    },
    {
      "entropy": 5.169706916809082,
      "epoch": 1.0124879923150816,
      "grad_norm": 1.25,
      "learning_rate": 0.0004905601637139436,
      "loss": 4.9501,
      "mean_token_accuracy": 0.21583596020936965,
      "num_tokens": 24163873.0,
      "step": 10540
    },
    {
      "entropy": 5.14606409072876,
      "epoch": 1.0129682997118155,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004905503349532365,
      "loss": 5.0548,
      "mean_token_accuracy": 0.2081949770450592,
      "num_tokens": 24175852.0,
      "step": 10545
    },
    {
      "entropy": 5.25450439453125,
      "epoch": 1.0134486071085496,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004905405011880906,
      "loss": 5.0792,
      "mean_token_accuracy": 0.20617119669914247,
      "num_tokens": 24187759.0,
      "step": 10550
    },
    {
      "entropy": 5.272856903076172,
      "epoch": 1.0139289145052834,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004905306624187343,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21113628149032593,
      "num_tokens": 24199725.0,
      "step": 10555
    },
    {
      "entropy": 5.276928234100342,
      "epoch": 1.0144092219020173,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004905208186453961,
      "loss": 5.1192,
      "mean_token_accuracy": 0.201485712826252,
      "num_tokens": 24210739.0,
      "step": 10560
    },
    {
      "entropy": 5.3130653381347654,
      "epoch": 1.0148895292987512,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004905109698683044,
      "loss": 5.1991,
      "mean_token_accuracy": 0.19676827937364577,
      "num_tokens": 24221873.0,
      "step": 10565
    },
    {
      "entropy": 5.239027881622315,
      "epoch": 1.015369836695485,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004905011160876878,
      "loss": 5.0554,
      "mean_token_accuracy": 0.20726247876882553,
      "num_tokens": 24233503.0,
      "step": 10570
    },
    {
      "entropy": 5.252388000488281,
      "epoch": 1.015850144092219,
      "grad_norm": 1.125,
      "learning_rate": 0.0004904912573037753,
      "loss": 5.0531,
      "mean_token_accuracy": 0.20353973805904388,
      "num_tokens": 24244465.0,
      "step": 10575
    },
    {
      "entropy": 5.175927019119262,
      "epoch": 1.0163304514889528,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004904813935167957,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21479705721139908,
      "num_tokens": 24255807.0,
      "step": 10580
    },
    {
      "entropy": 5.320192193984985,
      "epoch": 1.0168107588856867,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004904715247269779,
      "loss": 5.1381,
      "mean_token_accuracy": 0.19730682373046876,
      "num_tokens": 24268195.0,
      "step": 10585
    },
    {
      "entropy": 5.176008605957032,
      "epoch": 1.0172910662824208,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004904616509345514,
      "loss": 4.9152,
      "mean_token_accuracy": 0.21298785209655763,
      "num_tokens": 24278915.0,
      "step": 10590
    },
    {
      "entropy": 5.252929162979126,
      "epoch": 1.0177713736791547,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004904517721397449,
      "loss": 5.0986,
      "mean_token_accuracy": 0.2086629882454872,
      "num_tokens": 24289115.0,
      "step": 10595
    },
    {
      "entropy": 5.2012939453125,
      "epoch": 1.0182516810758886,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004904418883427881,
      "loss": 5.0559,
      "mean_token_accuracy": 0.20381117463111878,
      "num_tokens": 24300646.0,
      "step": 10600
    },
    {
      "entropy": 5.385745239257813,
      "epoch": 1.0187319884726225,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004904319995439104,
      "loss": 5.2048,
      "mean_token_accuracy": 0.20216926485300063,
      "num_tokens": 24312344.0,
      "step": 10605
    },
    {
      "entropy": 5.2061645030975345,
      "epoch": 1.0192122958693564,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004904221057433412,
      "loss": 4.999,
      "mean_token_accuracy": 0.20912384539842604,
      "num_tokens": 24323382.0,
      "step": 10610
    },
    {
      "entropy": 5.171212530136108,
      "epoch": 1.0196926032660902,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004904122069413105,
      "loss": 4.9917,
      "mean_token_accuracy": 0.2084407925605774,
      "num_tokens": 24336820.0,
      "step": 10615
    },
    {
      "entropy": 5.269424200057983,
      "epoch": 1.0201729106628241,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000490402303138048,
      "loss": 5.0256,
      "mean_token_accuracy": 0.2102523773908615,
      "num_tokens": 24348547.0,
      "step": 10620
    },
    {
      "entropy": 5.338338136672974,
      "epoch": 1.0206532180595582,
      "grad_norm": 1.25,
      "learning_rate": 0.0004903923943337836,
      "loss": 5.1929,
      "mean_token_accuracy": 0.20662245154380798,
      "num_tokens": 24359586.0,
      "step": 10625
    },
    {
      "entropy": 5.315977668762207,
      "epoch": 1.021133525456292,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004903824805287475,
      "loss": 5.1129,
      "mean_token_accuracy": 0.20428410023450852,
      "num_tokens": 24371647.0,
      "step": 10630
    },
    {
      "entropy": 5.270196199417114,
      "epoch": 1.021613832853026,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004903725617231696,
      "loss": 4.9709,
      "mean_token_accuracy": 0.209686715900898,
      "num_tokens": 24382308.0,
      "step": 10635
    },
    {
      "entropy": 5.220236968994141,
      "epoch": 1.0220941402497599,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004903626379172805,
      "loss": 5.0581,
      "mean_token_accuracy": 0.208523029088974,
      "num_tokens": 24393224.0,
      "step": 10640
    },
    {
      "entropy": 5.233209705352783,
      "epoch": 1.0225744476464937,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004903527091113102,
      "loss": 5.0932,
      "mean_token_accuracy": 0.20196254253387452,
      "num_tokens": 24405220.0,
      "step": 10645
    },
    {
      "entropy": 5.330379962921143,
      "epoch": 1.0230547550432276,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004903427753054897,
      "loss": 5.1115,
      "mean_token_accuracy": 0.20270898640155793,
      "num_tokens": 24416406.0,
      "step": 10650
    },
    {
      "entropy": 5.228357458114624,
      "epoch": 1.0235350624399615,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004903328365000492,
      "loss": 5.1229,
      "mean_token_accuracy": 0.2039305031299591,
      "num_tokens": 24428595.0,
      "step": 10655
    },
    {
      "entropy": 5.208801794052124,
      "epoch": 1.0240153698366954,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004903228926952199,
      "loss": 5.0178,
      "mean_token_accuracy": 0.21162311732769012,
      "num_tokens": 24439568.0,
      "step": 10660
    },
    {
      "entropy": 5.245709323883057,
      "epoch": 1.0244956772334295,
      "grad_norm": 1.25,
      "learning_rate": 0.0004903129438912322,
      "loss": 5.0695,
      "mean_token_accuracy": 0.20735363513231278,
      "num_tokens": 24451736.0,
      "step": 10665
    },
    {
      "entropy": 5.2331255912780765,
      "epoch": 1.0249759846301634,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004903029900883174,
      "loss": 5.0346,
      "mean_token_accuracy": 0.2127084791660309,
      "num_tokens": 24463463.0,
      "step": 10670
    },
    {
      "entropy": 5.246076965332032,
      "epoch": 1.0254562920268973,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004902930312867063,
      "loss": 5.057,
      "mean_token_accuracy": 0.2057100787758827,
      "num_tokens": 24474425.0,
      "step": 10675
    },
    {
      "entropy": 5.263680124282837,
      "epoch": 1.0259365994236311,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004902830674866306,
      "loss": 4.9802,
      "mean_token_accuracy": 0.21202512234449386,
      "num_tokens": 24486482.0,
      "step": 10680
    },
    {
      "entropy": 5.188517665863037,
      "epoch": 1.026416906820365,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004902730986883211,
      "loss": 5.0235,
      "mean_token_accuracy": 0.20175417214632035,
      "num_tokens": 24497709.0,
      "step": 10685
    },
    {
      "entropy": 5.181160497665405,
      "epoch": 1.026897214217099,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004902631248920096,
      "loss": 4.9836,
      "mean_token_accuracy": 0.21261143684387207,
      "num_tokens": 24508406.0,
      "step": 10690
    },
    {
      "entropy": 5.12087049484253,
      "epoch": 1.0273775216138328,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004902531460979274,
      "loss": 5.0506,
      "mean_token_accuracy": 0.20509957373142243,
      "num_tokens": 24520688.0,
      "step": 10695
    },
    {
      "entropy": 5.198111963272095,
      "epoch": 1.0278578290105667,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004902431623063065,
      "loss": 4.9728,
      "mean_token_accuracy": 0.21074305176734925,
      "num_tokens": 24531473.0,
      "step": 10700
    },
    {
      "entropy": 5.260646343231201,
      "epoch": 1.0283381364073008,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004902331735173785,
      "loss": 5.0716,
      "mean_token_accuracy": 0.19711542129516602,
      "num_tokens": 24543263.0,
      "step": 10705
    },
    {
      "entropy": 5.2504377365112305,
      "epoch": 1.0288184438040346,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004902231797313752,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21237300336360931,
      "num_tokens": 24554067.0,
      "step": 10710
    },
    {
      "entropy": 5.16482458114624,
      "epoch": 1.0292987512007685,
      "grad_norm": 1.125,
      "learning_rate": 0.0004902131809485288,
      "loss": 4.9803,
      "mean_token_accuracy": 0.2174995258450508,
      "num_tokens": 24566291.0,
      "step": 10715
    },
    {
      "entropy": 5.284157609939575,
      "epoch": 1.0297790585975024,
      "grad_norm": 1.125,
      "learning_rate": 0.0004902031771690713,
      "loss": 5.1206,
      "mean_token_accuracy": 0.20218359529972077,
      "num_tokens": 24579248.0,
      "step": 10720
    },
    {
      "entropy": 5.203498792648316,
      "epoch": 1.0302593659942363,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004901931683932352,
      "loss": 5.0019,
      "mean_token_accuracy": 0.2181095853447914,
      "num_tokens": 24591047.0,
      "step": 10725
    },
    {
      "entropy": 5.265077066421509,
      "epoch": 1.0307396733909702,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004901831546212526,
      "loss": 5.0689,
      "mean_token_accuracy": 0.20814187824726105,
      "num_tokens": 24602454.0,
      "step": 10730
    },
    {
      "entropy": 5.2441651821136475,
      "epoch": 1.031219980787704,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004901731358533562,
      "loss": 5.0812,
      "mean_token_accuracy": 0.20977521389722825,
      "num_tokens": 24614168.0,
      "step": 10735
    },
    {
      "entropy": 5.245899677276611,
      "epoch": 1.031700288184438,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004901631120897785,
      "loss": 5.0602,
      "mean_token_accuracy": 0.21061237156391144,
      "num_tokens": 24625588.0,
      "step": 10740
    },
    {
      "entropy": 5.19196047782898,
      "epoch": 1.032180595581172,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004901530833307522,
      "loss": 5.0436,
      "mean_token_accuracy": 0.2036419242620468,
      "num_tokens": 24637744.0,
      "step": 10745
    },
    {
      "entropy": 5.177149677276612,
      "epoch": 1.032660902977906,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004901430495765103,
      "loss": 4.9789,
      "mean_token_accuracy": 0.21069790571928024,
      "num_tokens": 24649631.0,
      "step": 10750
    },
    {
      "entropy": 5.238631963729858,
      "epoch": 1.0331412103746398,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004901330108272855,
      "loss": 5.0014,
      "mean_token_accuracy": 0.21060845106840134,
      "num_tokens": 24660622.0,
      "step": 10755
    },
    {
      "entropy": 5.26301851272583,
      "epoch": 1.0336215177713737,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004901229670833111,
      "loss": 5.0269,
      "mean_token_accuracy": 0.20819098502397537,
      "num_tokens": 24671072.0,
      "step": 10760
    },
    {
      "entropy": 5.233757972717285,
      "epoch": 1.0341018251681076,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004901129183448201,
      "loss": 5.0879,
      "mean_token_accuracy": 0.2040340319275856,
      "num_tokens": 24683160.0,
      "step": 10765
    },
    {
      "entropy": 5.291535329818726,
      "epoch": 1.0345821325648414,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004901028646120459,
      "loss": 5.103,
      "mean_token_accuracy": 0.20735848397016526,
      "num_tokens": 24694703.0,
      "step": 10770
    },
    {
      "entropy": 5.228680801391602,
      "epoch": 1.0350624399615753,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000490092805885222,
      "loss": 5.0153,
      "mean_token_accuracy": 0.20685346126556398,
      "num_tokens": 24706429.0,
      "step": 10775
    },
    {
      "entropy": 5.2095849990844725,
      "epoch": 1.0355427473583094,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004900827421645816,
      "loss": 5.0386,
      "mean_token_accuracy": 0.21020758599042894,
      "num_tokens": 24717682.0,
      "step": 10780
    },
    {
      "entropy": 5.325321197509766,
      "epoch": 1.0360230547550433,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004900726734503589,
      "loss": 5.114,
      "mean_token_accuracy": 0.20808326900005342,
      "num_tokens": 24728789.0,
      "step": 10785
    },
    {
      "entropy": 5.1388953685760494,
      "epoch": 1.0365033621517772,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004900625997427872,
      "loss": 4.924,
      "mean_token_accuracy": 0.2128748342394829,
      "num_tokens": 24739788.0,
      "step": 10790
    },
    {
      "entropy": 5.20656099319458,
      "epoch": 1.036983669548511,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004900525210421006,
      "loss": 5.0808,
      "mean_token_accuracy": 0.20767248123884202,
      "num_tokens": 24751828.0,
      "step": 10795
    },
    {
      "entropy": 5.261975908279419,
      "epoch": 1.037463976945245,
      "grad_norm": 1.25,
      "learning_rate": 0.0004900424373485329,
      "loss": 5.0077,
      "mean_token_accuracy": 0.21646483093500138,
      "num_tokens": 24762061.0,
      "step": 10800
    },
    {
      "entropy": 5.198897981643677,
      "epoch": 1.0379442843419788,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004900323486623185,
      "loss": 5.0227,
      "mean_token_accuracy": 0.2138120338320732,
      "num_tokens": 24774018.0,
      "step": 10805
    },
    {
      "entropy": 5.199652051925659,
      "epoch": 1.0384245917387127,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004900222549836914,
      "loss": 4.9977,
      "mean_token_accuracy": 0.21953998059034346,
      "num_tokens": 24785206.0,
      "step": 10810
    },
    {
      "entropy": 5.202074670791626,
      "epoch": 1.0389048991354466,
      "grad_norm": 1.328125,
      "learning_rate": 0.000490012156312886,
      "loss": 5.064,
      "mean_token_accuracy": 0.20815465450286866,
      "num_tokens": 24797730.0,
      "step": 10815
    },
    {
      "entropy": 5.285942316055298,
      "epoch": 1.0393852065321807,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004900020526501369,
      "loss": 5.1623,
      "mean_token_accuracy": 0.20548777133226395,
      "num_tokens": 24809267.0,
      "step": 10820
    },
    {
      "entropy": 5.323640298843384,
      "epoch": 1.0398655139289146,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004899919439956785,
      "loss": 5.0149,
      "mean_token_accuracy": 0.20243232399225236,
      "num_tokens": 24820461.0,
      "step": 10825
    },
    {
      "entropy": 5.137351226806641,
      "epoch": 1.0403458213256485,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004899818303497455,
      "loss": 5.0647,
      "mean_token_accuracy": 0.20617686212062836,
      "num_tokens": 24831959.0,
      "step": 10830
    },
    {
      "entropy": 5.212925910949707,
      "epoch": 1.0408261287223823,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004899717117125728,
      "loss": 5.0369,
      "mean_token_accuracy": 0.20992097556591033,
      "num_tokens": 24843655.0,
      "step": 10835
    },
    {
      "entropy": 5.257765245437622,
      "epoch": 1.0413064361191162,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004899615880843953,
      "loss": 4.9982,
      "mean_token_accuracy": 0.20728527754545212,
      "num_tokens": 24855643.0,
      "step": 10840
    },
    {
      "entropy": 5.193706226348877,
      "epoch": 1.04178674351585,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004899514594654481,
      "loss": 4.9615,
      "mean_token_accuracy": 0.2136431574821472,
      "num_tokens": 24865745.0,
      "step": 10845
    },
    {
      "entropy": 5.21293306350708,
      "epoch": 1.042267050912584,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004899413258559662,
      "loss": 4.9769,
      "mean_token_accuracy": 0.21370177417993547,
      "num_tokens": 24876567.0,
      "step": 10850
    },
    {
      "entropy": 5.155454540252686,
      "epoch": 1.0427473583093179,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004899311872561849,
      "loss": 4.9658,
      "mean_token_accuracy": 0.20952108949422837,
      "num_tokens": 24888294.0,
      "step": 10855
    },
    {
      "entropy": 5.250037145614624,
      "epoch": 1.043227665706052,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004899210436663398,
      "loss": 5.185,
      "mean_token_accuracy": 0.2029317781329155,
      "num_tokens": 24899559.0,
      "step": 10860
    },
    {
      "entropy": 5.281383514404297,
      "epoch": 1.0437079731027858,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004899108950866661,
      "loss": 5.0249,
      "mean_token_accuracy": 0.20818737596273423,
      "num_tokens": 24911205.0,
      "step": 10865
    },
    {
      "entropy": 5.22299599647522,
      "epoch": 1.0441882804995197,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004899007415173997,
      "loss": 5.0233,
      "mean_token_accuracy": 0.20931993424892426,
      "num_tokens": 24922635.0,
      "step": 10870
    },
    {
      "entropy": 5.166127109527588,
      "epoch": 1.0446685878962536,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004898905829587762,
      "loss": 5.0057,
      "mean_token_accuracy": 0.21393175125122071,
      "num_tokens": 24932941.0,
      "step": 10875
    },
    {
      "entropy": 5.258418607711792,
      "epoch": 1.0451488952929875,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004898804194110313,
      "loss": 5.0998,
      "mean_token_accuracy": 0.2095828041434288,
      "num_tokens": 24943980.0,
      "step": 10880
    },
    {
      "entropy": 5.259143209457397,
      "epoch": 1.0456292026897214,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004898702508744012,
      "loss": 5.0565,
      "mean_token_accuracy": 0.20374516993761063,
      "num_tokens": 24955233.0,
      "step": 10885
    },
    {
      "entropy": 5.2724817276000975,
      "epoch": 1.0461095100864553,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004898600773491221,
      "loss": 5.0842,
      "mean_token_accuracy": 0.20758722573518754,
      "num_tokens": 24966971.0,
      "step": 10890
    },
    {
      "entropy": 5.3097765922546385,
      "epoch": 1.0465898174831891,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004898498988354297,
      "loss": 5.0903,
      "mean_token_accuracy": 0.19847430735826493,
      "num_tokens": 24977515.0,
      "step": 10895
    },
    {
      "entropy": 5.250106525421143,
      "epoch": 1.0470701248799232,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004898397153335608,
      "loss": 5.0263,
      "mean_token_accuracy": 0.20932495445013047,
      "num_tokens": 24988131.0,
      "step": 10900
    },
    {
      "entropy": 5.2977838039398195,
      "epoch": 1.0475504322766571,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004898295268437517,
      "loss": 5.0911,
      "mean_token_accuracy": 0.20342385321855544,
      "num_tokens": 24999238.0,
      "step": 10905
    },
    {
      "entropy": 5.247994613647461,
      "epoch": 1.048030739673391,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004898193333662388,
      "loss": 5.1209,
      "mean_token_accuracy": 0.196574567258358,
      "num_tokens": 25011443.0,
      "step": 10910
    },
    {
      "entropy": 5.248688077926635,
      "epoch": 1.0485110470701249,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004898091349012588,
      "loss": 4.9624,
      "mean_token_accuracy": 0.21045973300933837,
      "num_tokens": 25022151.0,
      "step": 10915
    },
    {
      "entropy": 5.2520771503448485,
      "epoch": 1.0489913544668588,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004897989314490486,
      "loss": 5.0297,
      "mean_token_accuracy": 0.2160254970192909,
      "num_tokens": 25032730.0,
      "step": 10920
    },
    {
      "entropy": 5.122765350341797,
      "epoch": 1.0494716618635926,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004897887230098451,
      "loss": 5.0002,
      "mean_token_accuracy": 0.20753524005413054,
      "num_tokens": 25044455.0,
      "step": 10925
    },
    {
      "entropy": 5.171108150482178,
      "epoch": 1.0499519692603265,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004897785095838852,
      "loss": 4.9696,
      "mean_token_accuracy": 0.22201226353645326,
      "num_tokens": 25056647.0,
      "step": 10930
    },
    {
      "entropy": 5.178351974487304,
      "epoch": 1.0504322766570606,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004897682911714061,
      "loss": 4.9894,
      "mean_token_accuracy": 0.20792763382196428,
      "num_tokens": 25067618.0,
      "step": 10935
    },
    {
      "entropy": 5.200053596496582,
      "epoch": 1.0509125840537945,
      "grad_norm": 1.0625,
      "learning_rate": 0.000489758067772645,
      "loss": 4.9844,
      "mean_token_accuracy": 0.21017451584339142,
      "num_tokens": 25079753.0,
      "step": 10940
    },
    {
      "entropy": 5.215559291839599,
      "epoch": 1.0513928914505284,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004897478393878392,
      "loss": 5.0564,
      "mean_token_accuracy": 0.20782340914011002,
      "num_tokens": 25091023.0,
      "step": 10945
    },
    {
      "entropy": 5.233268356323242,
      "epoch": 1.0518731988472623,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004897376060172264,
      "loss": 5.0366,
      "mean_token_accuracy": 0.20589469671249389,
      "num_tokens": 25101396.0,
      "step": 10950
    },
    {
      "entropy": 5.23882155418396,
      "epoch": 1.0523535062439962,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004897273676610438,
      "loss": 5.0795,
      "mean_token_accuracy": 0.20385542511940002,
      "num_tokens": 25112562.0,
      "step": 10955
    },
    {
      "entropy": 5.258812808990479,
      "epoch": 1.05283381364073,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004897171243195295,
      "loss": 5.0884,
      "mean_token_accuracy": 0.20825523287057876,
      "num_tokens": 25124401.0,
      "step": 10960
    },
    {
      "entropy": 5.22129373550415,
      "epoch": 1.053314121037464,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000489706875992921,
      "loss": 5.0643,
      "mean_token_accuracy": 0.20664875507354735,
      "num_tokens": 25137345.0,
      "step": 10965
    },
    {
      "entropy": 5.276070785522461,
      "epoch": 1.0537944284341978,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004896966226814565,
      "loss": 5.0816,
      "mean_token_accuracy": 0.2024160921573639,
      "num_tokens": 25149300.0,
      "step": 10970
    },
    {
      "entropy": 5.261047220230102,
      "epoch": 1.054274735830932,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004896863643853739,
      "loss": 4.9984,
      "mean_token_accuracy": 0.20964281260967255,
      "num_tokens": 25159510.0,
      "step": 10975
    },
    {
      "entropy": 5.24853630065918,
      "epoch": 1.0547550432276658,
      "grad_norm": 1.25,
      "learning_rate": 0.0004896761011049114,
      "loss": 4.9803,
      "mean_token_accuracy": 0.21133439838886262,
      "num_tokens": 25172346.0,
      "step": 10980
    },
    {
      "entropy": 5.275299549102783,
      "epoch": 1.0552353506243997,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004896658328403074,
      "loss": 5.1464,
      "mean_token_accuracy": 0.2018449768424034,
      "num_tokens": 25183471.0,
      "step": 10985
    },
    {
      "entropy": 5.3128221988677975,
      "epoch": 1.0557156580211335,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004896555595918001,
      "loss": 5.0743,
      "mean_token_accuracy": 0.2136384963989258,
      "num_tokens": 25195067.0,
      "step": 10990
    },
    {
      "entropy": 5.252407884597778,
      "epoch": 1.0561959654178674,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004896452813596281,
      "loss": 5.0776,
      "mean_token_accuracy": 0.20362422615289688,
      "num_tokens": 25206165.0,
      "step": 10995
    },
    {
      "entropy": 5.1924701690673825,
      "epoch": 1.0566762728146013,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004896349981440301,
      "loss": 4.9962,
      "mean_token_accuracy": 0.21781055629253387,
      "num_tokens": 25217229.0,
      "step": 11000
    },
    {
      "entropy": 5.2920989990234375,
      "epoch": 1.0571565802113352,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004896247099452447,
      "loss": 5.1464,
      "mean_token_accuracy": 0.19886945188045502,
      "num_tokens": 25229349.0,
      "step": 11005
    },
    {
      "entropy": 5.26919469833374,
      "epoch": 1.057636887608069,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004896144167635108,
      "loss": 5.0356,
      "mean_token_accuracy": 0.20507772862911225,
      "num_tokens": 25240367.0,
      "step": 11010
    },
    {
      "entropy": 5.264904165267945,
      "epoch": 1.0581171950048032,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004896041185990675,
      "loss": 5.122,
      "mean_token_accuracy": 0.20194689482450484,
      "num_tokens": 25252025.0,
      "step": 11015
    },
    {
      "entropy": 5.230758619308472,
      "epoch": 1.058597502401537,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004895938154521538,
      "loss": 5.0282,
      "mean_token_accuracy": 0.2098274201154709,
      "num_tokens": 25264846.0,
      "step": 11020
    },
    {
      "entropy": 5.163432216644287,
      "epoch": 1.059077809798271,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004895835073230089,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2123186245560646,
      "num_tokens": 25276150.0,
      "step": 11025
    },
    {
      "entropy": 5.224885082244873,
      "epoch": 1.0595581171950048,
      "grad_norm": 1.125,
      "learning_rate": 0.0004895731942118722,
      "loss": 5.0143,
      "mean_token_accuracy": 0.20640959292650224,
      "num_tokens": 25287819.0,
      "step": 11030
    },
    {
      "entropy": 5.2742572784423825,
      "epoch": 1.0600384245917387,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004895628761189829,
      "loss": 5.0662,
      "mean_token_accuracy": 0.21121879816055297,
      "num_tokens": 25300467.0,
      "step": 11035
    },
    {
      "entropy": 5.171493864059448,
      "epoch": 1.0605187319884726,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004895525530445809,
      "loss": 5.0075,
      "mean_token_accuracy": 0.20629844218492507,
      "num_tokens": 25312447.0,
      "step": 11040
    },
    {
      "entropy": 5.253208208084106,
      "epoch": 1.0609990393852065,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004895422249889057,
      "loss": 5.1049,
      "mean_token_accuracy": 0.2049245521426201,
      "num_tokens": 25324160.0,
      "step": 11045
    },
    {
      "entropy": 5.2349621772766115,
      "epoch": 1.0614793467819403,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004895318919521971,
      "loss": 4.9534,
      "mean_token_accuracy": 0.21894117891788484,
      "num_tokens": 25335106.0,
      "step": 11050
    },
    {
      "entropy": 5.127588224411011,
      "epoch": 1.0619596541786744,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004895215539346949,
      "loss": 4.929,
      "mean_token_accuracy": 0.21181363761425018,
      "num_tokens": 25345246.0,
      "step": 11055
    },
    {
      "entropy": 5.21281099319458,
      "epoch": 1.0624399615754083,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004895112109366393,
      "loss": 5.0608,
      "mean_token_accuracy": 0.2061576247215271,
      "num_tokens": 25356665.0,
      "step": 11060
    },
    {
      "entropy": 5.227270174026489,
      "epoch": 1.0629202689721422,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004895008629582703,
      "loss": 5.0127,
      "mean_token_accuracy": 0.21041125804185867,
      "num_tokens": 25368128.0,
      "step": 11065
    },
    {
      "entropy": 5.223698472976684,
      "epoch": 1.063400576368876,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004894905099998283,
      "loss": 5.0176,
      "mean_token_accuracy": 0.21564193665981293,
      "num_tokens": 25379061.0,
      "step": 11070
    },
    {
      "entropy": 5.209757709503174,
      "epoch": 1.06388088376561,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004894801520615535,
      "loss": 5.0103,
      "mean_token_accuracy": 0.21384227126836777,
      "num_tokens": 25390157.0,
      "step": 11075
    },
    {
      "entropy": 5.165754270553589,
      "epoch": 1.0643611911623438,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004894697891436863,
      "loss": 5.0182,
      "mean_token_accuracy": 0.2041109472513199,
      "num_tokens": 25400446.0,
      "step": 11080
    },
    {
      "entropy": 5.228540325164795,
      "epoch": 1.0648414985590777,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004894594212464676,
      "loss": 5.057,
      "mean_token_accuracy": 0.2060192868113518,
      "num_tokens": 25412064.0,
      "step": 11085
    },
    {
      "entropy": 5.21215934753418,
      "epoch": 1.0653218059558118,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004894490483701381,
      "loss": 5.0475,
      "mean_token_accuracy": 0.20911297500133513,
      "num_tokens": 25422898.0,
      "step": 11090
    },
    {
      "entropy": 5.215141153335571,
      "epoch": 1.0658021133525457,
      "grad_norm": 1.125,
      "learning_rate": 0.0004894386705149382,
      "loss": 5.0288,
      "mean_token_accuracy": 0.2053459882736206,
      "num_tokens": 25434318.0,
      "step": 11095
    },
    {
      "entropy": 5.209214115142823,
      "epoch": 1.0662824207492796,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004894282876811093,
      "loss": 4.9959,
      "mean_token_accuracy": 0.21150643676519393,
      "num_tokens": 25444941.0,
      "step": 11100
    },
    {
      "entropy": 5.239191961288452,
      "epoch": 1.0667627281460135,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004894178998688921,
      "loss": 5.0166,
      "mean_token_accuracy": 0.20650841295719147,
      "num_tokens": 25455971.0,
      "step": 11105
    },
    {
      "entropy": 5.1472421169281,
      "epoch": 1.0672430355427474,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004894075070785281,
      "loss": 4.9854,
      "mean_token_accuracy": 0.20873002707958221,
      "num_tokens": 25467587.0,
      "step": 11110
    },
    {
      "entropy": 5.124123573303223,
      "epoch": 1.0677233429394812,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004893971093102585,
      "loss": 4.9921,
      "mean_token_accuracy": 0.2175788938999176,
      "num_tokens": 25479137.0,
      "step": 11115
    },
    {
      "entropy": 5.360449171066284,
      "epoch": 1.0682036503362151,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004893867065643245,
      "loss": 5.1536,
      "mean_token_accuracy": 0.20015736073255538,
      "num_tokens": 25491755.0,
      "step": 11120
    },
    {
      "entropy": 5.199368524551391,
      "epoch": 1.068683957732949,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004893762988409678,
      "loss": 4.9374,
      "mean_token_accuracy": 0.2175113782286644,
      "num_tokens": 25503581.0,
      "step": 11125
    },
    {
      "entropy": 5.1958105087280275,
      "epoch": 1.069164265129683,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004893658861404301,
      "loss": 5.0829,
      "mean_token_accuracy": 0.21035192906856537,
      "num_tokens": 25515032.0,
      "step": 11130
    },
    {
      "entropy": 5.195679807662964,
      "epoch": 1.069644572526417,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004893554684629529,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21170926094055176,
      "num_tokens": 25526347.0,
      "step": 11135
    },
    {
      "entropy": 5.285630607604981,
      "epoch": 1.0701248799231509,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004893450458087784,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20596026182174682,
      "num_tokens": 25538287.0,
      "step": 11140
    },
    {
      "entropy": 5.10020170211792,
      "epoch": 1.0706051873198847,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004893346181781483,
      "loss": 4.9907,
      "mean_token_accuracy": 0.2218176618218422,
      "num_tokens": 25548431.0,
      "step": 11145
    },
    {
      "entropy": 5.2269182205200195,
      "epoch": 1.0710854947166186,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004893241855713048,
      "loss": 5.1537,
      "mean_token_accuracy": 0.20656787902116774,
      "num_tokens": 25558706.0,
      "step": 11150
    },
    {
      "entropy": 5.324103832244873,
      "epoch": 1.0715658021133525,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004893137479884903,
      "loss": 5.0376,
      "mean_token_accuracy": 0.21219127327203752,
      "num_tokens": 25569920.0,
      "step": 11155
    },
    {
      "entropy": 5.18408842086792,
      "epoch": 1.0720461095100864,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004893033054299468,
      "loss": 4.9855,
      "mean_token_accuracy": 0.22083988189697265,
      "num_tokens": 25581340.0,
      "step": 11160
    },
    {
      "entropy": 5.223698997497559,
      "epoch": 1.0725264169068203,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000489292857895917,
      "loss": 5.0537,
      "mean_token_accuracy": 0.2028847187757492,
      "num_tokens": 25592477.0,
      "step": 11165
    },
    {
      "entropy": 5.205398082733154,
      "epoch": 1.0730067243035544,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004892824053866432,
      "loss": 5.0111,
      "mean_token_accuracy": 0.2127639412879944,
      "num_tokens": 25604583.0,
      "step": 11170
    },
    {
      "entropy": 5.3541834354400635,
      "epoch": 1.0734870317002883,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004892719479023683,
      "loss": 5.1483,
      "mean_token_accuracy": 0.19993159770965577,
      "num_tokens": 25614743.0,
      "step": 11175
    },
    {
      "entropy": 5.114918804168701,
      "epoch": 1.0739673390970221,
      "grad_norm": 1.234375,
      "learning_rate": 0.000489261485443335,
      "loss": 4.8775,
      "mean_token_accuracy": 0.2203249603509903,
      "num_tokens": 25626796.0,
      "step": 11180
    },
    {
      "entropy": 5.213155698776245,
      "epoch": 1.074447646493756,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004892510180097863,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20537566244602204,
      "num_tokens": 25638014.0,
      "step": 11185
    },
    {
      "entropy": 5.192825651168823,
      "epoch": 1.07492795389049,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004892405456019651,
      "loss": 5.034,
      "mean_token_accuracy": 0.21505061239004136,
      "num_tokens": 25649555.0,
      "step": 11190
    },
    {
      "entropy": 5.261461400985718,
      "epoch": 1.0754082612872238,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004892300682201147,
      "loss": 5.1488,
      "mean_token_accuracy": 0.2026241198182106,
      "num_tokens": 25662886.0,
      "step": 11195
    },
    {
      "entropy": 5.258253002166748,
      "epoch": 1.0758885686839577,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004892195858644782,
      "loss": 4.9886,
      "mean_token_accuracy": 0.2154127523303032,
      "num_tokens": 25673802.0,
      "step": 11200
    },
    {
      "entropy": 5.198265171051025,
      "epoch": 1.0763688760806915,
      "grad_norm": 1.46875,
      "learning_rate": 0.000489209098535299,
      "loss": 5.0392,
      "mean_token_accuracy": 0.21160761564970015,
      "num_tokens": 25685545.0,
      "step": 11205
    },
    {
      "entropy": 5.129959917068481,
      "epoch": 1.0768491834774256,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004891986062328205,
      "loss": 4.972,
      "mean_token_accuracy": 0.21262802928686142,
      "num_tokens": 25697955.0,
      "step": 11210
    },
    {
      "entropy": 5.235784816741943,
      "epoch": 1.0773294908741595,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004891881089572865,
      "loss": 5.0677,
      "mean_token_accuracy": 0.20539120882749556,
      "num_tokens": 25709707.0,
      "step": 11215
    },
    {
      "entropy": 5.303123235702515,
      "epoch": 1.0778097982708934,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004891776067089406,
      "loss": 5.0867,
      "mean_token_accuracy": 0.21078593134880066,
      "num_tokens": 25720576.0,
      "step": 11220
    },
    {
      "entropy": 5.210818147659301,
      "epoch": 1.0782901056676273,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004891670994880266,
      "loss": 4.9689,
      "mean_token_accuracy": 0.21445323079824447,
      "num_tokens": 25731387.0,
      "step": 11225
    },
    {
      "entropy": 5.168476152420044,
      "epoch": 1.0787704130643612,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004891565872947888,
      "loss": 4.9499,
      "mean_token_accuracy": 0.22056841254234313,
      "num_tokens": 25741813.0,
      "step": 11230
    },
    {
      "entropy": 5.22567892074585,
      "epoch": 1.079250720461095,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004891460701294706,
      "loss": 5.0417,
      "mean_token_accuracy": 0.20818577855825424,
      "num_tokens": 25753763.0,
      "step": 11235
    },
    {
      "entropy": 5.1633439540863035,
      "epoch": 1.079731027857829,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004891355479923167,
      "loss": 4.9847,
      "mean_token_accuracy": 0.20987370014190673,
      "num_tokens": 25765269.0,
      "step": 11240
    },
    {
      "entropy": 5.232744884490967,
      "epoch": 1.080211335254563,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004891250208835712,
      "loss": 5.0705,
      "mean_token_accuracy": 0.2110899955034256,
      "num_tokens": 25776841.0,
      "step": 11245
    },
    {
      "entropy": 5.290762424468994,
      "epoch": 1.080691642651297,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004891144888034784,
      "loss": 5.0937,
      "mean_token_accuracy": 0.19484081268310546,
      "num_tokens": 25788151.0,
      "step": 11250
    },
    {
      "entropy": 5.1890785694122314,
      "epoch": 1.0811719500480308,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004891039517522832,
      "loss": 4.977,
      "mean_token_accuracy": 0.21769267767667771,
      "num_tokens": 25799194.0,
      "step": 11255
    },
    {
      "entropy": 5.196116018295288,
      "epoch": 1.0816522574447647,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004890934097302299,
      "loss": 5.0113,
      "mean_token_accuracy": 0.2085309609770775,
      "num_tokens": 25809483.0,
      "step": 11260
    },
    {
      "entropy": 5.169402551651001,
      "epoch": 1.0821325648414986,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004890828627375632,
      "loss": 4.9959,
      "mean_token_accuracy": 0.19911896139383317,
      "num_tokens": 25820781.0,
      "step": 11265
    },
    {
      "entropy": 5.292964124679566,
      "epoch": 1.0826128722382324,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004890723107745283,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20198074877262115,
      "num_tokens": 25832111.0,
      "step": 11270
    },
    {
      "entropy": 5.24950852394104,
      "epoch": 1.0830931796349663,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004890617538413699,
      "loss": 5.0905,
      "mean_token_accuracy": 0.20630386471748352,
      "num_tokens": 25843300.0,
      "step": 11275
    },
    {
      "entropy": 5.235115003585816,
      "epoch": 1.0835734870317002,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004890511919383333,
      "loss": 5.0317,
      "mean_token_accuracy": 0.20131637156009674,
      "num_tokens": 25855337.0,
      "step": 11280
    },
    {
      "entropy": 5.3349052429199215,
      "epoch": 1.084053794428434,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004890406250656636,
      "loss": 5.1149,
      "mean_token_accuracy": 0.20222898274660112,
      "num_tokens": 25866683.0,
      "step": 11285
    },
    {
      "entropy": 5.183066892623901,
      "epoch": 1.0845341018251682,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004890300532236062,
      "loss": 5.0033,
      "mean_token_accuracy": 0.20709893852472305,
      "num_tokens": 25878041.0,
      "step": 11290
    },
    {
      "entropy": 5.215262126922608,
      "epoch": 1.085014409221902,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004890194764124064,
      "loss": 5.0198,
      "mean_token_accuracy": 0.20469354093074799,
      "num_tokens": 25889059.0,
      "step": 11295
    },
    {
      "entropy": 5.141601991653443,
      "epoch": 1.085494716618636,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004890088946323099,
      "loss": 4.9192,
      "mean_token_accuracy": 0.22135172188282012,
      "num_tokens": 25900062.0,
      "step": 11300
    },
    {
      "entropy": 5.249534749984742,
      "epoch": 1.0859750240153698,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004889983078835623,
      "loss": 5.1528,
      "mean_token_accuracy": 0.2062945395708084,
      "num_tokens": 25911813.0,
      "step": 11305
    },
    {
      "entropy": 5.212577438354492,
      "epoch": 1.0864553314121037,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004889877161664096,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21434760838747025,
      "num_tokens": 25922321.0,
      "step": 11310
    },
    {
      "entropy": 5.2093549251556395,
      "epoch": 1.0869356388088376,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004889771194810974,
      "loss": 5.0559,
      "mean_token_accuracy": 0.21490278989076614,
      "num_tokens": 25933631.0,
      "step": 11315
    },
    {
      "entropy": 5.281598234176636,
      "epoch": 1.0874159462055715,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004889665178278719,
      "loss": 5.0749,
      "mean_token_accuracy": 0.2093101292848587,
      "num_tokens": 25945190.0,
      "step": 11320
    },
    {
      "entropy": 5.223658561706543,
      "epoch": 1.0878962536023056,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004889559112069792,
      "loss": 5.0187,
      "mean_token_accuracy": 0.21777651607990264,
      "num_tokens": 25956817.0,
      "step": 11325
    },
    {
      "entropy": 5.302747201919556,
      "epoch": 1.0883765609990395,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004889452996186657,
      "loss": 5.0745,
      "mean_token_accuracy": 0.21182458102703094,
      "num_tokens": 25968134.0,
      "step": 11330
    },
    {
      "entropy": 5.172395086288452,
      "epoch": 1.0888568683957733,
      "grad_norm": 1.375,
      "learning_rate": 0.0004889346830631774,
      "loss": 5.0089,
      "mean_token_accuracy": 0.21319064497947693,
      "num_tokens": 25981825.0,
      "step": 11335
    },
    {
      "entropy": 5.273640537261963,
      "epoch": 1.0893371757925072,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000488924061540761,
      "loss": 5.1192,
      "mean_token_accuracy": 0.19374843388795854,
      "num_tokens": 25993962.0,
      "step": 11340
    },
    {
      "entropy": 5.229739046096801,
      "epoch": 1.089817483189241,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004889134350516633,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21363951861858368,
      "num_tokens": 26005705.0,
      "step": 11345
    },
    {
      "entropy": 5.199680805206299,
      "epoch": 1.090297790585975,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004889028035961308,
      "loss": 5.0354,
      "mean_token_accuracy": 0.2146790638566017,
      "num_tokens": 26017111.0,
      "step": 11350
    },
    {
      "entropy": 5.256040382385254,
      "epoch": 1.0907780979827089,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004888921671744103,
      "loss": 5.0851,
      "mean_token_accuracy": 0.20618789941072463,
      "num_tokens": 26028097.0,
      "step": 11355
    },
    {
      "entropy": 5.236567163467408,
      "epoch": 1.0912584053794427,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004888815257867488,
      "loss": 5.0097,
      "mean_token_accuracy": 0.20990807712078094,
      "num_tokens": 26039281.0,
      "step": 11360
    },
    {
      "entropy": 5.241058111190796,
      "epoch": 1.0917387127761768,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004888708794333934,
      "loss": 5.0851,
      "mean_token_accuracy": 0.20212292522192002,
      "num_tokens": 26051133.0,
      "step": 11365
    },
    {
      "entropy": 5.189146089553833,
      "epoch": 1.0922190201729107,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004888602281145913,
      "loss": 5.0707,
      "mean_token_accuracy": 0.20839928984642028,
      "num_tokens": 26062220.0,
      "step": 11370
    },
    {
      "entropy": 5.209463024139405,
      "epoch": 1.0926993275696446,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004888495718305897,
      "loss": 4.9734,
      "mean_token_accuracy": 0.22519829720258713,
      "num_tokens": 26072556.0,
      "step": 11375
    },
    {
      "entropy": 5.311297273635864,
      "epoch": 1.0931796349663785,
      "grad_norm": 1.359375,
      "learning_rate": 0.000488838910581636,
      "loss": 5.1076,
      "mean_token_accuracy": 0.2030702129006386,
      "num_tokens": 26084218.0,
      "step": 11380
    },
    {
      "entropy": 5.358635473251343,
      "epoch": 1.0936599423631124,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004888282443679777,
      "loss": 5.2159,
      "mean_token_accuracy": 0.20051150023937225,
      "num_tokens": 26096272.0,
      "step": 11385
    },
    {
      "entropy": 5.223580932617187,
      "epoch": 1.0941402497598463,
      "grad_norm": 2.125,
      "learning_rate": 0.0004888175731898627,
      "loss": 5.0876,
      "mean_token_accuracy": 0.20818673968315124,
      "num_tokens": 26109247.0,
      "step": 11390
    },
    {
      "entropy": 5.24344687461853,
      "epoch": 1.0946205571565801,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004888068970475384,
      "loss": 5.0138,
      "mean_token_accuracy": 0.21105274558067322,
      "num_tokens": 26121340.0,
      "step": 11395
    },
    {
      "entropy": 5.21724762916565,
      "epoch": 1.0951008645533142,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004887962159412529,
      "loss": 5.0422,
      "mean_token_accuracy": 0.20628211349248887,
      "num_tokens": 26133135.0,
      "step": 11400
    },
    {
      "entropy": 5.172627210617065,
      "epoch": 1.0955811719500481,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004887855298712541,
      "loss": 4.9445,
      "mean_token_accuracy": 0.21131701320409774,
      "num_tokens": 26145324.0,
      "step": 11405
    },
    {
      "entropy": 5.135566997528076,
      "epoch": 1.096061479346782,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048877483883779,
      "loss": 4.9509,
      "mean_token_accuracy": 0.2118778184056282,
      "num_tokens": 26156739.0,
      "step": 11410
    },
    {
      "entropy": 5.2167253494262695,
      "epoch": 1.0965417867435159,
      "grad_norm": 1.125,
      "learning_rate": 0.0004887641428411091,
      "loss": 5.051,
      "mean_token_accuracy": 0.20656709223985673,
      "num_tokens": 26169543.0,
      "step": 11415
    },
    {
      "entropy": 5.2406800270080565,
      "epoch": 1.0970220941402498,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004887534418814595,
      "loss": 5.0324,
      "mean_token_accuracy": 0.21196200847625732,
      "num_tokens": 26181317.0,
      "step": 11420
    },
    {
      "entropy": 5.261473274230957,
      "epoch": 1.0975024015369836,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004887427359590897,
      "loss": 5.1157,
      "mean_token_accuracy": 0.2026610180735588,
      "num_tokens": 26193402.0,
      "step": 11425
    },
    {
      "entropy": 5.19752197265625,
      "epoch": 1.0979827089337175,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004887320250742482,
      "loss": 5.0071,
      "mean_token_accuracy": 0.21291136294603347,
      "num_tokens": 26204660.0,
      "step": 11430
    },
    {
      "entropy": 5.311797189712524,
      "epoch": 1.0984630163304514,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004887213092271838,
      "loss": 5.0569,
      "mean_token_accuracy": 0.20815687179565429,
      "num_tokens": 26215593.0,
      "step": 11435
    },
    {
      "entropy": 5.184613609313965,
      "epoch": 1.0989433237271853,
      "grad_norm": 1.125,
      "learning_rate": 0.0004887105884181451,
      "loss": 4.9598,
      "mean_token_accuracy": 0.2183475077152252,
      "num_tokens": 26227481.0,
      "step": 11440
    },
    {
      "entropy": 5.256375646591186,
      "epoch": 1.0994236311239194,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004886998626473813,
      "loss": 5.1259,
      "mean_token_accuracy": 0.20083087533712388,
      "num_tokens": 26238260.0,
      "step": 11445
    },
    {
      "entropy": 5.274607563018799,
      "epoch": 1.0999039385206533,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004886891319151411,
      "loss": 5.0594,
      "mean_token_accuracy": 0.20587003082036973,
      "num_tokens": 26249749.0,
      "step": 11450
    },
    {
      "entropy": 5.225582933425903,
      "epoch": 1.1003842459173871,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004886783962216738,
      "loss": 5.0231,
      "mean_token_accuracy": 0.21181628853082657,
      "num_tokens": 26260639.0,
      "step": 11455
    },
    {
      "entropy": 5.20557222366333,
      "epoch": 1.100864553314121,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004886676555672287,
      "loss": 5.0324,
      "mean_token_accuracy": 0.20463857203722,
      "num_tokens": 26271797.0,
      "step": 11460
    },
    {
      "entropy": 5.235341119766235,
      "epoch": 1.101344860710855,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004886569099520551,
      "loss": 5.0407,
      "mean_token_accuracy": 0.20326029062271117,
      "num_tokens": 26283298.0,
      "step": 11465
    },
    {
      "entropy": 5.140804100036621,
      "epoch": 1.1018251681075888,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004886461593764024,
      "loss": 4.8768,
      "mean_token_accuracy": 0.21619092524051667,
      "num_tokens": 26293632.0,
      "step": 11470
    },
    {
      "entropy": 5.27526888847351,
      "epoch": 1.1023054755043227,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004886354038405204,
      "loss": 5.11,
      "mean_token_accuracy": 0.20375400334596633,
      "num_tokens": 26307053.0,
      "step": 11475
    },
    {
      "entropy": 5.226287412643432,
      "epoch": 1.1027857829010568,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004886246433446586,
      "loss": 4.947,
      "mean_token_accuracy": 0.2169455960392952,
      "num_tokens": 26317136.0,
      "step": 11480
    },
    {
      "entropy": 5.176907205581665,
      "epoch": 1.1032660902977907,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004886138778890669,
      "loss": 5.0549,
      "mean_token_accuracy": 0.21516386270523072,
      "num_tokens": 26328865.0,
      "step": 11485
    },
    {
      "entropy": 5.1918439865112305,
      "epoch": 1.1037463976945245,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004886031074739953,
      "loss": 5.0528,
      "mean_token_accuracy": 0.21228553503751754,
      "num_tokens": 26341010.0,
      "step": 11490
    },
    {
      "entropy": 5.3316321849823,
      "epoch": 1.1042267050912584,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004885923320996938,
      "loss": 5.1002,
      "mean_token_accuracy": 0.2051398828625679,
      "num_tokens": 26352203.0,
      "step": 11495
    },
    {
      "entropy": 5.247815465927124,
      "epoch": 1.1047070124879923,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004885815517664127,
      "loss": 5.0778,
      "mean_token_accuracy": 0.2059747874736786,
      "num_tokens": 26363096.0,
      "step": 11500
    },
    {
      "entropy": 5.158432292938232,
      "epoch": 1.1051873198847262,
      "grad_norm": 1.328125,
      "learning_rate": 0.000488570766474402,
      "loss": 4.9772,
      "mean_token_accuracy": 0.2166539430618286,
      "num_tokens": 26374776.0,
      "step": 11505
    },
    {
      "entropy": 5.175604820251465,
      "epoch": 1.10566762728146,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004885599762239124,
      "loss": 5.0772,
      "mean_token_accuracy": 0.211662220954895,
      "num_tokens": 26386654.0,
      "step": 11510
    },
    {
      "entropy": 5.24345178604126,
      "epoch": 1.106147934678194,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004885491810151943,
      "loss": 5.035,
      "mean_token_accuracy": 0.2083755299448967,
      "num_tokens": 26400022.0,
      "step": 11515
    },
    {
      "entropy": 5.1965940475463865,
      "epoch": 1.106628242074928,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004885383808484982,
      "loss": 5.02,
      "mean_token_accuracy": 0.2079739347100258,
      "num_tokens": 26412308.0,
      "step": 11520
    },
    {
      "entropy": 5.310696601867676,
      "epoch": 1.107108549471662,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004885275757240751,
      "loss": 5.1085,
      "mean_token_accuracy": 0.19413709938526152,
      "num_tokens": 26424506.0,
      "step": 11525
    },
    {
      "entropy": 5.262013483047485,
      "epoch": 1.1075888568683958,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004885167656421757,
      "loss": 5.0097,
      "mean_token_accuracy": 0.21698053628206254,
      "num_tokens": 26435020.0,
      "step": 11530
    },
    {
      "entropy": 5.063734102249145,
      "epoch": 1.1080691642651297,
      "grad_norm": 1.125,
      "learning_rate": 0.000488505950603051,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21516988426446915,
      "num_tokens": 26447952.0,
      "step": 11535
    },
    {
      "entropy": 5.240297079086304,
      "epoch": 1.1085494716618636,
      "grad_norm": 1.171875,
      "learning_rate": 0.000488495130606952,
      "loss": 5.0176,
      "mean_token_accuracy": 0.20866743773221968,
      "num_tokens": 26460001.0,
      "step": 11540
    },
    {
      "entropy": 5.19719762802124,
      "epoch": 1.1090297790585975,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004884843056541302,
      "loss": 5.0077,
      "mean_token_accuracy": 0.21265242397785186,
      "num_tokens": 26472139.0,
      "step": 11545
    },
    {
      "entropy": 5.189639568328857,
      "epoch": 1.1095100864553313,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004884734757448367,
      "loss": 5.0582,
      "mean_token_accuracy": 0.20886846929788588,
      "num_tokens": 26483855.0,
      "step": 11550
    },
    {
      "entropy": 5.2675727844238285,
      "epoch": 1.1099903938520654,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000488462640879323,
      "loss": 5.1292,
      "mean_token_accuracy": 0.20411465167999268,
      "num_tokens": 26496529.0,
      "step": 11555
    },
    {
      "entropy": 5.252080059051513,
      "epoch": 1.1104707012487993,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004884518010578405,
      "loss": 4.9836,
      "mean_token_accuracy": 0.21935612708330154,
      "num_tokens": 26506663.0,
      "step": 11560
    },
    {
      "entropy": 5.2437145709991455,
      "epoch": 1.1109510086455332,
      "grad_norm": 1.25,
      "learning_rate": 0.0004884409562806411,
      "loss": 5.1278,
      "mean_token_accuracy": 0.20246258974075318,
      "num_tokens": 26519091.0,
      "step": 11565
    },
    {
      "entropy": 5.228885507583618,
      "epoch": 1.111431316042267,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004884301065479765,
      "loss": 5.0586,
      "mean_token_accuracy": 0.20545517206192015,
      "num_tokens": 26530145.0,
      "step": 11570
    },
    {
      "entropy": 5.244892597198486,
      "epoch": 1.111911623439001,
      "grad_norm": 1.25,
      "learning_rate": 0.0004884192518600986,
      "loss": 4.991,
      "mean_token_accuracy": 0.20797066241502762,
      "num_tokens": 26541228.0,
      "step": 11575
    },
    {
      "entropy": 5.2725475311279295,
      "epoch": 1.1123919308357348,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004884083922172593,
      "loss": 5.0244,
      "mean_token_accuracy": 0.20872527211904526,
      "num_tokens": 26551359.0,
      "step": 11580
    },
    {
      "entropy": 5.112119483947754,
      "epoch": 1.1128722382324687,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004883975276197108,
      "loss": 5.0484,
      "mean_token_accuracy": 0.20813682675361633,
      "num_tokens": 26562911.0,
      "step": 11585
    },
    {
      "entropy": 5.299541187286377,
      "epoch": 1.1133525456292026,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004883866580677055,
      "loss": 5.0246,
      "mean_token_accuracy": 0.2140663743019104,
      "num_tokens": 26573198.0,
      "step": 11590
    },
    {
      "entropy": 5.245718288421631,
      "epoch": 1.1138328530259365,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004883757835614956,
      "loss": 5.0318,
      "mean_token_accuracy": 0.20570946335792542,
      "num_tokens": 26584932.0,
      "step": 11595
    },
    {
      "entropy": 5.309054660797119,
      "epoch": 1.1143131604226706,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004883649041013335,
      "loss": 5.176,
      "mean_token_accuracy": 0.2003674700856209,
      "num_tokens": 26596042.0,
      "step": 11600
    },
    {
      "entropy": 5.2267273426055905,
      "epoch": 1.1147934678194045,
      "grad_norm": 1.25,
      "learning_rate": 0.000488354019687472,
      "loss": 5.0756,
      "mean_token_accuracy": 0.20230771452188492,
      "num_tokens": 26607359.0,
      "step": 11605
    },
    {
      "entropy": 5.28858118057251,
      "epoch": 1.1152737752161384,
      "grad_norm": 1.25,
      "learning_rate": 0.0004883431303201636,
      "loss": 4.9973,
      "mean_token_accuracy": 0.2130118727684021,
      "num_tokens": 26617095.0,
      "step": 11610
    },
    {
      "entropy": 5.218375635147095,
      "epoch": 1.1157540826128722,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004883322359996613,
      "loss": 5.0796,
      "mean_token_accuracy": 0.20446051210165023,
      "num_tokens": 26628450.0,
      "step": 11615
    },
    {
      "entropy": 5.200555324554443,
      "epoch": 1.1162343900096061,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004883213367262179,
      "loss": 4.9821,
      "mean_token_accuracy": 0.21193538308143617,
      "num_tokens": 26639428.0,
      "step": 11620
    },
    {
      "entropy": 5.1242955207824705,
      "epoch": 1.11671469740634,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004883104325000866,
      "loss": 4.9713,
      "mean_token_accuracy": 0.21039544492959977,
      "num_tokens": 26650730.0,
      "step": 11625
    },
    {
      "entropy": 5.1885487079620365,
      "epoch": 1.1171950048030739,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004882995233215203,
      "loss": 4.9985,
      "mean_token_accuracy": 0.20951008945703506,
      "num_tokens": 26661745.0,
      "step": 11630
    },
    {
      "entropy": 5.1915202140808105,
      "epoch": 1.117675312199808,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004882886091907726,
      "loss": 4.9185,
      "mean_token_accuracy": 0.21715635508298875,
      "num_tokens": 26672976.0,
      "step": 11635
    },
    {
      "entropy": 5.275407075881958,
      "epoch": 1.1181556195965419,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048827769010809666,
      "loss": 5.0743,
      "mean_token_accuracy": 0.20475854426622392,
      "num_tokens": 26683503.0,
      "step": 11640
    },
    {
      "entropy": 5.237531566619873,
      "epoch": 1.1186359269932757,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048826676607374606,
      "loss": 5.0277,
      "mean_token_accuracy": 0.20738735049962997,
      "num_tokens": 26695160.0,
      "step": 11645
    },
    {
      "entropy": 5.139219951629639,
      "epoch": 1.1191162343900096,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048825583708797434,
      "loss": 5.0251,
      "mean_token_accuracy": 0.21477459818124772,
      "num_tokens": 26706514.0,
      "step": 11650
    },
    {
      "entropy": 5.1160496234893795,
      "epoch": 1.1195965417867435,
      "grad_norm": 1.375,
      "learning_rate": 0.0004882449031510354,
      "loss": 4.9191,
      "mean_token_accuracy": 0.21820003688335418,
      "num_tokens": 26717268.0,
      "step": 11655
    },
    {
      "entropy": 5.344860458374024,
      "epoch": 1.1200768491834774,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000488233964263183,
      "loss": 5.1864,
      "mean_token_accuracy": 0.19938425421714784,
      "num_tokens": 26728821.0,
      "step": 11660
    },
    {
      "entropy": 5.27126898765564,
      "epoch": 1.1205571565802113,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048822302042467115,
      "loss": 5.1211,
      "mean_token_accuracy": 0.2011058047413826,
      "num_tokens": 26740476.0,
      "step": 11665
    },
    {
      "entropy": 5.2651325225830075,
      "epoch": 1.1210374639769451,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004882120716357539,
      "loss": 5.1298,
      "mean_token_accuracy": 0.20450747162103652,
      "num_tokens": 26751289.0,
      "step": 11670
    },
    {
      "entropy": 5.203167343139649,
      "epoch": 1.1215177713736793,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004882011178966854,
      "loss": 5.0239,
      "mean_token_accuracy": 0.21840595752000808,
      "num_tokens": 26762115.0,
      "step": 11675
    },
    {
      "entropy": 5.150345325469971,
      "epoch": 1.1219980787704131,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004881901592077201,
      "loss": 4.9659,
      "mean_token_accuracy": 0.2084621950984001,
      "num_tokens": 26772837.0,
      "step": 11680
    },
    {
      "entropy": 5.270205640792847,
      "epoch": 1.122478386167147,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004881791955691123,
      "loss": 5.0553,
      "mean_token_accuracy": 0.20935554504394532,
      "num_tokens": 26783721.0,
      "step": 11685
    },
    {
      "entropy": 5.200037622451783,
      "epoch": 1.122958693563881,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048816822698111655,
      "loss": 5.0013,
      "mean_token_accuracy": 0.22180945724248885,
      "num_tokens": 26794320.0,
      "step": 11690
    },
    {
      "entropy": 5.250467014312744,
      "epoch": 1.1234390009606148,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048815725344398766,
      "loss": 5.1531,
      "mean_token_accuracy": 0.19589796364307405,
      "num_tokens": 26807053.0,
      "step": 11695
    },
    {
      "entropy": 5.326962566375732,
      "epoch": 1.1239193083573487,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048814627495798017,
      "loss": 5.0568,
      "mean_token_accuracy": 0.20563538670539855,
      "num_tokens": 26817245.0,
      "step": 11700
    },
    {
      "entropy": 5.208544683456421,
      "epoch": 1.1243996157540825,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004881352915233492,
      "loss": 5.0918,
      "mean_token_accuracy": 0.20330672115087509,
      "num_tokens": 26829498.0,
      "step": 11705
    },
    {
      "entropy": 5.199560737609863,
      "epoch": 1.1248799231508164,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048812430314034956,
      "loss": 4.9488,
      "mean_token_accuracy": 0.21759573221206666,
      "num_tokens": 26840534.0,
      "step": 11710
    },
    {
      "entropy": 5.231953382492065,
      "epoch": 1.1253602305475505,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004881133098092365,
      "loss": 5.035,
      "mean_token_accuracy": 0.20486913919448851,
      "num_tokens": 26851305.0,
      "step": 11715
    },
    {
      "entropy": 5.235578632354736,
      "epoch": 1.1258405379442844,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004881023115302652,
      "loss": 5.0319,
      "mean_token_accuracy": 0.2076029971241951,
      "num_tokens": 26862620.0,
      "step": 11720
    },
    {
      "entropy": 5.248742008209229,
      "epoch": 1.1263208453410183,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000488091308303691,
      "loss": 5.0477,
      "mean_token_accuracy": 0.2045100063085556,
      "num_tokens": 26874108.0,
      "step": 11725
    },
    {
      "entropy": 5.254848670959473,
      "epoch": 1.1268011527377522,
      "grad_norm": 1.25,
      "learning_rate": 0.0004880803001297694,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20787989348173141,
      "num_tokens": 26886523.0,
      "step": 11730
    },
    {
      "entropy": 5.196786642074585,
      "epoch": 1.127281460134486,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004880692870087559,
      "loss": 4.9729,
      "mean_token_accuracy": 0.21556005477905274,
      "num_tokens": 26897316.0,
      "step": 11735
    },
    {
      "entropy": 5.164411163330078,
      "epoch": 1.12776176753122,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048805826894090626,
      "loss": 5.0114,
      "mean_token_accuracy": 0.20621836334466934,
      "num_tokens": 26909807.0,
      "step": 11740
    },
    {
      "entropy": 5.280880498886108,
      "epoch": 1.1282420749279538,
      "grad_norm": 1.25,
      "learning_rate": 0.00048804724592647626,
      "loss": 5.1149,
      "mean_token_accuracy": 0.20533014386892318,
      "num_tokens": 26922054.0,
      "step": 11745
    },
    {
      "entropy": 5.226164817810059,
      "epoch": 1.1287223823246877,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004880362179657218,
      "loss": 4.9842,
      "mean_token_accuracy": 0.20833992213010788,
      "num_tokens": 26933838.0,
      "step": 11750
    },
    {
      "entropy": 5.193318128585815,
      "epoch": 1.1292026897214218,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048802518505889904,
      "loss": 5.0591,
      "mean_token_accuracy": 0.2035563752055168,
      "num_tokens": 26945396.0,
      "step": 11755
    },
    {
      "entropy": 5.275278806686401,
      "epoch": 1.1296829971181557,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048801414720626404,
      "loss": 5.1409,
      "mean_token_accuracy": 0.20095735788345337,
      "num_tokens": 26957809.0,
      "step": 11760
    },
    {
      "entropy": 5.2284914493560795,
      "epoch": 1.1301633045148896,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048800310440807294,
      "loss": 4.9788,
      "mean_token_accuracy": 0.21738197803497314,
      "num_tokens": 26969024.0,
      "step": 11765
    },
    {
      "entropy": 5.1334388256073,
      "epoch": 1.1306436119116234,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004879920566645823,
      "loss": 5.0446,
      "mean_token_accuracy": 0.20855933874845506,
      "num_tokens": 26980689.0,
      "step": 11770
    },
    {
      "entropy": 5.281749439239502,
      "epoch": 1.1311239193083573,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004879810039760486,
      "loss": 5.1025,
      "mean_token_accuracy": 0.19795218706130982,
      "num_tokens": 26992115.0,
      "step": 11775
    },
    {
      "entropy": 5.288855123519897,
      "epoch": 1.1316042267050912,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004879699463427284,
      "loss": 5.0357,
      "mean_token_accuracy": 0.2074179768562317,
      "num_tokens": 27003042.0,
      "step": 11780
    },
    {
      "entropy": 5.1747575283050535,
      "epoch": 1.132084534101825,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004879588837648785,
      "loss": 4.9706,
      "mean_token_accuracy": 0.2176398977637291,
      "num_tokens": 27013863.0,
      "step": 11785
    },
    {
      "entropy": 5.087520122528076,
      "epoch": 1.1325648414985592,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048794781624275554,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21497785598039626,
      "num_tokens": 27024423.0,
      "step": 11790
    },
    {
      "entropy": 5.138603782653808,
      "epoch": 1.133045148895293,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048793674377661664,
      "loss": 5.0342,
      "mean_token_accuracy": 0.20879024714231492,
      "num_tokens": 27035460.0,
      "step": 11795
    },
    {
      "entropy": 5.187189197540283,
      "epoch": 1.133525456292027,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048792566636671886,
      "loss": 5.044,
      "mean_token_accuracy": 0.20487866699695587,
      "num_tokens": 27046731.0,
      "step": 11800
    },
    {
      "entropy": 5.26811261177063,
      "epoch": 1.1340057636887608,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004879145840133194,
      "loss": 5.0315,
      "mean_token_accuracy": 0.21277425289154053,
      "num_tokens": 27057758.0,
      "step": 11805
    },
    {
      "entropy": 5.3031150817871096,
      "epoch": 1.1344860710854947,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004879034967166755,
      "loss": 5.0973,
      "mean_token_accuracy": 0.20476967245340347,
      "num_tokens": 27068055.0,
      "step": 11810
    },
    {
      "entropy": 5.180997562408447,
      "epoch": 1.1349663784822286,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004878924044770446,
      "loss": 4.9762,
      "mean_token_accuracy": 0.2072560265660286,
      "num_tokens": 27078602.0,
      "step": 11815
    },
    {
      "entropy": 5.151785182952881,
      "epoch": 1.1354466858789625,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004878813072946843,
      "loss": 5.0353,
      "mean_token_accuracy": 0.20635258555412292,
      "num_tokens": 27089944.0,
      "step": 11820
    },
    {
      "entropy": 5.223290538787841,
      "epoch": 1.1359269932756964,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048787020516985203,
      "loss": 5.0056,
      "mean_token_accuracy": 0.2173451825976372,
      "num_tokens": 27100486.0,
      "step": 11825
    },
    {
      "entropy": 5.223959970474243,
      "epoch": 1.1364073006724302,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048785909810280576,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20904099941253662,
      "num_tokens": 27111435.0,
      "step": 11830
    },
    {
      "entropy": 5.241550016403198,
      "epoch": 1.1368876080691643,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004878479860938033,
      "loss": 5.0836,
      "mean_token_accuracy": 0.2063623070716858,
      "num_tokens": 27123482.0,
      "step": 11835
    },
    {
      "entropy": 5.219643068313599,
      "epoch": 1.1373679154658982,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048783686914310266,
      "loss": 5.0003,
      "mean_token_accuracy": 0.21106838881969453,
      "num_tokens": 27135009.0,
      "step": 11840
    },
    {
      "entropy": 5.185359907150269,
      "epoch": 1.137848222862632,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004878257472509619,
      "loss": 4.9829,
      "mean_token_accuracy": 0.216535484790802,
      "num_tokens": 27145230.0,
      "step": 11845
    },
    {
      "entropy": 5.146259689331055,
      "epoch": 1.138328530259366,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004878146204176392,
      "loss": 5.0041,
      "mean_token_accuracy": 0.21105142086744308,
      "num_tokens": 27156490.0,
      "step": 11850
    },
    {
      "entropy": 5.213765287399292,
      "epoch": 1.1388088376560999,
      "grad_norm": 1.59375,
      "learning_rate": 0.000487803488643393,
      "loss": 5.0064,
      "mean_token_accuracy": 0.2123056322336197,
      "num_tokens": 27166727.0,
      "step": 11855
    },
    {
      "entropy": 5.2038267135620115,
      "epoch": 1.1392891450528337,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048779235192848166,
      "loss": 4.9641,
      "mean_token_accuracy": 0.20864419490098954,
      "num_tokens": 27177969.0,
      "step": 11860
    },
    {
      "entropy": 5.205870628356934,
      "epoch": 1.1397694524495678,
      "grad_norm": 1.25,
      "learning_rate": 0.0004877812102731638,
      "loss": 5.1126,
      "mean_token_accuracy": 0.21036774665117264,
      "num_tokens": 27189637.0,
      "step": 11865
    },
    {
      "entropy": 5.379373693466187,
      "epoch": 1.1402497598463017,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048777006367769804,
      "loss": 5.2138,
      "mean_token_accuracy": 0.20511647462844848,
      "num_tokens": 27201124.0,
      "step": 11870
    },
    {
      "entropy": 5.291901731491089,
      "epoch": 1.1407300672430356,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004877589121423432,
      "loss": 5.0858,
      "mean_token_accuracy": 0.20831723362207413,
      "num_tokens": 27212399.0,
      "step": 11875
    },
    {
      "entropy": 5.2092608451843265,
      "epoch": 1.1412103746397695,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004877477556673582,
      "loss": 5.0157,
      "mean_token_accuracy": 0.21417011171579362,
      "num_tokens": 27224469.0,
      "step": 11880
    },
    {
      "entropy": 5.219428968429566,
      "epoch": 1.1416906820365034,
      "grad_norm": 1.1875,
      "learning_rate": 0.000487736594253002,
      "loss": 5.0029,
      "mean_token_accuracy": 0.21346651911735534,
      "num_tokens": 27236026.0,
      "step": 11885
    },
    {
      "entropy": 5.192475175857544,
      "epoch": 1.1421709894332372,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048772542789953384,
      "loss": 5.0052,
      "mean_token_accuracy": 0.20912211984395981,
      "num_tokens": 27248332.0,
      "step": 11890
    },
    {
      "entropy": 5.224642705917359,
      "epoch": 1.1426512968299711,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048771425660721284,
      "loss": 4.9678,
      "mean_token_accuracy": 0.21154877543449402,
      "num_tokens": 27258701.0,
      "step": 11895
    },
    {
      "entropy": 5.2436291694641115,
      "epoch": 1.143131604226705,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048770308037629853,
      "loss": 5.1045,
      "mean_token_accuracy": 0.20591199696063994,
      "num_tokens": 27269808.0,
      "step": 11900
    },
    {
      "entropy": 5.251409816741943,
      "epoch": 1.143611911623439,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004876918992070502,
      "loss": 5.1241,
      "mean_token_accuracy": 0.20272381007671356,
      "num_tokens": 27282440.0,
      "step": 11905
    },
    {
      "entropy": 5.312520837783813,
      "epoch": 1.144092219020173,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004876807130997276,
      "loss": 5.1259,
      "mean_token_accuracy": 0.2114517942070961,
      "num_tokens": 27294430.0,
      "step": 11910
    },
    {
      "entropy": 5.269687557220459,
      "epoch": 1.1445725264169069,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004876695220545903,
      "loss": 5.1018,
      "mean_token_accuracy": 0.20862225145101548,
      "num_tokens": 27305896.0,
      "step": 11915
    },
    {
      "entropy": 5.324378967285156,
      "epoch": 1.1450528338136408,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048765832607189824,
      "loss": 5.0988,
      "mean_token_accuracy": 0.20127609074115754,
      "num_tokens": 27318195.0,
      "step": 11920
    },
    {
      "entropy": 5.175506639480591,
      "epoch": 1.1455331412103746,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048764712515191136,
      "loss": 5.0049,
      "mean_token_accuracy": 0.20919490605592728,
      "num_tokens": 27329055.0,
      "step": 11925
    },
    {
      "entropy": 5.177300453186035,
      "epoch": 1.1460134486071085,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048763591929488966,
      "loss": 5.0051,
      "mean_token_accuracy": 0.2174530118703842,
      "num_tokens": 27340296.0,
      "step": 11930
    },
    {
      "entropy": 5.267404413223266,
      "epoch": 1.1464937560038424,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004876247085010933,
      "loss": 5.0404,
      "mean_token_accuracy": 0.21036536991596222,
      "num_tokens": 27350995.0,
      "step": 11935
    },
    {
      "entropy": 5.221362781524658,
      "epoch": 1.1469740634005763,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048761349277078253,
      "loss": 5.0874,
      "mean_token_accuracy": 0.2018448144197464,
      "num_tokens": 27362463.0,
      "step": 11940
    },
    {
      "entropy": 5.222718954086304,
      "epoch": 1.1474543707973104,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048760227210421775,
      "loss": 4.9497,
      "mean_token_accuracy": 0.22092660516500473,
      "num_tokens": 27373579.0,
      "step": 11945
    },
    {
      "entropy": 5.201131677627563,
      "epoch": 1.1479346781940443,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004875910465016596,
      "loss": 4.9539,
      "mean_token_accuracy": 0.21881531327962875,
      "num_tokens": 27383813.0,
      "step": 11950
    },
    {
      "entropy": 5.175941801071167,
      "epoch": 1.1484149855907781,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004875798159633686,
      "loss": 5.0704,
      "mean_token_accuracy": 0.2024243652820587,
      "num_tokens": 27395651.0,
      "step": 11955
    },
    {
      "entropy": 5.266213846206665,
      "epoch": 1.148895292987512,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004875685804896055,
      "loss": 5.0691,
      "mean_token_accuracy": 0.20774777680635453,
      "num_tokens": 27407827.0,
      "step": 11960
    },
    {
      "entropy": 5.2271607398986815,
      "epoch": 1.149375600384246,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004875573400806312,
      "loss": 5.0829,
      "mean_token_accuracy": 0.20532586872577668,
      "num_tokens": 27418777.0,
      "step": 11965
    },
    {
      "entropy": 5.265883922576904,
      "epoch": 1.1498559077809798,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048754609473670654,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20701631754636765,
      "num_tokens": 27429626.0,
      "step": 11970
    },
    {
      "entropy": 5.291296482086182,
      "epoch": 1.1503362151777137,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004875348444580927,
      "loss": 5.2764,
      "mean_token_accuracy": 0.1975032925605774,
      "num_tokens": 27442304.0,
      "step": 11975
    },
    {
      "entropy": 5.2313751697540285,
      "epoch": 1.1508165225744476,
      "grad_norm": 1.203125,
      "learning_rate": 0.000487523589245051,
      "loss": 5.0191,
      "mean_token_accuracy": 0.21473737806081772,
      "num_tokens": 27454421.0,
      "step": 11980
    },
    {
      "entropy": 5.137607145309448,
      "epoch": 1.1512968299711814,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004875123290978425,
      "loss": 4.9102,
      "mean_token_accuracy": 0.21617841273546218,
      "num_tokens": 27465220.0,
      "step": 11985
    },
    {
      "entropy": 5.166693067550659,
      "epoch": 1.1517771373679155,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048750106401672876,
      "loss": 5.026,
      "mean_token_accuracy": 0.2057954788208008,
      "num_tokens": 27476542.0,
      "step": 11990
    },
    {
      "entropy": 5.182518482208252,
      "epoch": 1.1522574447646494,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048748979400197134,
      "loss": 4.9247,
      "mean_token_accuracy": 0.21806834936141967,
      "num_tokens": 27487721.0,
      "step": 11995
    },
    {
      "entropy": 5.208222246170044,
      "epoch": 1.1527377521613833,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048747851905383183,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20998998135328292,
      "num_tokens": 27498982.0,
      "step": 12000
    },
    {
      "epoch": 1.1527377521613833,
      "eval_entropy": 5.0481731802123395,
      "eval_loss": 5.137667179107666,
      "eval_mean_token_accuracy": 0.21272955124953472,
      "eval_num_tokens": 27498982.0,
      "eval_runtime": 26.4942,
      "eval_samples_per_second": 1238.571,
      "eval_steps_per_second": 154.826,
      "step": 12000
    },
    {
      "entropy": 5.1353850841522215,
      "epoch": 1.1532180595581172,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004874672391725721,
      "loss": 4.9261,
      "mean_token_accuracy": 0.21388213485479354,
      "num_tokens": 27509863.0,
      "step": 12005
    },
    {
      "entropy": 5.183210992813111,
      "epoch": 1.153698366954851,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004874559543584539,
      "loss": 5.0332,
      "mean_token_accuracy": 0.20671313107013703,
      "num_tokens": 27521061.0,
      "step": 12010
    },
    {
      "entropy": 5.211726665496826,
      "epoch": 1.154178674351585,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004874446646117394,
      "loss": 5.0404,
      "mean_token_accuracy": 0.20665457844734192,
      "num_tokens": 27532187.0,
      "step": 12015
    },
    {
      "entropy": 5.212849807739258,
      "epoch": 1.154658981748319,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004874333699326906,
      "loss": 5.027,
      "mean_token_accuracy": 0.21120043843984604,
      "num_tokens": 27543371.0,
      "step": 12020
    },
    {
      "entropy": 5.146895503997802,
      "epoch": 1.155139289145053,
      "grad_norm": 1.25,
      "learning_rate": 0.0004874220703215697,
      "loss": 4.9714,
      "mean_token_accuracy": 0.21074630320072174,
      "num_tokens": 27554423.0,
      "step": 12025
    },
    {
      "entropy": 5.207942485809326,
      "epoch": 1.1556195965417868,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004874107657786391,
      "loss": 5.0232,
      "mean_token_accuracy": 0.2101215198636055,
      "num_tokens": 27566397.0,
      "step": 12030
    },
    {
      "entropy": 5.2997599124908445,
      "epoch": 1.1560999039385207,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048739945630416124,
      "loss": 5.0772,
      "mean_token_accuracy": 0.21079109609127045,
      "num_tokens": 27578668.0,
      "step": 12035
    },
    {
      "entropy": 5.170396757125855,
      "epoch": 1.1565802113352546,
      "grad_norm": 1.25,
      "learning_rate": 0.0004873881418983987,
      "loss": 5.039,
      "mean_token_accuracy": 0.20226115733385086,
      "num_tokens": 27589345.0,
      "step": 12040
    },
    {
      "entropy": 5.101444911956787,
      "epoch": 1.1570605187319885,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004873768225616141,
      "loss": 4.9342,
      "mean_token_accuracy": 0.21675882190465928,
      "num_tokens": 27600482.0,
      "step": 12045
    },
    {
      "entropy": 5.258086347579956,
      "epoch": 1.1575408261287223,
      "grad_norm": 1.25,
      "learning_rate": 0.00048736549829407047,
      "loss": 5.0845,
      "mean_token_accuracy": 0.20765952169895172,
      "num_tokens": 27611543.0,
      "step": 12050
    },
    {
      "entropy": 5.281292915344238,
      "epoch": 1.1580211335254562,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004873541690960305,
      "loss": 5.0317,
      "mean_token_accuracy": 0.21217868179082872,
      "num_tokens": 27623503.0,
      "step": 12055
    },
    {
      "entropy": 5.116846132278442,
      "epoch": 1.15850144092219,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004873428349677573,
      "loss": 4.9128,
      "mean_token_accuracy": 0.22069915533065795,
      "num_tokens": 27633593.0,
      "step": 12060
    },
    {
      "entropy": 5.241214847564697,
      "epoch": 1.1589817483189242,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000487331495909514,
      "loss": 5.0855,
      "mean_token_accuracy": 0.20452489703893661,
      "num_tokens": 27646339.0,
      "step": 12065
    },
    {
      "entropy": 5.103432655334473,
      "epoch": 1.159462055715658,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048732015192156383,
      "loss": 4.837,
      "mean_token_accuracy": 0.22639833837747575,
      "num_tokens": 27657314.0,
      "step": 12070
    },
    {
      "entropy": 5.1560302734375,
      "epoch": 1.159942363112392,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048730880300417015,
      "loss": 5.0494,
      "mean_token_accuracy": 0.20824081152677537,
      "num_tokens": 27668680.0,
      "step": 12075
    },
    {
      "entropy": 5.192676019668579,
      "epoch": 1.1604226705091258,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048729744915759657,
      "loss": 5.033,
      "mean_token_accuracy": 0.2095215454697609,
      "num_tokens": 27679628.0,
      "step": 12080
    },
    {
      "entropy": 5.210277652740478,
      "epoch": 1.1609029779058597,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048728609038210655,
      "loss": 5.0836,
      "mean_token_accuracy": 0.21177640706300735,
      "num_tokens": 27691205.0,
      "step": 12085
    },
    {
      "entropy": 5.234472990036011,
      "epoch": 1.1613832853025936,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048727472667796395,
      "loss": 5.0244,
      "mean_token_accuracy": 0.2144126906991005,
      "num_tokens": 27702903.0,
      "step": 12090
    },
    {
      "entropy": 5.295352268218994,
      "epoch": 1.1618635926993275,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004872633580454325,
      "loss": 5.1342,
      "mean_token_accuracy": 0.1973687991499901,
      "num_tokens": 27716554.0,
      "step": 12095
    },
    {
      "entropy": 5.282321977615356,
      "epoch": 1.1623439000960616,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048725198448477616,
      "loss": 4.9929,
      "mean_token_accuracy": 0.21639626622200012,
      "num_tokens": 27727931.0,
      "step": 12100
    },
    {
      "entropy": 5.233769750595092,
      "epoch": 1.1628242074927955,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048724060599625893,
      "loss": 5.0894,
      "mean_token_accuracy": 0.20936337411403655,
      "num_tokens": 27739483.0,
      "step": 12105
    },
    {
      "entropy": 5.251340579986572,
      "epoch": 1.1633045148895294,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048722922258014506,
      "loss": 5.039,
      "mean_token_accuracy": 0.20441953986883163,
      "num_tokens": 27751157.0,
      "step": 12110
    },
    {
      "entropy": 5.314346170425415,
      "epoch": 1.1637848222862632,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004872178342366989,
      "loss": 5.1291,
      "mean_token_accuracy": 0.203353089094162,
      "num_tokens": 27762190.0,
      "step": 12115
    },
    {
      "entropy": 5.188182640075683,
      "epoch": 1.1642651296829971,
      "grad_norm": 1.375,
      "learning_rate": 0.00048720644096618475,
      "loss": 4.9951,
      "mean_token_accuracy": 0.20872483849525453,
      "num_tokens": 27773684.0,
      "step": 12120
    },
    {
      "entropy": 5.270470428466797,
      "epoch": 1.164745437079731,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004871950427688672,
      "loss": 5.0749,
      "mean_token_accuracy": 0.20176429599523543,
      "num_tokens": 27785388.0,
      "step": 12125
    },
    {
      "entropy": 5.17701473236084,
      "epoch": 1.1652257444764649,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048718363964501087,
      "loss": 4.9805,
      "mean_token_accuracy": 0.21482060253620147,
      "num_tokens": 27797603.0,
      "step": 12130
    },
    {
      "entropy": 5.270160245895386,
      "epoch": 1.1657060518731988,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004871722315948805,
      "loss": 5.1251,
      "mean_token_accuracy": 0.20826165974140168,
      "num_tokens": 27809444.0,
      "step": 12135
    },
    {
      "entropy": 5.285105228424072,
      "epoch": 1.1661863592699326,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004871608186187408,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20267860293388368,
      "num_tokens": 27820213.0,
      "step": 12140
    },
    {
      "entropy": 5.1990638256073,
      "epoch": 1.1666666666666667,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048714940071685703,
      "loss": 4.978,
      "mean_token_accuracy": 0.21349021941423416,
      "num_tokens": 27831564.0,
      "step": 12145
    },
    {
      "entropy": 5.284900283813476,
      "epoch": 1.1671469740634006,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048713797788949405,
      "loss": 5.0915,
      "mean_token_accuracy": 0.2030651569366455,
      "num_tokens": 27843686.0,
      "step": 12150
    },
    {
      "entropy": 5.326258134841919,
      "epoch": 1.1676272814601345,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048712655013691714,
      "loss": 5.0905,
      "mean_token_accuracy": 0.2048892229795456,
      "num_tokens": 27855428.0,
      "step": 12155
    },
    {
      "entropy": 5.3282177448272705,
      "epoch": 1.1681075888568684,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048711511745939165,
      "loss": 5.2125,
      "mean_token_accuracy": 0.20303212106227875,
      "num_tokens": 27866733.0,
      "step": 12160
    },
    {
      "entropy": 5.2372795104980465,
      "epoch": 1.1685878962536023,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000487103679857183,
      "loss": 5.0434,
      "mean_token_accuracy": 0.20275867432355882,
      "num_tokens": 27878639.0,
      "step": 12165
    },
    {
      "entropy": 5.148620700836181,
      "epoch": 1.1690682036503361,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004870922373305567,
      "loss": 5.0807,
      "mean_token_accuracy": 0.20348057001829148,
      "num_tokens": 27892139.0,
      "step": 12170
    },
    {
      "entropy": 5.320078420639038,
      "epoch": 1.1695485110470702,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048708078987977837,
      "loss": 5.12,
      "mean_token_accuracy": 0.20555612742900847,
      "num_tokens": 27904395.0,
      "step": 12175
    },
    {
      "entropy": 5.284656715393067,
      "epoch": 1.1700288184438041,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048706933750511394,
      "loss": 5.0748,
      "mean_token_accuracy": 0.20533101111650467,
      "num_tokens": 27916444.0,
      "step": 12180
    },
    {
      "entropy": 5.265898180007935,
      "epoch": 1.170509125840538,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004870578802068292,
      "loss": 5.0442,
      "mean_token_accuracy": 0.2167959377169609,
      "num_tokens": 27926914.0,
      "step": 12185
    },
    {
      "entropy": 5.228262519836425,
      "epoch": 1.170989433237272,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048704641798519006,
      "loss": 4.9678,
      "mean_token_accuracy": 0.21131121516227722,
      "num_tokens": 27938203.0,
      "step": 12190
    },
    {
      "entropy": 5.266461133956909,
      "epoch": 1.1714697406340058,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048703495084046286,
      "loss": 5.061,
      "mean_token_accuracy": 0.21243891268968582,
      "num_tokens": 27950463.0,
      "step": 12195
    },
    {
      "entropy": 5.21266508102417,
      "epoch": 1.1719500480307397,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004870234787729137,
      "loss": 4.9739,
      "mean_token_accuracy": 0.2062857836484909,
      "num_tokens": 27962312.0,
      "step": 12200
    },
    {
      "entropy": 5.1710998058319095,
      "epoch": 1.1724303554274735,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004870120017828089,
      "loss": 4.9898,
      "mean_token_accuracy": 0.21472591906785965,
      "num_tokens": 27973723.0,
      "step": 12205
    },
    {
      "entropy": 5.186633729934693,
      "epoch": 1.1729106628242074,
      "grad_norm": 1.34375,
      "learning_rate": 0.000487000519870415,
      "loss": 4.9505,
      "mean_token_accuracy": 0.21941281259059905,
      "num_tokens": 27983816.0,
      "step": 12210
    },
    {
      "entropy": 5.0969573020935055,
      "epoch": 1.1733909702209413,
      "grad_norm": 1.75,
      "learning_rate": 0.0004869890330359986,
      "loss": 5.0231,
      "mean_token_accuracy": 0.21480425149202348,
      "num_tokens": 27996516.0,
      "step": 12215
    },
    {
      "entropy": 5.190220832824707,
      "epoch": 1.1738712776176754,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004869775412798262,
      "loss": 5.0335,
      "mean_token_accuracy": 0.2105358585715294,
      "num_tokens": 28008256.0,
      "step": 12220
    },
    {
      "entropy": 5.160996389389038,
      "epoch": 1.1743515850144093,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048696604460216476,
      "loss": 4.8917,
      "mean_token_accuracy": 0.22466871589422227,
      "num_tokens": 28018882.0,
      "step": 12225
    },
    {
      "entropy": 5.2266583919525145,
      "epoch": 1.1748318924111432,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048695454300328123,
      "loss": 4.9985,
      "mean_token_accuracy": 0.21036388874053955,
      "num_tokens": 28030399.0,
      "step": 12230
    },
    {
      "entropy": 5.2568529605865475,
      "epoch": 1.175312199807877,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048694303648344256,
      "loss": 5.0618,
      "mean_token_accuracy": 0.21028077155351638,
      "num_tokens": 28041621.0,
      "step": 12235
    },
    {
      "entropy": 5.248664474487304,
      "epoch": 1.175792507204611,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048693152504291595,
      "loss": 4.9414,
      "mean_token_accuracy": 0.2103190928697586,
      "num_tokens": 28053198.0,
      "step": 12240
    },
    {
      "entropy": 5.173239469528198,
      "epoch": 1.1762728146013448,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004869200086819686,
      "loss": 4.9912,
      "mean_token_accuracy": 0.2124918982386589,
      "num_tokens": 28064095.0,
      "step": 12245
    },
    {
      "entropy": 5.221928071975708,
      "epoch": 1.1767531219980787,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048690848740086796,
      "loss": 5.0612,
      "mean_token_accuracy": 0.2014385148882866,
      "num_tokens": 28075665.0,
      "step": 12250
    },
    {
      "entropy": 5.259342050552368,
      "epoch": 1.1772334293948128,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004868969611998814,
      "loss": 5.0872,
      "mean_token_accuracy": 0.2091042011976242,
      "num_tokens": 28089166.0,
      "step": 12255
    },
    {
      "entropy": 5.338870573043823,
      "epoch": 1.1777137367915467,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004868854300792767,
      "loss": 5.106,
      "mean_token_accuracy": 0.20090546011924743,
      "num_tokens": 28101379.0,
      "step": 12260
    },
    {
      "entropy": 5.229467248916626,
      "epoch": 1.1781940441882806,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048687389403932144,
      "loss": 5.0514,
      "mean_token_accuracy": 0.20802178531885146,
      "num_tokens": 28112836.0,
      "step": 12265
    },
    {
      "entropy": 5.2618410110473635,
      "epoch": 1.1786743515850144,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004868623530802835,
      "loss": 5.1356,
      "mean_token_accuracy": 0.20263897329568864,
      "num_tokens": 28125039.0,
      "step": 12270
    },
    {
      "entropy": 5.251055955886841,
      "epoch": 1.1791546589817483,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048685080720243086,
      "loss": 4.9975,
      "mean_token_accuracy": 0.21049689054489135,
      "num_tokens": 28137504.0,
      "step": 12275
    },
    {
      "entropy": 5.248031091690064,
      "epoch": 1.1796349663784822,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004868392564060315,
      "loss": 5.0328,
      "mean_token_accuracy": 0.21265994310379027,
      "num_tokens": 28149483.0,
      "step": 12280
    },
    {
      "entropy": 5.232854175567627,
      "epoch": 1.180115273775216,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004868277006913537,
      "loss": 5.0294,
      "mean_token_accuracy": 0.20812107771635055,
      "num_tokens": 28161001.0,
      "step": 12285
    },
    {
      "entropy": 5.21848087310791,
      "epoch": 1.18059558117195,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004868161400586656,
      "loss": 5.0629,
      "mean_token_accuracy": 0.2031179279088974,
      "num_tokens": 28171146.0,
      "step": 12290
    },
    {
      "entropy": 5.3159027099609375,
      "epoch": 1.1810758885686838,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004868045745082357,
      "loss": 5.074,
      "mean_token_accuracy": 0.20277472585439682,
      "num_tokens": 28182870.0,
      "step": 12295
    },
    {
      "entropy": 5.216927814483642,
      "epoch": 1.181556195965418,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004867930040403326,
      "loss": 5.0853,
      "mean_token_accuracy": 0.21220019906759263,
      "num_tokens": 28194457.0,
      "step": 12300
    },
    {
      "entropy": 5.290600919723511,
      "epoch": 1.1820365033621518,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048678142865522475,
      "loss": 5.1097,
      "mean_token_accuracy": 0.20215620696544648,
      "num_tokens": 28205818.0,
      "step": 12305
    },
    {
      "entropy": 5.216201543807983,
      "epoch": 1.1825168107588857,
      "grad_norm": 1.328125,
      "learning_rate": 0.000486769848353181,
      "loss": 4.9493,
      "mean_token_accuracy": 0.21080254018306732,
      "num_tokens": 28216124.0,
      "step": 12310
    },
    {
      "entropy": 5.22677960395813,
      "epoch": 1.1829971181556196,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048675826313447027,
      "loss": 5.0495,
      "mean_token_accuracy": 0.21014672219753266,
      "num_tokens": 28227485.0,
      "step": 12315
    },
    {
      "entropy": 5.268592834472656,
      "epoch": 1.1834774255523535,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048674667299936135,
      "loss": 5.0825,
      "mean_token_accuracy": 0.20865130424499512,
      "num_tokens": 28239007.0,
      "step": 12320
    },
    {
      "entropy": 5.269042444229126,
      "epoch": 1.1839577329490873,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048673507794812356,
      "loss": 5.0527,
      "mean_token_accuracy": 0.2045225605368614,
      "num_tokens": 28250926.0,
      "step": 12325
    },
    {
      "entropy": 5.256142616271973,
      "epoch": 1.1844380403458212,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004867234779810259,
      "loss": 5.0646,
      "mean_token_accuracy": 0.20930347740650176,
      "num_tokens": 28263182.0,
      "step": 12330
    },
    {
      "entropy": 5.154686307907104,
      "epoch": 1.1849183477425553,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004867118730983378,
      "loss": 5.0056,
      "mean_token_accuracy": 0.21611176580190658,
      "num_tokens": 28274666.0,
      "step": 12335
    },
    {
      "entropy": 5.136308431625366,
      "epoch": 1.1853986551392892,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004867002633003286,
      "loss": 4.8964,
      "mean_token_accuracy": 0.2246379092335701,
      "num_tokens": 28286007.0,
      "step": 12340
    },
    {
      "entropy": 5.208309459686279,
      "epoch": 1.185878962536023,
      "grad_norm": 1.140625,
      "learning_rate": 0.000486688648587268,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21893275827169417,
      "num_tokens": 28298262.0,
      "step": 12345
    },
    {
      "entropy": 5.137412786483765,
      "epoch": 1.186359269932757,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004866770289594256,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21728730201721191,
      "num_tokens": 28309428.0,
      "step": 12350
    },
    {
      "entropy": 5.261605596542358,
      "epoch": 1.1868395773294909,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048666540441707107,
      "loss": 5.0972,
      "mean_token_accuracy": 0.20952664762735368,
      "num_tokens": 28319609.0,
      "step": 12355
    },
    {
      "entropy": 5.309325170516968,
      "epoch": 1.1873198847262247,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004866537749604744,
      "loss": 5.077,
      "mean_token_accuracy": 0.202881295979023,
      "num_tokens": 28330526.0,
      "step": 12360
    },
    {
      "entropy": 5.221873950958252,
      "epoch": 1.1878001921229586,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048664214058990546,
      "loss": 5.0362,
      "mean_token_accuracy": 0.20625924170017243,
      "num_tokens": 28343919.0,
      "step": 12365
    },
    {
      "entropy": 5.300599098205566,
      "epoch": 1.1882804995196925,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004866305013056346,
      "loss": 5.0442,
      "mean_token_accuracy": 0.20683213472366332,
      "num_tokens": 28354850.0,
      "step": 12370
    },
    {
      "entropy": 5.16650505065918,
      "epoch": 1.1887608069164266,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004866188571079318,
      "loss": 4.9504,
      "mean_token_accuracy": 0.21418039351701737,
      "num_tokens": 28365699.0,
      "step": 12375
    },
    {
      "entropy": 5.119503784179687,
      "epoch": 1.1892411143131605,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004866072079970676,
      "loss": 4.9032,
      "mean_token_accuracy": 0.22358543872833253,
      "num_tokens": 28377053.0,
      "step": 12380
    },
    {
      "entropy": 5.286639785766601,
      "epoch": 1.1897214217098944,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048659555397331236,
      "loss": 5.023,
      "mean_token_accuracy": 0.207548625767231,
      "num_tokens": 28387505.0,
      "step": 12385
    },
    {
      "entropy": 5.247905635833741,
      "epoch": 1.1902017291066282,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004865838950369366,
      "loss": 5.0295,
      "mean_token_accuracy": 0.20773934721946716,
      "num_tokens": 28399308.0,
      "step": 12390
    },
    {
      "entropy": 5.209375047683716,
      "epoch": 1.1906820365033621,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048657223118821116,
      "loss": 5.0041,
      "mean_token_accuracy": 0.21377101838588713,
      "num_tokens": 28410860.0,
      "step": 12395
    },
    {
      "entropy": 5.087126159667969,
      "epoch": 1.191162343900096,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048656056242740665,
      "loss": 4.8674,
      "mean_token_accuracy": 0.21621858775615693,
      "num_tokens": 28421867.0,
      "step": 12400
    },
    {
      "entropy": 5.17433762550354,
      "epoch": 1.19164265129683,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004865488887547942,
      "loss": 5.0012,
      "mean_token_accuracy": 0.21354222446680068,
      "num_tokens": 28433375.0,
      "step": 12405
    },
    {
      "entropy": 5.256328153610229,
      "epoch": 1.192122958693564,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004865372101706446,
      "loss": 5.0588,
      "mean_token_accuracy": 0.21048571914434433,
      "num_tokens": 28445242.0,
      "step": 12410
    },
    {
      "entropy": 5.164352035522461,
      "epoch": 1.1926032660902979,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004865255266752292,
      "loss": 4.987,
      "mean_token_accuracy": 0.21557822972536086,
      "num_tokens": 28455904.0,
      "step": 12415
    },
    {
      "entropy": 5.248293066024781,
      "epoch": 1.1930835734870318,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004865138382688191,
      "loss": 5.009,
      "mean_token_accuracy": 0.21178697794675827,
      "num_tokens": 28465768.0,
      "step": 12420
    },
    {
      "entropy": 5.2755701541900635,
      "epoch": 1.1935638808837656,
      "grad_norm": 1.375,
      "learning_rate": 0.0004865021449516859,
      "loss": 5.1062,
      "mean_token_accuracy": 0.2125866010785103,
      "num_tokens": 28476489.0,
      "step": 12425
    },
    {
      "entropy": 5.208784294128418,
      "epoch": 1.1940441882804995,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004864904467241008,
      "loss": 4.9623,
      "mean_token_accuracy": 0.21977111250162124,
      "num_tokens": 28487684.0,
      "step": 12430
    },
    {
      "entropy": 5.215552186965942,
      "epoch": 1.1945244956772334,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048647874358633556,
      "loss": 5.01,
      "mean_token_accuracy": 0.2085271418094635,
      "num_tokens": 28499083.0,
      "step": 12435
    },
    {
      "entropy": 5.171278190612793,
      "epoch": 1.1950048030739673,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048646703553866183,
      "loss": 5.0327,
      "mean_token_accuracy": 0.21017878949642183,
      "num_tokens": 28510057.0,
      "step": 12440
    },
    {
      "entropy": 5.146677398681641,
      "epoch": 1.1954851104707012,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004864553225813515,
      "loss": 4.9798,
      "mean_token_accuracy": 0.21541106700897217,
      "num_tokens": 28522787.0,
      "step": 12445
    },
    {
      "entropy": 5.3216673851013185,
      "epoch": 1.195965417867435,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004864436047146765,
      "loss": 5.0489,
      "mean_token_accuracy": 0.21266763359308244,
      "num_tokens": 28534353.0,
      "step": 12450
    },
    {
      "entropy": 5.2914228439331055,
      "epoch": 1.1964457252641691,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048643188193890874,
      "loss": 5.143,
      "mean_token_accuracy": 0.19696910232305526,
      "num_tokens": 28546079.0,
      "step": 12455
    },
    {
      "entropy": 5.26306381225586,
      "epoch": 1.196926032660903,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004864201542543206,
      "loss": 5.0898,
      "mean_token_accuracy": 0.2064315527677536,
      "num_tokens": 28557637.0,
      "step": 12460
    },
    {
      "entropy": 5.231850814819336,
      "epoch": 1.197406340057637,
      "grad_norm": 1.375,
      "learning_rate": 0.0004864084216611843,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21977874487638474,
      "num_tokens": 28568809.0,
      "step": 12465
    },
    {
      "entropy": 5.257216930389404,
      "epoch": 1.1978866474543708,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048639668415977207,
      "loss": 5.0817,
      "mean_token_accuracy": 0.21111681908369065,
      "num_tokens": 28579766.0,
      "step": 12470
    },
    {
      "entropy": 5.250015163421631,
      "epoch": 1.1983669548511047,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048638494175035665,
      "loss": 4.9811,
      "mean_token_accuracy": 0.2136036053299904,
      "num_tokens": 28590898.0,
      "step": 12475
    },
    {
      "entropy": 5.1615275859832765,
      "epoch": 1.1988472622478386,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004863731944332105,
      "loss": 4.9427,
      "mean_token_accuracy": 0.21254715770483018,
      "num_tokens": 28603380.0,
      "step": 12480
    },
    {
      "entropy": 5.232118940353393,
      "epoch": 1.1993275696445724,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004863614422086065,
      "loss": 5.1258,
      "mean_token_accuracy": 0.20304721593856812,
      "num_tokens": 28615509.0,
      "step": 12485
    },
    {
      "entropy": 5.3055541038513185,
      "epoch": 1.1998078770413065,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004863496850768174,
      "loss": 5.0507,
      "mean_token_accuracy": 0.20623885989189147,
      "num_tokens": 28627353.0,
      "step": 12490
    },
    {
      "entropy": 5.208082628250122,
      "epoch": 1.2002881844380404,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004863379230381162,
      "loss": 4.9085,
      "mean_token_accuracy": 0.2222481355071068,
      "num_tokens": 28638217.0,
      "step": 12495
    },
    {
      "entropy": 5.167784357070923,
      "epoch": 1.2007684918347743,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000486326156092776,
      "loss": 4.9701,
      "mean_token_accuracy": 0.2173997014760971,
      "num_tokens": 28648862.0,
      "step": 12500
    },
    {
      "entropy": 5.196328449249267,
      "epoch": 1.2012487992315082,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048631438424106985,
      "loss": 5.0487,
      "mean_token_accuracy": 0.21258731782436371,
      "num_tokens": 28661561.0,
      "step": 12505
    },
    {
      "entropy": 5.294944810867309,
      "epoch": 1.201729106628242,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048630260748327124,
      "loss": 5.0953,
      "mean_token_accuracy": 0.20518883764743806,
      "num_tokens": 28674851.0,
      "step": 12510
    },
    {
      "entropy": 5.225453805923462,
      "epoch": 1.202209414024976,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048629082581965355,
      "loss": 5.0561,
      "mean_token_accuracy": 0.2065837115049362,
      "num_tokens": 28688444.0,
      "step": 12515
    },
    {
      "entropy": 5.329750204086304,
      "epoch": 1.2026897214217098,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048627903925049033,
      "loss": 5.1048,
      "mean_token_accuracy": 0.20583407133817672,
      "num_tokens": 28699374.0,
      "step": 12520
    },
    {
      "entropy": 5.1880128383636475,
      "epoch": 1.2031700288184437,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048626724777605507,
      "loss": 5.0569,
      "mean_token_accuracy": 0.20676028579473496,
      "num_tokens": 28711283.0,
      "step": 12525
    },
    {
      "entropy": 5.221640539169312,
      "epoch": 1.2036503362151778,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004862554513966217,
      "loss": 5.0543,
      "mean_token_accuracy": 0.2035945698618889,
      "num_tokens": 28723445.0,
      "step": 12530
    },
    {
      "entropy": 5.210093259811401,
      "epoch": 1.2041306436119117,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048624365011246405,
      "loss": 5.0054,
      "mean_token_accuracy": 0.21303798109292985,
      "num_tokens": 28734505.0,
      "step": 12535
    },
    {
      "entropy": 5.117575836181641,
      "epoch": 1.2046109510086456,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004862318439238561,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21927265971899032,
      "num_tokens": 28745476.0,
      "step": 12540
    },
    {
      "entropy": 5.212747955322266,
      "epoch": 1.2050912584053795,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000486220032831072,
      "loss": 4.9118,
      "mean_token_accuracy": 0.2158356323838234,
      "num_tokens": 28755645.0,
      "step": 12545
    },
    {
      "entropy": 5.196612358093262,
      "epoch": 1.2055715658021133,
      "grad_norm": 1.25,
      "learning_rate": 0.0004862082168343859,
      "loss": 5.0287,
      "mean_token_accuracy": 0.2096465528011322,
      "num_tokens": 28767911.0,
      "step": 12550
    },
    {
      "entropy": 5.215525674819946,
      "epoch": 1.2060518731988472,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004861963959340722,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21145375370979308,
      "num_tokens": 28779418.0,
      "step": 12555
    },
    {
      "entropy": 5.286305236816406,
      "epoch": 1.206532180595581,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004861845701304053,
      "loss": 5.109,
      "mean_token_accuracy": 0.20508148074150084,
      "num_tokens": 28790857.0,
      "step": 12560
    },
    {
      "entropy": 5.280824089050293,
      "epoch": 1.2070124879923152,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048617273942365977,
      "loss": 5.0966,
      "mean_token_accuracy": 0.20530366152524948,
      "num_tokens": 28802975.0,
      "step": 12565
    },
    {
      "entropy": 5.207037353515625,
      "epoch": 1.207492795389049,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004861609038141103,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21204428672790526,
      "num_tokens": 28813874.0,
      "step": 12570
    },
    {
      "entropy": 5.209102725982666,
      "epoch": 1.207973102785783,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048614906330203165,
      "loss": 4.9939,
      "mean_token_accuracy": 0.20990731865167617,
      "num_tokens": 28826278.0,
      "step": 12575
    },
    {
      "entropy": 5.277382183074951,
      "epoch": 1.2084534101825168,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004861372178876987,
      "loss": 5.0884,
      "mean_token_accuracy": 0.20838196128606795,
      "num_tokens": 28838607.0,
      "step": 12580
    },
    {
      "entropy": 5.161143636703491,
      "epoch": 1.2089337175792507,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048612536757138653,
      "loss": 5.0183,
      "mean_token_accuracy": 0.21519987285137177,
      "num_tokens": 28849801.0,
      "step": 12585
    },
    {
      "entropy": 5.118639087677002,
      "epoch": 1.2094140249759846,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004861135123533702,
      "loss": 4.8945,
      "mean_token_accuracy": 0.2229227051138878,
      "num_tokens": 28861312.0,
      "step": 12590
    },
    {
      "entropy": 5.184957599639892,
      "epoch": 1.2098943323727185,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048610165223392503,
      "loss": 5.0425,
      "mean_token_accuracy": 0.20855659246444702,
      "num_tokens": 28872337.0,
      "step": 12595
    },
    {
      "entropy": 5.21443190574646,
      "epoch": 1.2103746397694524,
      "grad_norm": 1.375,
      "learning_rate": 0.0004860897872133263,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22272854298353195,
      "num_tokens": 28883263.0,
      "step": 12600
    },
    {
      "entropy": 5.219044160842896,
      "epoch": 1.2108549471661862,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004860779172918496,
      "loss": 5.0214,
      "mean_token_accuracy": 0.21036805361509323,
      "num_tokens": 28896166.0,
      "step": 12605
    },
    {
      "entropy": 5.2326911926269535,
      "epoch": 1.2113352545629203,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004860660424697704,
      "loss": 5.0631,
      "mean_token_accuracy": 0.20857544988393784,
      "num_tokens": 28906928.0,
      "step": 12610
    },
    {
      "entropy": 5.273232889175415,
      "epoch": 1.2118155619596542,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048605416274736434,
      "loss": 5.1357,
      "mean_token_accuracy": 0.19857011437416078,
      "num_tokens": 28918328.0,
      "step": 12615
    },
    {
      "entropy": 5.212222766876221,
      "epoch": 1.2122958693563881,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048604227812490744,
      "loss": 4.9972,
      "mean_token_accuracy": 0.21082258224487305,
      "num_tokens": 28930327.0,
      "step": 12620
    },
    {
      "entropy": 5.165164804458618,
      "epoch": 1.212776176753122,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048603038860267546,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20992020517587662,
      "num_tokens": 28942166.0,
      "step": 12625
    },
    {
      "entropy": 5.137792682647705,
      "epoch": 1.2132564841498559,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004860184941809445,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2244383603334427,
      "num_tokens": 28953159.0,
      "step": 12630
    },
    {
      "entropy": 5.178856229782104,
      "epoch": 1.2137367915465898,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048600659485999073,
      "loss": 4.9638,
      "mean_token_accuracy": 0.215152744948864,
      "num_tokens": 28963897.0,
      "step": 12635
    },
    {
      "entropy": 5.15120644569397,
      "epoch": 1.2142170989433236,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048599469064009027,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21606374084949492,
      "num_tokens": 28974452.0,
      "step": 12640
    },
    {
      "entropy": 5.091236591339111,
      "epoch": 1.2146974063400577,
      "grad_norm": 1.25,
      "learning_rate": 0.00048598278152151974,
      "loss": 4.91,
      "mean_token_accuracy": 0.2155248448252678,
      "num_tokens": 28984664.0,
      "step": 12645
    },
    {
      "entropy": 5.237447309494018,
      "epoch": 1.2151777137367916,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004859708675045555,
      "loss": 5.001,
      "mean_token_accuracy": 0.21790367364883423,
      "num_tokens": 28994992.0,
      "step": 12650
    },
    {
      "entropy": 5.305144929885865,
      "epoch": 1.2156580211335255,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004859589485894741,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20588420033454896,
      "num_tokens": 29006190.0,
      "step": 12655
    },
    {
      "entropy": 5.180465364456177,
      "epoch": 1.2161383285302594,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004859470247765524,
      "loss": 5.0022,
      "mean_token_accuracy": 0.20653647035360337,
      "num_tokens": 29017987.0,
      "step": 12660
    },
    {
      "entropy": 5.123558712005615,
      "epoch": 1.2166186359269933,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004859350960660671,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21601806730031967,
      "num_tokens": 29030012.0,
      "step": 12665
    },
    {
      "entropy": 5.2893301486969,
      "epoch": 1.2170989433237271,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004859231624582953,
      "loss": 5.0935,
      "mean_token_accuracy": 0.20400587022304534,
      "num_tokens": 29041684.0,
      "step": 12670
    },
    {
      "entropy": 5.232307243347168,
      "epoch": 1.217579250720461,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048591122395351394,
      "loss": 4.9662,
      "mean_token_accuracy": 0.21430334001779555,
      "num_tokens": 29052525.0,
      "step": 12675
    },
    {
      "entropy": 5.156563377380371,
      "epoch": 1.218059558117195,
      "grad_norm": 1.25,
      "learning_rate": 0.0004858992805520003,
      "loss": 4.9803,
      "mean_token_accuracy": 0.21755577027797698,
      "num_tokens": 29063471.0,
      "step": 12680
    },
    {
      "entropy": 5.13467583656311,
      "epoch": 1.218539865513929,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048588733225403153,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21439408659934997,
      "num_tokens": 29074616.0,
      "step": 12685
    },
    {
      "entropy": 5.173747587203979,
      "epoch": 1.219020172910663,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004858753790598851,
      "loss": 5.0475,
      "mean_token_accuracy": 0.20869094282388687,
      "num_tokens": 29086603.0,
      "step": 12690
    },
    {
      "entropy": 5.2212928295135494,
      "epoch": 1.2195004803073968,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004858634209698386,
      "loss": 4.9981,
      "mean_token_accuracy": 0.2189930111169815,
      "num_tokens": 29097232.0,
      "step": 12695
    },
    {
      "entropy": 5.338279867172242,
      "epoch": 1.2199807877041307,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048585145798416956,
      "loss": 5.1079,
      "mean_token_accuracy": 0.20422830283641816,
      "num_tokens": 29107453.0,
      "step": 12700
    },
    {
      "entropy": 5.144302606582642,
      "epoch": 1.2204610951008645,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004858394901031558,
      "loss": 5.0008,
      "mean_token_accuracy": 0.21367748230695724,
      "num_tokens": 29120255.0,
      "step": 12705
    },
    {
      "entropy": 5.2259269714355465,
      "epoch": 1.2209414024975984,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004858275173270751,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20858570784330369,
      "num_tokens": 29130892.0,
      "step": 12710
    },
    {
      "entropy": 5.306058883666992,
      "epoch": 1.2214217098943323,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048581553965620553,
      "loss": 5.0746,
      "mean_token_accuracy": 0.2050945580005646,
      "num_tokens": 29142613.0,
      "step": 12715
    },
    {
      "entropy": 5.2205894947052,
      "epoch": 1.2219020172910664,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048580355709082506,
      "loss": 4.9845,
      "mean_token_accuracy": 0.210810886323452,
      "num_tokens": 29153045.0,
      "step": 12720
    },
    {
      "entropy": 5.131451034545899,
      "epoch": 1.2223823246878003,
      "grad_norm": 1.171875,
      "learning_rate": 0.000485791569631212,
      "loss": 4.8956,
      "mean_token_accuracy": 0.2196899712085724,
      "num_tokens": 29163767.0,
      "step": 12725
    },
    {
      "entropy": 5.19201512336731,
      "epoch": 1.2228626320845342,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004857795772776446,
      "loss": 5.0414,
      "mean_token_accuracy": 0.21010521203279495,
      "num_tokens": 29174277.0,
      "step": 12730
    },
    {
      "entropy": 5.2668726444244385,
      "epoch": 1.223342939481268,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048576758003040127,
      "loss": 5.0918,
      "mean_token_accuracy": 0.20481030493974686,
      "num_tokens": 29186078.0,
      "step": 12735
    },
    {
      "entropy": 5.17754282951355,
      "epoch": 1.223823246878002,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00048575557788976066,
      "loss": 4.9628,
      "mean_token_accuracy": 0.2184329017996788,
      "num_tokens": 29198056.0,
      "step": 12740
    },
    {
      "entropy": 5.275543355941773,
      "epoch": 1.2243035542747358,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004857435708560013,
      "loss": 5.1529,
      "mean_token_accuracy": 0.1977175682783127,
      "num_tokens": 29209892.0,
      "step": 12745
    },
    {
      "entropy": 5.336450052261353,
      "epoch": 1.2247838616714697,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048573155892940204,
      "loss": 5.0404,
      "mean_token_accuracy": 0.20856212079524994,
      "num_tokens": 29220813.0,
      "step": 12750
    },
    {
      "entropy": 5.194973278045654,
      "epoch": 1.2252641690682036,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048571954211024164,
      "loss": 5.0461,
      "mean_token_accuracy": 0.20851152241230012,
      "num_tokens": 29232807.0,
      "step": 12755
    },
    {
      "entropy": 5.239838123321533,
      "epoch": 1.2257444764649374,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048570752039879924,
      "loss": 5.0247,
      "mean_token_accuracy": 0.2093370333313942,
      "num_tokens": 29243721.0,
      "step": 12760
    },
    {
      "entropy": 5.233508825302124,
      "epoch": 1.2262247838616716,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004856954937953539,
      "loss": 5.0434,
      "mean_token_accuracy": 0.21219601035118102,
      "num_tokens": 29255783.0,
      "step": 12765
    },
    {
      "entropy": 5.15324649810791,
      "epoch": 1.2267050912584054,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004856834623001848,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21612099558115005,
      "num_tokens": 29266857.0,
      "step": 12770
    },
    {
      "entropy": 5.098276424407959,
      "epoch": 1.2271853986551393,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004856714259135713,
      "loss": 4.9075,
      "mean_token_accuracy": 0.22160763591527938,
      "num_tokens": 29278479.0,
      "step": 12775
    },
    {
      "entropy": 5.268709802627564,
      "epoch": 1.2276657060518732,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004856593846357929,
      "loss": 4.9828,
      "mean_token_accuracy": 0.21758316308259965,
      "num_tokens": 29289369.0,
      "step": 12780
    },
    {
      "entropy": 5.187294960021973,
      "epoch": 1.228146013448607,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004856473384671291,
      "loss": 4.9886,
      "mean_token_accuracy": 0.20634982436895372,
      "num_tokens": 29301015.0,
      "step": 12785
    },
    {
      "entropy": 5.257769823074341,
      "epoch": 1.228626320845341,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048563528740785955,
      "loss": 5.0875,
      "mean_token_accuracy": 0.2078697681427002,
      "num_tokens": 29312638.0,
      "step": 12790
    },
    {
      "entropy": 5.261219310760498,
      "epoch": 1.2291066282420748,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048562323145826414,
      "loss": 5.0213,
      "mean_token_accuracy": 0.21456616520881652,
      "num_tokens": 29323540.0,
      "step": 12795
    },
    {
      "entropy": 5.1930254936218265,
      "epoch": 1.229586935638809,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004856111706186227,
      "loss": 5.0433,
      "mean_token_accuracy": 0.20860848277807237,
      "num_tokens": 29334287.0,
      "step": 12800
    },
    {
      "entropy": 5.200117683410644,
      "epoch": 1.2300672430355428,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048559910488921534,
      "loss": 4.9902,
      "mean_token_accuracy": 0.2059507504105568,
      "num_tokens": 29345278.0,
      "step": 12805
    },
    {
      "entropy": 5.252696943283081,
      "epoch": 1.2305475504322767,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000485587034270322,
      "loss": 5.057,
      "mean_token_accuracy": 0.20158623158931732,
      "num_tokens": 29357297.0,
      "step": 12810
    },
    {
      "entropy": 5.197402572631836,
      "epoch": 1.2310278578290106,
      "grad_norm": 1.1875,
      "learning_rate": 0.000485574958762223,
      "loss": 5.0166,
      "mean_token_accuracy": 0.2165746048092842,
      "num_tokens": 29368513.0,
      "step": 12815
    },
    {
      "entropy": 5.199039125442505,
      "epoch": 1.2315081652257445,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048556287836519886,
      "loss": 5.0254,
      "mean_token_accuracy": 0.205079285800457,
      "num_tokens": 29379227.0,
      "step": 12820
    },
    {
      "entropy": 5.122586107254028,
      "epoch": 1.2319884726224783,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004855507930795299,
      "loss": 5.0115,
      "mean_token_accuracy": 0.21328140199184417,
      "num_tokens": 29390048.0,
      "step": 12825
    },
    {
      "entropy": 5.245057535171509,
      "epoch": 1.2324687800192122,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048553870290549665,
      "loss": 5.0141,
      "mean_token_accuracy": 0.20793407559394836,
      "num_tokens": 29401252.0,
      "step": 12830
    },
    {
      "entropy": 5.2300293922424315,
      "epoch": 1.232949087415946,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048552660784338,
      "loss": 4.9784,
      "mean_token_accuracy": 0.21577298641204834,
      "num_tokens": 29413122.0,
      "step": 12835
    },
    {
      "entropy": 5.316891860961914,
      "epoch": 1.23342939481268,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004855145078934606,
      "loss": 5.085,
      "mean_token_accuracy": 0.2105143830180168,
      "num_tokens": 29423240.0,
      "step": 12840
    },
    {
      "entropy": 5.185998201370239,
      "epoch": 1.233909702209414,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004855024030560195,
      "loss": 4.9388,
      "mean_token_accuracy": 0.21454793214797974,
      "num_tokens": 29433818.0,
      "step": 12845
    },
    {
      "entropy": 5.217524909973145,
      "epoch": 1.234390009606148,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004854902933313376,
      "loss": 5.1062,
      "mean_token_accuracy": 0.20499005913734436,
      "num_tokens": 29446927.0,
      "step": 12850
    },
    {
      "entropy": 5.276363515853882,
      "epoch": 1.2348703170028819,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048547817871969607,
      "loss": 5.0581,
      "mean_token_accuracy": 0.2034383997321129,
      "num_tokens": 29458092.0,
      "step": 12855
    },
    {
      "entropy": 5.296837997436524,
      "epoch": 1.2353506243996157,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048546605922137633,
      "loss": 5.1266,
      "mean_token_accuracy": 0.19996122717857362,
      "num_tokens": 29470395.0,
      "step": 12860
    },
    {
      "entropy": 5.25747013092041,
      "epoch": 1.2358309317963496,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004854539348366596,
      "loss": 5.0773,
      "mean_token_accuracy": 0.20787269175052642,
      "num_tokens": 29481177.0,
      "step": 12865
    },
    {
      "entropy": 5.192795372009277,
      "epoch": 1.2363112391930835,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004854418055658274,
      "loss": 4.9743,
      "mean_token_accuracy": 0.2096899002790451,
      "num_tokens": 29492305.0,
      "step": 12870
    },
    {
      "entropy": 5.264670562744141,
      "epoch": 1.2367915465898176,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048542967140916134,
      "loss": 5.0317,
      "mean_token_accuracy": 0.21830671280622482,
      "num_tokens": 29503237.0,
      "step": 12875
    },
    {
      "entropy": 5.2417316913604735,
      "epoch": 1.2372718539865515,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004854175323669432,
      "loss": 5.0213,
      "mean_token_accuracy": 0.2089702308177948,
      "num_tokens": 29514195.0,
      "step": 12880
    },
    {
      "entropy": 5.254844427108765,
      "epoch": 1.2377521613832854,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004854053884394547,
      "loss": 5.1305,
      "mean_token_accuracy": 0.19987316727638244,
      "num_tokens": 29526110.0,
      "step": 12885
    },
    {
      "entropy": 5.197206592559814,
      "epoch": 1.2382324687800192,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048539323962697796,
      "loss": 5.0107,
      "mean_token_accuracy": 0.21437044590711593,
      "num_tokens": 29536912.0,
      "step": 12890
    },
    {
      "entropy": 5.279458951950073,
      "epoch": 1.2387127761767531,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004853810859297949,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20982509404420852,
      "num_tokens": 29548070.0,
      "step": 12895
    },
    {
      "entropy": 5.179880952835083,
      "epoch": 1.239193083573487,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048536892734818773,
      "loss": 5.0684,
      "mean_token_accuracy": 0.20430529713630677,
      "num_tokens": 29560713.0,
      "step": 12900
    },
    {
      "entropy": 5.196026945114136,
      "epoch": 1.239673390970221,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004853567638824387,
      "loss": 4.9809,
      "mean_token_accuracy": 0.22359374463558196,
      "num_tokens": 29571269.0,
      "step": 12905
    },
    {
      "entropy": 5.269033193588257,
      "epoch": 1.2401536983669548,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048534459553283026,
      "loss": 5.017,
      "mean_token_accuracy": 0.21566904038190843,
      "num_tokens": 29584024.0,
      "step": 12910
    },
    {
      "entropy": 5.125743961334228,
      "epoch": 1.2406340057636887,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004853324222996449,
      "loss": 4.866,
      "mean_token_accuracy": 0.22419926524162292,
      "num_tokens": 29595679.0,
      "step": 12915
    },
    {
      "entropy": 5.194203472137451,
      "epoch": 1.2411143131604228,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048532024418316525,
      "loss": 4.9923,
      "mean_token_accuracy": 0.21525548696517943,
      "num_tokens": 29607316.0,
      "step": 12920
    },
    {
      "entropy": 5.18575267791748,
      "epoch": 1.2415946205571566,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004853080611836741,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21795926988124847,
      "num_tokens": 29618929.0,
      "step": 12925
    },
    {
      "entropy": 5.268980264663696,
      "epoch": 1.2420749279538905,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048529587330145427,
      "loss": 5.1122,
      "mean_token_accuracy": 0.2052838012576103,
      "num_tokens": 29630253.0,
      "step": 12930
    },
    {
      "entropy": 5.206554841995239,
      "epoch": 1.2425552353506244,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048528368053678863,
      "loss": 5.0196,
      "mean_token_accuracy": 0.2098276674747467,
      "num_tokens": 29641676.0,
      "step": 12935
    },
    {
      "entropy": 5.19392580986023,
      "epoch": 1.2430355427473583,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004852714828899604,
      "loss": 5.036,
      "mean_token_accuracy": 0.2101952999830246,
      "num_tokens": 29651942.0,
      "step": 12940
    },
    {
      "entropy": 5.281698369979859,
      "epoch": 1.2435158501440922,
      "grad_norm": 1.25,
      "learning_rate": 0.00048525928036125264,
      "loss": 5.0869,
      "mean_token_accuracy": 0.206894651055336,
      "num_tokens": 29663094.0,
      "step": 12945
    },
    {
      "entropy": 5.215184593200684,
      "epoch": 1.243996157540826,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048524707295094884,
      "loss": 5.0479,
      "mean_token_accuracy": 0.2094126597046852,
      "num_tokens": 29675412.0,
      "step": 12950
    },
    {
      "entropy": 5.175008916854859,
      "epoch": 1.2444764649375601,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004852348606593322,
      "loss": 5.023,
      "mean_token_accuracy": 0.20618369728326796,
      "num_tokens": 29686684.0,
      "step": 12955
    },
    {
      "entropy": 5.202035903930664,
      "epoch": 1.244956772334294,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048522264348668646,
      "loss": 4.9708,
      "mean_token_accuracy": 0.21698465645313264,
      "num_tokens": 29697387.0,
      "step": 12960
    },
    {
      "entropy": 5.212286186218262,
      "epoch": 1.245437079731028,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004852104214332951,
      "loss": 4.916,
      "mean_token_accuracy": 0.22701910734176636,
      "num_tokens": 29708677.0,
      "step": 12965
    },
    {
      "entropy": 5.213592767715454,
      "epoch": 1.2459173871277618,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048519819449944205,
      "loss": 5.0903,
      "mean_token_accuracy": 0.2068006619811058,
      "num_tokens": 29720430.0,
      "step": 12970
    },
    {
      "entropy": 5.172637891769409,
      "epoch": 1.2463976945244957,
      "grad_norm": 1.1875,
      "learning_rate": 0.000485185962685411,
      "loss": 4.959,
      "mean_token_accuracy": 0.2224774643778801,
      "num_tokens": 29731370.0,
      "step": 12975
    },
    {
      "entropy": 5.324249887466431,
      "epoch": 1.2468780019212296,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000485173725991486,
      "loss": 5.058,
      "mean_token_accuracy": 0.20003255009651183,
      "num_tokens": 29743399.0,
      "step": 12980
    },
    {
      "entropy": 5.299694156646728,
      "epoch": 1.2473583093179634,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048516148441795124,
      "loss": 5.0345,
      "mean_token_accuracy": 0.20615182369947432,
      "num_tokens": 29755693.0,
      "step": 12985
    },
    {
      "entropy": 5.179002285003662,
      "epoch": 1.2478386167146973,
      "grad_norm": 2.890625,
      "learning_rate": 0.0004851492379650908,
      "loss": 5.0688,
      "mean_token_accuracy": 0.20135098546743393,
      "num_tokens": 29766576.0,
      "step": 12990
    },
    {
      "entropy": 5.115320539474487,
      "epoch": 1.2483189241114312,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004851369866331891,
      "loss": 4.9915,
      "mean_token_accuracy": 0.20593600869178771,
      "num_tokens": 29779255.0,
      "step": 12995
    },
    {
      "entropy": 5.152290916442871,
      "epoch": 1.2487992315081653,
      "grad_norm": 1.25,
      "learning_rate": 0.0004851247304225306,
      "loss": 4.9484,
      "mean_token_accuracy": 0.21889047771692277,
      "num_tokens": 29790172.0,
      "step": 13000
    },
    {
      "entropy": 5.250343704223633,
      "epoch": 1.2492795389048992,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004851124693333997,
      "loss": 5.0441,
      "mean_token_accuracy": 0.20099897384643556,
      "num_tokens": 29801206.0,
      "step": 13005
    },
    {
      "entropy": 5.3037238121032715,
      "epoch": 1.249759846301633,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004851002033660812,
      "loss": 5.0705,
      "mean_token_accuracy": 0.20247115045785904,
      "num_tokens": 29813917.0,
      "step": 13010
    },
    {
      "entropy": 5.125704765319824,
      "epoch": 1.250240153698367,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048508793252085994,
      "loss": 4.913,
      "mean_token_accuracy": 0.21437389999628068,
      "num_tokens": 29826036.0,
      "step": 13015
    },
    {
      "entropy": 5.122535133361817,
      "epoch": 1.2507204610951008,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004850756567980206,
      "loss": 4.8803,
      "mean_token_accuracy": 0.2239364728331566,
      "num_tokens": 29837901.0,
      "step": 13020
    },
    {
      "entropy": 5.114030408859253,
      "epoch": 1.2512007684918347,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048506337619784836,
      "loss": 4.9253,
      "mean_token_accuracy": 0.21681493818759917,
      "num_tokens": 29850074.0,
      "step": 13025
    },
    {
      "entropy": 5.206555509567261,
      "epoch": 1.2516810758885688,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004850510907206283,
      "loss": 4.9954,
      "mean_token_accuracy": 0.21088127940893173,
      "num_tokens": 29862447.0,
      "step": 13030
    },
    {
      "entropy": 5.1978973865509035,
      "epoch": 1.2521613832853027,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048503880036664555,
      "loss": 4.9939,
      "mean_token_accuracy": 0.2121051698923111,
      "num_tokens": 29873483.0,
      "step": 13035
    },
    {
      "entropy": 5.1342432498931885,
      "epoch": 1.2526416906820366,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004850265051361857,
      "loss": 4.9386,
      "mean_token_accuracy": 0.2179039388895035,
      "num_tokens": 29884807.0,
      "step": 13040
    },
    {
      "entropy": 5.179131507873535,
      "epoch": 1.2531219980787704,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004850142050295339,
      "loss": 4.984,
      "mean_token_accuracy": 0.20289032608270646,
      "num_tokens": 29895994.0,
      "step": 13045
    },
    {
      "entropy": 5.176501512527466,
      "epoch": 1.2536023054755043,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048500190004697595,
      "loss": 5.0495,
      "mean_token_accuracy": 0.21157522052526473,
      "num_tokens": 29907534.0,
      "step": 13050
    },
    {
      "entropy": 5.246091842651367,
      "epoch": 1.2540826128722382,
      "grad_norm": 1.25,
      "learning_rate": 0.0004849895901887974,
      "loss": 5.0131,
      "mean_token_accuracy": 0.21232030391693116,
      "num_tokens": 29919747.0,
      "step": 13055
    },
    {
      "entropy": 5.170240116119385,
      "epoch": 1.254562920268972,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004849772754552842,
      "loss": 5.0067,
      "mean_token_accuracy": 0.2150871694087982,
      "num_tokens": 29930254.0,
      "step": 13060
    },
    {
      "entropy": 5.304010820388794,
      "epoch": 1.255043227665706,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048496495584672214,
      "loss": 5.1277,
      "mean_token_accuracy": 0.2039937049150467,
      "num_tokens": 29941232.0,
      "step": 13065
    },
    {
      "entropy": 5.2707438468933105,
      "epoch": 1.2555235350624399,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048495263136339725,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2169503688812256,
      "num_tokens": 29952452.0,
      "step": 13070
    },
    {
      "entropy": 5.220221853256225,
      "epoch": 1.2560038424591737,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004849403020055956,
      "loss": 5.0077,
      "mean_token_accuracy": 0.21553540080785752,
      "num_tokens": 29963583.0,
      "step": 13075
    },
    {
      "entropy": 5.149609994888306,
      "epoch": 1.2564841498559078,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048492796777360373,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21285282224416732,
      "num_tokens": 29975186.0,
      "step": 13080
    },
    {
      "entropy": 5.163627672195434,
      "epoch": 1.2569644572526417,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048491562866770767,
      "loss": 4.9612,
      "mean_token_accuracy": 0.21754895895719528,
      "num_tokens": 29987474.0,
      "step": 13085
    },
    {
      "entropy": 5.180987167358398,
      "epoch": 1.2574447646493756,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048490328468819404,
      "loss": 4.9514,
      "mean_token_accuracy": 0.2201530158519745,
      "num_tokens": 29999254.0,
      "step": 13090
    },
    {
      "entropy": 5.216389226913452,
      "epoch": 1.2579250720461095,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048489093583534945,
      "loss": 5.0254,
      "mean_token_accuracy": 0.2093447118997574,
      "num_tokens": 30009901.0,
      "step": 13095
    },
    {
      "entropy": 5.2521946907043455,
      "epoch": 1.2584053794428434,
      "grad_norm": 1.375,
      "learning_rate": 0.0004848785821094606,
      "loss": 5.1092,
      "mean_token_accuracy": 0.20851932764053344,
      "num_tokens": 30022424.0,
      "step": 13100
    },
    {
      "entropy": 5.206987714767456,
      "epoch": 1.2588856868395775,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004848662235108142,
      "loss": 5.0383,
      "mean_token_accuracy": 0.2174348458647728,
      "num_tokens": 30034209.0,
      "step": 13105
    },
    {
      "entropy": 5.180519390106201,
      "epoch": 1.2593659942363113,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004848538600396973,
      "loss": 4.9552,
      "mean_token_accuracy": 0.21384225338697432,
      "num_tokens": 30044885.0,
      "step": 13110
    },
    {
      "entropy": 5.262897253036499,
      "epoch": 1.2598463016330452,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048484149169639694,
      "loss": 5.042,
      "mean_token_accuracy": 0.20745661556720735,
      "num_tokens": 30056592.0,
      "step": 13115
    },
    {
      "entropy": 5.354119539260864,
      "epoch": 1.260326609029779,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004848291184812003,
      "loss": 5.1549,
      "mean_token_accuracy": 0.19949035942554474,
      "num_tokens": 30068423.0,
      "step": 13120
    },
    {
      "entropy": 5.264368915557862,
      "epoch": 1.260806916426513,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004848167403943945,
      "loss": 5.0464,
      "mean_token_accuracy": 0.21190067231655121,
      "num_tokens": 30079132.0,
      "step": 13125
    },
    {
      "entropy": 5.114562511444092,
      "epoch": 1.2612872238232469,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048480435743626703,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22210379540920258,
      "num_tokens": 30090335.0,
      "step": 13130
    },
    {
      "entropy": 5.205622053146362,
      "epoch": 1.2617675312199808,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004847919696071054,
      "loss": 5.0193,
      "mean_token_accuracy": 0.2125892922282219,
      "num_tokens": 30101781.0,
      "step": 13135
    },
    {
      "entropy": 5.154063987731933,
      "epoch": 1.2622478386167146,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048477957690719716,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22134286016225815,
      "num_tokens": 30112486.0,
      "step": 13140
    },
    {
      "entropy": 5.206697750091553,
      "epoch": 1.2627281460134485,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004847671793368301,
      "loss": 5.014,
      "mean_token_accuracy": 0.21562036275863647,
      "num_tokens": 30123664.0,
      "step": 13145
    },
    {
      "entropy": 5.254877424240112,
      "epoch": 1.2632084534101824,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000484754776896292,
      "loss": 5.0733,
      "mean_token_accuracy": 0.21093373894691467,
      "num_tokens": 30134139.0,
      "step": 13150
    },
    {
      "entropy": 5.1566554546356205,
      "epoch": 1.2636887608069165,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004847423695858708,
      "loss": 5.0366,
      "mean_token_accuracy": 0.20727397203445436,
      "num_tokens": 30146127.0,
      "step": 13155
    },
    {
      "entropy": 5.2476564884185795,
      "epoch": 1.2641690682036504,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048472995740585456,
      "loss": 5.0228,
      "mean_token_accuracy": 0.211346735060215,
      "num_tokens": 30158177.0,
      "step": 13160
    },
    {
      "entropy": 5.247139978408813,
      "epoch": 1.2646493756003843,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004847175403565316,
      "loss": 5.0203,
      "mean_token_accuracy": 0.20765364021062852,
      "num_tokens": 30168930.0,
      "step": 13165
    },
    {
      "entropy": 5.280780839920044,
      "epoch": 1.2651296829971181,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048470511843818996,
      "loss": 5.0404,
      "mean_token_accuracy": 0.20737677961587905,
      "num_tokens": 30181206.0,
      "step": 13170
    },
    {
      "entropy": 5.245010757446289,
      "epoch": 1.265609990393852,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004846926916511182,
      "loss": 5.055,
      "mean_token_accuracy": 0.2142008513212204,
      "num_tokens": 30193093.0,
      "step": 13175
    },
    {
      "entropy": 5.181458473205566,
      "epoch": 1.266090297790586,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004846802599956048,
      "loss": 4.9315,
      "mean_token_accuracy": 0.21775425225496292,
      "num_tokens": 30204864.0,
      "step": 13180
    },
    {
      "entropy": 5.231187391281128,
      "epoch": 1.26657060518732,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048466782347193847,
      "loss": 5.0284,
      "mean_token_accuracy": 0.20964808911085128,
      "num_tokens": 30216312.0,
      "step": 13185
    },
    {
      "entropy": 5.261661243438721,
      "epoch": 1.267050912584054,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048465538208040775,
      "loss": 5.0821,
      "mean_token_accuracy": 0.20590133368968963,
      "num_tokens": 30227824.0,
      "step": 13190
    },
    {
      "entropy": 5.248310136795044,
      "epoch": 1.2675312199807878,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048464293582130166,
      "loss": 5.0759,
      "mean_token_accuracy": 0.2055477887392044,
      "num_tokens": 30239065.0,
      "step": 13195
    },
    {
      "entropy": 5.163929796218872,
      "epoch": 1.2680115273775217,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004846304846949091,
      "loss": 4.883,
      "mean_token_accuracy": 0.21213381439447404,
      "num_tokens": 30250684.0,
      "step": 13200
    },
    {
      "entropy": 5.17065806388855,
      "epoch": 1.2684918347742555,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048461802870151916,
      "loss": 5.0003,
      "mean_token_accuracy": 0.2156269982457161,
      "num_tokens": 30261376.0,
      "step": 13205
    },
    {
      "entropy": 5.231918096542358,
      "epoch": 1.2689721421709894,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048460556784142106,
      "loss": 4.9895,
      "mean_token_accuracy": 0.21724696308374405,
      "num_tokens": 30272441.0,
      "step": 13210
    },
    {
      "entropy": 5.151919507980347,
      "epoch": 1.2694524495677233,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048459310211490406,
      "loss": 4.9959,
      "mean_token_accuracy": 0.21814372390508652,
      "num_tokens": 30284446.0,
      "step": 13215
    },
    {
      "entropy": 5.248904848098755,
      "epoch": 1.2699327569644572,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004845806315222576,
      "loss": 5.0967,
      "mean_token_accuracy": 0.21224010288715361,
      "num_tokens": 30295666.0,
      "step": 13220
    },
    {
      "entropy": 5.267348718643189,
      "epoch": 1.270413064361191,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004845681560637711,
      "loss": 5.08,
      "mean_token_accuracy": 0.2102426454424858,
      "num_tokens": 30307663.0,
      "step": 13225
    },
    {
      "entropy": 5.267227077484131,
      "epoch": 1.270893371757925,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004845556757397344,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20916279256343842,
      "num_tokens": 30318914.0,
      "step": 13230
    },
    {
      "entropy": 5.2506834983825685,
      "epoch": 1.271373679154659,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004845431905504372,
      "loss": 5.0987,
      "mean_token_accuracy": 0.19959567487239838,
      "num_tokens": 30331499.0,
      "step": 13235
    },
    {
      "entropy": 5.2950389862060545,
      "epoch": 1.271853986551393,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048453070049616926,
      "loss": 5.0335,
      "mean_token_accuracy": 0.21158182322978974,
      "num_tokens": 30342186.0,
      "step": 13240
    },
    {
      "entropy": 5.259432601928711,
      "epoch": 1.2723342939481268,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048451820557722064,
      "loss": 5.0346,
      "mean_token_accuracy": 0.2108837068080902,
      "num_tokens": 30353423.0,
      "step": 13245
    },
    {
      "entropy": 5.172414541244507,
      "epoch": 1.2728146013448607,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004845057057938815,
      "loss": 4.9723,
      "mean_token_accuracy": 0.21434993147850037,
      "num_tokens": 30365008.0,
      "step": 13250
    },
    {
      "entropy": 5.099163627624511,
      "epoch": 1.2732949087415946,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048449320114644185,
      "loss": 4.9005,
      "mean_token_accuracy": 0.22391110211610793,
      "num_tokens": 30376589.0,
      "step": 13255
    },
    {
      "entropy": 5.159493684768677,
      "epoch": 1.2737752161383284,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004844806916351922,
      "loss": 5.0045,
      "mean_token_accuracy": 0.2077852413058281,
      "num_tokens": 30388228.0,
      "step": 13260
    },
    {
      "entropy": 5.204118394851685,
      "epoch": 1.2742555235350626,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004844681772604229,
      "loss": 5.0649,
      "mean_token_accuracy": 0.20871012806892394,
      "num_tokens": 30400225.0,
      "step": 13265
    },
    {
      "entropy": 5.210204792022705,
      "epoch": 1.2747358309317964,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048445565802242454,
      "loss": 5.0061,
      "mean_token_accuracy": 0.20563428848981857,
      "num_tokens": 30411913.0,
      "step": 13270
    },
    {
      "entropy": 5.200520849227905,
      "epoch": 1.2752161383285303,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004844431339214878,
      "loss": 4.9999,
      "mean_token_accuracy": 0.20750951766967773,
      "num_tokens": 30423891.0,
      "step": 13275
    },
    {
      "entropy": 5.251895999908447,
      "epoch": 1.2756964457252642,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004844306049579034,
      "loss": 5.0252,
      "mean_token_accuracy": 0.21057484149932862,
      "num_tokens": 30434144.0,
      "step": 13280
    },
    {
      "entropy": 5.166483402252197,
      "epoch": 1.276176753121998,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048441807113196216,
      "loss": 4.987,
      "mean_token_accuracy": 0.20822166204452514,
      "num_tokens": 30446628.0,
      "step": 13285
    },
    {
      "entropy": 5.267109060287476,
      "epoch": 1.276657060518732,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048440553244395517,
      "loss": 4.975,
      "mean_token_accuracy": 0.213594289124012,
      "num_tokens": 30456885.0,
      "step": 13290
    },
    {
      "entropy": 5.248937702178955,
      "epoch": 1.2771373679154658,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048439298889417357,
      "loss": 5.1062,
      "mean_token_accuracy": 0.20880619585514068,
      "num_tokens": 30468642.0,
      "step": 13295
    },
    {
      "entropy": 5.228408432006836,
      "epoch": 1.2776176753121997,
      "grad_norm": 1.375,
      "learning_rate": 0.00048438044048290847,
      "loss": 4.9941,
      "mean_token_accuracy": 0.2049667254090309,
      "num_tokens": 30480393.0,
      "step": 13300
    },
    {
      "entropy": 5.22806453704834,
      "epoch": 1.2780979827089336,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048436788721045135,
      "loss": 4.96,
      "mean_token_accuracy": 0.21500036716461182,
      "num_tokens": 30491918.0,
      "step": 13305
    },
    {
      "entropy": 5.21299786567688,
      "epoch": 1.2785782901056677,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004843553290770935,
      "loss": 4.9204,
      "mean_token_accuracy": 0.21900550574064254,
      "num_tokens": 30502250.0,
      "step": 13310
    },
    {
      "entropy": 5.162126874923706,
      "epoch": 1.2790585975024016,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004843427660831266,
      "loss": 4.966,
      "mean_token_accuracy": 0.2109062984585762,
      "num_tokens": 30514499.0,
      "step": 13315
    },
    {
      "entropy": 5.154655933380127,
      "epoch": 1.2795389048991355,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048433019822884235,
      "loss": 5.0053,
      "mean_token_accuracy": 0.20939105451107026,
      "num_tokens": 30525133.0,
      "step": 13320
    },
    {
      "entropy": 5.195331764221192,
      "epoch": 1.2800192122958693,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004843176255145325,
      "loss": 4.925,
      "mean_token_accuracy": 0.22028719633817673,
      "num_tokens": 30535778.0,
      "step": 13325
    },
    {
      "entropy": 5.174181270599365,
      "epoch": 1.2804995196926032,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004843050479404888,
      "loss": 4.9115,
      "mean_token_accuracy": 0.2227226972579956,
      "num_tokens": 30547711.0,
      "step": 13330
    },
    {
      "entropy": 5.1049683570861815,
      "epoch": 1.280979827089337,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048429246550700343,
      "loss": 4.882,
      "mean_token_accuracy": 0.2125895708799362,
      "num_tokens": 30558719.0,
      "step": 13335
    },
    {
      "entropy": 5.171171569824219,
      "epoch": 1.2814601344860712,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004842798782143686,
      "loss": 5.037,
      "mean_token_accuracy": 0.20505525916814804,
      "num_tokens": 30570173.0,
      "step": 13340
    },
    {
      "entropy": 5.231352186203003,
      "epoch": 1.281940441882805,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048426728606287627,
      "loss": 5.0039,
      "mean_token_accuracy": 0.21452445089817046,
      "num_tokens": 30580438.0,
      "step": 13345
    },
    {
      "entropy": 5.222430849075318,
      "epoch": 1.282420749279539,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004842546890528191,
      "loss": 4.9769,
      "mean_token_accuracy": 0.20783140361309052,
      "num_tokens": 30592191.0,
      "step": 13350
    },
    {
      "entropy": 5.182591104507447,
      "epoch": 1.2829010566762729,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004842420871844893,
      "loss": 5.0633,
      "mean_token_accuracy": 0.20939256250858307,
      "num_tokens": 30604041.0,
      "step": 13355
    },
    {
      "entropy": 5.135120248794555,
      "epoch": 1.2833813640730067,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004842294804581796,
      "loss": 4.9707,
      "mean_token_accuracy": 0.2216529667377472,
      "num_tokens": 30616087.0,
      "step": 13360
    },
    {
      "entropy": 5.219726514816284,
      "epoch": 1.2838616714697406,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048421686887418266,
      "loss": 4.9807,
      "mean_token_accuracy": 0.2148688942193985,
      "num_tokens": 30627150.0,
      "step": 13365
    },
    {
      "entropy": 5.22398567199707,
      "epoch": 1.2843419788664745,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004842042524327912,
      "loss": 4.9973,
      "mean_token_accuracy": 0.2107845887541771,
      "num_tokens": 30638627.0,
      "step": 13370
    },
    {
      "entropy": 5.236150693893433,
      "epoch": 1.2848222862632084,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004841916311342983,
      "loss": 5.0554,
      "mean_token_accuracy": 0.20471034646034242,
      "num_tokens": 30649257.0,
      "step": 13375
    },
    {
      "entropy": 5.169069099426269,
      "epoch": 1.2853025936599423,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004841790049789969,
      "loss": 5.0417,
      "mean_token_accuracy": 0.21622105985879897,
      "num_tokens": 30659765.0,
      "step": 13380
    },
    {
      "entropy": 5.188124179840088,
      "epoch": 1.2857829010566761,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048416637396718004,
      "loss": 4.9905,
      "mean_token_accuracy": 0.204777030646801,
      "num_tokens": 30672921.0,
      "step": 13385
    },
    {
      "entropy": 5.261342906951905,
      "epoch": 1.2862632084534102,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004841537380991411,
      "loss": 5.01,
      "mean_token_accuracy": 0.2119935542345047,
      "num_tokens": 30684994.0,
      "step": 13390
    },
    {
      "entropy": 5.229185009002686,
      "epoch": 1.2867435158501441,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048414109737517346,
      "loss": 5.0384,
      "mean_token_accuracy": 0.2086298167705536,
      "num_tokens": 30698094.0,
      "step": 13395
    },
    {
      "entropy": 5.25032753944397,
      "epoch": 1.287223823246878,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004841284517955706,
      "loss": 5.031,
      "mean_token_accuracy": 0.20857232809066772,
      "num_tokens": 30710716.0,
      "step": 13400
    },
    {
      "entropy": 5.214199781417847,
      "epoch": 1.2877041306436119,
      "grad_norm": 1.515625,
      "learning_rate": 0.000484115801360626,
      "loss": 5.035,
      "mean_token_accuracy": 0.2170042246580124,
      "num_tokens": 30721910.0,
      "step": 13405
    },
    {
      "entropy": 5.184518194198608,
      "epoch": 1.2881844380403458,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004841031460706335,
      "loss": 4.9875,
      "mean_token_accuracy": 0.21375347971916198,
      "num_tokens": 30734445.0,
      "step": 13410
    },
    {
      "entropy": 5.275124788284302,
      "epoch": 1.2886647454370797,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048409048592588683,
      "loss": 4.9473,
      "mean_token_accuracy": 0.22278327345848084,
      "num_tokens": 30745131.0,
      "step": 13415
    },
    {
      "entropy": 5.244984769821167,
      "epoch": 1.2891450528338138,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004840778209266799,
      "loss": 4.9927,
      "mean_token_accuracy": 0.2091956540942192,
      "num_tokens": 30757714.0,
      "step": 13420
    },
    {
      "entropy": 5.034538269042969,
      "epoch": 1.2896253602305476,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048406515107330685,
      "loss": 4.9578,
      "mean_token_accuracy": 0.2185475707054138,
      "num_tokens": 30768585.0,
      "step": 13425
    },
    {
      "entropy": 5.2477185249328615,
      "epoch": 1.2901056676272815,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048405247636606173,
      "loss": 5.0447,
      "mean_token_accuracy": 0.21114134937524795,
      "num_tokens": 30779866.0,
      "step": 13430
    },
    {
      "entropy": 5.247416639328003,
      "epoch": 1.2905859750240154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048403979680523894,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21192098557949066,
      "num_tokens": 30791252.0,
      "step": 13435
    },
    {
      "entropy": 5.211714458465576,
      "epoch": 1.2910662824207493,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004840271123911328,
      "loss": 5.0129,
      "mean_token_accuracy": 0.22117493599653243,
      "num_tokens": 30802938.0,
      "step": 13440
    },
    {
      "entropy": 5.278520298004151,
      "epoch": 1.2915465898174832,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004840144231240377,
      "loss": 5.1549,
      "mean_token_accuracy": 0.19686158150434493,
      "num_tokens": 30816382.0,
      "step": 13445
    },
    {
      "entropy": 5.279440021514892,
      "epoch": 1.292026897214217,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004840017290042484,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21663313508033752,
      "num_tokens": 30826402.0,
      "step": 13450
    },
    {
      "entropy": 5.20681791305542,
      "epoch": 1.292507204610951,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048398903003205957,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21727296561002732,
      "num_tokens": 30837862.0,
      "step": 13455
    },
    {
      "entropy": 5.20816593170166,
      "epoch": 1.2929875120076848,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048397632620776604,
      "loss": 4.926,
      "mean_token_accuracy": 0.21321765333414078,
      "num_tokens": 30848351.0,
      "step": 13460
    },
    {
      "entropy": 5.145204877853393,
      "epoch": 1.293467819404419,
      "grad_norm": 1.875,
      "learning_rate": 0.00048396361753166276,
      "loss": 4.991,
      "mean_token_accuracy": 0.21544851660728453,
      "num_tokens": 30858854.0,
      "step": 13465
    },
    {
      "entropy": 5.162298345565796,
      "epoch": 1.2939481268011528,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048395090400404466,
      "loss": 5.0192,
      "mean_token_accuracy": 0.21126462519168854,
      "num_tokens": 30869977.0,
      "step": 13470
    },
    {
      "entropy": 5.2936337947845455,
      "epoch": 1.2944284341978867,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048393818562520715,
      "loss": 5.0196,
      "mean_token_accuracy": 0.2177215427160263,
      "num_tokens": 30881620.0,
      "step": 13475
    },
    {
      "entropy": 5.226119756698608,
      "epoch": 1.2949087415946205,
      "grad_norm": 1.25,
      "learning_rate": 0.00048392546239544535,
      "loss": 5.0543,
      "mean_token_accuracy": 0.21232413798570632,
      "num_tokens": 30893575.0,
      "step": 13480
    },
    {
      "entropy": 5.15901746749878,
      "epoch": 1.2953890489913544,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004839127343150547,
      "loss": 4.9938,
      "mean_token_accuracy": 0.21345060616731643,
      "num_tokens": 30904489.0,
      "step": 13485
    },
    {
      "entropy": 5.270200920104981,
      "epoch": 1.2958693563880883,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004839000013843307,
      "loss": 5.0723,
      "mean_token_accuracy": 0.20590703785419465,
      "num_tokens": 30915513.0,
      "step": 13490
    },
    {
      "entropy": 5.269288063049316,
      "epoch": 1.2963496637848224,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048388726360356894,
      "loss": 5.0063,
      "mean_token_accuracy": 0.21073528081178666,
      "num_tokens": 30927633.0,
      "step": 13495
    },
    {
      "entropy": 5.325243234634399,
      "epoch": 1.2968299711815563,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004838745209730653,
      "loss": 5.1021,
      "mean_token_accuracy": 0.1983416572213173,
      "num_tokens": 30938937.0,
      "step": 13500
    },
    {
      "entropy": 5.2199643611907955,
      "epoch": 1.2973102785782902,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048386177349311535,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21419948935508729,
      "num_tokens": 30950847.0,
      "step": 13505
    },
    {
      "entropy": 5.190974378585816,
      "epoch": 1.297790585975024,
      "grad_norm": 1.25,
      "learning_rate": 0.0004838490211640153,
      "loss": 4.9408,
      "mean_token_accuracy": 0.2141892433166504,
      "num_tokens": 30962583.0,
      "step": 13510
    },
    {
      "entropy": 5.234346485137939,
      "epoch": 1.298270893371758,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004838362639860611,
      "loss": 5.017,
      "mean_token_accuracy": 0.20485191494226457,
      "num_tokens": 30973487.0,
      "step": 13515
    },
    {
      "entropy": 5.164180660247803,
      "epoch": 1.2987512007684918,
      "grad_norm": 1.171875,
      "learning_rate": 0.000483823501959549,
      "loss": 4.911,
      "mean_token_accuracy": 0.21933772712945937,
      "num_tokens": 30985176.0,
      "step": 13520
    },
    {
      "entropy": 5.234972190856934,
      "epoch": 1.2992315081652257,
      "grad_norm": 1.25,
      "learning_rate": 0.00048381073508477527,
      "loss": 5.1061,
      "mean_token_accuracy": 0.20437309443950652,
      "num_tokens": 30997239.0,
      "step": 13525
    },
    {
      "entropy": 5.271120500564575,
      "epoch": 1.2997118155619596,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048379796336203625,
      "loss": 5.0579,
      "mean_token_accuracy": 0.20211298018693924,
      "num_tokens": 31009381.0,
      "step": 13530
    },
    {
      "entropy": 5.2370641231536865,
      "epoch": 1.3001921229586935,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004837851867916286,
      "loss": 4.9349,
      "mean_token_accuracy": 0.21570398658514023,
      "num_tokens": 31020728.0,
      "step": 13535
    },
    {
      "entropy": 5.204510927200317,
      "epoch": 1.3006724303554273,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004837724053738488,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20896351784467698,
      "num_tokens": 31032933.0,
      "step": 13540
    },
    {
      "entropy": 5.242821741104126,
      "epoch": 1.3011527377521614,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048375961910899373,
      "loss": 5.0358,
      "mean_token_accuracy": 0.20853658020496368,
      "num_tokens": 31044919.0,
      "step": 13545
    },
    {
      "entropy": 5.250010824203491,
      "epoch": 1.3016330451488953,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004837468279973602,
      "loss": 5.0298,
      "mean_token_accuracy": 0.20970372557640077,
      "num_tokens": 31057140.0,
      "step": 13550
    },
    {
      "entropy": 5.174522161483765,
      "epoch": 1.3021133525456292,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004837340320392451,
      "loss": 4.9556,
      "mean_token_accuracy": 0.21604088693857193,
      "num_tokens": 31068035.0,
      "step": 13555
    },
    {
      "entropy": 5.206137704849243,
      "epoch": 1.302593659942363,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048372123123494563,
      "loss": 5.0153,
      "mean_token_accuracy": 0.21016810536384584,
      "num_tokens": 31079819.0,
      "step": 13560
    },
    {
      "entropy": 5.107056665420532,
      "epoch": 1.303073967339097,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000483708425584759,
      "loss": 4.9148,
      "mean_token_accuracy": 0.21521454751491548,
      "num_tokens": 31091272.0,
      "step": 13565
    },
    {
      "entropy": 5.207752704620361,
      "epoch": 1.3035542747358309,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004836956150889825,
      "loss": 5.0285,
      "mean_token_accuracy": 0.20764566957950592,
      "num_tokens": 31101513.0,
      "step": 13570
    },
    {
      "entropy": 5.243871259689331,
      "epoch": 1.304034582132565,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004836827997479134,
      "loss": 5.0711,
      "mean_token_accuracy": 0.20300604403018951,
      "num_tokens": 31112937.0,
      "step": 13575
    },
    {
      "entropy": 5.210528326034546,
      "epoch": 1.3045148895292988,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004836699795618494,
      "loss": 5.0589,
      "mean_token_accuracy": 0.21165550649166107,
      "num_tokens": 31123445.0,
      "step": 13580
    },
    {
      "entropy": 5.223980379104614,
      "epoch": 1.3049951969260327,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004836571545310881,
      "loss": 5.0021,
      "mean_token_accuracy": 0.2154447391629219,
      "num_tokens": 31135016.0,
      "step": 13585
    },
    {
      "entropy": 5.139055633544922,
      "epoch": 1.3054755043227666,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048364432465592723,
      "loss": 4.9727,
      "mean_token_accuracy": 0.20944690555334092,
      "num_tokens": 31146772.0,
      "step": 13590
    },
    {
      "entropy": 5.193195247650147,
      "epoch": 1.3059558117195005,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004836314899366647,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21997501850128173,
      "num_tokens": 31157977.0,
      "step": 13595
    },
    {
      "entropy": 5.17445387840271,
      "epoch": 1.3064361191162344,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048361865037359846,
      "loss": 4.9775,
      "mean_token_accuracy": 0.21851352602243423,
      "num_tokens": 31170693.0,
      "step": 13600
    },
    {
      "entropy": 5.232918787002563,
      "epoch": 1.3069164265129682,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048360580596702664,
      "loss": 5.0035,
      "mean_token_accuracy": 0.21215256303548813,
      "num_tokens": 31181276.0,
      "step": 13605
    },
    {
      "entropy": 5.2662091732025145,
      "epoch": 1.3073967339097021,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048359295671724744,
      "loss": 5.109,
      "mean_token_accuracy": 0.20561769604682922,
      "num_tokens": 31192740.0,
      "step": 13610
    },
    {
      "entropy": 5.245311260223389,
      "epoch": 1.307877041306436,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004835801026245592,
      "loss": 5.0032,
      "mean_token_accuracy": 0.20504884868860246,
      "num_tokens": 31204075.0,
      "step": 13615
    },
    {
      "entropy": 5.262555456161499,
      "epoch": 1.30835734870317,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048356724368926035,
      "loss": 5.08,
      "mean_token_accuracy": 0.20021516531705857,
      "num_tokens": 31215590.0,
      "step": 13620
    },
    {
      "entropy": 5.2641185283660885,
      "epoch": 1.308837656099904,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048355437991164937,
      "loss": 5.0709,
      "mean_token_accuracy": 0.2019081935286522,
      "num_tokens": 31228427.0,
      "step": 13625
    },
    {
      "entropy": 5.258098936080932,
      "epoch": 1.3093179634966379,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000483541511292025,
      "loss": 5.0798,
      "mean_token_accuracy": 0.2146941527724266,
      "num_tokens": 31240348.0,
      "step": 13630
    },
    {
      "entropy": 5.287716341018677,
      "epoch": 1.3097982708933718,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048352863783068594,
      "loss": 5.0967,
      "mean_token_accuracy": 0.2075056254863739,
      "num_tokens": 31253028.0,
      "step": 13635
    },
    {
      "entropy": 5.28109221458435,
      "epoch": 1.3102785782901056,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048351575952793117,
      "loss": 5.0577,
      "mean_token_accuracy": 0.208870966732502,
      "num_tokens": 31265242.0,
      "step": 13640
    },
    {
      "entropy": 5.197099018096924,
      "epoch": 1.3107588856868395,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004835028763840595,
      "loss": 5.0047,
      "mean_token_accuracy": 0.2099291354417801,
      "num_tokens": 31277248.0,
      "step": 13645
    },
    {
      "entropy": 5.195971345901489,
      "epoch": 1.3112391930835736,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004834899883993703,
      "loss": 5.072,
      "mean_token_accuracy": 0.20533291548490523,
      "num_tokens": 31289631.0,
      "step": 13650
    },
    {
      "entropy": 5.134831285476684,
      "epoch": 1.3117195004803075,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048347709557416263,
      "loss": 4.8959,
      "mean_token_accuracy": 0.21984872668981553,
      "num_tokens": 31301006.0,
      "step": 13655
    },
    {
      "entropy": 5.2645245552062985,
      "epoch": 1.3121998078770414,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004834641979087359,
      "loss": 5.0916,
      "mean_token_accuracy": 0.20365192294120787,
      "num_tokens": 31312621.0,
      "step": 13660
    },
    {
      "entropy": 5.296423053741455,
      "epoch": 1.3126801152737753,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004834512954033894,
      "loss": 5.0684,
      "mean_token_accuracy": 0.21249050498008729,
      "num_tokens": 31324173.0,
      "step": 13665
    },
    {
      "entropy": 5.230805253982544,
      "epoch": 1.3131604226705091,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048343838805842284,
      "loss": 5.0064,
      "mean_token_accuracy": 0.21248940825462342,
      "num_tokens": 31336872.0,
      "step": 13670
    },
    {
      "entropy": 5.249256801605225,
      "epoch": 1.313640730067243,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048342547587413583,
      "loss": 5.0432,
      "mean_token_accuracy": 0.20969412475824356,
      "num_tokens": 31348760.0,
      "step": 13675
    },
    {
      "entropy": 5.278874301910401,
      "epoch": 1.314121037463977,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004834125588508282,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21372607052326204,
      "num_tokens": 31360591.0,
      "step": 13680
    },
    {
      "entropy": 5.250448751449585,
      "epoch": 1.3146013448607108,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004833996369887998,
      "loss": 5.1982,
      "mean_token_accuracy": 0.20440512001514435,
      "num_tokens": 31372273.0,
      "step": 13685
    },
    {
      "entropy": 5.211995458602905,
      "epoch": 1.3150816522574447,
      "grad_norm": 1.25,
      "learning_rate": 0.00048338671028835063,
      "loss": 4.9886,
      "mean_token_accuracy": 0.21652114242315293,
      "num_tokens": 31382569.0,
      "step": 13690
    },
    {
      "entropy": 5.32315993309021,
      "epoch": 1.3155619596541785,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048337377874978086,
      "loss": 5.0851,
      "mean_token_accuracy": 0.20504068583250046,
      "num_tokens": 31393913.0,
      "step": 13695
    },
    {
      "entropy": 5.205372333526611,
      "epoch": 1.3160422670509127,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048336084237339067,
      "loss": 4.952,
      "mean_token_accuracy": 0.21446582525968552,
      "num_tokens": 31405645.0,
      "step": 13700
    },
    {
      "entropy": 5.118508815765381,
      "epoch": 1.3165225744476465,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004833479011594804,
      "loss": 4.9271,
      "mean_token_accuracy": 0.2186314731836319,
      "num_tokens": 31418728.0,
      "step": 13705
    },
    {
      "entropy": 5.189344167709351,
      "epoch": 1.3170028818443804,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048333495510835057,
      "loss": 4.9646,
      "mean_token_accuracy": 0.21466858685016632,
      "num_tokens": 31430226.0,
      "step": 13710
    },
    {
      "entropy": 5.22680435180664,
      "epoch": 1.3174831892411143,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048332200422030163,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21814581006765366,
      "num_tokens": 31441751.0,
      "step": 13715
    },
    {
      "entropy": 5.170705413818359,
      "epoch": 1.3179634966378482,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004833090484956345,
      "loss": 5.028,
      "mean_token_accuracy": 0.21439666748046876,
      "num_tokens": 31455413.0,
      "step": 13720
    },
    {
      "entropy": 5.233014106750488,
      "epoch": 1.318443804034582,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048329608793464966,
      "loss": 4.8937,
      "mean_token_accuracy": 0.22648588716983795,
      "num_tokens": 31466486.0,
      "step": 13725
    },
    {
      "entropy": 5.181014776229858,
      "epoch": 1.3189241114313162,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004832831225376482,
      "loss": 5.0251,
      "mean_token_accuracy": 0.21297899931669234,
      "num_tokens": 31477048.0,
      "step": 13730
    },
    {
      "entropy": 5.2522327423095705,
      "epoch": 1.31940441882805,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004832701523049312,
      "loss": 5.0423,
      "mean_token_accuracy": 0.20735066384077072,
      "num_tokens": 31488015.0,
      "step": 13735
    },
    {
      "entropy": 5.177381563186645,
      "epoch": 1.319884726224784,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004832571772367997,
      "loss": 4.9339,
      "mean_token_accuracy": 0.21644928604364394,
      "num_tokens": 31499735.0,
      "step": 13740
    },
    {
      "entropy": 5.086191082000733,
      "epoch": 1.3203650336215178,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048324419733355485,
      "loss": 4.7745,
      "mean_token_accuracy": 0.22439781874418258,
      "num_tokens": 31511970.0,
      "step": 13745
    },
    {
      "entropy": 5.258924102783203,
      "epoch": 1.3208453410182517,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048323121259549805,
      "loss": 5.0241,
      "mean_token_accuracy": 0.21613921523094176,
      "num_tokens": 31522780.0,
      "step": 13750
    },
    {
      "entropy": 5.228280925750733,
      "epoch": 1.3213256484149856,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048321822302293095,
      "loss": 5.0255,
      "mean_token_accuracy": 0.2109622076153755,
      "num_tokens": 31535578.0,
      "step": 13755
    },
    {
      "entropy": 5.236905670166015,
      "epoch": 1.3218059558117194,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004832052286161549,
      "loss": 5.1063,
      "mean_token_accuracy": 0.19888232201337813,
      "num_tokens": 31546623.0,
      "step": 13760
    },
    {
      "entropy": 5.304514980316162,
      "epoch": 1.3222862632084533,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048319222937547176,
      "loss": 5.1726,
      "mean_token_accuracy": 0.205259270966053,
      "num_tokens": 31559334.0,
      "step": 13765
    },
    {
      "entropy": 5.287638854980469,
      "epoch": 1.3227665706051872,
      "grad_norm": 1.375,
      "learning_rate": 0.00048317922530118323,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21478652209043503,
      "num_tokens": 31570431.0,
      "step": 13770
    },
    {
      "entropy": 5.138346004486084,
      "epoch": 1.3232468780019213,
      "grad_norm": 1.25,
      "learning_rate": 0.0004831662163935912,
      "loss": 4.8619,
      "mean_token_accuracy": 0.22110362648963927,
      "num_tokens": 31580314.0,
      "step": 13775
    },
    {
      "entropy": 5.161703729629517,
      "epoch": 1.3237271853986552,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048315320265299784,
      "loss": 4.9438,
      "mean_token_accuracy": 0.2206251084804535,
      "num_tokens": 31592071.0,
      "step": 13780
    },
    {
      "entropy": 5.285171556472778,
      "epoch": 1.324207492795389,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048314018407970516,
      "loss": 5.1159,
      "mean_token_accuracy": 0.2036852404475212,
      "num_tokens": 31603143.0,
      "step": 13785
    },
    {
      "entropy": 5.269954872131348,
      "epoch": 1.324687800192123,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048312716067401535,
      "loss": 5.1162,
      "mean_token_accuracy": 0.20816019624471666,
      "num_tokens": 31613979.0,
      "step": 13790
    },
    {
      "entropy": 5.180557441711426,
      "epoch": 1.3251681075888568,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000483114132436231,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21332663148641587,
      "num_tokens": 31625206.0,
      "step": 13795
    },
    {
      "entropy": 5.2473814487457275,
      "epoch": 1.3256484149855907,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004831010993666545,
      "loss": 5.0992,
      "mean_token_accuracy": 0.2111356794834137,
      "num_tokens": 31637709.0,
      "step": 13800
    },
    {
      "entropy": 5.273555183410645,
      "epoch": 1.3261287223823248,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004830880614655884,
      "loss": 5.03,
      "mean_token_accuracy": 0.20575864464044571,
      "num_tokens": 31648754.0,
      "step": 13805
    },
    {
      "entropy": 5.200221586227417,
      "epoch": 1.3266090297790587,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048307501873333527,
      "loss": 4.9007,
      "mean_token_accuracy": 0.22089982777833939,
      "num_tokens": 31660177.0,
      "step": 13810
    },
    {
      "entropy": 5.192464923858642,
      "epoch": 1.3270893371757926,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004830619711701982,
      "loss": 5.0457,
      "mean_token_accuracy": 0.21462684720754624,
      "num_tokens": 31671467.0,
      "step": 13815
    },
    {
      "entropy": 5.290485000610351,
      "epoch": 1.3275696445725265,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004830489187764799,
      "loss": 5.0337,
      "mean_token_accuracy": 0.2115232452750206,
      "num_tokens": 31683073.0,
      "step": 13820
    },
    {
      "entropy": 5.162988138198853,
      "epoch": 1.3280499519692603,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004830358615524835,
      "loss": 4.9568,
      "mean_token_accuracy": 0.2122074097394943,
      "num_tokens": 31694661.0,
      "step": 13825
    },
    {
      "entropy": 5.146875333786011,
      "epoch": 1.3285302593659942,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048302279949851215,
      "loss": 4.9986,
      "mean_token_accuracy": 0.2104497179389,
      "num_tokens": 31706904.0,
      "step": 13830
    },
    {
      "entropy": 5.243095207214355,
      "epoch": 1.329010566762728,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048300973261486906,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21630474478006362,
      "num_tokens": 31718087.0,
      "step": 13835
    },
    {
      "entropy": 5.147808933258057,
      "epoch": 1.329490874159462,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004829966609018577,
      "loss": 4.8615,
      "mean_token_accuracy": 0.21689556390047074,
      "num_tokens": 31728960.0,
      "step": 13840
    },
    {
      "entropy": 5.233026647567749,
      "epoch": 1.3299711815561959,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048298358435978146,
      "loss": 5.0263,
      "mean_token_accuracy": 0.2107004478573799,
      "num_tokens": 31741461.0,
      "step": 13845
    },
    {
      "entropy": 5.1558229446411135,
      "epoch": 1.3304514889529298,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048297050298894394,
      "loss": 4.9787,
      "mean_token_accuracy": 0.20693062245845795,
      "num_tokens": 31753582.0,
      "step": 13850
    },
    {
      "entropy": 5.119613695144653,
      "epoch": 1.3309317963496639,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004829574167896489,
      "loss": 4.9245,
      "mean_token_accuracy": 0.21973290294408798,
      "num_tokens": 31764354.0,
      "step": 13855
    },
    {
      "entropy": 5.204716777801513,
      "epoch": 1.3314121037463977,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048294432576220027,
      "loss": 4.9756,
      "mean_token_accuracy": 0.2088870570063591,
      "num_tokens": 31775860.0,
      "step": 13860
    },
    {
      "entropy": 5.261440181732178,
      "epoch": 1.3318924111431316,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004829312299069017,
      "loss": 4.9983,
      "mean_token_accuracy": 0.2129855439066887,
      "num_tokens": 31787606.0,
      "step": 13865
    },
    {
      "entropy": 5.19459981918335,
      "epoch": 1.3323727185398655,
      "grad_norm": 1.25,
      "learning_rate": 0.00048291812922405755,
      "loss": 4.9854,
      "mean_token_accuracy": 0.2124458596110344,
      "num_tokens": 31798601.0,
      "step": 13870
    },
    {
      "entropy": 5.204279994964599,
      "epoch": 1.3328530259365994,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004829050237139717,
      "loss": 4.9303,
      "mean_token_accuracy": 0.2194435179233551,
      "num_tokens": 31809843.0,
      "step": 13875
    },
    {
      "entropy": 5.236581516265869,
      "epoch": 1.3333333333333333,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004828919133769486,
      "loss": 5.0544,
      "mean_token_accuracy": 0.20692742317914964,
      "num_tokens": 31821309.0,
      "step": 13880
    },
    {
      "entropy": 5.214200639724732,
      "epoch": 1.3338136407300674,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004828787982132926,
      "loss": 4.9612,
      "mean_token_accuracy": 0.2103517845273018,
      "num_tokens": 31832215.0,
      "step": 13885
    },
    {
      "entropy": 5.160177850723267,
      "epoch": 1.3342939481268012,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048286567822330815,
      "loss": 4.9356,
      "mean_token_accuracy": 0.2206076666712761,
      "num_tokens": 31843254.0,
      "step": 13890
    },
    {
      "entropy": 5.10599422454834,
      "epoch": 1.3347742555235351,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004828525534072999,
      "loss": 4.9535,
      "mean_token_accuracy": 0.21035333424806596,
      "num_tokens": 31855678.0,
      "step": 13895
    },
    {
      "entropy": 5.254456281661987,
      "epoch": 1.335254562920269,
      "grad_norm": 1.25,
      "learning_rate": 0.00048283942376557254,
      "loss": 5.0429,
      "mean_token_accuracy": 0.21277134269475936,
      "num_tokens": 31867264.0,
      "step": 13900
    },
    {
      "entropy": 5.139927387237549,
      "epoch": 1.3357348703170029,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048282628929843097,
      "loss": 4.9005,
      "mean_token_accuracy": 0.2252553090453148,
      "num_tokens": 31878908.0,
      "step": 13905
    },
    {
      "entropy": 5.195985317230225,
      "epoch": 1.3362151777137368,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048281315000617996,
      "loss": 5.009,
      "mean_token_accuracy": 0.2161973625421524,
      "num_tokens": 31890299.0,
      "step": 13910
    },
    {
      "entropy": 5.200396919250489,
      "epoch": 1.3366954851104706,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004828000058891248,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2154739335179329,
      "num_tokens": 31901279.0,
      "step": 13915
    },
    {
      "entropy": 5.211345100402832,
      "epoch": 1.3371757925072045,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004827868569475706,
      "loss": 4.9946,
      "mean_token_accuracy": 0.21455213129520417,
      "num_tokens": 31912851.0,
      "step": 13920
    },
    {
      "entropy": 5.188769102096558,
      "epoch": 1.3376560999039384,
      "grad_norm": 2.359375,
      "learning_rate": 0.00048277370318182243,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21996437907218933,
      "num_tokens": 31923479.0,
      "step": 13925
    },
    {
      "entropy": 5.207550668716431,
      "epoch": 1.3381364073006723,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048276054459218596,
      "loss": 5.0695,
      "mean_token_accuracy": 0.2104545548558235,
      "num_tokens": 31935035.0,
      "step": 13930
    },
    {
      "entropy": 5.259209775924683,
      "epoch": 1.3386167146974064,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048274738117896643,
      "loss": 5.0515,
      "mean_token_accuracy": 0.20499828457832336,
      "num_tokens": 31946140.0,
      "step": 13935
    },
    {
      "entropy": 5.188251924514771,
      "epoch": 1.3390970220941403,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048273421294246966,
      "loss": 4.9642,
      "mean_token_accuracy": 0.20899737626314163,
      "num_tokens": 31957315.0,
      "step": 13940
    },
    {
      "entropy": 5.135606336593628,
      "epoch": 1.3395773294908742,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048272103988300134,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21634371131658553,
      "num_tokens": 31967773.0,
      "step": 13945
    },
    {
      "entropy": 5.315240287780762,
      "epoch": 1.340057636887608,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004827078620008672,
      "loss": 5.1038,
      "mean_token_accuracy": 0.20027837306261062,
      "num_tokens": 31980112.0,
      "step": 13950
    },
    {
      "entropy": 5.196237277984619,
      "epoch": 1.340537944284342,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048269467929637337,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2225482925772667,
      "num_tokens": 31991655.0,
      "step": 13955
    },
    {
      "entropy": 5.109450483322144,
      "epoch": 1.341018251681076,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048268149176982576,
      "loss": 4.9333,
      "mean_token_accuracy": 0.22459781169891357,
      "num_tokens": 32003551.0,
      "step": 13960
    },
    {
      "entropy": 5.108917379379273,
      "epoch": 1.34149855907781,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048266829942153055,
      "loss": 4.881,
      "mean_token_accuracy": 0.2192325174808502,
      "num_tokens": 32015096.0,
      "step": 13965
    },
    {
      "entropy": 5.238538646697998,
      "epoch": 1.3419788664745438,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048265510225179413,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20859594643115997,
      "num_tokens": 32026219.0,
      "step": 13970
    },
    {
      "entropy": 5.210604858398438,
      "epoch": 1.3424591738712777,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004826419002609229,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2123058870434761,
      "num_tokens": 32037301.0,
      "step": 13975
    },
    {
      "entropy": 5.153656911849976,
      "epoch": 1.3429394812680115,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048262869344922326,
      "loss": 4.9218,
      "mean_token_accuracy": 0.22116306871175767,
      "num_tokens": 32048707.0,
      "step": 13980
    },
    {
      "entropy": 5.320045757293701,
      "epoch": 1.3434197886647454,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048261548181700186,
      "loss": 5.0768,
      "mean_token_accuracy": 0.1979660466313362,
      "num_tokens": 32061008.0,
      "step": 13985
    },
    {
      "entropy": 5.142725229263306,
      "epoch": 1.3439000960614793,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004826022653645655,
      "loss": 4.8748,
      "mean_token_accuracy": 0.22682179808616637,
      "num_tokens": 32071360.0,
      "step": 13990
    },
    {
      "entropy": 5.117513418197632,
      "epoch": 1.3443804034582132,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000482589044092221,
      "loss": 4.9617,
      "mean_token_accuracy": 0.2133632093667984,
      "num_tokens": 32083441.0,
      "step": 13995
    },
    {
      "entropy": 5.168340063095092,
      "epoch": 1.344860710854947,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048257581800027527,
      "loss": 4.9865,
      "mean_token_accuracy": 0.21795926988124847,
      "num_tokens": 32094773.0,
      "step": 14000
    },
    {
      "entropy": 5.169700813293457,
      "epoch": 1.345341018251681,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004825625870890354,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21260569989681244,
      "num_tokens": 32107076.0,
      "step": 14005
    },
    {
      "entropy": 5.355350923538208,
      "epoch": 1.345821325648415,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004825493513588086,
      "loss": 5.1313,
      "mean_token_accuracy": 0.19574850052595139,
      "num_tokens": 32118781.0,
      "step": 14010
    },
    {
      "entropy": 5.177445936203003,
      "epoch": 1.346301633045149,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048253611080990226,
      "loss": 4.9368,
      "mean_token_accuracy": 0.2191154956817627,
      "num_tokens": 32130181.0,
      "step": 14015
    },
    {
      "entropy": 5.203184986114502,
      "epoch": 1.3467819404418828,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004825228654426236,
      "loss": 4.9667,
      "mean_token_accuracy": 0.21455437391996385,
      "num_tokens": 32142576.0,
      "step": 14020
    },
    {
      "entropy": 5.201105737686158,
      "epoch": 1.3472622478386167,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004825096152572803,
      "loss": 4.9677,
      "mean_token_accuracy": 0.21448068618774413,
      "num_tokens": 32152413.0,
      "step": 14025
    },
    {
      "entropy": 5.186262798309326,
      "epoch": 1.3477425552353506,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048249636025417974,
      "loss": 4.9717,
      "mean_token_accuracy": 0.21499240696430205,
      "num_tokens": 32163539.0,
      "step": 14030
    },
    {
      "entropy": 5.275635766983032,
      "epoch": 1.3482228626320845,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048248310043362997,
      "loss": 4.9712,
      "mean_token_accuracy": 0.21079451888799666,
      "num_tokens": 32174662.0,
      "step": 14035
    },
    {
      "entropy": 5.133161973953247,
      "epoch": 1.3487031700288186,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004824698357959386,
      "loss": 4.9686,
      "mean_token_accuracy": 0.21750828325748445,
      "num_tokens": 32185950.0,
      "step": 14040
    },
    {
      "entropy": 5.204601240158081,
      "epoch": 1.3491834774255524,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048245656634141385,
      "loss": 5.0478,
      "mean_token_accuracy": 0.21049775779247284,
      "num_tokens": 32198076.0,
      "step": 14045
    },
    {
      "entropy": 5.280143928527832,
      "epoch": 1.3496637848222863,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048244329207036354,
      "loss": 5.0056,
      "mean_token_accuracy": 0.21039787530899048,
      "num_tokens": 32208739.0,
      "step": 14050
    },
    {
      "entropy": 5.269425296783448,
      "epoch": 1.3501440922190202,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048243001298309604,
      "loss": 5.161,
      "mean_token_accuracy": 0.20239224433898925,
      "num_tokens": 32221494.0,
      "step": 14055
    },
    {
      "entropy": 5.18841290473938,
      "epoch": 1.350624399615754,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048241672907991954,
      "loss": 4.9412,
      "mean_token_accuracy": 0.21473922729492187,
      "num_tokens": 32232991.0,
      "step": 14060
    },
    {
      "entropy": 5.219390726089477,
      "epoch": 1.351104707012488,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004824034403611424,
      "loss": 5.0292,
      "mean_token_accuracy": 0.2088583678007126,
      "num_tokens": 32244022.0,
      "step": 14065
    },
    {
      "entropy": 5.162661790847778,
      "epoch": 1.3515850144092219,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004823901468270733,
      "loss": 4.971,
      "mean_token_accuracy": 0.20842716246843337,
      "num_tokens": 32255779.0,
      "step": 14070
    },
    {
      "entropy": 5.170685052871704,
      "epoch": 1.3520653218059557,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004823768484780209,
      "loss": 4.913,
      "mean_token_accuracy": 0.21686336994171143,
      "num_tokens": 32267505.0,
      "step": 14075
    },
    {
      "entropy": 5.178246021270752,
      "epoch": 1.3525456292026896,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048236354531429375,
      "loss": 5.0104,
      "mean_token_accuracy": 0.21996340453624724,
      "num_tokens": 32278675.0,
      "step": 14080
    },
    {
      "entropy": 5.145812034606934,
      "epoch": 1.3530259365994235,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004823502373362009,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21581885665655137,
      "num_tokens": 32289372.0,
      "step": 14085
    },
    {
      "entropy": 5.174611568450928,
      "epoch": 1.3535062439961576,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004823369245440512,
      "loss": 4.9643,
      "mean_token_accuracy": 0.2155382066965103,
      "num_tokens": 32301281.0,
      "step": 14090
    },
    {
      "entropy": 5.266873502731324,
      "epoch": 1.3539865513928915,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048232360693815387,
      "loss": 5.0009,
      "mean_token_accuracy": 0.21678151339292526,
      "num_tokens": 32312204.0,
      "step": 14095
    },
    {
      "entropy": 5.21032452583313,
      "epoch": 1.3544668587896254,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048231028451881786,
      "loss": 4.9666,
      "mean_token_accuracy": 0.2161950558423996,
      "num_tokens": 32323607.0,
      "step": 14100
    },
    {
      "entropy": 5.202493286132812,
      "epoch": 1.3549471661863592,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004822969572863527,
      "loss": 5.0017,
      "mean_token_accuracy": 0.22105452567338943,
      "num_tokens": 32334864.0,
      "step": 14105
    },
    {
      "entropy": 5.171344232559204,
      "epoch": 1.3554274735830931,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048228362524106776,
      "loss": 5.0128,
      "mean_token_accuracy": 0.20835835188627244,
      "num_tokens": 32345664.0,
      "step": 14110
    },
    {
      "entropy": 5.249145126342773,
      "epoch": 1.3559077809798272,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048227028838327253,
      "loss": 4.966,
      "mean_token_accuracy": 0.2264058396220207,
      "num_tokens": 32355727.0,
      "step": 14115
    },
    {
      "entropy": 5.18209319114685,
      "epoch": 1.356388088376561,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048225694671327665,
      "loss": 4.8499,
      "mean_token_accuracy": 0.22090059518814087,
      "num_tokens": 32367132.0,
      "step": 14120
    },
    {
      "entropy": 5.1642967700958256,
      "epoch": 1.356868395773295,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004822436002313899,
      "loss": 4.867,
      "mean_token_accuracy": 0.21960778683423995,
      "num_tokens": 32377599.0,
      "step": 14125
    },
    {
      "entropy": 5.221139860153198,
      "epoch": 1.3573487031700289,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004822302489379222,
      "loss": 5.0537,
      "mean_token_accuracy": 0.20366920977830888,
      "num_tokens": 32389659.0,
      "step": 14130
    },
    {
      "entropy": 5.183340167999267,
      "epoch": 1.3578290105667628,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048221689283318335,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21909718364477157,
      "num_tokens": 32400573.0,
      "step": 14135
    },
    {
      "entropy": 5.224713134765625,
      "epoch": 1.3583093179634966,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004822035319174837,
      "loss": 4.964,
      "mean_token_accuracy": 0.2175660640001297,
      "num_tokens": 32411009.0,
      "step": 14140
    },
    {
      "entropy": 5.1596277236938475,
      "epoch": 1.3587896253602305,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004821901661911332,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2167072281241417,
      "num_tokens": 32422489.0,
      "step": 14145
    },
    {
      "entropy": 5.12332239151001,
      "epoch": 1.3592699327569644,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004821767956544423,
      "loss": 4.8512,
      "mean_token_accuracy": 0.21995280385017396,
      "num_tokens": 32432613.0,
      "step": 14150
    },
    {
      "entropy": 5.130664348602295,
      "epoch": 1.3597502401536983,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004821634203077214,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21551697701215744,
      "num_tokens": 32444556.0,
      "step": 14155
    },
    {
      "entropy": 5.2179999351501465,
      "epoch": 1.3602305475504322,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000482150040151281,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22001786530017853,
      "num_tokens": 32455143.0,
      "step": 14160
    },
    {
      "entropy": 5.166575622558594,
      "epoch": 1.3607108549471663,
      "grad_norm": 1.125,
      "learning_rate": 0.0004821366551854318,
      "loss": 4.992,
      "mean_token_accuracy": 0.21432398706674577,
      "num_tokens": 32466966.0,
      "step": 14165
    },
    {
      "entropy": 5.118087530136108,
      "epoch": 1.3611911623439001,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004821232654104845,
      "loss": 4.9097,
      "mean_token_accuracy": 0.2189925104379654,
      "num_tokens": 32479435.0,
      "step": 14170
    },
    {
      "entropy": 5.148586893081665,
      "epoch": 1.361671469740634,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048210987082675005,
      "loss": 4.991,
      "mean_token_accuracy": 0.2165745496749878,
      "num_tokens": 32490923.0,
      "step": 14175
    },
    {
      "entropy": 5.26572003364563,
      "epoch": 1.362151777137368,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048209647143453946,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21697789281606675,
      "num_tokens": 32500939.0,
      "step": 14180
    },
    {
      "entropy": 5.095041751861572,
      "epoch": 1.3626320845341018,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048208306723416356,
      "loss": 4.9145,
      "mean_token_accuracy": 0.21775346547365187,
      "num_tokens": 32513260.0,
      "step": 14185
    },
    {
      "entropy": 5.27839674949646,
      "epoch": 1.3631123919308357,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004820696582259339,
      "loss": 5.0545,
      "mean_token_accuracy": 0.21027806550264358,
      "num_tokens": 32524813.0,
      "step": 14190
    },
    {
      "entropy": 5.230822229385376,
      "epoch": 1.3635926993275698,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004820562444101616,
      "loss": 5.0003,
      "mean_token_accuracy": 0.2081260398030281,
      "num_tokens": 32536904.0,
      "step": 14195
    },
    {
      "entropy": 5.131050539016724,
      "epoch": 1.3640730067243036,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004820428257871581,
      "loss": 4.9031,
      "mean_token_accuracy": 0.217434361577034,
      "num_tokens": 32547419.0,
      "step": 14200
    },
    {
      "entropy": 5.100382137298584,
      "epoch": 1.3645533141210375,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004820294023572351,
      "loss": 4.8447,
      "mean_token_accuracy": 0.2238169178366661,
      "num_tokens": 32558644.0,
      "step": 14205
    },
    {
      "entropy": 5.153950548171997,
      "epoch": 1.3650336215177714,
      "grad_norm": 1.25,
      "learning_rate": 0.000482015974120704,
      "loss": 4.9167,
      "mean_token_accuracy": 0.22043030560016633,
      "num_tokens": 32569629.0,
      "step": 14210
    },
    {
      "entropy": 5.142817115783691,
      "epoch": 1.3655139289145053,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048200254107787677,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21136997640132904,
      "num_tokens": 32581850.0,
      "step": 14215
    },
    {
      "entropy": 5.324521112442016,
      "epoch": 1.3659942363112392,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048198910322906516,
      "loss": 5.0715,
      "mean_token_accuracy": 0.20727096647024154,
      "num_tokens": 32592478.0,
      "step": 14220
    },
    {
      "entropy": 5.299382877349854,
      "epoch": 1.366474543707973,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048197566057458125,
      "loss": 5.1711,
      "mean_token_accuracy": 0.19924191981554032,
      "num_tokens": 32603099.0,
      "step": 14225
    },
    {
      "entropy": 5.284470462799073,
      "epoch": 1.366954851104707,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000481962213114737,
      "loss": 5.0611,
      "mean_token_accuracy": 0.20557418912649156,
      "num_tokens": 32613415.0,
      "step": 14230
    },
    {
      "entropy": 5.1805455684661865,
      "epoch": 1.3674351585014408,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004819487608498448,
      "loss": 5.0196,
      "mean_token_accuracy": 0.2177915096282959,
      "num_tokens": 32625348.0,
      "step": 14235
    },
    {
      "entropy": 5.18522310256958,
      "epoch": 1.3679154658981747,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048193530378021687,
      "loss": 4.9197,
      "mean_token_accuracy": 0.21621484011411668,
      "num_tokens": 32636878.0,
      "step": 14240
    },
    {
      "entropy": 5.210684299468994,
      "epoch": 1.3683957732949088,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048192184190616567,
      "loss": 5.1046,
      "mean_token_accuracy": 0.20881797075271608,
      "num_tokens": 32647808.0,
      "step": 14245
    },
    {
      "entropy": 5.434156227111816,
      "epoch": 1.3688760806916427,
      "grad_norm": 1.25,
      "learning_rate": 0.0004819083752280037,
      "loss": 5.163,
      "mean_token_accuracy": 0.19552553594112396,
      "num_tokens": 32659425.0,
      "step": 14250
    },
    {
      "entropy": 5.2306708812713625,
      "epoch": 1.3693563880883766,
      "grad_norm": 1.125,
      "learning_rate": 0.00048189490374604373,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21178983747959138,
      "num_tokens": 32671797.0,
      "step": 14255
    },
    {
      "entropy": 5.155837774276733,
      "epoch": 1.3698366954851104,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004818814274605983,
      "loss": 4.9849,
      "mean_token_accuracy": 0.20952397733926773,
      "num_tokens": 32683506.0,
      "step": 14260
    },
    {
      "entropy": 5.240473318099975,
      "epoch": 1.3703170028818443,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004818679463719805,
      "loss": 5.0961,
      "mean_token_accuracy": 0.19705576598644256,
      "num_tokens": 32695047.0,
      "step": 14265
    },
    {
      "entropy": 5.18463830947876,
      "epoch": 1.3707973102785782,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004818544604805033,
      "loss": 4.9261,
      "mean_token_accuracy": 0.22047822624444963,
      "num_tokens": 32705433.0,
      "step": 14270
    },
    {
      "entropy": 5.211135911941528,
      "epoch": 1.3712776176753123,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004818409697864798,
      "loss": 5.0697,
      "mean_token_accuracy": 0.20451985597610473,
      "num_tokens": 32717315.0,
      "step": 14275
    },
    {
      "entropy": 5.203352355957032,
      "epoch": 1.3717579250720462,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048182747429022303,
      "loss": 4.9363,
      "mean_token_accuracy": 0.2213026240468025,
      "num_tokens": 32728817.0,
      "step": 14280
    },
    {
      "entropy": 5.246301984786987,
      "epoch": 1.37223823246878,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004818139739920465,
      "loss": 5.084,
      "mean_token_accuracy": 0.2083810895681381,
      "num_tokens": 32741030.0,
      "step": 14285
    },
    {
      "entropy": 5.243307781219483,
      "epoch": 1.372718539865514,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004818004688922637,
      "loss": 5.1112,
      "mean_token_accuracy": 0.20372393876314163,
      "num_tokens": 32752212.0,
      "step": 14290
    },
    {
      "entropy": 5.218288898468018,
      "epoch": 1.3731988472622478,
      "grad_norm": 1.265625,
      "learning_rate": 0.000481786958991188,
      "loss": 5.0,
      "mean_token_accuracy": 0.2096698150038719,
      "num_tokens": 32764581.0,
      "step": 14295
    },
    {
      "entropy": 5.231195068359375,
      "epoch": 1.3736791546589817,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048177344428913316,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21354733407497406,
      "num_tokens": 32775992.0,
      "step": 14300
    },
    {
      "entropy": 5.222694540023804,
      "epoch": 1.3741594620557156,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048175992478641293,
      "loss": 5.0441,
      "mean_token_accuracy": 0.2150271236896515,
      "num_tokens": 32787164.0,
      "step": 14305
    },
    {
      "entropy": 5.119797325134277,
      "epoch": 1.3746397694524495,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004817464004833412,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22109542936086654,
      "num_tokens": 32798190.0,
      "step": 14310
    },
    {
      "entropy": 5.159321594238281,
      "epoch": 1.3751200768491834,
      "grad_norm": 1.75,
      "learning_rate": 0.00048173287138023204,
      "loss": 4.9184,
      "mean_token_accuracy": 0.22451429665088654,
      "num_tokens": 32809559.0,
      "step": 14315
    },
    {
      "entropy": 5.2917036533355715,
      "epoch": 1.3756003842459175,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004817193374773993,
      "loss": 5.0268,
      "mean_token_accuracy": 0.20981132835149766,
      "num_tokens": 32820799.0,
      "step": 14320
    },
    {
      "entropy": 5.190317916870117,
      "epoch": 1.3760806916426513,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048170579877515753,
      "loss": 4.9788,
      "mean_token_accuracy": 0.21351646780967712,
      "num_tokens": 32833971.0,
      "step": 14325
    },
    {
      "entropy": 5.119227409362793,
      "epoch": 1.3765609990393852,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004816922552738209,
      "loss": 4.921,
      "mean_token_accuracy": 0.22124264538288116,
      "num_tokens": 32845237.0,
      "step": 14330
    },
    {
      "entropy": 5.241313552856445,
      "epoch": 1.377041306436119,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048167870697370373,
      "loss": 5.0492,
      "mean_token_accuracy": 0.20772060453891755,
      "num_tokens": 32857960.0,
      "step": 14335
    },
    {
      "entropy": 5.133769941329956,
      "epoch": 1.377521613832853,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004816651538751207,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22245627641677856,
      "num_tokens": 32868801.0,
      "step": 14340
    },
    {
      "entropy": 5.219274282455444,
      "epoch": 1.3780019212295869,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048165159597838664,
      "loss": 4.9914,
      "mean_token_accuracy": 0.21078670471906663,
      "num_tokens": 32880544.0,
      "step": 14345
    },
    {
      "entropy": 5.305929708480835,
      "epoch": 1.378482228626321,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000481638033283816,
      "loss": 5.0466,
      "mean_token_accuracy": 0.20440396517515183,
      "num_tokens": 32892179.0,
      "step": 14350
    },
    {
      "entropy": 5.276677465438842,
      "epoch": 1.3789625360230549,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048162446579172387,
      "loss": 5.0414,
      "mean_token_accuracy": 0.2159889444708824,
      "num_tokens": 32903854.0,
      "step": 14355
    },
    {
      "entropy": 5.229605770111084,
      "epoch": 1.3794428434197887,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004816108935024252,
      "loss": 5.0251,
      "mean_token_accuracy": 0.21733158975839614,
      "num_tokens": 32914731.0,
      "step": 14360
    },
    {
      "entropy": 5.0980428695678714,
      "epoch": 1.3799231508165226,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048159731641623507,
      "loss": 4.9026,
      "mean_token_accuracy": 0.21410641074180603,
      "num_tokens": 32925894.0,
      "step": 14365
    },
    {
      "entropy": 5.113473749160766,
      "epoch": 1.3804034582132565,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004815837345334687,
      "loss": 4.9737,
      "mean_token_accuracy": 0.21785761266946793,
      "num_tokens": 32937463.0,
      "step": 14370
    },
    {
      "entropy": 5.288600778579712,
      "epoch": 1.3808837656099904,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004815701478544415,
      "loss": 5.0475,
      "mean_token_accuracy": 0.20696077048778533,
      "num_tokens": 32949635.0,
      "step": 14375
    },
    {
      "entropy": 5.222473812103272,
      "epoch": 1.3813640730067243,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048155655637946876,
      "loss": 5.0223,
      "mean_token_accuracy": 0.21494010835886002,
      "num_tokens": 32961043.0,
      "step": 14380
    },
    {
      "entropy": 5.221520328521729,
      "epoch": 1.3818443804034581,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004815429601088662,
      "loss": 5.0633,
      "mean_token_accuracy": 0.21154215335845947,
      "num_tokens": 32972218.0,
      "step": 14385
    },
    {
      "entropy": 5.154136419296265,
      "epoch": 1.382324687800192,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004815293590429494,
      "loss": 4.8818,
      "mean_token_accuracy": 0.21617092341184616,
      "num_tokens": 32983157.0,
      "step": 14390
    },
    {
      "entropy": 5.239258289337158,
      "epoch": 1.382804995196926,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048151575318203417,
      "loss": 5.05,
      "mean_token_accuracy": 0.21567121744155884,
      "num_tokens": 32994524.0,
      "step": 14395
    },
    {
      "entropy": 5.250635623931885,
      "epoch": 1.38328530259366,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048150214252643637,
      "loss": 5.0527,
      "mean_token_accuracy": 0.20703865736722946,
      "num_tokens": 33006477.0,
      "step": 14400
    },
    {
      "entropy": 5.266903305053711,
      "epoch": 1.3837656099903939,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000481488527076472,
      "loss": 5.0141,
      "mean_token_accuracy": 0.21121742129325866,
      "num_tokens": 33018067.0,
      "step": 14405
    },
    {
      "entropy": 5.224885272979736,
      "epoch": 1.3842459173871278,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004814749068324571,
      "loss": 5.0459,
      "mean_token_accuracy": 0.21172730028629302,
      "num_tokens": 33031927.0,
      "step": 14410
    },
    {
      "entropy": 5.255479717254639,
      "epoch": 1.3847262247838616,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048146128179470804,
      "loss": 5.0512,
      "mean_token_accuracy": 0.210753932595253,
      "num_tokens": 33042516.0,
      "step": 14415
    },
    {
      "entropy": 5.252200555801392,
      "epoch": 1.3852065321805955,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004814476519635411,
      "loss": 4.9179,
      "mean_token_accuracy": 0.21413985043764114,
      "num_tokens": 33054931.0,
      "step": 14420
    },
    {
      "entropy": 5.2543628215789795,
      "epoch": 1.3856868395773294,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048143401733927274,
      "loss": 5.0992,
      "mean_token_accuracy": 0.201175194978714,
      "num_tokens": 33067333.0,
      "step": 14425
    },
    {
      "entropy": 5.180438232421875,
      "epoch": 1.3861671469740635,
      "grad_norm": 1.25,
      "learning_rate": 0.00048142037792221943,
      "loss": 4.9661,
      "mean_token_accuracy": 0.21888195127248763,
      "num_tokens": 33078469.0,
      "step": 14430
    },
    {
      "entropy": 5.233856964111328,
      "epoch": 1.3866474543707974,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004814067337126978,
      "loss": 4.9784,
      "mean_token_accuracy": 0.21540186554193497,
      "num_tokens": 33089993.0,
      "step": 14435
    },
    {
      "entropy": 5.24426908493042,
      "epoch": 1.3871277617675313,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004813930847110248,
      "loss": 5.0442,
      "mean_token_accuracy": 0.20659275650978087,
      "num_tokens": 33101837.0,
      "step": 14440
    },
    {
      "entropy": 5.2071277618408205,
      "epoch": 1.3876080691642652,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004813794309175173,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21476538181304933,
      "num_tokens": 33113036.0,
      "step": 14445
    },
    {
      "entropy": 5.155508708953858,
      "epoch": 1.388088376560999,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048136577233249205,
      "loss": 4.9506,
      "mean_token_accuracy": 0.21820630580186845,
      "num_tokens": 33124826.0,
      "step": 14450
    },
    {
      "entropy": 5.140398740768433,
      "epoch": 1.388568683957733,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004813521089562666,
      "loss": 4.8181,
      "mean_token_accuracy": 0.2221426010131836,
      "num_tokens": 33135878.0,
      "step": 14455
    },
    {
      "entropy": 5.186589622497559,
      "epoch": 1.3890489913544668,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004813384407891577,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21590172201395036,
      "num_tokens": 33146938.0,
      "step": 14460
    },
    {
      "entropy": 5.193923997879028,
      "epoch": 1.3895292987512007,
      "grad_norm": 1.171875,
      "learning_rate": 0.000481324767831483,
      "loss": 4.9746,
      "mean_token_accuracy": 0.20940403789281845,
      "num_tokens": 33159223.0,
      "step": 14465
    },
    {
      "entropy": 5.210604524612426,
      "epoch": 1.3900096061479346,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004813110900835598,
      "loss": 5.101,
      "mean_token_accuracy": 0.2111714169383049,
      "num_tokens": 33170235.0,
      "step": 14470
    },
    {
      "entropy": 5.195747232437133,
      "epoch": 1.3904899135446687,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004812974075457058,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22175109535455703,
      "num_tokens": 33181193.0,
      "step": 14475
    },
    {
      "entropy": 5.206734895706177,
      "epoch": 1.3909702209414025,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048128372021823845,
      "loss": 4.9629,
      "mean_token_accuracy": 0.2090666264295578,
      "num_tokens": 33192505.0,
      "step": 14480
    },
    {
      "entropy": 5.1848443984985355,
      "epoch": 1.3914505283381364,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048127002810147574,
      "loss": 4.9613,
      "mean_token_accuracy": 0.21469399183988572,
      "num_tokens": 33204113.0,
      "step": 14485
    },
    {
      "entropy": 5.370704317092896,
      "epoch": 1.3919308357348703,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004812563311957355,
      "loss": 5.1736,
      "mean_token_accuracy": 0.20272520035505295,
      "num_tokens": 33216175.0,
      "step": 14490
    },
    {
      "entropy": 5.287360000610351,
      "epoch": 1.3924111431316042,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004812426295013356,
      "loss": 5.0072,
      "mean_token_accuracy": 0.2139357402920723,
      "num_tokens": 33226916.0,
      "step": 14495
    },
    {
      "entropy": 5.2050987720489506,
      "epoch": 1.392891450528338,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048122892301859433,
      "loss": 5.0495,
      "mean_token_accuracy": 0.21303432881832124,
      "num_tokens": 33238020.0,
      "step": 14500
    },
    {
      "entropy": 5.189413595199585,
      "epoch": 1.3933717579250722,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048121521174782983,
      "loss": 4.8965,
      "mean_token_accuracy": 0.21707932651042938,
      "num_tokens": 33248343.0,
      "step": 14505
    },
    {
      "entropy": 5.289135456085205,
      "epoch": 1.393852065321806,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048120149568936044,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21099590361118317,
      "num_tokens": 33259206.0,
      "step": 14510
    },
    {
      "entropy": 5.210521984100342,
      "epoch": 1.39433237271854,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004811877748435046,
      "loss": 5.0576,
      "mean_token_accuracy": 0.20826054960489274,
      "num_tokens": 33271020.0,
      "step": 14515
    },
    {
      "entropy": 5.1733119010925295,
      "epoch": 1.3948126801152738,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004811740492105809,
      "loss": 4.9426,
      "mean_token_accuracy": 0.2198547676205635,
      "num_tokens": 33282603.0,
      "step": 14520
    },
    {
      "entropy": 5.2782834529876705,
      "epoch": 1.3952929875120077,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000481160318790908,
      "loss": 5.032,
      "mean_token_accuracy": 0.21469722986221312,
      "num_tokens": 33292877.0,
      "step": 14525
    },
    {
      "entropy": 5.3352720737457275,
      "epoch": 1.3957732949087416,
      "grad_norm": 1.125,
      "learning_rate": 0.00048114658358480467,
      "loss": 5.2142,
      "mean_token_accuracy": 0.20414863377809525,
      "num_tokens": 33304835.0,
      "step": 14530
    },
    {
      "entropy": 5.285925006866455,
      "epoch": 1.3962536023054755,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048113284359258977,
      "loss": 5.0378,
      "mean_token_accuracy": 0.20788490921258926,
      "num_tokens": 33315028.0,
      "step": 14535
    },
    {
      "entropy": 5.175967645645142,
      "epoch": 1.3967339097022093,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048111909881458234,
      "loss": 4.9719,
      "mean_token_accuracy": 0.21415939629077912,
      "num_tokens": 33325321.0,
      "step": 14540
    },
    {
      "entropy": 5.221309518814087,
      "epoch": 1.3972142170989432,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048110534925110146,
      "loss": 5.0544,
      "mean_token_accuracy": 0.21494554728269577,
      "num_tokens": 33337677.0,
      "step": 14545
    },
    {
      "entropy": 5.166701793670654,
      "epoch": 1.397694524495677,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004810915949024664,
      "loss": 4.9739,
      "mean_token_accuracy": 0.21374745965003966,
      "num_tokens": 33347613.0,
      "step": 14550
    },
    {
      "entropy": 5.180025053024292,
      "epoch": 1.3981748318924112,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004810778357689965,
      "loss": 4.9567,
      "mean_token_accuracy": 0.21305174827575685,
      "num_tokens": 33358912.0,
      "step": 14555
    },
    {
      "entropy": 5.298856735229492,
      "epoch": 1.398655139289145,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048106407185101116,
      "loss": 5.1018,
      "mean_token_accuracy": 0.2127239316701889,
      "num_tokens": 33371635.0,
      "step": 14560
    },
    {
      "entropy": 5.20993971824646,
      "epoch": 1.399135446685879,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048105030314883,
      "loss": 4.8786,
      "mean_token_accuracy": 0.22549621760845184,
      "num_tokens": 33383521.0,
      "step": 14565
    },
    {
      "entropy": 5.125545310974121,
      "epoch": 1.3996157540826129,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004810365296627725,
      "loss": 4.9576,
      "mean_token_accuracy": 0.21008861809968948,
      "num_tokens": 33395297.0,
      "step": 14570
    },
    {
      "entropy": 5.129734754562378,
      "epoch": 1.4000960614793467,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004810227513931587,
      "loss": 4.854,
      "mean_token_accuracy": 0.23019299656152725,
      "num_tokens": 33407168.0,
      "step": 14575
    },
    {
      "entropy": 5.159085702896118,
      "epoch": 1.4005763688760806,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004810089683403084,
      "loss": 4.8892,
      "mean_token_accuracy": 0.22039234042167663,
      "num_tokens": 33418568.0,
      "step": 14580
    },
    {
      "entropy": 5.178706502914428,
      "epoch": 1.4010566762728147,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004809951805045415,
      "loss": 5.0157,
      "mean_token_accuracy": 0.20913531631231308,
      "num_tokens": 33428822.0,
      "step": 14585
    },
    {
      "entropy": 5.166435289382934,
      "epoch": 1.4015369836695486,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048098138788617815,
      "loss": 4.9574,
      "mean_token_accuracy": 0.22104396522045136,
      "num_tokens": 33441547.0,
      "step": 14590
    },
    {
      "entropy": 5.2178719997406,
      "epoch": 1.4020172910662825,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004809675904855387,
      "loss": 4.9922,
      "mean_token_accuracy": 0.21890699118375778,
      "num_tokens": 33452445.0,
      "step": 14595
    },
    {
      "entropy": 5.163801288604736,
      "epoch": 1.4024975984630164,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048095378830294343,
      "loss": 4.9921,
      "mean_token_accuracy": 0.21575697213411332,
      "num_tokens": 33463845.0,
      "step": 14600
    },
    {
      "entropy": 5.2120081901550295,
      "epoch": 1.4029779058597502,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048093998133871276,
      "loss": 4.9587,
      "mean_token_accuracy": 0.2152189791202545,
      "num_tokens": 33474950.0,
      "step": 14605
    },
    {
      "entropy": 5.133649778366089,
      "epoch": 1.4034582132564841,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004809261695931671,
      "loss": 4.9189,
      "mean_token_accuracy": 0.21753356456756592,
      "num_tokens": 33486225.0,
      "step": 14610
    },
    {
      "entropy": 5.217970705032348,
      "epoch": 1.403938520653218,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004809123530666273,
      "loss": 4.9878,
      "mean_token_accuracy": 0.21210669875144958,
      "num_tokens": 33498125.0,
      "step": 14615
    },
    {
      "entropy": 5.173962259292603,
      "epoch": 1.4044188280499519,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004808985317594142,
      "loss": 4.8633,
      "mean_token_accuracy": 0.22013083696365357,
      "num_tokens": 33508731.0,
      "step": 14620
    },
    {
      "entropy": 5.146500015258789,
      "epoch": 1.4048991354466858,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048088470567184854,
      "loss": 4.9003,
      "mean_token_accuracy": 0.22141543924808502,
      "num_tokens": 33520356.0,
      "step": 14625
    },
    {
      "entropy": 5.238033294677734,
      "epoch": 1.4053794428434199,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048087087480425133,
      "loss": 5.0655,
      "mean_token_accuracy": 0.21416936367750167,
      "num_tokens": 33533179.0,
      "step": 14630
    },
    {
      "entropy": 5.114054107666016,
      "epoch": 1.4058597502401537,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004808570391569437,
      "loss": 4.8819,
      "mean_token_accuracy": 0.21909631192684173,
      "num_tokens": 33544700.0,
      "step": 14635
    },
    {
      "entropy": 5.219782400131225,
      "epoch": 1.4063400576368876,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048084319873024694,
      "loss": 5.1301,
      "mean_token_accuracy": 0.20895856469869614,
      "num_tokens": 33558781.0,
      "step": 14640
    },
    {
      "entropy": 5.255144453048706,
      "epoch": 1.4068203650336215,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004808293535244823,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21659952402114868,
      "num_tokens": 33569508.0,
      "step": 14645
    },
    {
      "entropy": 5.192801141738892,
      "epoch": 1.4073006724303554,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004808155035399712,
      "loss": 4.9394,
      "mean_token_accuracy": 0.21387868523597717,
      "num_tokens": 33580977.0,
      "step": 14650
    },
    {
      "entropy": 5.205459547042847,
      "epoch": 1.4077809798270893,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004808016487770354,
      "loss": 5.0285,
      "mean_token_accuracy": 0.21216964572668076,
      "num_tokens": 33593757.0,
      "step": 14655
    },
    {
      "entropy": 5.2906494140625,
      "epoch": 1.4082612872238234,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048078778923599637,
      "loss": 5.1102,
      "mean_token_accuracy": 0.20616656839847564,
      "num_tokens": 33604903.0,
      "step": 14660
    },
    {
      "entropy": 5.20711932182312,
      "epoch": 1.4087415946205573,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048077392491717593,
      "loss": 4.9451,
      "mean_token_accuracy": 0.22091505974531173,
      "num_tokens": 33616976.0,
      "step": 14665
    },
    {
      "entropy": 5.221698760986328,
      "epoch": 1.4092219020172911,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048076005582089597,
      "loss": 4.9746,
      "mean_token_accuracy": 0.21916515231132508,
      "num_tokens": 33627654.0,
      "step": 14670
    },
    {
      "entropy": 5.162622785568237,
      "epoch": 1.409702209414025,
      "grad_norm": 1.5,
      "learning_rate": 0.00048074618194747845,
      "loss": 4.9498,
      "mean_token_accuracy": 0.22189666330814362,
      "num_tokens": 33638377.0,
      "step": 14675
    },
    {
      "entropy": 5.223220920562744,
      "epoch": 1.410182516810759,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004807323032972456,
      "loss": 4.9596,
      "mean_token_accuracy": 0.21743300706148147,
      "num_tokens": 33650959.0,
      "step": 14680
    },
    {
      "entropy": 5.266449928283691,
      "epoch": 1.4106628242074928,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004807184198705196,
      "loss": 5.0671,
      "mean_token_accuracy": 0.20646697878837586,
      "num_tokens": 33663069.0,
      "step": 14685
    },
    {
      "entropy": 5.174374055862427,
      "epoch": 1.4111431316042267,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004807045316676226,
      "loss": 5.018,
      "mean_token_accuracy": 0.21160861551761628,
      "num_tokens": 33674868.0,
      "step": 14690
    },
    {
      "entropy": 5.14315276145935,
      "epoch": 1.4116234390009605,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004806906386888773,
      "loss": 4.9544,
      "mean_token_accuracy": 0.218818736076355,
      "num_tokens": 33685482.0,
      "step": 14695
    },
    {
      "entropy": 5.174482965469361,
      "epoch": 1.4121037463976944,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048067674093460607,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21084309816360475,
      "num_tokens": 33697518.0,
      "step": 14700
    },
    {
      "entropy": 5.191282606124878,
      "epoch": 1.4125840537944283,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048066283840513175,
      "loss": 4.9541,
      "mean_token_accuracy": 0.21277682334184647,
      "num_tokens": 33708827.0,
      "step": 14705
    },
    {
      "entropy": 5.171807956695557,
      "epoch": 1.4130643611911624,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004806489311007769,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2136704981327057,
      "num_tokens": 33719452.0,
      "step": 14710
    },
    {
      "entropy": 5.233827257156372,
      "epoch": 1.4135446685878963,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048063501902186463,
      "loss": 5.0671,
      "mean_token_accuracy": 0.21293904185295104,
      "num_tokens": 33730630.0,
      "step": 14715
    },
    {
      "entropy": 5.155137252807617,
      "epoch": 1.4140249759846302,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048062110216871775,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21682412922382355,
      "num_tokens": 33742696.0,
      "step": 14720
    },
    {
      "entropy": 5.16945595741272,
      "epoch": 1.414505283381364,
      "grad_norm": 1.375,
      "learning_rate": 0.00048060718054165945,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2166140481829643,
      "num_tokens": 33753818.0,
      "step": 14725
    },
    {
      "entropy": 5.310275602340698,
      "epoch": 1.414985590778098,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000480593254141013,
      "loss": 5.1778,
      "mean_token_accuracy": 0.2063077598810196,
      "num_tokens": 33766847.0,
      "step": 14730
    },
    {
      "entropy": 5.258056879043579,
      "epoch": 1.4154658981748318,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048057932296710165,
      "loss": 4.9449,
      "mean_token_accuracy": 0.21450569182634355,
      "num_tokens": 33777070.0,
      "step": 14735
    },
    {
      "entropy": 5.212310886383056,
      "epoch": 1.415946205571566,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004805653870202489,
      "loss": 4.9657,
      "mean_token_accuracy": 0.22121918052434922,
      "num_tokens": 33788826.0,
      "step": 14740
    },
    {
      "entropy": 5.095027780532837,
      "epoch": 1.4164265129682998,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048055144630077825,
      "loss": 4.9416,
      "mean_token_accuracy": 0.21797083765268327,
      "num_tokens": 33801564.0,
      "step": 14745
    },
    {
      "entropy": 5.206371259689331,
      "epoch": 1.4169068203650337,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048053750080901336,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21631144881248474,
      "num_tokens": 33813389.0,
      "step": 14750
    },
    {
      "entropy": 5.234323501586914,
      "epoch": 1.4173871277617676,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048052355054527794,
      "loss": 4.9648,
      "mean_token_accuracy": 0.2106672167778015,
      "num_tokens": 33824585.0,
      "step": 14755
    },
    {
      "entropy": 5.233028173446655,
      "epoch": 1.4178674351585014,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048050959550989606,
      "loss": 5.0485,
      "mean_token_accuracy": 0.20341713428497316,
      "num_tokens": 33837995.0,
      "step": 14760
    },
    {
      "entropy": 5.170246982574463,
      "epoch": 1.4183477425552353,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004804956357031916,
      "loss": 4.952,
      "mean_token_accuracy": 0.22394883036613464,
      "num_tokens": 33848849.0,
      "step": 14765
    },
    {
      "entropy": 5.174532175064087,
      "epoch": 1.4188280499519692,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048048167112548873,
      "loss": 4.9352,
      "mean_token_accuracy": 0.2236589953303337,
      "num_tokens": 33859711.0,
      "step": 14770
    },
    {
      "entropy": 5.210760164260864,
      "epoch": 1.419308357348703,
      "grad_norm": 1.25,
      "learning_rate": 0.00048046770177711157,
      "loss": 5.0318,
      "mean_token_accuracy": 0.20741466879844667,
      "num_tokens": 33872613.0,
      "step": 14775
    },
    {
      "entropy": 5.093054723739624,
      "epoch": 1.419788664745437,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004804537276583844,
      "loss": 4.8736,
      "mean_token_accuracy": 0.21994310468435288,
      "num_tokens": 33883759.0,
      "step": 14780
    },
    {
      "entropy": 5.247519397735596,
      "epoch": 1.420268972142171,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004804397487696319,
      "loss": 5.0215,
      "mean_token_accuracy": 0.21717611104249954,
      "num_tokens": 33894448.0,
      "step": 14785
    },
    {
      "entropy": 5.25812931060791,
      "epoch": 1.420749279538905,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004804257651111783,
      "loss": 5.0585,
      "mean_token_accuracy": 0.21176890134811402,
      "num_tokens": 33906976.0,
      "step": 14790
    },
    {
      "entropy": 5.151072931289673,
      "epoch": 1.4212295869356388,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048041177668334853,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21503363102674483,
      "num_tokens": 33919384.0,
      "step": 14795
    },
    {
      "entropy": 5.152737236022949,
      "epoch": 1.4217098943323727,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004803977834864672,
      "loss": 4.8865,
      "mean_token_accuracy": 0.21984277069568633,
      "num_tokens": 33931080.0,
      "step": 14800
    },
    {
      "entropy": 5.151620006561279,
      "epoch": 1.4221902017291066,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048038378552085927,
      "loss": 4.958,
      "mean_token_accuracy": 0.2133355513215065,
      "num_tokens": 33943021.0,
      "step": 14805
    },
    {
      "entropy": 5.224500226974487,
      "epoch": 1.4226705091258405,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048036978278684974,
      "loss": 4.9762,
      "mean_token_accuracy": 0.21871453821659087,
      "num_tokens": 33953609.0,
      "step": 14810
    },
    {
      "entropy": 5.210731410980225,
      "epoch": 1.4231508165225746,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004803557752847636,
      "loss": 5.0236,
      "mean_token_accuracy": 0.2084116667509079,
      "num_tokens": 33966985.0,
      "step": 14815
    },
    {
      "entropy": 5.187292766571045,
      "epoch": 1.4236311239193085,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048034176301492616,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21701618880033494,
      "num_tokens": 33976536.0,
      "step": 14820
    },
    {
      "entropy": 5.179544830322266,
      "epoch": 1.4241114313160423,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004803277459776628,
      "loss": 4.9671,
      "mean_token_accuracy": 0.2153777301311493,
      "num_tokens": 33988000.0,
      "step": 14825
    },
    {
      "entropy": 5.211529874801636,
      "epoch": 1.4245917387127762,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048031372417329875,
      "loss": 5.0756,
      "mean_token_accuracy": 0.21460383832454683,
      "num_tokens": 33998944.0,
      "step": 14830
    },
    {
      "entropy": 5.256339645385742,
      "epoch": 1.42507204610951,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004802996976021598,
      "loss": 5.0569,
      "mean_token_accuracy": 0.21368759125471115,
      "num_tokens": 34010440.0,
      "step": 14835
    },
    {
      "entropy": 5.239995527267456,
      "epoch": 1.425552353506244,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048028566626457145,
      "loss": 5.0308,
      "mean_token_accuracy": 0.21028411239385605,
      "num_tokens": 34023222.0,
      "step": 14840
    },
    {
      "entropy": 5.171328163146972,
      "epoch": 1.4260326609029779,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048027163016085947,
      "loss": 4.8804,
      "mean_token_accuracy": 0.22000223100185395,
      "num_tokens": 34035413.0,
      "step": 14845
    },
    {
      "entropy": 5.18064250946045,
      "epoch": 1.4265129682997117,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048025758929134976,
      "loss": 4.9932,
      "mean_token_accuracy": 0.21662225872278212,
      "num_tokens": 34046700.0,
      "step": 14850
    },
    {
      "entropy": 5.195637941360474,
      "epoch": 1.4269932756964456,
      "grad_norm": 1.25,
      "learning_rate": 0.0004802435436563684,
      "loss": 5.0087,
      "mean_token_accuracy": 0.21236393302679063,
      "num_tokens": 34058039.0,
      "step": 14855
    },
    {
      "entropy": 5.168593406677246,
      "epoch": 1.4274735830931795,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048022949325624134,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20887863785028457,
      "num_tokens": 34069602.0,
      "step": 14860
    },
    {
      "entropy": 5.26692214012146,
      "epoch": 1.4279538904899136,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048021543809129483,
      "loss": 5.0255,
      "mean_token_accuracy": 0.2055478110909462,
      "num_tokens": 34080514.0,
      "step": 14865
    },
    {
      "entropy": 5.19987211227417,
      "epoch": 1.4284341978866475,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004802013781618552,
      "loss": 4.9405,
      "mean_token_accuracy": 0.21859416365623474,
      "num_tokens": 34091416.0,
      "step": 14870
    },
    {
      "entropy": 5.2315025806427,
      "epoch": 1.4289145052833814,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048018731346824895,
      "loss": 4.9785,
      "mean_token_accuracy": 0.2110746070742607,
      "num_tokens": 34102153.0,
      "step": 14875
    },
    {
      "entropy": 5.224228048324585,
      "epoch": 1.4293948126801153,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004801732440108026,
      "loss": 5.0334,
      "mean_token_accuracy": 0.21679691076278687,
      "num_tokens": 34111917.0,
      "step": 14880
    },
    {
      "entropy": 5.220215892791748,
      "epoch": 1.4298751200768491,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004801591697898427,
      "loss": 5.0226,
      "mean_token_accuracy": 0.20932268351316452,
      "num_tokens": 34124038.0,
      "step": 14885
    },
    {
      "entropy": 5.271049737930298,
      "epoch": 1.430355427473583,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004801450908056961,
      "loss": 5.0676,
      "mean_token_accuracy": 0.20845692455768586,
      "num_tokens": 34134503.0,
      "step": 14890
    },
    {
      "entropy": 5.1690514087677,
      "epoch": 1.4308357348703171,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004801310070586896,
      "loss": 4.893,
      "mean_token_accuracy": 0.21585622280836106,
      "num_tokens": 34146145.0,
      "step": 14895
    },
    {
      "entropy": 5.167447185516357,
      "epoch": 1.431316042267051,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004801169185491503,
      "loss": 4.9607,
      "mean_token_accuracy": 0.2130047559738159,
      "num_tokens": 34156672.0,
      "step": 14900
    },
    {
      "entropy": 5.25537748336792,
      "epoch": 1.4317963496637849,
      "grad_norm": 2.703125,
      "learning_rate": 0.00048010282527740516,
      "loss": 5.1295,
      "mean_token_accuracy": 0.20089703649282456,
      "num_tokens": 34168817.0,
      "step": 14905
    },
    {
      "entropy": 5.225982522964477,
      "epoch": 1.4322766570605188,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048008872724378146,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2209317907691002,
      "num_tokens": 34179804.0,
      "step": 14910
    },
    {
      "entropy": 5.301078510284424,
      "epoch": 1.4327569644572526,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004800746244486065,
      "loss": 5.1958,
      "mean_token_accuracy": 0.19929109364748002,
      "num_tokens": 34192293.0,
      "step": 14915
    },
    {
      "entropy": 5.257237911224365,
      "epoch": 1.4332372718539865,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004800605168922077,
      "loss": 5.0272,
      "mean_token_accuracy": 0.21058556884527208,
      "num_tokens": 34203331.0,
      "step": 14920
    },
    {
      "entropy": 5.2653662204742435,
      "epoch": 1.4337175792507204,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048004640457491267,
      "loss": 4.9423,
      "mean_token_accuracy": 0.21585774272680283,
      "num_tokens": 34213554.0,
      "step": 14925
    },
    {
      "entropy": 5.151469802856445,
      "epoch": 1.4341978866474543,
      "grad_norm": 1.40625,
      "learning_rate": 0.000480032287497049,
      "loss": 4.9338,
      "mean_token_accuracy": 0.21965181231498718,
      "num_tokens": 34225824.0,
      "step": 14930
    },
    {
      "entropy": 5.161107969284058,
      "epoch": 1.4346781940441882,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048001816565894427,
      "loss": 5.0031,
      "mean_token_accuracy": 0.21323498338460922,
      "num_tokens": 34237479.0,
      "step": 14935
    },
    {
      "entropy": 5.216752386093139,
      "epoch": 1.435158501440922,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004800040390609267,
      "loss": 5.0489,
      "mean_token_accuracy": 0.20657793134450914,
      "num_tokens": 34248706.0,
      "step": 14940
    },
    {
      "entropy": 5.342103147506714,
      "epoch": 1.4356388088376562,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047998990770332396,
      "loss": 5.073,
      "mean_token_accuracy": 0.21256547123193742,
      "num_tokens": 34258957.0,
      "step": 14945
    },
    {
      "entropy": 5.179416179656982,
      "epoch": 1.43611911623439,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004799757715864643,
      "loss": 5.0057,
      "mean_token_accuracy": 0.21079510301351548,
      "num_tokens": 34270635.0,
      "step": 14950
    },
    {
      "entropy": 5.269561624526977,
      "epoch": 1.436599423631124,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004799616307106759,
      "loss": 4.9842,
      "mean_token_accuracy": 0.2172810450196266,
      "num_tokens": 34282491.0,
      "step": 14955
    },
    {
      "entropy": 5.214649724960327,
      "epoch": 1.4370797310278578,
      "grad_norm": 1.203125,
      "learning_rate": 0.000479947485076287,
      "loss": 4.908,
      "mean_token_accuracy": 0.22197681069374084,
      "num_tokens": 34294666.0,
      "step": 14960
    },
    {
      "entropy": 5.15767240524292,
      "epoch": 1.4375600384245917,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047993333468362607,
      "loss": 4.9089,
      "mean_token_accuracy": 0.2162582531571388,
      "num_tokens": 34305065.0,
      "step": 14965
    },
    {
      "entropy": 5.252692651748657,
      "epoch": 1.4380403458213258,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047991917953302173,
      "loss": 5.0332,
      "mean_token_accuracy": 0.2052166134119034,
      "num_tokens": 34315955.0,
      "step": 14970
    },
    {
      "entropy": 5.205110740661621,
      "epoch": 1.4385206532180597,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00047990501962480236,
      "loss": 4.9128,
      "mean_token_accuracy": 0.2206428363919258,
      "num_tokens": 34327011.0,
      "step": 14975
    },
    {
      "entropy": 5.1544841766357425,
      "epoch": 1.4390009606147935,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000479890854959297,
      "loss": 4.9262,
      "mean_token_accuracy": 0.2173188716173172,
      "num_tokens": 34338011.0,
      "step": 14980
    },
    {
      "entropy": 5.142746543884277,
      "epoch": 1.4394812680115274,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004798766855368344,
      "loss": 4.8649,
      "mean_token_accuracy": 0.22042015045881272,
      "num_tokens": 34348957.0,
      "step": 14985
    },
    {
      "entropy": 5.17303581237793,
      "epoch": 1.4399615754082613,
      "grad_norm": 1.25,
      "learning_rate": 0.00047986251135774343,
      "loss": 4.9051,
      "mean_token_accuracy": 0.2164036050438881,
      "num_tokens": 34359442.0,
      "step": 14990
    },
    {
      "entropy": 5.161843967437744,
      "epoch": 1.4404418828049952,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004798483324223533,
      "loss": 4.982,
      "mean_token_accuracy": 0.21637497395277022,
      "num_tokens": 34371129.0,
      "step": 14995
    },
    {
      "entropy": 5.167888069152832,
      "epoch": 1.440922190201729,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004798341487309932,
      "loss": 4.9884,
      "mean_token_accuracy": 0.21421175301074982,
      "num_tokens": 34381565.0,
      "step": 15000
    },
    {
      "epoch": 1.440922190201729,
      "eval_entropy": 5.079185588019235,
      "eval_loss": 5.0769805908203125,
      "eval_mean_token_accuracy": 0.21766199918054244,
      "eval_num_tokens": 34381565.0,
      "eval_runtime": 27.3767,
      "eval_samples_per_second": 1198.646,
      "eval_steps_per_second": 149.835,
      "step": 15000
    },
    {
      "entropy": 5.237632179260254,
      "epoch": 1.441402497598463,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047981996028399233,
      "loss": 4.9117,
      "mean_token_accuracy": 0.2205919474363327,
      "num_tokens": 34392145.0,
      "step": 15005
    },
    {
      "entropy": 5.2222918510437015,
      "epoch": 1.4418828049951968,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004798057670816802,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20997018367052078,
      "num_tokens": 34403193.0,
      "step": 15010
    },
    {
      "entropy": 5.181120300292969,
      "epoch": 1.4423631123919307,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004797915691243863,
      "loss": 4.9047,
      "mean_token_accuracy": 0.21174042522907258,
      "num_tokens": 34414289.0,
      "step": 15015
    },
    {
      "entropy": 5.253153848648071,
      "epoch": 1.4428434197886648,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004797773664124403,
      "loss": 5.0667,
      "mean_token_accuracy": 0.20874678939580918,
      "num_tokens": 34424439.0,
      "step": 15020
    },
    {
      "entropy": 5.16706862449646,
      "epoch": 1.4433237271853987,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047976315894617195,
      "loss": 4.8951,
      "mean_token_accuracy": 0.21594517230987548,
      "num_tokens": 34437299.0,
      "step": 15025
    },
    {
      "entropy": 5.151429271697998,
      "epoch": 1.4438040345821326,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000479748946725911,
      "loss": 4.8347,
      "mean_token_accuracy": 0.22693199664354324,
      "num_tokens": 34448397.0,
      "step": 15030
    },
    {
      "entropy": 5.121964788436889,
      "epoch": 1.4442843419788665,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004797347297519875,
      "loss": 4.901,
      "mean_token_accuracy": 0.2159625694155693,
      "num_tokens": 34459389.0,
      "step": 15035
    },
    {
      "entropy": 5.243630743026733,
      "epoch": 1.4447646493756003,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047972050802473154,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21555271446704866,
      "num_tokens": 34470610.0,
      "step": 15040
    },
    {
      "entropy": 5.150574588775635,
      "epoch": 1.4452449567723342,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004797062815444733,
      "loss": 4.9342,
      "mean_token_accuracy": 0.21940028220415114,
      "num_tokens": 34482100.0,
      "step": 15045
    },
    {
      "entropy": 5.135382747650146,
      "epoch": 1.4457252641690683,
      "grad_norm": 1.28125,
      "learning_rate": 0.000479692050311543,
      "loss": 4.9531,
      "mean_token_accuracy": 0.20916046500205993,
      "num_tokens": 34493086.0,
      "step": 15050
    },
    {
      "entropy": 5.120890712738037,
      "epoch": 1.4462055715658022,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004796778143262711,
      "loss": 4.8759,
      "mean_token_accuracy": 0.22374433130025864,
      "num_tokens": 34505030.0,
      "step": 15055
    },
    {
      "entropy": 5.263736581802368,
      "epoch": 1.446685878962536,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004796635735889882,
      "loss": 5.0537,
      "mean_token_accuracy": 0.20703519731760026,
      "num_tokens": 34516790.0,
      "step": 15060
    },
    {
      "entropy": 5.298580360412598,
      "epoch": 1.44716618635927,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047964932810002476,
      "loss": 5.067,
      "mean_token_accuracy": 0.20643970370292664,
      "num_tokens": 34530333.0,
      "step": 15065
    },
    {
      "entropy": 5.206116342544556,
      "epoch": 1.4476464937560038,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004796350778597117,
      "loss": 4.9657,
      "mean_token_accuracy": 0.21884484589099884,
      "num_tokens": 34540945.0,
      "step": 15070
    },
    {
      "entropy": 5.252432060241699,
      "epoch": 1.4481268011527377,
      "grad_norm": 1.25,
      "learning_rate": 0.0004796208228683796,
      "loss": 4.966,
      "mean_token_accuracy": 0.21029854118824004,
      "num_tokens": 34551449.0,
      "step": 15075
    },
    {
      "entropy": 5.223735713958741,
      "epoch": 1.4486071085494716,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047960656312635977,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20698425471782683,
      "num_tokens": 34562841.0,
      "step": 15080
    },
    {
      "entropy": 5.2001410007476805,
      "epoch": 1.4490874159462055,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004795922986339831,
      "loss": 4.9874,
      "mean_token_accuracy": 0.21529216319322586,
      "num_tokens": 34574408.0,
      "step": 15085
    },
    {
      "entropy": 5.261038303375244,
      "epoch": 1.4495677233429394,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047957802939158057,
      "loss": 5.0193,
      "mean_token_accuracy": 0.21423302143812178,
      "num_tokens": 34586243.0,
      "step": 15090
    },
    {
      "entropy": 5.256597185134888,
      "epoch": 1.4500480307396733,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004795637553994838,
      "loss": 4.9711,
      "mean_token_accuracy": 0.21437820941209793,
      "num_tokens": 34596631.0,
      "step": 15095
    },
    {
      "entropy": 5.150667285919189,
      "epoch": 1.4505283381364074,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047954947665802404,
      "loss": 4.9377,
      "mean_token_accuracy": 0.22520441710948944,
      "num_tokens": 34607952.0,
      "step": 15100
    },
    {
      "entropy": 5.111911106109619,
      "epoch": 1.4510086455331412,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004795351931675329,
      "loss": 4.8929,
      "mean_token_accuracy": 0.2204710900783539,
      "num_tokens": 34618355.0,
      "step": 15105
    },
    {
      "entropy": 5.194798278808594,
      "epoch": 1.4514889529298751,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004795209049283419,
      "loss": 4.9519,
      "mean_token_accuracy": 0.21247511953115464,
      "num_tokens": 34629402.0,
      "step": 15110
    },
    {
      "entropy": 5.298788166046142,
      "epoch": 1.451969260326609,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004795066119407827,
      "loss": 5.084,
      "mean_token_accuracy": 0.2148731216788292,
      "num_tokens": 34642523.0,
      "step": 15115
    },
    {
      "entropy": 5.100717067718506,
      "epoch": 1.4524495677233429,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004794923142051873,
      "loss": 4.8796,
      "mean_token_accuracy": 0.22591916471719742,
      "num_tokens": 34653714.0,
      "step": 15120
    },
    {
      "entropy": 5.132459259033203,
      "epoch": 1.452929875120077,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047947801172188755,
      "loss": 4.9478,
      "mean_token_accuracy": 0.21783642768859862,
      "num_tokens": 34664899.0,
      "step": 15125
    },
    {
      "entropy": 5.263347339630127,
      "epoch": 1.4534101825168109,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004794637044912155,
      "loss": 4.9883,
      "mean_token_accuracy": 0.2141053393483162,
      "num_tokens": 34674888.0,
      "step": 15130
    },
    {
      "entropy": 5.113697671890259,
      "epoch": 1.4538904899135447,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004794493925135034,
      "loss": 4.8862,
      "mean_token_accuracy": 0.22564695328474044,
      "num_tokens": 34686446.0,
      "step": 15135
    },
    {
      "entropy": 5.210177564620972,
      "epoch": 1.4543707973102786,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047943507578908357,
      "loss": 5.0846,
      "mean_token_accuracy": 0.21561945527791976,
      "num_tokens": 34697592.0,
      "step": 15140
    },
    {
      "entropy": 5.22652645111084,
      "epoch": 1.4548511047070125,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004794207543182883,
      "loss": 4.9795,
      "mean_token_accuracy": 0.21546683311462403,
      "num_tokens": 34709305.0,
      "step": 15145
    },
    {
      "entropy": 5.250665235519409,
      "epoch": 1.4553314121037464,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047940642810145005,
      "loss": 5.0341,
      "mean_token_accuracy": 0.21591262966394426,
      "num_tokens": 34721606.0,
      "step": 15150
    },
    {
      "entropy": 5.211900520324707,
      "epoch": 1.4558117195004803,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047939209713890156,
      "loss": 4.931,
      "mean_token_accuracy": 0.216531440615654,
      "num_tokens": 34732125.0,
      "step": 15155
    },
    {
      "entropy": 5.114988374710083,
      "epoch": 1.4562920268972142,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047937776143097547,
      "loss": 4.9471,
      "mean_token_accuracy": 0.21610984802246094,
      "num_tokens": 34743317.0,
      "step": 15160
    },
    {
      "entropy": 5.136571788787842,
      "epoch": 1.456772334293948,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004793634209780047,
      "loss": 4.9453,
      "mean_token_accuracy": 0.212065427005291,
      "num_tokens": 34754546.0,
      "step": 15165
    },
    {
      "entropy": 5.204717922210693,
      "epoch": 1.457252641690682,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004793490757803221,
      "loss": 4.8764,
      "mean_token_accuracy": 0.22004846781492232,
      "num_tokens": 34766027.0,
      "step": 15170
    },
    {
      "entropy": 5.189595699310303,
      "epoch": 1.457732949087416,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047933472583826063,
      "loss": 4.9453,
      "mean_token_accuracy": 0.21505171954631805,
      "num_tokens": 34777750.0,
      "step": 15175
    },
    {
      "entropy": 5.151628828048706,
      "epoch": 1.45821325648415,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004793203711521537,
      "loss": 5.0242,
      "mean_token_accuracy": 0.2145438924431801,
      "num_tokens": 34788924.0,
      "step": 15180
    },
    {
      "entropy": 5.218436574935913,
      "epoch": 1.4586935638808838,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047930601172233446,
      "loss": 4.9373,
      "mean_token_accuracy": 0.22205283790826796,
      "num_tokens": 34800153.0,
      "step": 15185
    },
    {
      "entropy": 5.157160520553589,
      "epoch": 1.4591738712776177,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047929164754913624,
      "loss": 4.9794,
      "mean_token_accuracy": 0.21759857088327408,
      "num_tokens": 34811597.0,
      "step": 15190
    },
    {
      "entropy": 5.184308481216431,
      "epoch": 1.4596541786743515,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004792772786328926,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22128719389438628,
      "num_tokens": 34822661.0,
      "step": 15195
    },
    {
      "entropy": 5.252554178237915,
      "epoch": 1.4601344860710854,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047926290497393714,
      "loss": 5.0408,
      "mean_token_accuracy": 0.21350204646587373,
      "num_tokens": 34833554.0,
      "step": 15200
    },
    {
      "entropy": 5.161551856994629,
      "epoch": 1.4606147934678195,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004792485265726036,
      "loss": 5.0106,
      "mean_token_accuracy": 0.21851711422204972,
      "num_tokens": 34846060.0,
      "step": 15205
    },
    {
      "entropy": 5.164449882507324,
      "epoch": 1.4610951008645534,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004792341434292257,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21443849354982375,
      "num_tokens": 34857324.0,
      "step": 15210
    },
    {
      "entropy": 5.187251091003418,
      "epoch": 1.4615754082612873,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004792197555441374,
      "loss": 4.9217,
      "mean_token_accuracy": 0.22372396737337114,
      "num_tokens": 34869317.0,
      "step": 15215
    },
    {
      "entropy": 5.16807336807251,
      "epoch": 1.4620557156580212,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004792053629176729,
      "loss": 4.9445,
      "mean_token_accuracy": 0.2145458087325096,
      "num_tokens": 34880961.0,
      "step": 15220
    },
    {
      "entropy": 5.186451578140259,
      "epoch": 1.462536023054755,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004791909655501662,
      "loss": 4.9522,
      "mean_token_accuracy": 0.21926966458559036,
      "num_tokens": 34893518.0,
      "step": 15225
    },
    {
      "entropy": 5.204141616821289,
      "epoch": 1.463016330451489,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004791765634419516,
      "loss": 5.0324,
      "mean_token_accuracy": 0.2133212849497795,
      "num_tokens": 34906201.0,
      "step": 15230
    },
    {
      "entropy": 5.249502277374267,
      "epoch": 1.4634966378482228,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047916215659336343,
      "loss": 4.9502,
      "mean_token_accuracy": 0.21960481107234955,
      "num_tokens": 34916841.0,
      "step": 15235
    },
    {
      "entropy": 5.246438694000244,
      "epoch": 1.4639769452449567,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004791477450047363,
      "loss": 5.0307,
      "mean_token_accuracy": 0.20862113684415817,
      "num_tokens": 34928081.0,
      "step": 15240
    },
    {
      "entropy": 5.1865509986877445,
      "epoch": 1.4644572526416906,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047913332867640464,
      "loss": 4.9486,
      "mean_token_accuracy": 0.2162141665816307,
      "num_tokens": 34939072.0,
      "step": 15245
    },
    {
      "entropy": 5.192448329925537,
      "epoch": 1.4649375600384245,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004791189076087033,
      "loss": 4.9628,
      "mean_token_accuracy": 0.20894254893064498,
      "num_tokens": 34950898.0,
      "step": 15250
    },
    {
      "entropy": 5.299640941619873,
      "epoch": 1.4654178674351586,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047910448180196703,
      "loss": 5.0125,
      "mean_token_accuracy": 0.21086450964212416,
      "num_tokens": 34961560.0,
      "step": 15255
    },
    {
      "entropy": 5.184868288040161,
      "epoch": 1.4658981748318924,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004790900512565307,
      "loss": 5.0054,
      "mean_token_accuracy": 0.22377318143844604,
      "num_tokens": 34973604.0,
      "step": 15260
    },
    {
      "entropy": 5.17895131111145,
      "epoch": 1.4663784822286263,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004790756159727294,
      "loss": 5.0458,
      "mean_token_accuracy": 0.2112097218632698,
      "num_tokens": 34985369.0,
      "step": 15265
    },
    {
      "entropy": 5.336804389953613,
      "epoch": 1.4668587896253602,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047906117595089835,
      "loss": 5.0599,
      "mean_token_accuracy": 0.2033305808901787,
      "num_tokens": 34996746.0,
      "step": 15270
    },
    {
      "entropy": 5.171100234985351,
      "epoch": 1.467339097022094,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004790467311913727,
      "loss": 4.9887,
      "mean_token_accuracy": 0.21788059920072556,
      "num_tokens": 35008032.0,
      "step": 15275
    },
    {
      "entropy": 5.169945621490479,
      "epoch": 1.4678194044188282,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004790322816944879,
      "loss": 5.0474,
      "mean_token_accuracy": 0.21190387457609178,
      "num_tokens": 35019423.0,
      "step": 15280
    },
    {
      "entropy": 5.1857232570648195,
      "epoch": 1.468299711815562,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004790178274605793,
      "loss": 4.8755,
      "mean_token_accuracy": 0.22400282621383666,
      "num_tokens": 35029985.0,
      "step": 15285
    },
    {
      "entropy": 5.249120235443115,
      "epoch": 1.468780019212296,
      "grad_norm": 1.375,
      "learning_rate": 0.00047900336848998254,
      "loss": 5.008,
      "mean_token_accuracy": 0.21788189113140105,
      "num_tokens": 35040580.0,
      "step": 15290
    },
    {
      "entropy": 5.206815195083618,
      "epoch": 1.4692603266090298,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004789889047830334,
      "loss": 5.029,
      "mean_token_accuracy": 0.21605148911476135,
      "num_tokens": 35051414.0,
      "step": 15295
    },
    {
      "entropy": 5.183011484146118,
      "epoch": 1.4697406340057637,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047897443634006766,
      "loss": 4.9595,
      "mean_token_accuracy": 0.21441274732351304,
      "num_tokens": 35062483.0,
      "step": 15300
    },
    {
      "entropy": 5.193024778366089,
      "epoch": 1.4702209414024976,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004789599631614211,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2195652812719345,
      "num_tokens": 35073536.0,
      "step": 15305
    },
    {
      "entropy": 5.276510906219483,
      "epoch": 1.4707012487992315,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004789454852474298,
      "loss": 4.9766,
      "mean_token_accuracy": 0.2108702078461647,
      "num_tokens": 35085636.0,
      "step": 15310
    },
    {
      "entropy": 5.1820910453796385,
      "epoch": 1.4711815561959654,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004789310025984299,
      "loss": 4.9309,
      "mean_token_accuracy": 0.21217521131038666,
      "num_tokens": 35096660.0,
      "step": 15315
    },
    {
      "entropy": 5.180193614959717,
      "epoch": 1.4716618635926992,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047891651521475776,
      "loss": 5.0724,
      "mean_token_accuracy": 0.21113442331552507,
      "num_tokens": 35109657.0,
      "step": 15320
    },
    {
      "entropy": 5.229466962814331,
      "epoch": 1.4721421709894331,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047890202309674963,
      "loss": 4.9124,
      "mean_token_accuracy": 0.2109218120574951,
      "num_tokens": 35120399.0,
      "step": 15325
    },
    {
      "entropy": 5.143092107772827,
      "epoch": 1.4726224783861672,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047888752624474195,
      "loss": 4.8732,
      "mean_token_accuracy": 0.2195092886686325,
      "num_tokens": 35130105.0,
      "step": 15330
    },
    {
      "entropy": 5.125137853622436,
      "epoch": 1.473102785782901,
      "grad_norm": 1.125,
      "learning_rate": 0.0004788730246590714,
      "loss": 4.934,
      "mean_token_accuracy": 0.2206098258495331,
      "num_tokens": 35142547.0,
      "step": 15335
    },
    {
      "entropy": 5.169401216506958,
      "epoch": 1.473583093179635,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047885851834007456,
      "loss": 4.8558,
      "mean_token_accuracy": 0.225422403216362,
      "num_tokens": 35154603.0,
      "step": 15340
    },
    {
      "entropy": 5.158287620544433,
      "epoch": 1.4740634005763689,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047884400728808824,
      "loss": 4.9115,
      "mean_token_accuracy": 0.22038723528385162,
      "num_tokens": 35165727.0,
      "step": 15345
    },
    {
      "entropy": 5.133256816864014,
      "epoch": 1.4745437079731027,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004788294915034494,
      "loss": 5.0003,
      "mean_token_accuracy": 0.2177749752998352,
      "num_tokens": 35177642.0,
      "step": 15350
    },
    {
      "entropy": 5.176407718658448,
      "epoch": 1.4750240153698366,
      "grad_norm": 1.21875,
      "learning_rate": 0.000478814970986495,
      "loss": 4.9669,
      "mean_token_accuracy": 0.224184550344944,
      "num_tokens": 35189377.0,
      "step": 15355
    },
    {
      "entropy": 5.1742919921875,
      "epoch": 1.4755043227665707,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047880044573756213,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2125691995024681,
      "num_tokens": 35201244.0,
      "step": 15360
    },
    {
      "entropy": 5.235166549682617,
      "epoch": 1.4759846301633046,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047878591575698816,
      "loss": 4.9863,
      "mean_token_accuracy": 0.22294313162565232,
      "num_tokens": 35212491.0,
      "step": 15365
    },
    {
      "entropy": 5.156862735748291,
      "epoch": 1.4764649375600385,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004787713810451103,
      "loss": 4.9297,
      "mean_token_accuracy": 0.2169858604669571,
      "num_tokens": 35225552.0,
      "step": 15370
    },
    {
      "entropy": 5.16960825920105,
      "epoch": 1.4769452449567724,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047875684160226606,
      "loss": 4.9296,
      "mean_token_accuracy": 0.21951996237039567,
      "num_tokens": 35237961.0,
      "step": 15375
    },
    {
      "entropy": 5.155251836776733,
      "epoch": 1.4774255523535063,
      "grad_norm": 1.359375,
      "learning_rate": 0.000478742297428793,
      "loss": 4.9849,
      "mean_token_accuracy": 0.21524409651756288,
      "num_tokens": 35249502.0,
      "step": 15380
    },
    {
      "entropy": 5.21789026260376,
      "epoch": 1.4779058597502401,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047872774852502877,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21286043226718904,
      "num_tokens": 35262978.0,
      "step": 15385
    },
    {
      "entropy": 5.233468246459961,
      "epoch": 1.478386167146974,
      "grad_norm": 1.828125,
      "learning_rate": 0.0004787131948913112,
      "loss": 5.0735,
      "mean_token_accuracy": 0.21489138901233673,
      "num_tokens": 35273289.0,
      "step": 15390
    },
    {
      "entropy": 5.209843397140503,
      "epoch": 1.478866474543708,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047869863652797806,
      "loss": 4.984,
      "mean_token_accuracy": 0.21049595773220062,
      "num_tokens": 35283827.0,
      "step": 15395
    },
    {
      "entropy": 5.271842098236084,
      "epoch": 1.4793467819404418,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004786840734353675,
      "loss": 5.0043,
      "mean_token_accuracy": 0.216803178191185,
      "num_tokens": 35293264.0,
      "step": 15400
    },
    {
      "entropy": 5.146999740600586,
      "epoch": 1.4798270893371757,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047866950561381756,
      "loss": 4.9496,
      "mean_token_accuracy": 0.22027858942747117,
      "num_tokens": 35304380.0,
      "step": 15405
    },
    {
      "entropy": 5.188980627059936,
      "epoch": 1.4803073967339098,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004786549330636665,
      "loss": 4.92,
      "mean_token_accuracy": 0.22153494507074356,
      "num_tokens": 35316805.0,
      "step": 15410
    },
    {
      "entropy": 5.165693569183349,
      "epoch": 1.4807877041306436,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047864035578525256,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21513320207595826,
      "num_tokens": 35328645.0,
      "step": 15415
    },
    {
      "entropy": 5.174641895294189,
      "epoch": 1.4812680115273775,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004786257737789143,
      "loss": 4.9764,
      "mean_token_accuracy": 0.21259803771972657,
      "num_tokens": 35339983.0,
      "step": 15420
    },
    {
      "entropy": 5.295954084396362,
      "epoch": 1.4817483189241114,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004786111870449902,
      "loss": 5.0722,
      "mean_token_accuracy": 0.2064305305480957,
      "num_tokens": 35350720.0,
      "step": 15425
    },
    {
      "entropy": 5.266617107391357,
      "epoch": 1.4822286263208453,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047859659558381894,
      "loss": 5.0296,
      "mean_token_accuracy": 0.21185945421457292,
      "num_tokens": 35361435.0,
      "step": 15430
    },
    {
      "entropy": 5.174959993362426,
      "epoch": 1.4827089337175792,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047858199939573935,
      "loss": 5.0713,
      "mean_token_accuracy": 0.20945972055196763,
      "num_tokens": 35372829.0,
      "step": 15435
    },
    {
      "entropy": 5.18730525970459,
      "epoch": 1.4831892411143133,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047856739848109014,
      "loss": 4.9724,
      "mean_token_accuracy": 0.21238687336444856,
      "num_tokens": 35383270.0,
      "step": 15440
    },
    {
      "entropy": 5.346497917175293,
      "epoch": 1.4836695485110472,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047855279284021046,
      "loss": 5.071,
      "mean_token_accuracy": 0.2070194289088249,
      "num_tokens": 35396308.0,
      "step": 15445
    },
    {
      "entropy": 5.224224328994751,
      "epoch": 1.484149855907781,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047853818247343933,
      "loss": 4.9966,
      "mean_token_accuracy": 0.21232093274593353,
      "num_tokens": 35408944.0,
      "step": 15450
    },
    {
      "entropy": 5.202389430999756,
      "epoch": 1.484630163304515,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047852356738111606,
      "loss": 5.0554,
      "mean_token_accuracy": 0.2056686744093895,
      "num_tokens": 35422227.0,
      "step": 15455
    },
    {
      "entropy": 5.234705924987793,
      "epoch": 1.4851104707012488,
      "grad_norm": 1.25,
      "learning_rate": 0.0004785089475635799,
      "loss": 5.0273,
      "mean_token_accuracy": 0.20966014117002488,
      "num_tokens": 35434110.0,
      "step": 15460
    },
    {
      "entropy": 5.240862035751343,
      "epoch": 1.4855907780979827,
      "grad_norm": 1.25,
      "learning_rate": 0.00047849432302117024,
      "loss": 4.9835,
      "mean_token_accuracy": 0.22062526196241378,
      "num_tokens": 35445194.0,
      "step": 15465
    },
    {
      "entropy": 5.278164005279541,
      "epoch": 1.4860710854947166,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047847969375422656,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21505406945943834,
      "num_tokens": 35455953.0,
      "step": 15470
    },
    {
      "entropy": 5.182776069641113,
      "epoch": 1.4865513928914504,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004784650597630887,
      "loss": 4.8851,
      "mean_token_accuracy": 0.22381227165460588,
      "num_tokens": 35466932.0,
      "step": 15475
    },
    {
      "entropy": 5.133643245697021,
      "epoch": 1.4870317002881843,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047845042104809635,
      "loss": 4.9393,
      "mean_token_accuracy": 0.2171243280172348,
      "num_tokens": 35477755.0,
      "step": 15480
    },
    {
      "entropy": 5.123105812072754,
      "epoch": 1.4875120076849184,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004784357776095892,
      "loss": 4.914,
      "mean_token_accuracy": 0.22364708334207534,
      "num_tokens": 35489913.0,
      "step": 15485
    },
    {
      "entropy": 5.157333278656006,
      "epoch": 1.4879923150816523,
      "grad_norm": 1.5,
      "learning_rate": 0.0004784211294479075,
      "loss": 5.0102,
      "mean_token_accuracy": 0.22068246603012084,
      "num_tokens": 35500314.0,
      "step": 15490
    },
    {
      "entropy": 5.289217948913574,
      "epoch": 1.4884726224783862,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004784064765633912,
      "loss": 5.0818,
      "mean_token_accuracy": 0.20678599178791046,
      "num_tokens": 35514717.0,
      "step": 15495
    },
    {
      "entropy": 5.235702085494995,
      "epoch": 1.48895292987512,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047839181895638057,
      "loss": 4.967,
      "mean_token_accuracy": 0.2128964975476265,
      "num_tokens": 35526122.0,
      "step": 15500
    },
    {
      "entropy": 5.215676593780517,
      "epoch": 1.489433237271854,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047837715662721575,
      "loss": 4.9909,
      "mean_token_accuracy": 0.21167586296796798,
      "num_tokens": 35538270.0,
      "step": 15505
    },
    {
      "entropy": 5.274493265151977,
      "epoch": 1.4899135446685878,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004783624895762372,
      "loss": 5.056,
      "mean_token_accuracy": 0.20970784723758698,
      "num_tokens": 35549598.0,
      "step": 15510
    },
    {
      "entropy": 5.240608072280883,
      "epoch": 1.490393852065322,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047834781780378563,
      "loss": 4.9753,
      "mean_token_accuracy": 0.21392974108457566,
      "num_tokens": 35559623.0,
      "step": 15515
    },
    {
      "entropy": 5.177938508987427,
      "epoch": 1.4908741594620558,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004783331413102015,
      "loss": 4.9521,
      "mean_token_accuracy": 0.2154352053999901,
      "num_tokens": 35571578.0,
      "step": 15520
    },
    {
      "entropy": 5.208767890930176,
      "epoch": 1.4913544668587897,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047831846009582557,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20873095095157623,
      "num_tokens": 35583570.0,
      "step": 15525
    },
    {
      "entropy": 5.253358221054077,
      "epoch": 1.4918347742555236,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004783037741609988,
      "loss": 5.0107,
      "mean_token_accuracy": 0.21408694982528687,
      "num_tokens": 35595360.0,
      "step": 15530
    },
    {
      "entropy": 5.1934398174285885,
      "epoch": 1.4923150816522575,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004782890835060621,
      "loss": 4.9684,
      "mean_token_accuracy": 0.2158981055021286,
      "num_tokens": 35607326.0,
      "step": 15535
    },
    {
      "entropy": 5.255344343185425,
      "epoch": 1.4927953890489913,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004782743881313564,
      "loss": 5.01,
      "mean_token_accuracy": 0.20556861609220506,
      "num_tokens": 35618504.0,
      "step": 15540
    },
    {
      "entropy": 5.2002874374389645,
      "epoch": 1.4932756964457252,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047825968803722315,
      "loss": 4.953,
      "mean_token_accuracy": 0.21954639106988907,
      "num_tokens": 35631111.0,
      "step": 15545
    },
    {
      "entropy": 5.213309907913208,
      "epoch": 1.493756003842459,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004782449832240035,
      "loss": 4.976,
      "mean_token_accuracy": 0.22199355214834213,
      "num_tokens": 35643434.0,
      "step": 15550
    },
    {
      "entropy": 5.2878014087677006,
      "epoch": 1.494236311239193,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004782302736920387,
      "loss": 5.0424,
      "mean_token_accuracy": 0.20822387486696242,
      "num_tokens": 35654531.0,
      "step": 15555
    },
    {
      "entropy": 5.200317287445069,
      "epoch": 1.4947166186359269,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004782155594416705,
      "loss": 4.9558,
      "mean_token_accuracy": 0.21570693254470824,
      "num_tokens": 35665488.0,
      "step": 15560
    },
    {
      "entropy": 5.266824865341187,
      "epoch": 1.495196926032661,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047820084047324045,
      "loss": 5.0211,
      "mean_token_accuracy": 0.2090889558196068,
      "num_tokens": 35677110.0,
      "step": 15565
    },
    {
      "entropy": 5.337477970123291,
      "epoch": 1.4956772334293948,
      "grad_norm": 1.65625,
      "learning_rate": 0.00047818611678709027,
      "loss": 5.0498,
      "mean_token_accuracy": 0.21185137182474137,
      "num_tokens": 35688694.0,
      "step": 15570
    },
    {
      "entropy": 5.142055225372315,
      "epoch": 1.4961575408261287,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004781713883835618,
      "loss": 4.9594,
      "mean_token_accuracy": 0.21244597434997559,
      "num_tokens": 35699497.0,
      "step": 15575
    },
    {
      "entropy": 5.240458869934082,
      "epoch": 1.4966378482228626,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047815665526299695,
      "loss": 5.0106,
      "mean_token_accuracy": 0.21614381223917006,
      "num_tokens": 35711034.0,
      "step": 15580
    },
    {
      "entropy": 5.221970939636231,
      "epoch": 1.4971181556195965,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004781419174257378,
      "loss": 4.9092,
      "mean_token_accuracy": 0.2229369357228279,
      "num_tokens": 35721804.0,
      "step": 15585
    },
    {
      "entropy": 5.180604791641235,
      "epoch": 1.4975984630163304,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004781271748721266,
      "loss": 4.9198,
      "mean_token_accuracy": 0.2167037159204483,
      "num_tokens": 35732897.0,
      "step": 15590
    },
    {
      "entropy": 5.049993848800659,
      "epoch": 1.4980787704130645,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004781124276025055,
      "loss": 4.8342,
      "mean_token_accuracy": 0.23009953051805496,
      "num_tokens": 35743792.0,
      "step": 15595
    },
    {
      "entropy": 5.238765239715576,
      "epoch": 1.4985590778097984,
      "grad_norm": 1.390625,
      "learning_rate": 0.000478097675617217,
      "loss": 5.0003,
      "mean_token_accuracy": 0.209477436542511,
      "num_tokens": 35754659.0,
      "step": 15600
    },
    {
      "entropy": 5.282638645172119,
      "epoch": 1.4990393852065322,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047808291891660357,
      "loss": 4.9736,
      "mean_token_accuracy": 0.22076985687017442,
      "num_tokens": 35766269.0,
      "step": 15605
    },
    {
      "entropy": 5.245017004013062,
      "epoch": 1.4995196926032661,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047806815750100774,
      "loss": 5.011,
      "mean_token_accuracy": 0.21388485729694368,
      "num_tokens": 35777634.0,
      "step": 15610
    },
    {
      "entropy": 5.18912181854248,
      "epoch": 1.5,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004780533913707723,
      "loss": 5.0194,
      "mean_token_accuracy": 0.2132517158985138,
      "num_tokens": 35788689.0,
      "step": 15615
    },
    {
      "entropy": 5.167920112609863,
      "epoch": 1.5004803073967339,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047803862052624006,
      "loss": 4.9264,
      "mean_token_accuracy": 0.21909511387348174,
      "num_tokens": 35799365.0,
      "step": 15620
    },
    {
      "entropy": 5.188027477264404,
      "epoch": 1.5009606147934678,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047802384496775397,
      "loss": 4.888,
      "mean_token_accuracy": 0.22602555751800538,
      "num_tokens": 35810682.0,
      "step": 15625
    },
    {
      "entropy": 5.207361793518066,
      "epoch": 1.5014409221902016,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004780090646956571,
      "loss": 5.0004,
      "mean_token_accuracy": 0.20565223842859268,
      "num_tokens": 35822193.0,
      "step": 15630
    },
    {
      "entropy": 5.152439975738526,
      "epoch": 1.5019212295869355,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047799427971029245,
      "loss": 5.0138,
      "mean_token_accuracy": 0.2142855703830719,
      "num_tokens": 35835486.0,
      "step": 15635
    },
    {
      "entropy": 5.228617715835571,
      "epoch": 1.5024015369836694,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004779794900120034,
      "loss": 4.9904,
      "mean_token_accuracy": 0.2074977844953537,
      "num_tokens": 35846245.0,
      "step": 15640
    },
    {
      "entropy": 5.266604280471801,
      "epoch": 1.5028818443804035,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004779646956011334,
      "loss": 5.1052,
      "mean_token_accuracy": 0.21231913268566133,
      "num_tokens": 35857914.0,
      "step": 15645
    },
    {
      "entropy": 5.2517180919647215,
      "epoch": 1.5033621517771374,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047794989647802574,
      "loss": 5.0139,
      "mean_token_accuracy": 0.2074258178472519,
      "num_tokens": 35870235.0,
      "step": 15650
    },
    {
      "entropy": 5.258316087722778,
      "epoch": 1.5038424591738713,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047793509264302424,
      "loss": 5.0404,
      "mean_token_accuracy": 0.21442606300115585,
      "num_tokens": 35882154.0,
      "step": 15655
    },
    {
      "entropy": 5.222346687316895,
      "epoch": 1.5043227665706052,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047792028409647237,
      "loss": 4.9717,
      "mean_token_accuracy": 0.22582461386919023,
      "num_tokens": 35894192.0,
      "step": 15660
    },
    {
      "entropy": 5.215750789642334,
      "epoch": 1.5048030739673393,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047790547083871414,
      "loss": 4.9851,
      "mean_token_accuracy": 0.2158886209130287,
      "num_tokens": 35906017.0,
      "step": 15665
    },
    {
      "entropy": 5.252561759948731,
      "epoch": 1.5052833813640731,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047789065287009335,
      "loss": 5.0498,
      "mean_token_accuracy": 0.20963285118341446,
      "num_tokens": 35917671.0,
      "step": 15670
    },
    {
      "entropy": 5.124835777282715,
      "epoch": 1.505763688760807,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004778758301909542,
      "loss": 4.9088,
      "mean_token_accuracy": 0.22070486843585968,
      "num_tokens": 35928421.0,
      "step": 15675
    },
    {
      "entropy": 5.244018316268921,
      "epoch": 1.506243996157541,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004778610028016405,
      "loss": 5.0506,
      "mean_token_accuracy": 0.20935219824314116,
      "num_tokens": 35939689.0,
      "step": 15680
    },
    {
      "entropy": 5.264218950271607,
      "epoch": 1.5067243035542748,
      "grad_norm": 1.375,
      "learning_rate": 0.0004778461707024967,
      "loss": 5.0693,
      "mean_token_accuracy": 0.21339828819036483,
      "num_tokens": 35952342.0,
      "step": 15685
    },
    {
      "entropy": 5.355706691741943,
      "epoch": 1.5072046109510087,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004778313338938672,
      "loss": 5.0988,
      "mean_token_accuracy": 0.20041017532348632,
      "num_tokens": 35964726.0,
      "step": 15690
    },
    {
      "entropy": 5.233206462860108,
      "epoch": 1.5076849183477425,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047781649237609643,
      "loss": 4.9598,
      "mean_token_accuracy": 0.22016313225030898,
      "num_tokens": 35976932.0,
      "step": 15695
    },
    {
      "entropy": 5.126315975189209,
      "epoch": 1.5081652257444764,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004778016461495289,
      "loss": 4.8962,
      "mean_token_accuracy": 0.22394221723079683,
      "num_tokens": 35988371.0,
      "step": 15700
    },
    {
      "entropy": 5.100627088546753,
      "epoch": 1.5086455331412103,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004777867952145094,
      "loss": 4.9068,
      "mean_token_accuracy": 0.22676490843296052,
      "num_tokens": 36001310.0,
      "step": 15705
    },
    {
      "entropy": 5.308777284622193,
      "epoch": 1.5091258405379442,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004777719395713826,
      "loss": 5.0953,
      "mean_token_accuracy": 0.2028970018029213,
      "num_tokens": 36011975.0,
      "step": 15710
    },
    {
      "entropy": 5.259601831436157,
      "epoch": 1.509606147934678,
      "grad_norm": 1.375,
      "learning_rate": 0.00047775707922049354,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21105975210666655,
      "num_tokens": 36024987.0,
      "step": 15715
    },
    {
      "entropy": 5.2028594493865965,
      "epoch": 1.510086455331412,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004777422141621871,
      "loss": 4.9765,
      "mean_token_accuracy": 0.22186583429574966,
      "num_tokens": 36036003.0,
      "step": 15720
    },
    {
      "entropy": 5.247424936294555,
      "epoch": 1.510566762728146,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004777273443968085,
      "loss": 5.0309,
      "mean_token_accuracy": 0.20934086441993713,
      "num_tokens": 36047948.0,
      "step": 15725
    },
    {
      "entropy": 5.20718903541565,
      "epoch": 1.51104707012488,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004777124699247029,
      "loss": 4.9221,
      "mean_token_accuracy": 0.22265944480895997,
      "num_tokens": 36059197.0,
      "step": 15730
    },
    {
      "entropy": 5.31180329322815,
      "epoch": 1.5115273775216138,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004776975907462157,
      "loss": 5.0689,
      "mean_token_accuracy": 0.20789443105459213,
      "num_tokens": 36072167.0,
      "step": 15735
    },
    {
      "entropy": 5.291225957870483,
      "epoch": 1.5120076849183477,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004776827068616924,
      "loss": 5.1217,
      "mean_token_accuracy": 0.20822081267833709,
      "num_tokens": 36083805.0,
      "step": 15740
    },
    {
      "entropy": 5.175408792495728,
      "epoch": 1.5124879923150818,
      "grad_norm": 1.5,
      "learning_rate": 0.0004776678182714785,
      "loss": 4.941,
      "mean_token_accuracy": 0.2204638674855232,
      "num_tokens": 36095450.0,
      "step": 15745
    },
    {
      "entropy": 5.176610803604126,
      "epoch": 1.5129682997118157,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047765292497591955,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22696231454610824,
      "num_tokens": 36106319.0,
      "step": 15750
    },
    {
      "entropy": 5.1658721446990965,
      "epoch": 1.5134486071085496,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047763802697536146,
      "loss": 4.7904,
      "mean_token_accuracy": 0.23046525716781616,
      "num_tokens": 36117015.0,
      "step": 15755
    },
    {
      "entropy": 5.161894273757935,
      "epoch": 1.5139289145052834,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047762312427015015,
      "loss": 4.9808,
      "mean_token_accuracy": 0.2110310509800911,
      "num_tokens": 36127308.0,
      "step": 15760
    },
    {
      "entropy": 5.144322776794434,
      "epoch": 1.5144092219020173,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047760821686063153,
      "loss": 4.8763,
      "mean_token_accuracy": 0.22216299325227737,
      "num_tokens": 36139811.0,
      "step": 15765
    },
    {
      "entropy": 5.236377143859864,
      "epoch": 1.5148895292987512,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047759330474715173,
      "loss": 5.0275,
      "mean_token_accuracy": 0.2083413526415825,
      "num_tokens": 36151330.0,
      "step": 15770
    },
    {
      "entropy": 5.182797956466675,
      "epoch": 1.515369836695485,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047757838793005704,
      "loss": 4.9417,
      "mean_token_accuracy": 0.21746316701173782,
      "num_tokens": 36162436.0,
      "step": 15775
    },
    {
      "entropy": 5.313698863983154,
      "epoch": 1.515850144092219,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047756346640969366,
      "loss": 5.1723,
      "mean_token_accuracy": 0.19481663256883622,
      "num_tokens": 36175642.0,
      "step": 15780
    },
    {
      "entropy": 5.235282707214355,
      "epoch": 1.5163304514889528,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047754854018640803,
      "loss": 4.9177,
      "mean_token_accuracy": 0.21664471924304962,
      "num_tokens": 36186925.0,
      "step": 15785
    },
    {
      "entropy": 5.1779684066772464,
      "epoch": 1.5168107588856867,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047753360926054684,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21144023090600966,
      "num_tokens": 36199280.0,
      "step": 15790
    },
    {
      "entropy": 5.1393238544464115,
      "epoch": 1.5172910662824206,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047751867363245653,
      "loss": 4.9068,
      "mean_token_accuracy": 0.2260114625096321,
      "num_tokens": 36210276.0,
      "step": 15795
    },
    {
      "entropy": 5.283250999450684,
      "epoch": 1.5177713736791547,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004775037333024841,
      "loss": 5.0122,
      "mean_token_accuracy": 0.21384208500385285,
      "num_tokens": 36221211.0,
      "step": 15800
    },
    {
      "entropy": 5.162764644622802,
      "epoch": 1.5182516810758886,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004774887882709762,
      "loss": 4.8561,
      "mean_token_accuracy": 0.22608301639556885,
      "num_tokens": 36233465.0,
      "step": 15805
    },
    {
      "entropy": 5.177501535415649,
      "epoch": 1.5187319884726225,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047747383853827995,
      "loss": 4.9298,
      "mean_token_accuracy": 0.2170847699046135,
      "num_tokens": 36245594.0,
      "step": 15810
    },
    {
      "entropy": 5.235691499710083,
      "epoch": 1.5192122958693564,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004774588841047424,
      "loss": 5.0809,
      "mean_token_accuracy": 0.20554232150316237,
      "num_tokens": 36257776.0,
      "step": 15815
    },
    {
      "entropy": 5.190318441390991,
      "epoch": 1.5196926032660905,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004774439249707108,
      "loss": 4.9304,
      "mean_token_accuracy": 0.20821536779403688,
      "num_tokens": 36269056.0,
      "step": 15820
    },
    {
      "entropy": 5.196181344985962,
      "epoch": 1.5201729106628243,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004774289611365323,
      "loss": 4.9483,
      "mean_token_accuracy": 0.2159997284412384,
      "num_tokens": 36280700.0,
      "step": 15825
    },
    {
      "entropy": 5.243834018707275,
      "epoch": 1.5206532180595582,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047741399260255434,
      "loss": 5.0214,
      "mean_token_accuracy": 0.21227672100067138,
      "num_tokens": 36291708.0,
      "step": 15830
    },
    {
      "entropy": 5.223207235336304,
      "epoch": 1.521133525456292,
      "grad_norm": 1.25,
      "learning_rate": 0.00047739901936912467,
      "loss": 4.9339,
      "mean_token_accuracy": 0.21783231049776078,
      "num_tokens": 36303830.0,
      "step": 15835
    },
    {
      "entropy": 5.170607089996338,
      "epoch": 1.521613832853026,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004773840414365907,
      "loss": 4.9624,
      "mean_token_accuracy": 0.21529533118009567,
      "num_tokens": 36316514.0,
      "step": 15840
    },
    {
      "entropy": 5.298376560211182,
      "epoch": 1.5220941402497599,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047736905880530026,
      "loss": 5.0121,
      "mean_token_accuracy": 0.21417759954929352,
      "num_tokens": 36328234.0,
      "step": 15845
    },
    {
      "entropy": 5.174363374710083,
      "epoch": 1.5225744476464937,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004773540714756012,
      "loss": 4.9525,
      "mean_token_accuracy": 0.21775961816310882,
      "num_tokens": 36340383.0,
      "step": 15850
    },
    {
      "entropy": 5.136829090118408,
      "epoch": 1.5230547550432276,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047733907944784144,
      "loss": 4.9517,
      "mean_token_accuracy": 0.21839896738529205,
      "num_tokens": 36351970.0,
      "step": 15855
    },
    {
      "entropy": 5.181724166870117,
      "epoch": 1.5235350624399615,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004773240827223691,
      "loss": 4.9679,
      "mean_token_accuracy": 0.2163769781589508,
      "num_tokens": 36363163.0,
      "step": 15860
    },
    {
      "entropy": 5.288920259475708,
      "epoch": 1.5240153698366954,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004773090812995323,
      "loss": 5.0218,
      "mean_token_accuracy": 0.2156849056482315,
      "num_tokens": 36375079.0,
      "step": 15865
    },
    {
      "entropy": 5.24628210067749,
      "epoch": 1.5244956772334293,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00047729407517967945,
      "loss": 4.9263,
      "mean_token_accuracy": 0.21710618436336518,
      "num_tokens": 36386402.0,
      "step": 15870
    },
    {
      "entropy": 5.113162517547607,
      "epoch": 1.5249759846301632,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047727906436315884,
      "loss": 4.8813,
      "mean_token_accuracy": 0.22310798317193986,
      "num_tokens": 36397973.0,
      "step": 15875
    },
    {
      "entropy": 5.2043804168701175,
      "epoch": 1.5254562920268973,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047726404885031895,
      "loss": 5.0598,
      "mean_token_accuracy": 0.2078785181045532,
      "num_tokens": 36409941.0,
      "step": 15880
    },
    {
      "entropy": 5.286999845504761,
      "epoch": 1.5259365994236311,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047724902864150845,
      "loss": 5.0387,
      "mean_token_accuracy": 0.20836753994226456,
      "num_tokens": 36420609.0,
      "step": 15885
    },
    {
      "entropy": 5.186961364746094,
      "epoch": 1.526416906820365,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047723400373707607,
      "loss": 4.9386,
      "mean_token_accuracy": 0.22394360303878785,
      "num_tokens": 36430854.0,
      "step": 15890
    },
    {
      "entropy": 5.054891920089721,
      "epoch": 1.526897214217099,
      "grad_norm": 2.484375,
      "learning_rate": 0.0004772189741373707,
      "loss": 4.8786,
      "mean_token_accuracy": 0.21826124787330628,
      "num_tokens": 36442225.0,
      "step": 15895
    },
    {
      "entropy": 5.344627428054809,
      "epoch": 1.527377521613833,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047720393984274117,
      "loss": 5.1013,
      "mean_token_accuracy": 0.2073631688952446,
      "num_tokens": 36454488.0,
      "step": 15900
    },
    {
      "entropy": 5.247308588027954,
      "epoch": 1.5278578290105669,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047718890085353654,
      "loss": 4.8978,
      "mean_token_accuracy": 0.2186833694577217,
      "num_tokens": 36466197.0,
      "step": 15905
    },
    {
      "entropy": 5.222841358184814,
      "epoch": 1.5283381364073008,
      "grad_norm": 1.125,
      "learning_rate": 0.000477173857170106,
      "loss": 5.0504,
      "mean_token_accuracy": 0.20534132719039916,
      "num_tokens": 36478159.0,
      "step": 15910
    },
    {
      "entropy": 5.218592119216919,
      "epoch": 1.5288184438040346,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047715880879279894,
      "loss": 5.0904,
      "mean_token_accuracy": 0.21452660113573074,
      "num_tokens": 36489621.0,
      "step": 15915
    },
    {
      "entropy": 5.218483352661133,
      "epoch": 1.5292987512007685,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004771437557219646,
      "loss": 5.0482,
      "mean_token_accuracy": 0.21390313357114793,
      "num_tokens": 36500629.0,
      "step": 15920
    },
    {
      "entropy": 5.303373098373413,
      "epoch": 1.5297790585975024,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004771286979579524,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21668098866939545,
      "num_tokens": 36512110.0,
      "step": 15925
    },
    {
      "entropy": 5.115925073623657,
      "epoch": 1.5302593659942363,
      "grad_norm": 1.125,
      "learning_rate": 0.0004771136355011121,
      "loss": 4.8493,
      "mean_token_accuracy": 0.22711279541254042,
      "num_tokens": 36522499.0,
      "step": 15930
    },
    {
      "entropy": 5.179617643356323,
      "epoch": 1.5307396733909702,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047709856835179333,
      "loss": 5.0067,
      "mean_token_accuracy": 0.2063772425055504,
      "num_tokens": 36533465.0,
      "step": 15935
    },
    {
      "entropy": 5.178196716308594,
      "epoch": 1.531219980787704,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047708349651034586,
      "loss": 4.951,
      "mean_token_accuracy": 0.21656466126441956,
      "num_tokens": 36545213.0,
      "step": 15940
    },
    {
      "entropy": 5.096554756164551,
      "epoch": 1.531700288184438,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047706841997711974,
      "loss": 4.8413,
      "mean_token_accuracy": 0.2273925006389618,
      "num_tokens": 36555656.0,
      "step": 15945
    },
    {
      "entropy": 5.1885477066040036,
      "epoch": 1.5321805955811718,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047705333875246495,
      "loss": 4.967,
      "mean_token_accuracy": 0.21013035923242568,
      "num_tokens": 36568027.0,
      "step": 15950
    },
    {
      "entropy": 5.163483285903931,
      "epoch": 1.532660902977906,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047703825283673153,
      "loss": 4.905,
      "mean_token_accuracy": 0.2153702959418297,
      "num_tokens": 36579434.0,
      "step": 15955
    },
    {
      "entropy": 5.22842230796814,
      "epoch": 1.5331412103746398,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004770231622302699,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21248690783977509,
      "num_tokens": 36591068.0,
      "step": 15960
    },
    {
      "entropy": 5.197845554351806,
      "epoch": 1.5336215177713737,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047700806693343016,
      "loss": 4.9295,
      "mean_token_accuracy": 0.2170954242348671,
      "num_tokens": 36603013.0,
      "step": 15965
    },
    {
      "entropy": 5.244511938095092,
      "epoch": 1.5341018251681076,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047699296694656316,
      "loss": 5.059,
      "mean_token_accuracy": 0.202449931204319,
      "num_tokens": 36614565.0,
      "step": 15970
    },
    {
      "entropy": 5.207989835739136,
      "epoch": 1.5345821325648417,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004769778622700192,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21845265179872514,
      "num_tokens": 36625685.0,
      "step": 15975
    },
    {
      "entropy": 5.097838544845581,
      "epoch": 1.5350624399615755,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047696275290414885,
      "loss": 4.885,
      "mean_token_accuracy": 0.22171718031167983,
      "num_tokens": 36637278.0,
      "step": 15980
    },
    {
      "entropy": 5.153434371948242,
      "epoch": 1.5355427473583094,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047694763884930324,
      "loss": 4.9486,
      "mean_token_accuracy": 0.21606606394052505,
      "num_tokens": 36649527.0,
      "step": 15985
    },
    {
      "entropy": 5.227485990524292,
      "epoch": 1.5360230547550433,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047693252010583314,
      "loss": 4.9454,
      "mean_token_accuracy": 0.21462788730859755,
      "num_tokens": 36660653.0,
      "step": 15990
    },
    {
      "entropy": 5.19086217880249,
      "epoch": 1.5365033621517772,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004769173966740895,
      "loss": 4.951,
      "mean_token_accuracy": 0.21711477041244506,
      "num_tokens": 36672340.0,
      "step": 15995
    },
    {
      "entropy": 5.269045448303222,
      "epoch": 1.536983669548511,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047690226855442346,
      "loss": 5.0653,
      "mean_token_accuracy": 0.20662891119718552,
      "num_tokens": 36684468.0,
      "step": 16000
    },
    {
      "entropy": 5.222882461547852,
      "epoch": 1.537463976945245,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004768871357471863,
      "loss": 4.956,
      "mean_token_accuracy": 0.21230894178152085,
      "num_tokens": 36695921.0,
      "step": 16005
    },
    {
      "entropy": 5.076692390441894,
      "epoch": 1.5379442843419788,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047687199825272936,
      "loss": 4.8792,
      "mean_token_accuracy": 0.21670906096696854,
      "num_tokens": 36707029.0,
      "step": 16010
    },
    {
      "entropy": 5.075561904907227,
      "epoch": 1.5384245917387127,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047685685607140403,
      "loss": 4.8566,
      "mean_token_accuracy": 0.22038549035787583,
      "num_tokens": 36717899.0,
      "step": 16015
    },
    {
      "entropy": 5.194740104675293,
      "epoch": 1.5389048991354466,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047684170920356185,
      "loss": 4.8912,
      "mean_token_accuracy": 0.21530256867408754,
      "num_tokens": 36728691.0,
      "step": 16020
    },
    {
      "entropy": 5.246204948425293,
      "epoch": 1.5393852065321805,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004768265576495546,
      "loss": 5.053,
      "mean_token_accuracy": 0.21174314320087434,
      "num_tokens": 36741122.0,
      "step": 16025
    },
    {
      "entropy": 5.276718425750732,
      "epoch": 1.5398655139289144,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047681140140973396,
      "loss": 5.092,
      "mean_token_accuracy": 0.21147883534431458,
      "num_tokens": 36752841.0,
      "step": 16030
    },
    {
      "entropy": 5.253206825256347,
      "epoch": 1.5403458213256485,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004767962404844517,
      "loss": 4.9777,
      "mean_token_accuracy": 0.22381875663995743,
      "num_tokens": 36763030.0,
      "step": 16035
    },
    {
      "entropy": 5.274652576446533,
      "epoch": 1.5408261287223823,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047678107487406015,
      "loss": 5.0712,
      "mean_token_accuracy": 0.20782736986875533,
      "num_tokens": 36774442.0,
      "step": 16040
    },
    {
      "entropy": 5.212074089050293,
      "epoch": 1.5413064361191162,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047676590457891116,
      "loss": 4.9843,
      "mean_token_accuracy": 0.21144439578056334,
      "num_tokens": 36785650.0,
      "step": 16045
    },
    {
      "entropy": 5.178154468536377,
      "epoch": 1.54178674351585,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004767507295993569,
      "loss": 4.9103,
      "mean_token_accuracy": 0.21515233665704728,
      "num_tokens": 36797021.0,
      "step": 16050
    },
    {
      "entropy": 5.196694326400757,
      "epoch": 1.5422670509125842,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004767355499357498,
      "loss": 4.9492,
      "mean_token_accuracy": 0.21749195009469985,
      "num_tokens": 36807593.0,
      "step": 16055
    },
    {
      "entropy": 5.158777284622192,
      "epoch": 1.542747358309318,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004767203655884423,
      "loss": 4.931,
      "mean_token_accuracy": 0.21572162061929703,
      "num_tokens": 36820185.0,
      "step": 16060
    },
    {
      "entropy": 5.138628339767456,
      "epoch": 1.543227665706052,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004767051765577869,
      "loss": 4.8476,
      "mean_token_accuracy": 0.22345304489135742,
      "num_tokens": 36831350.0,
      "step": 16065
    },
    {
      "entropy": 5.1023742198944095,
      "epoch": 1.5437079731027858,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047668998284413624,
      "loss": 4.8357,
      "mean_token_accuracy": 0.22024931758642197,
      "num_tokens": 36842475.0,
      "step": 16070
    },
    {
      "entropy": 5.163780879974365,
      "epoch": 1.5441882804995197,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047667478444784306,
      "loss": 4.928,
      "mean_token_accuracy": 0.22085826694965363,
      "num_tokens": 36853385.0,
      "step": 16075
    },
    {
      "entropy": 5.15146312713623,
      "epoch": 1.5446685878962536,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004766595813692602,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2146839678287506,
      "num_tokens": 36866339.0,
      "step": 16080
    },
    {
      "entropy": 5.2222068309783936,
      "epoch": 1.5451488952929875,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047664437360874076,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21685247421264647,
      "num_tokens": 36877943.0,
      "step": 16085
    },
    {
      "entropy": 5.23546667098999,
      "epoch": 1.5456292026897214,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047662916116663766,
      "loss": 4.977,
      "mean_token_accuracy": 0.21543146073818206,
      "num_tokens": 36889343.0,
      "step": 16090
    },
    {
      "entropy": 5.2024623394012455,
      "epoch": 1.5461095100864553,
      "grad_norm": 1.25,
      "learning_rate": 0.00047661394404330417,
      "loss": 4.9803,
      "mean_token_accuracy": 0.21335113495588304,
      "num_tokens": 36900727.0,
      "step": 16095
    },
    {
      "entropy": 5.205817413330078,
      "epoch": 1.5465898174831891,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047659872223909357,
      "loss": 4.938,
      "mean_token_accuracy": 0.2207164227962494,
      "num_tokens": 36911959.0,
      "step": 16100
    },
    {
      "entropy": 5.165293169021607,
      "epoch": 1.547070124879923,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004765834957543592,
      "loss": 4.8618,
      "mean_token_accuracy": 0.22186000049114227,
      "num_tokens": 36924964.0,
      "step": 16105
    },
    {
      "entropy": 5.1673260688781735,
      "epoch": 1.547550432276657,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047656826458945475,
      "loss": 5.0645,
      "mean_token_accuracy": 0.20457442849874496,
      "num_tokens": 36937033.0,
      "step": 16110
    },
    {
      "entropy": 5.173828268051148,
      "epoch": 1.548030739673391,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047655302874473365,
      "loss": 4.9136,
      "mean_token_accuracy": 0.2201266497373581,
      "num_tokens": 36948745.0,
      "step": 16115
    },
    {
      "entropy": 5.21050181388855,
      "epoch": 1.5485110470701249,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004765377882205498,
      "loss": 4.8794,
      "mean_token_accuracy": 0.22163072675466539,
      "num_tokens": 36961210.0,
      "step": 16120
    },
    {
      "entropy": 5.217321825027466,
      "epoch": 1.5489913544668588,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004765225430172568,
      "loss": 5.016,
      "mean_token_accuracy": 0.21420252919197083,
      "num_tokens": 36973720.0,
      "step": 16125
    },
    {
      "entropy": 5.261472129821778,
      "epoch": 1.5494716618635929,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004765072931352089,
      "loss": 5.0607,
      "mean_token_accuracy": 0.21540989279747008,
      "num_tokens": 36986899.0,
      "step": 16130
    },
    {
      "entropy": 5.187014579772949,
      "epoch": 1.5499519692603267,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047649203857476,
      "loss": 4.9474,
      "mean_token_accuracy": 0.21313762068748474,
      "num_tokens": 36998324.0,
      "step": 16135
    },
    {
      "entropy": 5.238290023803711,
      "epoch": 1.5504322766570606,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047647677933626423,
      "loss": 5.0022,
      "mean_token_accuracy": 0.21799693256616592,
      "num_tokens": 37009779.0,
      "step": 16140
    },
    {
      "entropy": 5.23692307472229,
      "epoch": 1.5509125840537945,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047646151542007583,
      "loss": 5.0318,
      "mean_token_accuracy": 0.2163985028862953,
      "num_tokens": 37022473.0,
      "step": 16145
    },
    {
      "entropy": 5.281168365478516,
      "epoch": 1.5513928914505284,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004764462468265494,
      "loss": 5.0774,
      "mean_token_accuracy": 0.21095783710479737,
      "num_tokens": 37033456.0,
      "step": 16150
    },
    {
      "entropy": 5.213506841659546,
      "epoch": 1.5518731988472623,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047643097355603913,
      "loss": 5.0016,
      "mean_token_accuracy": 0.21093750298023223,
      "num_tokens": 37044983.0,
      "step": 16155
    },
    {
      "entropy": 5.2459046840667725,
      "epoch": 1.5523535062439962,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004764156956088999,
      "loss": 4.9257,
      "mean_token_accuracy": 0.2162885919213295,
      "num_tokens": 37055992.0,
      "step": 16160
    },
    {
      "entropy": 5.213085269927978,
      "epoch": 1.55283381364073,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004764004129854863,
      "loss": 4.9644,
      "mean_token_accuracy": 0.22350261062383653,
      "num_tokens": 37068227.0,
      "step": 16165
    },
    {
      "entropy": 5.133727169036865,
      "epoch": 1.553314121037464,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047638512568615307,
      "loss": 4.9826,
      "mean_token_accuracy": 0.21540612280368804,
      "num_tokens": 37080221.0,
      "step": 16170
    },
    {
      "entropy": 5.2076140403747555,
      "epoch": 1.5537944284341978,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004763698337112553,
      "loss": 4.9597,
      "mean_token_accuracy": 0.21362027227878572,
      "num_tokens": 37092223.0,
      "step": 16175
    },
    {
      "entropy": 5.274347734451294,
      "epoch": 1.5542747358309317,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004763545370611479,
      "loss": 4.9981,
      "mean_token_accuracy": 0.20787416249513627,
      "num_tokens": 37104169.0,
      "step": 16180
    },
    {
      "entropy": 5.221208333969116,
      "epoch": 1.5547550432276656,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047633923573618605,
      "loss": 5.0247,
      "mean_token_accuracy": 0.21744638234376906,
      "num_tokens": 37115784.0,
      "step": 16185
    },
    {
      "entropy": 5.1707789421081545,
      "epoch": 1.5552353506243997,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000476323929736725,
      "loss": 4.8782,
      "mean_token_accuracy": 0.2215387210249901,
      "num_tokens": 37128018.0,
      "step": 16190
    },
    {
      "entropy": 5.224068737030029,
      "epoch": 1.5557156580211335,
      "grad_norm": 1.25,
      "learning_rate": 0.00047630861906312004,
      "loss": 4.9289,
      "mean_token_accuracy": 0.21138983219861984,
      "num_tokens": 37139330.0,
      "step": 16195
    },
    {
      "entropy": 5.096193552017212,
      "epoch": 1.5561959654178674,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004762933037157268,
      "loss": 4.8233,
      "mean_token_accuracy": 0.22867411375045776,
      "num_tokens": 37151531.0,
      "step": 16200
    },
    {
      "entropy": 5.245355653762817,
      "epoch": 1.5566762728146013,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047627798369490076,
      "loss": 4.9976,
      "mean_token_accuracy": 0.21494047343730927,
      "num_tokens": 37162442.0,
      "step": 16205
    },
    {
      "entropy": 5.206206369400024,
      "epoch": 1.5571565802113354,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047626265900099757,
      "loss": 4.9775,
      "mean_token_accuracy": 0.21960482448339463,
      "num_tokens": 37173737.0,
      "step": 16210
    },
    {
      "entropy": 5.131538534164429,
      "epoch": 1.5576368876080693,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047624732963437314,
      "loss": 4.9252,
      "mean_token_accuracy": 0.21751572340726852,
      "num_tokens": 37185983.0,
      "step": 16215
    },
    {
      "entropy": 5.280227994918823,
      "epoch": 1.5581171950048032,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047623199559538324,
      "loss": 5.0361,
      "mean_token_accuracy": 0.20776247680187226,
      "num_tokens": 37197119.0,
      "step": 16220
    },
    {
      "entropy": 5.3747467517852785,
      "epoch": 1.558597502401537,
      "grad_norm": 1.40625,
      "learning_rate": 0.000476216656884384,
      "loss": 5.1192,
      "mean_token_accuracy": 0.2094549924135208,
      "num_tokens": 37209516.0,
      "step": 16225
    },
    {
      "entropy": 5.0878071784973145,
      "epoch": 1.559077809798271,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047620131350173135,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21942342966794967,
      "num_tokens": 37220559.0,
      "step": 16230
    },
    {
      "entropy": 5.233538007736206,
      "epoch": 1.5595581171950048,
      "grad_norm": 1.375,
      "learning_rate": 0.0004761859654477817,
      "loss": 4.9923,
      "mean_token_accuracy": 0.21255081444978713,
      "num_tokens": 37232232.0,
      "step": 16235
    },
    {
      "entropy": 5.134554386138916,
      "epoch": 1.5600384245917387,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004761706127228914,
      "loss": 4.8599,
      "mean_token_accuracy": 0.22222913652658463,
      "num_tokens": 37242819.0,
      "step": 16240
    },
    {
      "entropy": 5.226075601577759,
      "epoch": 1.5605187319884726,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004761552553274168,
      "loss": 4.9892,
      "mean_token_accuracy": 0.21356966942548752,
      "num_tokens": 37254940.0,
      "step": 16245
    },
    {
      "entropy": 5.238306951522827,
      "epoch": 1.5609990393852065,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004761398932617144,
      "loss": 5.0249,
      "mean_token_accuracy": 0.21183981001377106,
      "num_tokens": 37266220.0,
      "step": 16250
    },
    {
      "entropy": 5.271004724502563,
      "epoch": 1.5614793467819403,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000476124526526141,
      "loss": 5.0231,
      "mean_token_accuracy": 0.20968729853630066,
      "num_tokens": 37278136.0,
      "step": 16255
    },
    {
      "entropy": 5.215725421905518,
      "epoch": 1.5619596541786742,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047610915512105327,
      "loss": 4.962,
      "mean_token_accuracy": 0.21382360756397248,
      "num_tokens": 37290358.0,
      "step": 16260
    },
    {
      "entropy": 5.276764726638794,
      "epoch": 1.562439961575408,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004760937790468082,
      "loss": 5.0378,
      "mean_token_accuracy": 0.21809940040111542,
      "num_tokens": 37300719.0,
      "step": 16265
    },
    {
      "entropy": 5.167660093307495,
      "epoch": 1.5629202689721422,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004760783983037627,
      "loss": 5.0006,
      "mean_token_accuracy": 0.2116939067840576,
      "num_tokens": 37311917.0,
      "step": 16270
    },
    {
      "entropy": 5.220437860488891,
      "epoch": 1.563400576368876,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004760630128922738,
      "loss": 4.9523,
      "mean_token_accuracy": 0.2214370086789131,
      "num_tokens": 37325076.0,
      "step": 16275
    },
    {
      "entropy": 5.246109247207642,
      "epoch": 1.56388088376561,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004760476228126989,
      "loss": 4.9723,
      "mean_token_accuracy": 0.2120165705680847,
      "num_tokens": 37336064.0,
      "step": 16280
    },
    {
      "entropy": 5.21238899230957,
      "epoch": 1.5643611911623438,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004760322280653951,
      "loss": 4.9827,
      "mean_token_accuracy": 0.21337343603372574,
      "num_tokens": 37347343.0,
      "step": 16285
    },
    {
      "entropy": 5.194455909729004,
      "epoch": 1.564841498559078,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004760168286507199,
      "loss": 4.9289,
      "mean_token_accuracy": 0.22189528197050096,
      "num_tokens": 37359067.0,
      "step": 16290
    },
    {
      "entropy": 5.201845932006836,
      "epoch": 1.5653218059558118,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047600142456903085,
      "loss": 4.9623,
      "mean_token_accuracy": 0.21071673929691315,
      "num_tokens": 37369811.0,
      "step": 16295
    },
    {
      "entropy": 5.115090799331665,
      "epoch": 1.5658021133525457,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047598601582068555,
      "loss": 4.8212,
      "mean_token_accuracy": 0.22707246392965316,
      "num_tokens": 37380944.0,
      "step": 16300
    },
    {
      "entropy": 5.216742134094238,
      "epoch": 1.5662824207492796,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004759706024060418,
      "loss": 4.9849,
      "mean_token_accuracy": 0.20313621759414674,
      "num_tokens": 37392099.0,
      "step": 16305
    },
    {
      "entropy": 5.160527658462525,
      "epoch": 1.5667627281460135,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004759551843254575,
      "loss": 4.9603,
      "mean_token_accuracy": 0.21864845156669616,
      "num_tokens": 37404254.0,
      "step": 16310
    },
    {
      "entropy": 5.140524530410767,
      "epoch": 1.5672430355427474,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047593976157929034,
      "loss": 4.898,
      "mean_token_accuracy": 0.22447916567325593,
      "num_tokens": 37416112.0,
      "step": 16315
    },
    {
      "entropy": 5.182991886138916,
      "epoch": 1.5677233429394812,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004759243341678987,
      "loss": 4.909,
      "mean_token_accuracy": 0.21476919054985047,
      "num_tokens": 37428350.0,
      "step": 16320
    },
    {
      "entropy": 5.205784273147583,
      "epoch": 1.5682036503362151,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004759089020916407,
      "loss": 5.01,
      "mean_token_accuracy": 0.21999542713165282,
      "num_tokens": 37440059.0,
      "step": 16325
    },
    {
      "entropy": 5.26274094581604,
      "epoch": 1.568683957732949,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047589346535087444,
      "loss": 5.0166,
      "mean_token_accuracy": 0.2096041738986969,
      "num_tokens": 37450880.0,
      "step": 16330
    },
    {
      "entropy": 5.187830924987793,
      "epoch": 1.5691642651296829,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004758780239459586,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22152020931243896,
      "num_tokens": 37463484.0,
      "step": 16335
    },
    {
      "entropy": 5.1897376537322994,
      "epoch": 1.5696445725264168,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004758625778772514,
      "loss": 4.9547,
      "mean_token_accuracy": 0.2186115562915802,
      "num_tokens": 37474952.0,
      "step": 16340
    },
    {
      "entropy": 5.233623313903808,
      "epoch": 1.5701248799231509,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047584712714511166,
      "loss": 5.0676,
      "mean_token_accuracy": 0.20902722775936128,
      "num_tokens": 37486858.0,
      "step": 16345
    },
    {
      "entropy": 5.236107397079468,
      "epoch": 1.5706051873198847,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047583167174989797,
      "loss": 4.8903,
      "mean_token_accuracy": 0.21899403333663942,
      "num_tokens": 37497852.0,
      "step": 16350
    },
    {
      "entropy": 5.257089567184448,
      "epoch": 1.5710854947166186,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004758162116919692,
      "loss": 5.0057,
      "mean_token_accuracy": 0.2129534587264061,
      "num_tokens": 37509051.0,
      "step": 16355
    },
    {
      "entropy": 5.256408596038819,
      "epoch": 1.5715658021133525,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047580074697168434,
      "loss": 5.104,
      "mean_token_accuracy": 0.21249212473630905,
      "num_tokens": 37520415.0,
      "step": 16360
    },
    {
      "entropy": 5.223194980621338,
      "epoch": 1.5720461095100866,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047578527758940236,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2147620514035225,
      "num_tokens": 37532087.0,
      "step": 16365
    },
    {
      "entropy": 5.217049312591553,
      "epoch": 1.5725264169068205,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004757698035454825,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2100052624940872,
      "num_tokens": 37545219.0,
      "step": 16370
    },
    {
      "entropy": 5.2135436058044435,
      "epoch": 1.5730067243035544,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004757543248402839,
      "loss": 4.9814,
      "mean_token_accuracy": 0.21640931218862533,
      "num_tokens": 37558138.0,
      "step": 16375
    },
    {
      "entropy": 5.20058274269104,
      "epoch": 1.5734870317002883,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047573884147416597,
      "loss": 4.8618,
      "mean_token_accuracy": 0.21841758340597153,
      "num_tokens": 37568975.0,
      "step": 16380
    },
    {
      "entropy": 5.162148952484131,
      "epoch": 1.5739673390970221,
      "grad_norm": 2.078125,
      "learning_rate": 0.0004757233534474883,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22167371958494186,
      "num_tokens": 37580774.0,
      "step": 16385
    },
    {
      "entropy": 5.107549953460693,
      "epoch": 1.574447646493756,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004757078607606103,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2185952588915825,
      "num_tokens": 37592858.0,
      "step": 16390
    },
    {
      "entropy": 5.200916862487793,
      "epoch": 1.57492795389049,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004756923634138918,
      "loss": 4.9932,
      "mean_token_accuracy": 0.21158214658498764,
      "num_tokens": 37604005.0,
      "step": 16395
    },
    {
      "entropy": 5.220621585845947,
      "epoch": 1.5754082612872238,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047567686140769264,
      "loss": 4.9279,
      "mean_token_accuracy": 0.21968565434217452,
      "num_tokens": 37614629.0,
      "step": 16400
    },
    {
      "entropy": 5.163049602508545,
      "epoch": 1.5758885686839577,
      "grad_norm": 1.703125,
      "learning_rate": 0.00047566135474237247,
      "loss": 4.9308,
      "mean_token_accuracy": 0.22496608048677444,
      "num_tokens": 37626118.0,
      "step": 16405
    },
    {
      "entropy": 5.274559116363525,
      "epoch": 1.5763688760806915,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047564584341829166,
      "loss": 5.0519,
      "mean_token_accuracy": 0.20112663954496385,
      "num_tokens": 37638514.0,
      "step": 16410
    },
    {
      "entropy": 5.27506012916565,
      "epoch": 1.5768491834774254,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047563032743581,
      "loss": 5.0402,
      "mean_token_accuracy": 0.21117023229599,
      "num_tokens": 37649602.0,
      "step": 16415
    },
    {
      "entropy": 5.21573543548584,
      "epoch": 1.5773294908741593,
      "grad_norm": 1.3125,
      "learning_rate": 0.00047561480679528804,
      "loss": 5.0075,
      "mean_token_accuracy": 0.21156248599290847,
      "num_tokens": 37661428.0,
      "step": 16420
    },
    {
      "entropy": 5.264469957351684,
      "epoch": 1.5778097982708934,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004755992814970859,
      "loss": 4.9809,
      "mean_token_accuracy": 0.21753261089324952,
      "num_tokens": 37673608.0,
      "step": 16425
    },
    {
      "entropy": 5.314175653457641,
      "epoch": 1.5782901056676273,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000475583751541564,
      "loss": 5.1199,
      "mean_token_accuracy": 0.21422826498746872,
      "num_tokens": 37684957.0,
      "step": 16430
    },
    {
      "entropy": 5.165050983428955,
      "epoch": 1.5787704130643612,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047556821692908315,
      "loss": 4.9941,
      "mean_token_accuracy": 0.21423042714595794,
      "num_tokens": 37697773.0,
      "step": 16435
    },
    {
      "entropy": 5.196795701980591,
      "epoch": 1.579250720461095,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004755526776600038,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21462689489126205,
      "num_tokens": 37708315.0,
      "step": 16440
    },
    {
      "entropy": 5.200553369522095,
      "epoch": 1.5797310278578292,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047553713373468684,
      "loss": 4.9759,
      "mean_token_accuracy": 0.21728957146406175,
      "num_tokens": 37720440.0,
      "step": 16445
    },
    {
      "entropy": 5.1274253845214846,
      "epoch": 1.580211335254563,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047552158515349306,
      "loss": 4.8655,
      "mean_token_accuracy": 0.2221350699663162,
      "num_tokens": 37732332.0,
      "step": 16450
    },
    {
      "entropy": 5.158447980880737,
      "epoch": 1.580691642651297,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047550603191678356,
      "loss": 4.9588,
      "mean_token_accuracy": 0.212063567340374,
      "num_tokens": 37744770.0,
      "step": 16455
    },
    {
      "entropy": 5.1647241592407225,
      "epoch": 1.5811719500480308,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004754904740249194,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22328755259513855,
      "num_tokens": 37755958.0,
      "step": 16460
    },
    {
      "entropy": 5.220455503463745,
      "epoch": 1.5816522574447647,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047547491147826156,
      "loss": 4.9251,
      "mean_token_accuracy": 0.22103270888328552,
      "num_tokens": 37767182.0,
      "step": 16465
    },
    {
      "entropy": 5.180071401596069,
      "epoch": 1.5821325648414986,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004754593442771718,
      "loss": 5.0088,
      "mean_token_accuracy": 0.2134843185544014,
      "num_tokens": 37778299.0,
      "step": 16470
    },
    {
      "entropy": 5.3074744701385494,
      "epoch": 1.5826128722382324,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00047544377242201115,
      "loss": 5.0231,
      "mean_token_accuracy": 0.21542281955480574,
      "num_tokens": 37789879.0,
      "step": 16475
    },
    {
      "entropy": 5.272194480895996,
      "epoch": 1.5830931796349663,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047542819591314136,
      "loss": 5.0188,
      "mean_token_accuracy": 0.212533038854599,
      "num_tokens": 37800726.0,
      "step": 16480
    },
    {
      "entropy": 5.186125612258911,
      "epoch": 1.5835734870317002,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004754126147509241,
      "loss": 4.9675,
      "mean_token_accuracy": 0.21615018397569657,
      "num_tokens": 37812298.0,
      "step": 16485
    },
    {
      "entropy": 5.128080272674561,
      "epoch": 1.584053794428434,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047539702893572086,
      "loss": 4.8679,
      "mean_token_accuracy": 0.23003831058740615,
      "num_tokens": 37824433.0,
      "step": 16490
    },
    {
      "entropy": 5.225295734405518,
      "epoch": 1.584534101825168,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047538143846789376,
      "loss": 4.9691,
      "mean_token_accuracy": 0.2206984981894493,
      "num_tokens": 37835647.0,
      "step": 16495
    },
    {
      "entropy": 5.252954530715942,
      "epoch": 1.585014409221902,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004753658433478047,
      "loss": 5.0422,
      "mean_token_accuracy": 0.21227143555879593,
      "num_tokens": 37847721.0,
      "step": 16500
    },
    {
      "entropy": 5.208762693405151,
      "epoch": 1.585494716618636,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047535024357581564,
      "loss": 4.8934,
      "mean_token_accuracy": 0.22291746139526367,
      "num_tokens": 37860111.0,
      "step": 16505
    },
    {
      "entropy": 5.192453622817993,
      "epoch": 1.5859750240153698,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004753346391522889,
      "loss": 5.0377,
      "mean_token_accuracy": 0.2114849090576172,
      "num_tokens": 37872321.0,
      "step": 16510
    },
    {
      "entropy": 5.326900005340576,
      "epoch": 1.5864553314121037,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047531903007758667,
      "loss": 5.0894,
      "mean_token_accuracy": 0.2090721145272255,
      "num_tokens": 37885195.0,
      "step": 16515
    },
    {
      "entropy": 5.185902452468872,
      "epoch": 1.5869356388088378,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004753034163520714,
      "loss": 4.9163,
      "mean_token_accuracy": 0.22317426800727844,
      "num_tokens": 37895487.0,
      "step": 16520
    },
    {
      "entropy": 5.097846508026123,
      "epoch": 1.5874159462055717,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047528779797610557,
      "loss": 5.0206,
      "mean_token_accuracy": 0.21476031243801116,
      "num_tokens": 37908344.0,
      "step": 16525
    },
    {
      "entropy": 5.18268985748291,
      "epoch": 1.5878962536023056,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047527217495005184,
      "loss": 4.9455,
      "mean_token_accuracy": 0.21450677365064622,
      "num_tokens": 37920220.0,
      "step": 16530
    },
    {
      "entropy": 5.20910611152649,
      "epoch": 1.5883765609990395,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047525654727427285,
      "loss": 4.8032,
      "mean_token_accuracy": 0.22643448561429977,
      "num_tokens": 37930874.0,
      "step": 16535
    },
    {
      "entropy": 5.140932655334472,
      "epoch": 1.5888568683957733,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004752409149491315,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21765501499176027,
      "num_tokens": 37942501.0,
      "step": 16540
    },
    {
      "entropy": 5.1464025497436525,
      "epoch": 1.5893371757925072,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047522527797499075,
      "loss": 4.9615,
      "mean_token_accuracy": 0.21340625435113908,
      "num_tokens": 37954408.0,
      "step": 16545
    },
    {
      "entropy": 5.214100360870361,
      "epoch": 1.589817483189241,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004752096363522135,
      "loss": 5.0065,
      "mean_token_accuracy": 0.2071237489581108,
      "num_tokens": 37966088.0,
      "step": 16550
    },
    {
      "entropy": 5.242699670791626,
      "epoch": 1.590297790585975,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047519399008116305,
      "loss": 4.9759,
      "mean_token_accuracy": 0.20768681764602662,
      "num_tokens": 37977726.0,
      "step": 16555
    },
    {
      "entropy": 5.2041003704071045,
      "epoch": 1.5907780979827089,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004751783391622026,
      "loss": 5.0047,
      "mean_token_accuracy": 0.2108057200908661,
      "num_tokens": 37989024.0,
      "step": 16560
    },
    {
      "entropy": 5.288478231430053,
      "epoch": 1.5912584053794427,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004751626835956955,
      "loss": 5.0146,
      "mean_token_accuracy": 0.20912586897611618,
      "num_tokens": 37999870.0,
      "step": 16565
    },
    {
      "entropy": 5.223183679580688,
      "epoch": 1.5917387127761766,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004751470233820053,
      "loss": 5.0406,
      "mean_token_accuracy": 0.20820825845003127,
      "num_tokens": 38010945.0,
      "step": 16570
    },
    {
      "entropy": 5.133132600784302,
      "epoch": 1.5922190201729105,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004751313585214955,
      "loss": 4.9291,
      "mean_token_accuracy": 0.2216455817222595,
      "num_tokens": 38021584.0,
      "step": 16575
    },
    {
      "entropy": 5.232112073898316,
      "epoch": 1.5926993275696446,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004751156890145298,
      "loss": 5.0116,
      "mean_token_accuracy": 0.21161716282367707,
      "num_tokens": 38034407.0,
      "step": 16580
    },
    {
      "entropy": 5.307591915130615,
      "epoch": 1.5931796349663785,
      "grad_norm": 1.265625,
      "learning_rate": 0.000475100014861472,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20845135003328324,
      "num_tokens": 38047061.0,
      "step": 16585
    },
    {
      "entropy": 5.292564392089844,
      "epoch": 1.5936599423631124,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004750843360626861,
      "loss": 5.0711,
      "mean_token_accuracy": 0.21227787137031556,
      "num_tokens": 38059168.0,
      "step": 16590
    },
    {
      "entropy": 5.167920112609863,
      "epoch": 1.5941402497598463,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004750686526185359,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21823298633098603,
      "num_tokens": 38070514.0,
      "step": 16595
    },
    {
      "entropy": 5.31266827583313,
      "epoch": 1.5946205571565804,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047505296452938584,
      "loss": 5.0625,
      "mean_token_accuracy": 0.21305903792381287,
      "num_tokens": 38081995.0,
      "step": 16600
    },
    {
      "entropy": 5.170768451690674,
      "epoch": 1.5951008645533142,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047503727179559995,
      "loss": 4.9143,
      "mean_token_accuracy": 0.22254307568073273,
      "num_tokens": 38094371.0,
      "step": 16605
    },
    {
      "entropy": 5.1151838302612305,
      "epoch": 1.5955811719500481,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047502157441754256,
      "loss": 4.8809,
      "mean_token_accuracy": 0.21849231868982316,
      "num_tokens": 38106340.0,
      "step": 16610
    },
    {
      "entropy": 5.217184972763062,
      "epoch": 1.596061479346782,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004750058723955781,
      "loss": 4.9291,
      "mean_token_accuracy": 0.2166606068611145,
      "num_tokens": 38117641.0,
      "step": 16615
    },
    {
      "entropy": 5.238893795013428,
      "epoch": 1.5965417867435159,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004749901657300713,
      "loss": 4.9369,
      "mean_token_accuracy": 0.21517444550991058,
      "num_tokens": 38129628.0,
      "step": 16620
    },
    {
      "entropy": 5.133989334106445,
      "epoch": 1.5970220941402498,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047497445442138667,
      "loss": 4.911,
      "mean_token_accuracy": 0.21797181516885758,
      "num_tokens": 38140843.0,
      "step": 16625
    },
    {
      "entropy": 5.173162269592285,
      "epoch": 1.5975024015369836,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047495873846988896,
      "loss": 4.8926,
      "mean_token_accuracy": 0.22584536224603652,
      "num_tokens": 38152666.0,
      "step": 16630
    },
    {
      "entropy": 5.141886472702026,
      "epoch": 1.5979827089337175,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004749430178759431,
      "loss": 4.863,
      "mean_token_accuracy": 0.22337797582149505,
      "num_tokens": 38164102.0,
      "step": 16635
    },
    {
      "entropy": 5.142643737792969,
      "epoch": 1.5984630163304514,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047492729263991413,
      "loss": 4.9459,
      "mean_token_accuracy": 0.22133104503154755,
      "num_tokens": 38175211.0,
      "step": 16640
    },
    {
      "entropy": 5.153062152862549,
      "epoch": 1.5989433237271853,
      "grad_norm": 1.25,
      "learning_rate": 0.00047491156276216695,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22370103001594543,
      "num_tokens": 38186968.0,
      "step": 16645
    },
    {
      "entropy": 5.197846555709839,
      "epoch": 1.5994236311239192,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047489582824306704,
      "loss": 4.9616,
      "mean_token_accuracy": 0.21329808384180068,
      "num_tokens": 38198740.0,
      "step": 16650
    },
    {
      "entropy": 5.163088321685791,
      "epoch": 1.5999039385206533,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047488008908297955,
      "loss": 4.8812,
      "mean_token_accuracy": 0.226847368478775,
      "num_tokens": 38209110.0,
      "step": 16655
    },
    {
      "entropy": 5.142639398574829,
      "epoch": 1.6003842459173871,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004748643452822699,
      "loss": 4.8621,
      "mean_token_accuracy": 0.22405126839876174,
      "num_tokens": 38221840.0,
      "step": 16660
    },
    {
      "entropy": 5.246883201599121,
      "epoch": 1.600864553314121,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004748485968413036,
      "loss": 5.0058,
      "mean_token_accuracy": 0.21522666662931442,
      "num_tokens": 38232752.0,
      "step": 16665
    },
    {
      "entropy": 5.168501853942871,
      "epoch": 1.601344860710855,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047483284376044634,
      "loss": 4.9295,
      "mean_token_accuracy": 0.21313532441854477,
      "num_tokens": 38243394.0,
      "step": 16670
    },
    {
      "entropy": 5.19124550819397,
      "epoch": 1.601825168107589,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004748170860400638,
      "loss": 5.0366,
      "mean_token_accuracy": 0.21428718715906142,
      "num_tokens": 38254317.0,
      "step": 16675
    },
    {
      "entropy": 5.166198587417602,
      "epoch": 1.602305475504323,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047480132368052185,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22433751225471496,
      "num_tokens": 38266886.0,
      "step": 16680
    },
    {
      "entropy": 5.259693670272827,
      "epoch": 1.6027857829010568,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047478555668218643,
      "loss": 5.0962,
      "mean_token_accuracy": 0.20623203366994858,
      "num_tokens": 38279975.0,
      "step": 16685
    },
    {
      "entropy": 5.243435001373291,
      "epoch": 1.6032660902977907,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004747697850454237,
      "loss": 5.0443,
      "mean_token_accuracy": 0.20509886294603347,
      "num_tokens": 38290237.0,
      "step": 16690
    },
    {
      "entropy": 5.26441798210144,
      "epoch": 1.6037463976945245,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004747540087705997,
      "loss": 4.8926,
      "mean_token_accuracy": 0.2192753419280052,
      "num_tokens": 38300833.0,
      "step": 16695
    },
    {
      "entropy": 5.090855932235717,
      "epoch": 1.6042267050912584,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004747382278580808,
      "loss": 4.8599,
      "mean_token_accuracy": 0.22707237005233766,
      "num_tokens": 38311926.0,
      "step": 16700
    },
    {
      "entropy": 5.227768993377685,
      "epoch": 1.6047070124879923,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004747224423082333,
      "loss": 5.0827,
      "mean_token_accuracy": 0.2084112659096718,
      "num_tokens": 38323053.0,
      "step": 16705
    },
    {
      "entropy": 5.227802515029907,
      "epoch": 1.6051873198847262,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047470665212142384,
      "loss": 5.0075,
      "mean_token_accuracy": 0.210758538544178,
      "num_tokens": 38334959.0,
      "step": 16710
    },
    {
      "entropy": 5.237191152572632,
      "epoch": 1.60566762728146,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047469085729801887,
      "loss": 5.0325,
      "mean_token_accuracy": 0.2084895506501198,
      "num_tokens": 38346402.0,
      "step": 16715
    },
    {
      "entropy": 5.229112815856934,
      "epoch": 1.606147934678194,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047467505783838515,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21257033944129944,
      "num_tokens": 38358549.0,
      "step": 16720
    },
    {
      "entropy": 5.198215436935425,
      "epoch": 1.6066282420749278,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004746592537428895,
      "loss": 4.9687,
      "mean_token_accuracy": 0.2161519557237625,
      "num_tokens": 38370840.0,
      "step": 16725
    },
    {
      "entropy": 5.214632749557495,
      "epoch": 1.6071085494716617,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047464344501189877,
      "loss": 4.97,
      "mean_token_accuracy": 0.2140952005982399,
      "num_tokens": 38382980.0,
      "step": 16730
    },
    {
      "entropy": 5.215225458145142,
      "epoch": 1.6075888568683958,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047462763164578015,
      "loss": 4.8909,
      "mean_token_accuracy": 0.22906397581100463,
      "num_tokens": 38393836.0,
      "step": 16735
    },
    {
      "entropy": 5.105970335006714,
      "epoch": 1.6080691642651297,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004746118136449007,
      "loss": 4.8775,
      "mean_token_accuracy": 0.22095937430858612,
      "num_tokens": 38405770.0,
      "step": 16740
    },
    {
      "entropy": 5.221188735961914,
      "epoch": 1.6085494716618636,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004745959910096276,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20926397293806076,
      "num_tokens": 38418847.0,
      "step": 16745
    },
    {
      "entropy": 5.15964150428772,
      "epoch": 1.6090297790585975,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047458016374032837,
      "loss": 4.887,
      "mean_token_accuracy": 0.221133254468441,
      "num_tokens": 38431532.0,
      "step": 16750
    },
    {
      "entropy": 5.3003456592559814,
      "epoch": 1.6095100864553316,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004745643318373703,
      "loss": 5.1263,
      "mean_token_accuracy": 0.2126389726996422,
      "num_tokens": 38442071.0,
      "step": 16755
    },
    {
      "entropy": 5.187539005279541,
      "epoch": 1.6099903938520654,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047454849530112106,
      "loss": 5.0208,
      "mean_token_accuracy": 0.2113887995481491,
      "num_tokens": 38454979.0,
      "step": 16760
    },
    {
      "entropy": 5.1524640083312985,
      "epoch": 1.6104707012487993,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047453265413194826,
      "loss": 4.8632,
      "mean_token_accuracy": 0.22795895785093306,
      "num_tokens": 38465579.0,
      "step": 16765
    },
    {
      "entropy": 5.136258935928344,
      "epoch": 1.6109510086455332,
      "grad_norm": 1.125,
      "learning_rate": 0.00047451680833021973,
      "loss": 4.9399,
      "mean_token_accuracy": 0.21665653735399246,
      "num_tokens": 38477139.0,
      "step": 16770
    },
    {
      "entropy": 5.147571849822998,
      "epoch": 1.611431316042267,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004745009578963034,
      "loss": 4.8658,
      "mean_token_accuracy": 0.22467368990182876,
      "num_tokens": 38487638.0,
      "step": 16775
    },
    {
      "entropy": 5.07762565612793,
      "epoch": 1.611911623439001,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047448510283056716,
      "loss": 4.7774,
      "mean_token_accuracy": 0.23332076221704484,
      "num_tokens": 38498395.0,
      "step": 16780
    },
    {
      "entropy": 5.177888011932373,
      "epoch": 1.6123919308357348,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047446924313337925,
      "loss": 4.8788,
      "mean_token_accuracy": 0.2227744609117508,
      "num_tokens": 38509372.0,
      "step": 16785
    },
    {
      "entropy": 5.190420770645142,
      "epoch": 1.6128722382324687,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047445337880510773,
      "loss": 4.914,
      "mean_token_accuracy": 0.21360821574926375,
      "num_tokens": 38520651.0,
      "step": 16790
    },
    {
      "entropy": 5.212368726730347,
      "epoch": 1.6133525456292026,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004744375098461211,
      "loss": 5.0245,
      "mean_token_accuracy": 0.20779217034578323,
      "num_tokens": 38531461.0,
      "step": 16795
    },
    {
      "entropy": 5.13183479309082,
      "epoch": 1.6138328530259365,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004744216362567876,
      "loss": 4.8483,
      "mean_token_accuracy": 0.22859577983617782,
      "num_tokens": 38542707.0,
      "step": 16800
    },
    {
      "entropy": 5.119955587387085,
      "epoch": 1.6143131604226704,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047440575803747595,
      "loss": 4.8633,
      "mean_token_accuracy": 0.2269836589694023,
      "num_tokens": 38553664.0,
      "step": 16805
    },
    {
      "entropy": 5.239011907577515,
      "epoch": 1.6147934678194045,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047438987518855463,
      "loss": 4.9431,
      "mean_token_accuracy": 0.21409208625555037,
      "num_tokens": 38564891.0,
      "step": 16810
    },
    {
      "entropy": 5.225918102264404,
      "epoch": 1.6152737752161384,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004743739877103926,
      "loss": 4.9436,
      "mean_token_accuracy": 0.2142092153429985,
      "num_tokens": 38576535.0,
      "step": 16815
    },
    {
      "entropy": 5.183576917648315,
      "epoch": 1.6157540826128722,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004743580956033585,
      "loss": 4.945,
      "mean_token_accuracy": 0.2154086321592331,
      "num_tokens": 38588160.0,
      "step": 16820
    },
    {
      "entropy": 5.121462869644165,
      "epoch": 1.6162343900096061,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047434219886782135,
      "loss": 4.8632,
      "mean_token_accuracy": 0.22172888964414597,
      "num_tokens": 38599281.0,
      "step": 16825
    },
    {
      "entropy": 5.195735836029053,
      "epoch": 1.6167146974063402,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004743262975041504,
      "loss": 4.955,
      "mean_token_accuracy": 0.21920662075281144,
      "num_tokens": 38611823.0,
      "step": 16830
    },
    {
      "entropy": 5.149890565872193,
      "epoch": 1.617195004803074,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004743103915127146,
      "loss": 4.9538,
      "mean_token_accuracy": 0.21739039570093155,
      "num_tokens": 38623181.0,
      "step": 16835
    },
    {
      "entropy": 5.200158739089966,
      "epoch": 1.617675312199808,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047429448089388336,
      "loss": 4.9696,
      "mean_token_accuracy": 0.2187592789530754,
      "num_tokens": 38633149.0,
      "step": 16840
    },
    {
      "entropy": 5.221182060241699,
      "epoch": 1.6181556195965419,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047427856564802605,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2120251700282097,
      "num_tokens": 38644204.0,
      "step": 16845
    },
    {
      "entropy": 5.230680799484253,
      "epoch": 1.6186359269932757,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004742626457755122,
      "loss": 4.9378,
      "mean_token_accuracy": 0.21604713052511215,
      "num_tokens": 38655287.0,
      "step": 16850
    },
    {
      "entropy": 5.113589954376221,
      "epoch": 1.6191162343900096,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004742467212767114,
      "loss": 4.8251,
      "mean_token_accuracy": 0.219390632212162,
      "num_tokens": 38665732.0,
      "step": 16855
    },
    {
      "entropy": 5.223013973236084,
      "epoch": 1.6195965417867435,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004742307921519933,
      "loss": 5.0406,
      "mean_token_accuracy": 0.20536922067403793,
      "num_tokens": 38677289.0,
      "step": 16860
    },
    {
      "entropy": 5.2103111267089846,
      "epoch": 1.6200768491834774,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047421485840172794,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22298752069473265,
      "num_tokens": 38689581.0,
      "step": 16865
    },
    {
      "entropy": 5.220411396026611,
      "epoch": 1.6205571565802113,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004741989200262851,
      "loss": 4.9867,
      "mean_token_accuracy": 0.20872607231140136,
      "num_tokens": 38701327.0,
      "step": 16870
    },
    {
      "entropy": 5.14514126777649,
      "epoch": 1.6210374639769451,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004741829770260347,
      "loss": 4.9804,
      "mean_token_accuracy": 0.215652734041214,
      "num_tokens": 38713431.0,
      "step": 16875
    },
    {
      "entropy": 5.251890325546265,
      "epoch": 1.621517771373679,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047416702940134714,
      "loss": 4.9948,
      "mean_token_accuracy": 0.2162155494093895,
      "num_tokens": 38724091.0,
      "step": 16880
    },
    {
      "entropy": 5.2016918659210205,
      "epoch": 1.621998078770413,
      "grad_norm": 1.125,
      "learning_rate": 0.00047415107715259255,
      "loss": 5.0467,
      "mean_token_accuracy": 0.21599478870630265,
      "num_tokens": 38736219.0,
      "step": 16885
    },
    {
      "entropy": 5.177992057800293,
      "epoch": 1.622478386167147,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047413512028014125,
      "loss": 4.912,
      "mean_token_accuracy": 0.21863125264644623,
      "num_tokens": 38747512.0,
      "step": 16890
    },
    {
      "entropy": 5.235227012634278,
      "epoch": 1.622958693563881,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004741191587843638,
      "loss": 5.0535,
      "mean_token_accuracy": 0.207265405356884,
      "num_tokens": 38758777.0,
      "step": 16895
    },
    {
      "entropy": 5.240120792388916,
      "epoch": 1.6234390009606148,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004741031926656308,
      "loss": 4.9862,
      "mean_token_accuracy": 0.22030759155750274,
      "num_tokens": 38770094.0,
      "step": 16900
    },
    {
      "entropy": 5.231153726577759,
      "epoch": 1.6239193083573487,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004740872219243128,
      "loss": 4.9434,
      "mean_token_accuracy": 0.2218146428465843,
      "num_tokens": 38781419.0,
      "step": 16905
    },
    {
      "entropy": 5.178551816940308,
      "epoch": 1.6243996157540828,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004740712465607807,
      "loss": 4.9366,
      "mean_token_accuracy": 0.2185610607266426,
      "num_tokens": 38794303.0,
      "step": 16910
    },
    {
      "entropy": 5.17335844039917,
      "epoch": 1.6248799231508166,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004740552665754054,
      "loss": 4.9657,
      "mean_token_accuracy": 0.22210513651371003,
      "num_tokens": 38805387.0,
      "step": 16915
    },
    {
      "entropy": 5.192090368270874,
      "epoch": 1.6253602305475505,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047403928196855776,
      "loss": 4.9461,
      "mean_token_accuracy": 0.2186659425497055,
      "num_tokens": 38815307.0,
      "step": 16920
    },
    {
      "entropy": 5.228534889221192,
      "epoch": 1.6258405379442844,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047402329274060916,
      "loss": 4.9618,
      "mean_token_accuracy": 0.22098185122013092,
      "num_tokens": 38826585.0,
      "step": 16925
    },
    {
      "entropy": 5.169501161575317,
      "epoch": 1.6263208453410183,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004740072988919306,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22761445492506027,
      "num_tokens": 38838406.0,
      "step": 16930
    },
    {
      "entropy": 5.1789051532745365,
      "epoch": 1.6268011527377522,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004739913004228936,
      "loss": 4.8992,
      "mean_token_accuracy": 0.22279890179634093,
      "num_tokens": 38849952.0,
      "step": 16935
    },
    {
      "entropy": 5.210028076171875,
      "epoch": 1.627281460134486,
      "grad_norm": 1.125,
      "learning_rate": 0.0004739752973338694,
      "loss": 5.0136,
      "mean_token_accuracy": 0.21209951490163803,
      "num_tokens": 38862761.0,
      "step": 16940
    },
    {
      "entropy": 5.16680908203125,
      "epoch": 1.62776176753122,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047395928962522965,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2145182102918625,
      "num_tokens": 38874422.0,
      "step": 16945
    },
    {
      "entropy": 5.202334213256836,
      "epoch": 1.6282420749279538,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047394327729734595,
      "loss": 4.9856,
      "mean_token_accuracy": 0.21239533871412278,
      "num_tokens": 38885751.0,
      "step": 16950
    },
    {
      "entropy": 5.178429508209229,
      "epoch": 1.6287223823246877,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004739272603505901,
      "loss": 4.9453,
      "mean_token_accuracy": 0.22557273656129836,
      "num_tokens": 38897539.0,
      "step": 16955
    },
    {
      "entropy": 5.200405740737915,
      "epoch": 1.6292026897214216,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000473911238785334,
      "loss": 4.9277,
      "mean_token_accuracy": 0.215556238591671,
      "num_tokens": 38908439.0,
      "step": 16960
    },
    {
      "entropy": 5.1607146739959715,
      "epoch": 1.6296829971181557,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004738952126019496,
      "loss": 4.9186,
      "mean_token_accuracy": 0.21244364380836486,
      "num_tokens": 38920881.0,
      "step": 16965
    },
    {
      "entropy": 5.183220863342285,
      "epoch": 1.6301633045148896,
      "grad_norm": 1.125,
      "learning_rate": 0.0004738791818008089,
      "loss": 4.9401,
      "mean_token_accuracy": 0.22676119953393936,
      "num_tokens": 38932097.0,
      "step": 16970
    },
    {
      "entropy": 5.27858247756958,
      "epoch": 1.6306436119116234,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004738631463822841,
      "loss": 5.1211,
      "mean_token_accuracy": 0.2101389467716217,
      "num_tokens": 38945122.0,
      "step": 16975
    },
    {
      "entropy": 5.135886716842651,
      "epoch": 1.6311239193083573,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047384710634674766,
      "loss": 4.9744,
      "mean_token_accuracy": 0.22530607432126998,
      "num_tokens": 38956136.0,
      "step": 16980
    },
    {
      "entropy": 5.2493969917297365,
      "epoch": 1.6316042267050914,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047383106169457184,
      "loss": 4.9443,
      "mean_token_accuracy": 0.21589515805244447,
      "num_tokens": 38968650.0,
      "step": 16985
    },
    {
      "entropy": 5.1400614261627195,
      "epoch": 1.6320845341018253,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004738150124261292,
      "loss": 4.8874,
      "mean_token_accuracy": 0.22474647015333177,
      "num_tokens": 38979834.0,
      "step": 16990
    },
    {
      "entropy": 5.092259693145752,
      "epoch": 1.6325648414985592,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047379895854179226,
      "loss": 4.873,
      "mean_token_accuracy": 0.22775078266859056,
      "num_tokens": 38991548.0,
      "step": 16995
    },
    {
      "entropy": 5.178805875778198,
      "epoch": 1.633045148895293,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004737829000419338,
      "loss": 4.9371,
      "mean_token_accuracy": 0.21949717700481414,
      "num_tokens": 39003344.0,
      "step": 17000
    },
    {
      "entropy": 5.197120285034179,
      "epoch": 1.633525456292027,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047376683692692666,
      "loss": 4.9658,
      "mean_token_accuracy": 0.221602226793766,
      "num_tokens": 39015278.0,
      "step": 17005
    },
    {
      "entropy": 5.201696586608887,
      "epoch": 1.6340057636887608,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004737507691971439,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21613511592149734,
      "num_tokens": 39027634.0,
      "step": 17010
    },
    {
      "entropy": 5.2033178329467775,
      "epoch": 1.6344860710854947,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047373469685295833,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21693381518125535,
      "num_tokens": 39038226.0,
      "step": 17015
    },
    {
      "entropy": 5.218664073944092,
      "epoch": 1.6349663784822286,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047371861989474326,
      "loss": 5.0085,
      "mean_token_accuracy": 0.2104928031563759,
      "num_tokens": 39050627.0,
      "step": 17020
    },
    {
      "entropy": 5.221230792999267,
      "epoch": 1.6354466858789625,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004737025383228719,
      "loss": 4.9687,
      "mean_token_accuracy": 0.21166858822107315,
      "num_tokens": 39061734.0,
      "step": 17025
    },
    {
      "entropy": 5.14799394607544,
      "epoch": 1.6359269932756964,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047368645213771764,
      "loss": 5.0062,
      "mean_token_accuracy": 0.2070786565542221,
      "num_tokens": 39074288.0,
      "step": 17030
    },
    {
      "entropy": 5.1601263046264645,
      "epoch": 1.6364073006724302,
      "grad_norm": 1.140625,
      "learning_rate": 0.000473670361339654,
      "loss": 4.9984,
      "mean_token_accuracy": 0.20869170427322387,
      "num_tokens": 39084967.0,
      "step": 17035
    },
    {
      "entropy": 5.214772272109985,
      "epoch": 1.6368876080691641,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004736542659290544,
      "loss": 4.8992,
      "mean_token_accuracy": 0.226063272356987,
      "num_tokens": 39095093.0,
      "step": 17040
    },
    {
      "entropy": 5.213840675354004,
      "epoch": 1.6373679154658982,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004736381659062927,
      "loss": 4.9828,
      "mean_token_accuracy": 0.22120442390441894,
      "num_tokens": 39106952.0,
      "step": 17045
    },
    {
      "entropy": 5.127022695541382,
      "epoch": 1.637848222862632,
      "grad_norm": 1.125,
      "learning_rate": 0.00047362206127174255,
      "loss": 4.8654,
      "mean_token_accuracy": 0.22036115527153016,
      "num_tokens": 39117682.0,
      "step": 17050
    },
    {
      "entropy": 5.227935409545898,
      "epoch": 1.638328530259366,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047360595202577786,
      "loss": 5.0397,
      "mean_token_accuracy": 0.2137133926153183,
      "num_tokens": 39129767.0,
      "step": 17055
    },
    {
      "entropy": 5.1464762687683105,
      "epoch": 1.6388088376560999,
      "grad_norm": 1.125,
      "learning_rate": 0.00047358983816877284,
      "loss": 4.9398,
      "mean_token_accuracy": 0.22097924798727037,
      "num_tokens": 39140906.0,
      "step": 17060
    },
    {
      "entropy": 5.239005613327026,
      "epoch": 1.639289145052834,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004735737197011015,
      "loss": 4.8921,
      "mean_token_accuracy": 0.21217268854379653,
      "num_tokens": 39151884.0,
      "step": 17065
    },
    {
      "entropy": 5.0779197216033936,
      "epoch": 1.6397694524495678,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047355759662313793,
      "loss": 4.8985,
      "mean_token_accuracy": 0.2180098295211792,
      "num_tokens": 39162986.0,
      "step": 17070
    },
    {
      "entropy": 5.227844715118408,
      "epoch": 1.6402497598463017,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004735414689352566,
      "loss": 4.9758,
      "mean_token_accuracy": 0.22047170251607895,
      "num_tokens": 39174280.0,
      "step": 17075
    },
    {
      "entropy": 5.218602323532105,
      "epoch": 1.6407300672430356,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004735253366378318,
      "loss": 4.9692,
      "mean_token_accuracy": 0.2209726244211197,
      "num_tokens": 39185499.0,
      "step": 17080
    },
    {
      "entropy": 5.107168197631836,
      "epoch": 1.6412103746397695,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004735091997312383,
      "loss": 4.874,
      "mean_token_accuracy": 0.21981294602155685,
      "num_tokens": 39196322.0,
      "step": 17085
    },
    {
      "entropy": 5.170092058181763,
      "epoch": 1.6416906820365034,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047349305821585067,
      "loss": 4.9521,
      "mean_token_accuracy": 0.21500749588012696,
      "num_tokens": 39207332.0,
      "step": 17090
    },
    {
      "entropy": 5.2344396114349365,
      "epoch": 1.6421709894332372,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004734769120920435,
      "loss": 5.011,
      "mean_token_accuracy": 0.20959337502717973,
      "num_tokens": 39217305.0,
      "step": 17095
    },
    {
      "entropy": 5.186958837509155,
      "epoch": 1.6426512968299711,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004734607613601919,
      "loss": 4.9066,
      "mean_token_accuracy": 0.21929243355989456,
      "num_tokens": 39228895.0,
      "step": 17100
    },
    {
      "entropy": 5.16012716293335,
      "epoch": 1.643131604226705,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047344460602067077,
      "loss": 4.9734,
      "mean_token_accuracy": 0.212332022190094,
      "num_tokens": 39240859.0,
      "step": 17105
    },
    {
      "entropy": 5.12966890335083,
      "epoch": 1.643611911623439,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000473428446073855,
      "loss": 4.8382,
      "mean_token_accuracy": 0.2324831336736679,
      "num_tokens": 39251556.0,
      "step": 17110
    },
    {
      "entropy": 5.122574043273926,
      "epoch": 1.6440922190201728,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047341228152012003,
      "loss": 4.8207,
      "mean_token_accuracy": 0.2206296965479851,
      "num_tokens": 39263919.0,
      "step": 17115
    },
    {
      "entropy": 5.2395096778869625,
      "epoch": 1.6445725264169067,
      "grad_norm": 1.171875,
      "learning_rate": 0.000473396112359841,
      "loss": 5.1162,
      "mean_token_accuracy": 0.21277490854263306,
      "num_tokens": 39276873.0,
      "step": 17120
    },
    {
      "entropy": 5.1469810009002686,
      "epoch": 1.6450528338136408,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047337993859339334,
      "loss": 4.8605,
      "mean_token_accuracy": 0.2282344475388527,
      "num_tokens": 39288912.0,
      "step": 17125
    },
    {
      "entropy": 5.128656625747681,
      "epoch": 1.6455331412103746,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047336376022115255,
      "loss": 4.9098,
      "mean_token_accuracy": 0.21901513040065765,
      "num_tokens": 39301115.0,
      "step": 17130
    },
    {
      "entropy": 5.190395832061768,
      "epoch": 1.6460134486071085,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047334757724349437,
      "loss": 5.0093,
      "mean_token_accuracy": 0.21030217707157134,
      "num_tokens": 39313535.0,
      "step": 17135
    },
    {
      "entropy": 5.2210915088653564,
      "epoch": 1.6464937560038426,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004733313896607943,
      "loss": 4.9789,
      "mean_token_accuracy": 0.21241505295038224,
      "num_tokens": 39325289.0,
      "step": 17140
    },
    {
      "entropy": 5.208249616622925,
      "epoch": 1.6469740634005765,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004733151974734284,
      "loss": 4.908,
      "mean_token_accuracy": 0.22152280658483506,
      "num_tokens": 39337382.0,
      "step": 17145
    },
    {
      "entropy": 5.137962436676025,
      "epoch": 1.6474543707973104,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047329900068177245,
      "loss": 4.9075,
      "mean_token_accuracy": 0.2145901009440422,
      "num_tokens": 39348548.0,
      "step": 17150
    },
    {
      "entropy": 5.157047128677368,
      "epoch": 1.6479346781940443,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047328279928620244,
      "loss": 4.9333,
      "mean_token_accuracy": 0.21921810358762742,
      "num_tokens": 39359788.0,
      "step": 17155
    },
    {
      "entropy": 5.163013887405396,
      "epoch": 1.6484149855907781,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004732665932870947,
      "loss": 4.9216,
      "mean_token_accuracy": 0.22413923889398574,
      "num_tokens": 39370533.0,
      "step": 17160
    },
    {
      "entropy": 5.175518369674682,
      "epoch": 1.648895292987512,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047325038268482544,
      "loss": 4.9606,
      "mean_token_accuracy": 0.21602587401866913,
      "num_tokens": 39381157.0,
      "step": 17165
    },
    {
      "entropy": 5.231780481338501,
      "epoch": 1.649375600384246,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004732341674797709,
      "loss": 4.9338,
      "mean_token_accuracy": 0.21494247913360595,
      "num_tokens": 39391280.0,
      "step": 17170
    },
    {
      "entropy": 5.06181001663208,
      "epoch": 1.6498559077809798,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047321794767230766,
      "loss": 4.8278,
      "mean_token_accuracy": 0.2198556125164032,
      "num_tokens": 39402097.0,
      "step": 17175
    },
    {
      "entropy": 5.19902925491333,
      "epoch": 1.6503362151777137,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047320172326281224,
      "loss": 5.0401,
      "mean_token_accuracy": 0.21755478233098985,
      "num_tokens": 39414590.0,
      "step": 17180
    },
    {
      "entropy": 5.185191679000854,
      "epoch": 1.6508165225744476,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047318549425166134,
      "loss": 4.9068,
      "mean_token_accuracy": 0.21633596420288087,
      "num_tokens": 39425933.0,
      "step": 17185
    },
    {
      "entropy": 5.154320430755615,
      "epoch": 1.6512968299711814,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004731692606392318,
      "loss": 4.9343,
      "mean_token_accuracy": 0.22334263175725938,
      "num_tokens": 39437059.0,
      "step": 17190
    },
    {
      "entropy": 5.2015008449554445,
      "epoch": 1.6517771373679153,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004731530224259004,
      "loss": 4.9897,
      "mean_token_accuracy": 0.21624634116888047,
      "num_tokens": 39449582.0,
      "step": 17195
    },
    {
      "entropy": 5.277575874328614,
      "epoch": 1.6522574447646494,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004731367796120442,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21190531849861144,
      "num_tokens": 39461147.0,
      "step": 17200
    },
    {
      "entropy": 5.215583848953247,
      "epoch": 1.6527377521613833,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004731205321980404,
      "loss": 4.9746,
      "mean_token_accuracy": 0.2155717730522156,
      "num_tokens": 39471881.0,
      "step": 17205
    },
    {
      "entropy": 5.137748193740845,
      "epoch": 1.6532180595581172,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047310428018426616,
      "loss": 4.9991,
      "mean_token_accuracy": 0.21469996720552445,
      "num_tokens": 39483632.0,
      "step": 17210
    },
    {
      "entropy": 5.213284111022949,
      "epoch": 1.653698366954851,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004730880235710987,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21498021632432937,
      "num_tokens": 39495587.0,
      "step": 17215
    },
    {
      "entropy": 5.180560970306397,
      "epoch": 1.6541786743515852,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004730717623589155,
      "loss": 4.9529,
      "mean_token_accuracy": 0.2163096159696579,
      "num_tokens": 39505503.0,
      "step": 17220
    },
    {
      "entropy": 5.194420623779297,
      "epoch": 1.654658981748319,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004730554965480942,
      "loss": 4.9335,
      "mean_token_accuracy": 0.2232720896601677,
      "num_tokens": 39516681.0,
      "step": 17225
    },
    {
      "entropy": 5.142379474639893,
      "epoch": 1.655139289145053,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004730392261390124,
      "loss": 4.8876,
      "mean_token_accuracy": 0.22180719971656798,
      "num_tokens": 39528451.0,
      "step": 17230
    },
    {
      "entropy": 5.174005651473999,
      "epoch": 1.6556195965417868,
      "grad_norm": 1.25,
      "learning_rate": 0.0004730229511320478,
      "loss": 4.9715,
      "mean_token_accuracy": 0.21471776366233825,
      "num_tokens": 39540306.0,
      "step": 17235
    },
    {
      "entropy": 5.208211851119995,
      "epoch": 1.6560999039385207,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047300667152757827,
      "loss": 4.8616,
      "mean_token_accuracy": 0.2205296650528908,
      "num_tokens": 39551260.0,
      "step": 17240
    },
    {
      "entropy": 5.235602188110351,
      "epoch": 1.6565802113352546,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047299038732598184,
      "loss": 5.0131,
      "mean_token_accuracy": 0.20846610367298127,
      "num_tokens": 39562991.0,
      "step": 17245
    },
    {
      "entropy": 5.133588790893555,
      "epoch": 1.6570605187319885,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047297409852763644,
      "loss": 4.8776,
      "mean_token_accuracy": 0.2240191653370857,
      "num_tokens": 39574629.0,
      "step": 17250
    },
    {
      "entropy": 5.11315336227417,
      "epoch": 1.6575408261287223,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004729578051329204,
      "loss": 5.0093,
      "mean_token_accuracy": 0.21511300653219223,
      "num_tokens": 39585599.0,
      "step": 17255
    },
    {
      "entropy": 5.172319173812866,
      "epoch": 1.6580211335254562,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047294150714221185,
      "loss": 4.9567,
      "mean_token_accuracy": 0.21769025921821594,
      "num_tokens": 39597225.0,
      "step": 17260
    },
    {
      "entropy": 5.2303026676177975,
      "epoch": 1.65850144092219,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004729252045558894,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21479110270738602,
      "num_tokens": 39609296.0,
      "step": 17265
    },
    {
      "entropy": 5.104281997680664,
      "epoch": 1.658981748318924,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047290889737433133,
      "loss": 4.7992,
      "mean_token_accuracy": 0.226961749792099,
      "num_tokens": 39619369.0,
      "step": 17270
    },
    {
      "entropy": 5.130922079086304,
      "epoch": 1.6594620557156579,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047289258559791633,
      "loss": 4.8816,
      "mean_token_accuracy": 0.2231891483068466,
      "num_tokens": 39630438.0,
      "step": 17275
    },
    {
      "entropy": 5.190590143203735,
      "epoch": 1.659942363112392,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047287626922702317,
      "loss": 4.9523,
      "mean_token_accuracy": 0.2177107259631157,
      "num_tokens": 39643339.0,
      "step": 17280
    },
    {
      "entropy": 5.150815439224243,
      "epoch": 1.6604226705091258,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047285994826203054,
      "loss": 4.9342,
      "mean_token_accuracy": 0.22329081147909163,
      "num_tokens": 39654495.0,
      "step": 17285
    },
    {
      "entropy": 5.157508802413941,
      "epoch": 1.6609029779058597,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004728436227033175,
      "loss": 4.9131,
      "mean_token_accuracy": 0.22402867674827576,
      "num_tokens": 39665870.0,
      "step": 17290
    },
    {
      "entropy": 5.144074010848999,
      "epoch": 1.6613832853025938,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047282729255126294,
      "loss": 4.8765,
      "mean_token_accuracy": 0.22250173091888428,
      "num_tokens": 39676761.0,
      "step": 17295
    },
    {
      "entropy": 5.196831130981446,
      "epoch": 1.6618635926993277,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004728109578062461,
      "loss": 4.9242,
      "mean_token_accuracy": 0.2150837540626526,
      "num_tokens": 39687515.0,
      "step": 17300
    },
    {
      "entropy": 5.131394958496093,
      "epoch": 1.6623439000960616,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047279461846864626,
      "loss": 4.9038,
      "mean_token_accuracy": 0.2260887786746025,
      "num_tokens": 39699599.0,
      "step": 17305
    },
    {
      "entropy": 5.146971082687378,
      "epoch": 1.6628242074927955,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047277827453884265,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22930624932050706,
      "num_tokens": 39710122.0,
      "step": 17310
    },
    {
      "entropy": 5.214857578277588,
      "epoch": 1.6633045148895294,
      "grad_norm": 1.125,
      "learning_rate": 0.00047276192601721477,
      "loss": 5.0037,
      "mean_token_accuracy": 0.21599909663200378,
      "num_tokens": 39722884.0,
      "step": 17315
    },
    {
      "entropy": 5.1511125564575195,
      "epoch": 1.6637848222862632,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004727455729041422,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2238711640238762,
      "num_tokens": 39734464.0,
      "step": 17320
    },
    {
      "entropy": 5.167579317092896,
      "epoch": 1.6642651296829971,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047272921520000465,
      "loss": 4.9403,
      "mean_token_accuracy": 0.21940008401870728,
      "num_tokens": 39745948.0,
      "step": 17325
    },
    {
      "entropy": 5.168394899368286,
      "epoch": 1.664745437079731,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004727128529051819,
      "loss": 4.9331,
      "mean_token_accuracy": 0.21809136718511582,
      "num_tokens": 39757024.0,
      "step": 17330
    },
    {
      "entropy": 5.11815447807312,
      "epoch": 1.6652257444764649,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004726964860200537,
      "loss": 4.8674,
      "mean_token_accuracy": 0.22152062505483627,
      "num_tokens": 39767205.0,
      "step": 17335
    },
    {
      "entropy": 5.207518625259399,
      "epoch": 1.6657060518731988,
      "grad_norm": 1.25,
      "learning_rate": 0.0004726801145450002,
      "loss": 4.9499,
      "mean_token_accuracy": 0.21363518089056016,
      "num_tokens": 39779543.0,
      "step": 17340
    },
    {
      "entropy": 5.121877574920655,
      "epoch": 1.6661863592699326,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004726637384804014,
      "loss": 4.8591,
      "mean_token_accuracy": 0.22580553889274596,
      "num_tokens": 39790827.0,
      "step": 17345
    },
    {
      "entropy": 5.0485070705413815,
      "epoch": 1.6666666666666665,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004726473578266375,
      "loss": 4.8101,
      "mean_token_accuracy": 0.2236025184392929,
      "num_tokens": 39801995.0,
      "step": 17350
    },
    {
      "entropy": 5.151128673553467,
      "epoch": 1.6671469740634006,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047263097258408893,
      "loss": 4.9364,
      "mean_token_accuracy": 0.22019630521535874,
      "num_tokens": 39813319.0,
      "step": 17355
    },
    {
      "entropy": 5.193193864822388,
      "epoch": 1.6676272814601345,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004726145827531359,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2251357078552246,
      "num_tokens": 39824178.0,
      "step": 17360
    },
    {
      "entropy": 5.088051462173462,
      "epoch": 1.6681075888568684,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047259818833415916,
      "loss": 4.8813,
      "mean_token_accuracy": 0.2182392492890358,
      "num_tokens": 39835452.0,
      "step": 17365
    },
    {
      "entropy": 5.081688070297242,
      "epoch": 1.6685878962536023,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047258178932753917,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21993949711322786,
      "num_tokens": 39848083.0,
      "step": 17370
    },
    {
      "entropy": 5.255282211303711,
      "epoch": 1.6690682036503364,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047256538573365675,
      "loss": 4.9413,
      "mean_token_accuracy": 0.21243193745613098,
      "num_tokens": 39859132.0,
      "step": 17375
    },
    {
      "entropy": 5.183927822113037,
      "epoch": 1.6695485110470702,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004725489775528928,
      "loss": 4.9125,
      "mean_token_accuracy": 0.21751776188611985,
      "num_tokens": 39870237.0,
      "step": 17380
    },
    {
      "entropy": 5.107290601730346,
      "epoch": 1.6700288184438041,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047253256478562805,
      "loss": 4.8966,
      "mean_token_accuracy": 0.22015711069107055,
      "num_tokens": 39883009.0,
      "step": 17385
    },
    {
      "entropy": 5.14463324546814,
      "epoch": 1.670509125840538,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047251614743224374,
      "loss": 4.9348,
      "mean_token_accuracy": 0.21654152870178223,
      "num_tokens": 39894444.0,
      "step": 17390
    },
    {
      "entropy": 5.2430681705474855,
      "epoch": 1.670989433237272,
      "grad_norm": 1.25,
      "learning_rate": 0.00047249972549312107,
      "loss": 4.9759,
      "mean_token_accuracy": 0.2145759254693985,
      "num_tokens": 39905312.0,
      "step": 17395
    },
    {
      "entropy": 5.2056762218475345,
      "epoch": 1.6714697406340058,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004724832989686411,
      "loss": 4.9727,
      "mean_token_accuracy": 0.21948964148759842,
      "num_tokens": 39916724.0,
      "step": 17400
    },
    {
      "entropy": 5.209683036804199,
      "epoch": 1.6719500480307397,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047246686785918545,
      "loss": 4.9588,
      "mean_token_accuracy": 0.22007214426994323,
      "num_tokens": 39928919.0,
      "step": 17405
    },
    {
      "entropy": 5.210084724426269,
      "epoch": 1.6724303554274735,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047245043216513546,
      "loss": 5.056,
      "mean_token_accuracy": 0.20960242301225662,
      "num_tokens": 39940301.0,
      "step": 17410
    },
    {
      "entropy": 5.142881965637207,
      "epoch": 1.6729106628242074,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004724339918868727,
      "loss": 4.8792,
      "mean_token_accuracy": 0.22467657029628754,
      "num_tokens": 39951269.0,
      "step": 17415
    },
    {
      "entropy": 5.246879768371582,
      "epoch": 1.6733909702209413,
      "grad_norm": 1.125,
      "learning_rate": 0.000472417547024779,
      "loss": 4.9911,
      "mean_token_accuracy": 0.21876149177551268,
      "num_tokens": 39963057.0,
      "step": 17420
    },
    {
      "entropy": 5.297744655609131,
      "epoch": 1.6738712776176752,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047240109757923593,
      "loss": 5.0319,
      "mean_token_accuracy": 0.21296219229698182,
      "num_tokens": 39973876.0,
      "step": 17425
    },
    {
      "entropy": 5.234021139144898,
      "epoch": 1.674351585014409,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004723846435506256,
      "loss": 4.9801,
      "mean_token_accuracy": 0.20887034833431245,
      "num_tokens": 39984477.0,
      "step": 17430
    },
    {
      "entropy": 5.136983108520508,
      "epoch": 1.6748318924111432,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047236818493932994,
      "loss": 4.9322,
      "mean_token_accuracy": 0.21904544830322265,
      "num_tokens": 39996266.0,
      "step": 17435
    },
    {
      "entropy": 5.198496341705322,
      "epoch": 1.675312199807877,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004723517217457311,
      "loss": 4.9879,
      "mean_token_accuracy": 0.21927057355642318,
      "num_tokens": 40008618.0,
      "step": 17440
    },
    {
      "entropy": 5.223957967758179,
      "epoch": 1.675792507204611,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004723352539702113,
      "loss": 4.8867,
      "mean_token_accuracy": 0.21995180249214172,
      "num_tokens": 40020290.0,
      "step": 17445
    },
    {
      "entropy": 5.143985843658447,
      "epoch": 1.6762728146013448,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004723187816131529,
      "loss": 4.8337,
      "mean_token_accuracy": 0.22527581155300141,
      "num_tokens": 40031167.0,
      "step": 17450
    },
    {
      "entropy": 5.100689220428467,
      "epoch": 1.676753121998079,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004723023046749383,
      "loss": 4.9422,
      "mean_token_accuracy": 0.2256269261240959,
      "num_tokens": 40043278.0,
      "step": 17455
    },
    {
      "entropy": 5.174584102630615,
      "epoch": 1.6772334293948128,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047228582315595,
      "loss": 4.9083,
      "mean_token_accuracy": 0.22240625470876693,
      "num_tokens": 40055022.0,
      "step": 17460
    },
    {
      "entropy": 5.1908957958221436,
      "epoch": 1.6777137367915467,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004722693370565708,
      "loss": 4.9901,
      "mean_token_accuracy": 0.219715815782547,
      "num_tokens": 40066556.0,
      "step": 17465
    },
    {
      "entropy": 5.301260280609131,
      "epoch": 1.6781940441882806,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047225284637718323,
      "loss": 5.0555,
      "mean_token_accuracy": 0.20598280131816865,
      "num_tokens": 40078015.0,
      "step": 17470
    },
    {
      "entropy": 5.116975736618042,
      "epoch": 1.6786743515850144,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004722363511181703,
      "loss": 4.9238,
      "mean_token_accuracy": 0.21710177659988403,
      "num_tokens": 40090061.0,
      "step": 17475
    },
    {
      "entropy": 5.234718608856201,
      "epoch": 1.6791546589817483,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000472219851279915,
      "loss": 4.9874,
      "mean_token_accuracy": 0.2144838660955429,
      "num_tokens": 40102569.0,
      "step": 17480
    },
    {
      "entropy": 5.2285964012146,
      "epoch": 1.6796349663784822,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004722033468628004,
      "loss": 4.9319,
      "mean_token_accuracy": 0.2161531552672386,
      "num_tokens": 40114883.0,
      "step": 17485
    },
    {
      "entropy": 5.08643593788147,
      "epoch": 1.680115273775216,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004721868378672098,
      "loss": 4.8533,
      "mean_token_accuracy": 0.22618663609027861,
      "num_tokens": 40126337.0,
      "step": 17490
    },
    {
      "entropy": 5.157539749145508,
      "epoch": 1.68059558117195,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004721703242935261,
      "loss": 4.8984,
      "mean_token_accuracy": 0.2228461652994156,
      "num_tokens": 40137438.0,
      "step": 17495
    },
    {
      "entropy": 5.191547393798828,
      "epoch": 1.6810758885686838,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004721538061421331,
      "loss": 4.9162,
      "mean_token_accuracy": 0.21829658895730972,
      "num_tokens": 40148061.0,
      "step": 17500
    },
    {
      "entropy": 5.207766056060791,
      "epoch": 1.6815561959654177,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047213728341341407,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21853085309267045,
      "num_tokens": 40159118.0,
      "step": 17505
    },
    {
      "entropy": 5.139745998382568,
      "epoch": 1.6820365033621518,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004721207561077527,
      "loss": 4.8821,
      "mean_token_accuracy": 0.21980861723423004,
      "num_tokens": 40170027.0,
      "step": 17510
    },
    {
      "entropy": 5.075328588485718,
      "epoch": 1.6825168107588857,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004721042242255327,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21771486848592758,
      "num_tokens": 40181819.0,
      "step": 17515
    },
    {
      "entropy": 5.217421913146973,
      "epoch": 1.6829971181556196,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047208768776713805,
      "loss": 4.8515,
      "mean_token_accuracy": 0.22088800370693207,
      "num_tokens": 40191870.0,
      "step": 17520
    },
    {
      "entropy": 5.168523836135864,
      "epoch": 1.6834774255523535,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004720711467329523,
      "loss": 4.9202,
      "mean_token_accuracy": 0.21939578503370286,
      "num_tokens": 40203191.0,
      "step": 17525
    },
    {
      "entropy": 5.2416441440582275,
      "epoch": 1.6839577329490876,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004720546011233599,
      "loss": 5.0535,
      "mean_token_accuracy": 0.20682528018951415,
      "num_tokens": 40214217.0,
      "step": 17530
    },
    {
      "entropy": 5.170129060745239,
      "epoch": 1.6844380403458215,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004720380509387446,
      "loss": 4.961,
      "mean_token_accuracy": 0.21959460228681565,
      "num_tokens": 40225711.0,
      "step": 17535
    },
    {
      "entropy": 5.263588285446167,
      "epoch": 1.6849183477425553,
      "grad_norm": 1.234375,
      "learning_rate": 0.000472021496179491,
      "loss": 5.0621,
      "mean_token_accuracy": 0.21372593492269515,
      "num_tokens": 40235748.0,
      "step": 17540
    },
    {
      "entropy": 5.2267053604125975,
      "epoch": 1.6853986551392892,
      "grad_norm": 2.46875,
      "learning_rate": 0.00047200493684598316,
      "loss": 5.0048,
      "mean_token_accuracy": 0.20580882877111434,
      "num_tokens": 40246090.0,
      "step": 17545
    },
    {
      "entropy": 5.217617464065552,
      "epoch": 1.685878962536023,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047198837293860573,
      "loss": 4.9924,
      "mean_token_accuracy": 0.21832366436719894,
      "num_tokens": 40257709.0,
      "step": 17550
    },
    {
      "entropy": 5.203021669387818,
      "epoch": 1.686359269932757,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004719718044577432,
      "loss": 4.95,
      "mean_token_accuracy": 0.21834967136383057,
      "num_tokens": 40268670.0,
      "step": 17555
    },
    {
      "entropy": 5.144425535202027,
      "epoch": 1.6868395773294909,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047195523140378034,
      "loss": 4.8467,
      "mean_token_accuracy": 0.22830918431282043,
      "num_tokens": 40279289.0,
      "step": 17560
    },
    {
      "entropy": 5.219867277145386,
      "epoch": 1.6873198847262247,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047193865377710177,
      "loss": 5.0711,
      "mean_token_accuracy": 0.21139773428440095,
      "num_tokens": 40289953.0,
      "step": 17565
    },
    {
      "entropy": 5.299340629577637,
      "epoch": 1.6878001921229586,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047192207157809246,
      "loss": 4.9904,
      "mean_token_accuracy": 0.21408141404390335,
      "num_tokens": 40302180.0,
      "step": 17570
    },
    {
      "entropy": 5.145879220962525,
      "epoch": 1.6882804995196925,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047190548480713736,
      "loss": 4.861,
      "mean_token_accuracy": 0.224839748442173,
      "num_tokens": 40312556.0,
      "step": 17575
    },
    {
      "entropy": 5.119893217086792,
      "epoch": 1.6887608069164264,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047188889346462163,
      "loss": 4.9627,
      "mean_token_accuracy": 0.21624465733766557,
      "num_tokens": 40323213.0,
      "step": 17580
    },
    {
      "entropy": 5.198696947097778,
      "epoch": 1.6892411143131603,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047187229755093037,
      "loss": 5.0032,
      "mean_token_accuracy": 0.2118437558412552,
      "num_tokens": 40334567.0,
      "step": 17585
    },
    {
      "entropy": 5.284020042419433,
      "epoch": 1.6897214217098944,
      "grad_norm": 1.484375,
      "learning_rate": 0.000471855697066449,
      "loss": 4.9719,
      "mean_token_accuracy": 0.21612982749938964,
      "num_tokens": 40345606.0,
      "step": 17590
    },
    {
      "entropy": 5.20478892326355,
      "epoch": 1.6902017291066282,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047183909201156297,
      "loss": 4.9586,
      "mean_token_accuracy": 0.21463811546564102,
      "num_tokens": 40356366.0,
      "step": 17595
    },
    {
      "entropy": 5.065234518051147,
      "epoch": 1.6906820365033621,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004718224823866576,
      "loss": 4.8628,
      "mean_token_accuracy": 0.22374972105026245,
      "num_tokens": 40367451.0,
      "step": 17600
    },
    {
      "entropy": 5.071690702438355,
      "epoch": 1.691162343900096,
      "grad_norm": 1.25,
      "learning_rate": 0.0004718058681921186,
      "loss": 4.8161,
      "mean_token_accuracy": 0.2273660957813263,
      "num_tokens": 40379377.0,
      "step": 17605
    },
    {
      "entropy": 5.20514407157898,
      "epoch": 1.6916426512968301,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047178924942833185,
      "loss": 4.9524,
      "mean_token_accuracy": 0.2166668578982353,
      "num_tokens": 40390615.0,
      "step": 17610
    },
    {
      "entropy": 5.211584186553955,
      "epoch": 1.692122958693564,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004717726260956831,
      "loss": 4.9895,
      "mean_token_accuracy": 0.21267928630113603,
      "num_tokens": 40402475.0,
      "step": 17615
    },
    {
      "entropy": 5.215222120285034,
      "epoch": 1.6926032660902979,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004717559981945581,
      "loss": 4.9457,
      "mean_token_accuracy": 0.21044566184282304,
      "num_tokens": 40414311.0,
      "step": 17620
    },
    {
      "entropy": 5.133531761169434,
      "epoch": 1.6930835734870318,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004717393657253432,
      "loss": 4.8535,
      "mean_token_accuracy": 0.22564322054386138,
      "num_tokens": 40425859.0,
      "step": 17625
    },
    {
      "entropy": 5.257960176467895,
      "epoch": 1.6935638808837656,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004717227286884243,
      "loss": 4.9873,
      "mean_token_accuracy": 0.21294575184583664,
      "num_tokens": 40436679.0,
      "step": 17630
    },
    {
      "entropy": 5.263258695602417,
      "epoch": 1.6940441882804995,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004717060870841879,
      "loss": 5.0298,
      "mean_token_accuracy": 0.21023935526609422,
      "num_tokens": 40446985.0,
      "step": 17635
    },
    {
      "entropy": 5.150032043457031,
      "epoch": 1.6945244956772334,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004716894409130202,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22042877674102784,
      "num_tokens": 40458777.0,
      "step": 17640
    },
    {
      "entropy": 5.244808864593506,
      "epoch": 1.6950048030739673,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004716727901753078,
      "loss": 5.0138,
      "mean_token_accuracy": 0.20943745374679565,
      "num_tokens": 40469493.0,
      "step": 17645
    },
    {
      "entropy": 5.202561950683593,
      "epoch": 1.6954851104707012,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004716561348714371,
      "loss": 5.0121,
      "mean_token_accuracy": 0.22005030512809753,
      "num_tokens": 40480889.0,
      "step": 17650
    },
    {
      "entropy": 5.161393690109253,
      "epoch": 1.695965417867435,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047163947500179494,
      "loss": 4.8345,
      "mean_token_accuracy": 0.21635822057724,
      "num_tokens": 40491630.0,
      "step": 17655
    },
    {
      "entropy": 5.314888334274292,
      "epoch": 1.696445725264169,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004716228105667681,
      "loss": 5.0903,
      "mean_token_accuracy": 0.1983462765812874,
      "num_tokens": 40503204.0,
      "step": 17660
    },
    {
      "entropy": 5.164434957504272,
      "epoch": 1.696926032660903,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004716061415667435,
      "loss": 4.935,
      "mean_token_accuracy": 0.21733080297708512,
      "num_tokens": 40513917.0,
      "step": 17665
    },
    {
      "entropy": 5.195461082458496,
      "epoch": 1.697406340057637,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000471589468002108,
      "loss": 5.0176,
      "mean_token_accuracy": 0.21355461776256562,
      "num_tokens": 40526501.0,
      "step": 17670
    },
    {
      "entropy": 5.115848875045776,
      "epoch": 1.6978866474543708,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004715727898732488,
      "loss": 4.8685,
      "mean_token_accuracy": 0.2183568984270096,
      "num_tokens": 40536641.0,
      "step": 17675
    },
    {
      "entropy": 5.151384925842285,
      "epoch": 1.6983669548511047,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047155610718055315,
      "loss": 4.9394,
      "mean_token_accuracy": 0.22387948185205458,
      "num_tokens": 40547497.0,
      "step": 17680
    },
    {
      "entropy": 5.180646371841431,
      "epoch": 1.6988472622478388,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047153941992440833,
      "loss": 4.8991,
      "mean_token_accuracy": 0.21631496250629426,
      "num_tokens": 40559212.0,
      "step": 17685
    },
    {
      "entropy": 5.096219110488891,
      "epoch": 1.6993275696445727,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004715227281052018,
      "loss": 4.8574,
      "mean_token_accuracy": 0.21974581778049468,
      "num_tokens": 40571692.0,
      "step": 17690
    },
    {
      "entropy": 5.156823396682739,
      "epoch": 1.6998078770413065,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004715060317233211,
      "loss": 4.9393,
      "mean_token_accuracy": 0.21674936413764953,
      "num_tokens": 40583751.0,
      "step": 17695
    },
    {
      "entropy": 5.281472826004029,
      "epoch": 1.7002881844380404,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004714893307791538,
      "loss": 4.9267,
      "mean_token_accuracy": 0.2216561868786812,
      "num_tokens": 40594782.0,
      "step": 17700
    },
    {
      "entropy": 5.180476665496826,
      "epoch": 1.7007684918347743,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00047147262527308766,
      "loss": 4.9345,
      "mean_token_accuracy": 0.2192664310336113,
      "num_tokens": 40604583.0,
      "step": 17705
    },
    {
      "entropy": 5.134631013870239,
      "epoch": 1.7012487992315082,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004714559152055106,
      "loss": 4.9706,
      "mean_token_accuracy": 0.209435398876667,
      "num_tokens": 40617957.0,
      "step": 17710
    },
    {
      "entropy": 5.252760362625122,
      "epoch": 1.701729106628242,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004714392005768106,
      "loss": 4.9233,
      "mean_token_accuracy": 0.21750059574842454,
      "num_tokens": 40628016.0,
      "step": 17715
    },
    {
      "entropy": 5.277926778793335,
      "epoch": 1.702209414024976,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004714224813873756,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20990212857723237,
      "num_tokens": 40639339.0,
      "step": 17720
    },
    {
      "entropy": 5.200560426712036,
      "epoch": 1.7026897214217098,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047140575763759393,
      "loss": 5.0191,
      "mean_token_accuracy": 0.21855507791042328,
      "num_tokens": 40651634.0,
      "step": 17725
    },
    {
      "entropy": 5.217713785171509,
      "epoch": 1.7031700288184437,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047138902932785363,
      "loss": 4.979,
      "mean_token_accuracy": 0.21410482078790666,
      "num_tokens": 40662742.0,
      "step": 17730
    },
    {
      "entropy": 5.21391658782959,
      "epoch": 1.7036503362151776,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047137229645854333,
      "loss": 4.9924,
      "mean_token_accuracy": 0.21396541595458984,
      "num_tokens": 40673586.0,
      "step": 17735
    },
    {
      "entropy": 5.152993297576904,
      "epoch": 1.7041306436119115,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004713555590300513,
      "loss": 4.8341,
      "mean_token_accuracy": 0.22104510217905043,
      "num_tokens": 40684238.0,
      "step": 17740
    },
    {
      "entropy": 5.216054058074951,
      "epoch": 1.7046109510086456,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004713388170427664,
      "loss": 5.0667,
      "mean_token_accuracy": 0.2095320776104927,
      "num_tokens": 40696154.0,
      "step": 17745
    },
    {
      "entropy": 5.183490943908692,
      "epoch": 1.7050912584053795,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004713220704970771,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21786237210035325,
      "num_tokens": 40708328.0,
      "step": 17750
    },
    {
      "entropy": 5.266184234619141,
      "epoch": 1.7055715658021133,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047130531939337236,
      "loss": 5.049,
      "mean_token_accuracy": 0.21456747949123384,
      "num_tokens": 40719067.0,
      "step": 17755
    },
    {
      "entropy": 5.084443759918213,
      "epoch": 1.7060518731988472,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047128856373204086,
      "loss": 4.7517,
      "mean_token_accuracy": 0.23450069278478622,
      "num_tokens": 40730638.0,
      "step": 17760
    },
    {
      "entropy": 5.230572271347046,
      "epoch": 1.7065321805955813,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047127180351347184,
      "loss": 4.9287,
      "mean_token_accuracy": 0.2171478882431984,
      "num_tokens": 40743245.0,
      "step": 17765
    },
    {
      "entropy": 5.204093408584595,
      "epoch": 1.7070124879923152,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004712550387380544,
      "loss": 4.9326,
      "mean_token_accuracy": 0.22030990123748778,
      "num_tokens": 40753769.0,
      "step": 17770
    },
    {
      "entropy": 5.099897384643555,
      "epoch": 1.707492795389049,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004712382694061776,
      "loss": 4.9509,
      "mean_token_accuracy": 0.22105333656072618,
      "num_tokens": 40765470.0,
      "step": 17775
    },
    {
      "entropy": 5.155581617355347,
      "epoch": 1.707973102785783,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047122149551823096,
      "loss": 4.9681,
      "mean_token_accuracy": 0.21430449783802033,
      "num_tokens": 40776427.0,
      "step": 17780
    },
    {
      "entropy": 5.182538223266602,
      "epoch": 1.7084534101825168,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004712047170746039,
      "loss": 4.9103,
      "mean_token_accuracy": 0.21651101112365723,
      "num_tokens": 40788707.0,
      "step": 17785
    },
    {
      "entropy": 5.285484361648559,
      "epoch": 1.7089337175792507,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047118793407568586,
      "loss": 5.0605,
      "mean_token_accuracy": 0.2113065540790558,
      "num_tokens": 40800092.0,
      "step": 17790
    },
    {
      "entropy": 5.133542919158936,
      "epoch": 1.7094140249759846,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047117114652186657,
      "loss": 4.8634,
      "mean_token_accuracy": 0.22110250741243362,
      "num_tokens": 40811389.0,
      "step": 17795
    },
    {
      "entropy": 5.184970092773438,
      "epoch": 1.7098943323727185,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047115435441353573,
      "loss": 4.9307,
      "mean_token_accuracy": 0.21759785413742067,
      "num_tokens": 40821438.0,
      "step": 17800
    },
    {
      "entropy": 5.137134313583374,
      "epoch": 1.7103746397694524,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047113755775108333,
      "loss": 4.8881,
      "mean_token_accuracy": 0.21693666279315948,
      "num_tokens": 40833335.0,
      "step": 17805
    },
    {
      "entropy": 5.2518003463745115,
      "epoch": 1.7108549471661862,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047112075653489913,
      "loss": 5.0466,
      "mean_token_accuracy": 0.21543618589639663,
      "num_tokens": 40844090.0,
      "step": 17810
    },
    {
      "entropy": 5.181757497787475,
      "epoch": 1.7113352545629201,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004711039507653734,
      "loss": 4.9079,
      "mean_token_accuracy": 0.214246067404747,
      "num_tokens": 40855176.0,
      "step": 17815
    },
    {
      "entropy": 5.182609748840332,
      "epoch": 1.7118155619596542,
      "grad_norm": 1.375,
      "learning_rate": 0.0004710871404428961,
      "loss": 5.0407,
      "mean_token_accuracy": 0.214334174990654,
      "num_tokens": 40866190.0,
      "step": 17820
    },
    {
      "entropy": 5.161588287353515,
      "epoch": 1.7122958693563881,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047107032556785786,
      "loss": 4.9115,
      "mean_token_accuracy": 0.21330410838127137,
      "num_tokens": 40876273.0,
      "step": 17825
    },
    {
      "entropy": 5.150371789932251,
      "epoch": 1.712776176753122,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047105350614064874,
      "loss": 4.8145,
      "mean_token_accuracy": 0.2292025625705719,
      "num_tokens": 40887843.0,
      "step": 17830
    },
    {
      "entropy": 5.171097373962402,
      "epoch": 1.7132564841498559,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047103668216165944,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21923250555992127,
      "num_tokens": 40898648.0,
      "step": 17835
    },
    {
      "entropy": 5.089150714874267,
      "epoch": 1.71373679154659,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047101985363128045,
      "loss": 4.9019,
      "mean_token_accuracy": 0.22062649130821227,
      "num_tokens": 40910817.0,
      "step": 17840
    },
    {
      "entropy": 5.214507579803467,
      "epoch": 1.7142170989433239,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047100302054990255,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21721866130828857,
      "num_tokens": 40922520.0,
      "step": 17845
    },
    {
      "entropy": 5.260058403015137,
      "epoch": 1.7146974063400577,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004709861829179165,
      "loss": 4.9892,
      "mean_token_accuracy": 0.21929923743009566,
      "num_tokens": 40934188.0,
      "step": 17850
    },
    {
      "entropy": 5.220877552032471,
      "epoch": 1.7151777137367916,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047096934073571325,
      "loss": 4.9546,
      "mean_token_accuracy": 0.21677092462778091,
      "num_tokens": 40944520.0,
      "step": 17855
    },
    {
      "entropy": 5.2322766304016115,
      "epoch": 1.7156580211335255,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047095249400368384,
      "loss": 4.9653,
      "mean_token_accuracy": 0.20631475299596785,
      "num_tokens": 40955874.0,
      "step": 17860
    },
    {
      "entropy": 5.169892597198486,
      "epoch": 1.7161383285302594,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047093564272221927,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22091901898384095,
      "num_tokens": 40966697.0,
      "step": 17865
    },
    {
      "entropy": 5.0956440448760985,
      "epoch": 1.7166186359269933,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047091878689171105,
      "loss": 4.9462,
      "mean_token_accuracy": 0.21731471717357637,
      "num_tokens": 40978107.0,
      "step": 17870
    },
    {
      "entropy": 5.244924736022949,
      "epoch": 1.7170989433237271,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004709019265125502,
      "loss": 4.9451,
      "mean_token_accuracy": 0.21523994356393814,
      "num_tokens": 40991535.0,
      "step": 17875
    },
    {
      "entropy": 5.27036190032959,
      "epoch": 1.717579250720461,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047088506158512837,
      "loss": 4.9594,
      "mean_token_accuracy": 0.21793600171804428,
      "num_tokens": 41001333.0,
      "step": 17880
    },
    {
      "entropy": 5.094172525405884,
      "epoch": 1.718059558117195,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047086819210983714,
      "loss": 4.8327,
      "mean_token_accuracy": 0.2214152991771698,
      "num_tokens": 41012854.0,
      "step": 17885
    },
    {
      "entropy": 5.102037811279297,
      "epoch": 1.7185398655139288,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047085131808706813,
      "loss": 4.8782,
      "mean_token_accuracy": 0.22039142102003098,
      "num_tokens": 41024041.0,
      "step": 17890
    },
    {
      "entropy": 5.219194936752319,
      "epoch": 1.7190201729106627,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004708344395172129,
      "loss": 4.9316,
      "mean_token_accuracy": 0.2194577068090439,
      "num_tokens": 41034788.0,
      "step": 17895
    },
    {
      "entropy": 5.2273260116577145,
      "epoch": 1.7195004803073968,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004708175564006636,
      "loss": 4.9472,
      "mean_token_accuracy": 0.2231810972094536,
      "num_tokens": 41046513.0,
      "step": 17900
    },
    {
      "entropy": 5.228028392791748,
      "epoch": 1.7199807877041307,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004708006687378121,
      "loss": 5.0609,
      "mean_token_accuracy": 0.20346369594335556,
      "num_tokens": 41059054.0,
      "step": 17905
    },
    {
      "entropy": 5.204518795013428,
      "epoch": 1.7204610951008645,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004707837765290505,
      "loss": 4.9131,
      "mean_token_accuracy": 0.2163485288619995,
      "num_tokens": 41070171.0,
      "step": 17910
    },
    {
      "entropy": 5.117541742324829,
      "epoch": 1.7209414024975984,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004707668797747709,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22116096168756486,
      "num_tokens": 41080963.0,
      "step": 17915
    },
    {
      "entropy": 5.169916820526123,
      "epoch": 1.7214217098943325,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004707499784753657,
      "loss": 4.9772,
      "mean_token_accuracy": 0.21681641638278962,
      "num_tokens": 41092658.0,
      "step": 17920
    },
    {
      "entropy": 5.21237359046936,
      "epoch": 1.7219020172910664,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004707330726312273,
      "loss": 4.9764,
      "mean_token_accuracy": 0.21580159962177276,
      "num_tokens": 41104521.0,
      "step": 17925
    },
    {
      "entropy": 5.174707317352295,
      "epoch": 1.7223823246878003,
      "grad_norm": 1.25,
      "learning_rate": 0.00047071616224274803,
      "loss": 4.9293,
      "mean_token_accuracy": 0.21993063986301423,
      "num_tokens": 41116420.0,
      "step": 17930
    },
    {
      "entropy": 5.185934782028198,
      "epoch": 1.7228626320845342,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004706992473103207,
      "loss": 4.9423,
      "mean_token_accuracy": 0.21858857572078705,
      "num_tokens": 41127747.0,
      "step": 17935
    },
    {
      "entropy": 5.211478137969971,
      "epoch": 1.723342939481268,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047068232783433806,
      "loss": 4.8953,
      "mean_token_accuracy": 0.22366634458303453,
      "num_tokens": 41138498.0,
      "step": 17940
    },
    {
      "entropy": 5.194902038574218,
      "epoch": 1.723823246878002,
      "grad_norm": 1.125,
      "learning_rate": 0.0004706654038151927,
      "loss": 5.0088,
      "mean_token_accuracy": 0.21217814832925797,
      "num_tokens": 41149781.0,
      "step": 17945
    },
    {
      "entropy": 5.205005788803101,
      "epoch": 1.7243035542747358,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004706484752532777,
      "loss": 4.96,
      "mean_token_accuracy": 0.22322595566511155,
      "num_tokens": 41160421.0,
      "step": 17950
    },
    {
      "entropy": 5.182406616210938,
      "epoch": 1.7247838616714697,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004706315421489861,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22752329409122468,
      "num_tokens": 41171192.0,
      "step": 17955
    },
    {
      "entropy": 5.128621912002563,
      "epoch": 1.7252641690682036,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004706146045027109,
      "loss": 4.9197,
      "mean_token_accuracy": 0.22375346720218658,
      "num_tokens": 41182396.0,
      "step": 17960
    },
    {
      "entropy": 5.2468804836273195,
      "epoch": 1.7257444764649374,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004705976623148455,
      "loss": 5.0593,
      "mean_token_accuracy": 0.20294710993766785,
      "num_tokens": 41194939.0,
      "step": 17965
    },
    {
      "entropy": 5.209125185012818,
      "epoch": 1.7262247838616713,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047058071558578324,
      "loss": 4.9252,
      "mean_token_accuracy": 0.21660442799329757,
      "num_tokens": 41206670.0,
      "step": 17970
    },
    {
      "entropy": 5.260093832015992,
      "epoch": 1.7267050912584054,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004705637643159175,
      "loss": 5.1333,
      "mean_token_accuracy": 0.2126966819167137,
      "num_tokens": 41220200.0,
      "step": 17975
    },
    {
      "entropy": 5.207899570465088,
      "epoch": 1.7271853986551393,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047054680850564185,
      "loss": 4.9029,
      "mean_token_accuracy": 0.22000434696674348,
      "num_tokens": 41231697.0,
      "step": 17980
    },
    {
      "entropy": 5.166058969497681,
      "epoch": 1.7276657060518732,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004705298481553499,
      "loss": 4.9571,
      "mean_token_accuracy": 0.2158450573682785,
      "num_tokens": 41243835.0,
      "step": 17985
    },
    {
      "entropy": 5.203484106063843,
      "epoch": 1.728146013448607,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047051288326543553,
      "loss": 4.9751,
      "mean_token_accuracy": 0.2187863752245903,
      "num_tokens": 41255147.0,
      "step": 17990
    },
    {
      "entropy": 5.302144956588745,
      "epoch": 1.7286263208453412,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047049591383629247,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20399728566408157,
      "num_tokens": 41267863.0,
      "step": 17995
    },
    {
      "entropy": 5.269688940048217,
      "epoch": 1.729106628242075,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047047893986831493,
      "loss": 4.9625,
      "mean_token_accuracy": 0.22040341347455977,
      "num_tokens": 41278506.0,
      "step": 18000
    },
    {
      "epoch": 1.729106628242075,
      "eval_entropy": 5.008384053036389,
      "eval_loss": 5.022619724273682,
      "eval_mean_token_accuracy": 0.22223934777142768,
      "eval_num_tokens": 41278506.0,
      "eval_runtime": 27.1543,
      "eval_samples_per_second": 1208.465,
      "eval_steps_per_second": 151.063,
      "step": 18000
    },
    {
      "entropy": 5.189908981323242,
      "epoch": 1.729586935638809,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047046196136189686,
      "loss": 4.9614,
      "mean_token_accuracy": 0.21455368101596833,
      "num_tokens": 41289825.0,
      "step": 18005
    },
    {
      "entropy": 5.172221899032593,
      "epoch": 1.7300672430355428,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004704449783174323,
      "loss": 4.9198,
      "mean_token_accuracy": 0.21491539776325225,
      "num_tokens": 41300451.0,
      "step": 18010
    },
    {
      "entropy": 5.170941638946533,
      "epoch": 1.7305475504322767,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004704279907353158,
      "loss": 4.8812,
      "mean_token_accuracy": 0.22620732188224793,
      "num_tokens": 41312762.0,
      "step": 18015
    },
    {
      "entropy": 5.106926393508911,
      "epoch": 1.7310278578290106,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047041099861594167,
      "loss": 4.9405,
      "mean_token_accuracy": 0.2183744117617607,
      "num_tokens": 41323464.0,
      "step": 18020
    },
    {
      "entropy": 5.171554803848267,
      "epoch": 1.7315081652257445,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004703940019597044,
      "loss": 4.9432,
      "mean_token_accuracy": 0.218562813103199,
      "num_tokens": 41335361.0,
      "step": 18025
    },
    {
      "entropy": 5.203702163696289,
      "epoch": 1.7319884726224783,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047037700076699857,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21696866154670716,
      "num_tokens": 41345353.0,
      "step": 18030
    },
    {
      "entropy": 5.18451418876648,
      "epoch": 1.7324687800192122,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000470359995038219,
      "loss": 5.0054,
      "mean_token_accuracy": 0.21888639330863952,
      "num_tokens": 41356578.0,
      "step": 18035
    },
    {
      "entropy": 5.2541642665863035,
      "epoch": 1.732949087415946,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004703429847737604,
      "loss": 4.9489,
      "mean_token_accuracy": 0.22214508503675462,
      "num_tokens": 41367668.0,
      "step": 18040
    },
    {
      "entropy": 5.1678791522979735,
      "epoch": 1.73342939481268,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004703259699740177,
      "loss": 4.9116,
      "mean_token_accuracy": 0.22133572697639464,
      "num_tokens": 41379746.0,
      "step": 18045
    },
    {
      "entropy": 5.201324319839477,
      "epoch": 1.7339097022094139,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047030895063938607,
      "loss": 5.0106,
      "mean_token_accuracy": 0.21498912423849106,
      "num_tokens": 41390251.0,
      "step": 18050
    },
    {
      "entropy": 5.2157799243927006,
      "epoch": 1.734390009606148,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047029192677026043,
      "loss": 4.9142,
      "mean_token_accuracy": 0.21945614367723465,
      "num_tokens": 41401689.0,
      "step": 18055
    },
    {
      "entropy": 5.203940582275391,
      "epoch": 1.7348703170028819,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004702748983670363,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21389392763376236,
      "num_tokens": 41414001.0,
      "step": 18060
    },
    {
      "entropy": 5.279625320434571,
      "epoch": 1.7353506243996157,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004702578654301088,
      "loss": 5.0494,
      "mean_token_accuracy": 0.2168555811047554,
      "num_tokens": 41426012.0,
      "step": 18065
    },
    {
      "entropy": 5.235019016265869,
      "epoch": 1.7358309317963496,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004702408279598734,
      "loss": 5.0304,
      "mean_token_accuracy": 0.21296436488628387,
      "num_tokens": 41436799.0,
      "step": 18070
    },
    {
      "entropy": 5.188997459411621,
      "epoch": 1.7363112391930837,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004702237859567258,
      "loss": 4.9136,
      "mean_token_accuracy": 0.2276575818657875,
      "num_tokens": 41448802.0,
      "step": 18075
    },
    {
      "entropy": 5.105773305892944,
      "epoch": 1.7367915465898176,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004702067394210616,
      "loss": 4.8363,
      "mean_token_accuracy": 0.22819617986679078,
      "num_tokens": 41459576.0,
      "step": 18080
    },
    {
      "entropy": 5.233019828796387,
      "epoch": 1.7372718539865515,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047018968835327643,
      "loss": 4.9529,
      "mean_token_accuracy": 0.21334681063890457,
      "num_tokens": 41471105.0,
      "step": 18085
    },
    {
      "entropy": 5.197208118438721,
      "epoch": 1.7377521613832854,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004701726327537664,
      "loss": 4.9309,
      "mean_token_accuracy": 0.22097699344158173,
      "num_tokens": 41482747.0,
      "step": 18090
    },
    {
      "entropy": 5.148146152496338,
      "epoch": 1.7382324687800192,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004701555726229274,
      "loss": 4.9727,
      "mean_token_accuracy": 0.2093600407242775,
      "num_tokens": 41494268.0,
      "step": 18095
    },
    {
      "entropy": 5.183890056610108,
      "epoch": 1.7387127761767531,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004701385079611555,
      "loss": 4.937,
      "mean_token_accuracy": 0.21966792941093444,
      "num_tokens": 41506144.0,
      "step": 18100
    },
    {
      "entropy": 5.146788787841797,
      "epoch": 1.739193083573487,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047012143876884677,
      "loss": 4.8177,
      "mean_token_accuracy": 0.23037531822919846,
      "num_tokens": 41518165.0,
      "step": 18105
    },
    {
      "entropy": 5.173630714416504,
      "epoch": 1.739673390970221,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004701043650463977,
      "loss": 4.9086,
      "mean_token_accuracy": 0.22266392260789872,
      "num_tokens": 41529644.0,
      "step": 18110
    },
    {
      "entropy": 5.134193992614746,
      "epoch": 1.7401536983669548,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004700872867942046,
      "loss": 4.8914,
      "mean_token_accuracy": 0.22757913917303085,
      "num_tokens": 41540540.0,
      "step": 18115
    },
    {
      "entropy": 5.169078969955445,
      "epoch": 1.7406340057636887,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000470070204012664,
      "loss": 4.9095,
      "mean_token_accuracy": 0.2146324098110199,
      "num_tokens": 41551937.0,
      "step": 18120
    },
    {
      "entropy": 5.140912675857544,
      "epoch": 1.7411143131604225,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047005311670217256,
      "loss": 4.8595,
      "mean_token_accuracy": 0.22714108377695083,
      "num_tokens": 41561630.0,
      "step": 18125
    },
    {
      "entropy": 5.121124219894409,
      "epoch": 1.7415946205571564,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047003602486312687,
      "loss": 4.8494,
      "mean_token_accuracy": 0.2231152132153511,
      "num_tokens": 41573152.0,
      "step": 18130
    },
    {
      "entropy": 5.221160221099853,
      "epoch": 1.7420749279538905,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004700189284959238,
      "loss": 4.9783,
      "mean_token_accuracy": 0.20664760917425157,
      "num_tokens": 41585280.0,
      "step": 18135
    },
    {
      "entropy": 5.177599668502808,
      "epoch": 1.7425552353506244,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047000182760096037,
      "loss": 4.8969,
      "mean_token_accuracy": 0.22130205631256103,
      "num_tokens": 41596308.0,
      "step": 18140
    },
    {
      "entropy": 5.10551872253418,
      "epoch": 1.7430355427473583,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004699847221786335,
      "loss": 4.9074,
      "mean_token_accuracy": 0.21892183125019074,
      "num_tokens": 41607479.0,
      "step": 18145
    },
    {
      "entropy": 5.103967523574829,
      "epoch": 1.7435158501440924,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004699676122293403,
      "loss": 4.8842,
      "mean_token_accuracy": 0.2148337781429291,
      "num_tokens": 41619832.0,
      "step": 18150
    },
    {
      "entropy": 5.228288221359253,
      "epoch": 1.7439961575408263,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004699504977534782,
      "loss": 5.0367,
      "mean_token_accuracy": 0.2169421136379242,
      "num_tokens": 41631715.0,
      "step": 18155
    },
    {
      "entropy": 5.249570417404175,
      "epoch": 1.7444764649375601,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004699333787514444,
      "loss": 4.9988,
      "mean_token_accuracy": 0.21617266535758972,
      "num_tokens": 41643454.0,
      "step": 18160
    },
    {
      "entropy": 5.203645801544189,
      "epoch": 1.744956772334294,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004699162552236363,
      "loss": 4.945,
      "mean_token_accuracy": 0.22019296288490295,
      "num_tokens": 41655319.0,
      "step": 18165
    },
    {
      "entropy": 5.167347478866577,
      "epoch": 1.745437079731028,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00046989912717045165,
      "loss": 4.9744,
      "mean_token_accuracy": 0.21599013358354568,
      "num_tokens": 41667082.0,
      "step": 18170
    },
    {
      "entropy": 5.233082294464111,
      "epoch": 1.7459173871277618,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046988199459228793,
      "loss": 4.9118,
      "mean_token_accuracy": 0.21897320002317427,
      "num_tokens": 41678281.0,
      "step": 18175
    },
    {
      "entropy": 5.171620273590088,
      "epoch": 1.7463976945244957,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004698648574895429,
      "loss": 4.9043,
      "mean_token_accuracy": 0.2208816662430763,
      "num_tokens": 41690596.0,
      "step": 18180
    },
    {
      "entropy": 5.158304119110108,
      "epoch": 1.7468780019212296,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046984771586261465,
      "loss": 4.9488,
      "mean_token_accuracy": 0.21384944021701813,
      "num_tokens": 41702441.0,
      "step": 18185
    },
    {
      "entropy": 5.148635053634644,
      "epoch": 1.7473583093179634,
      "grad_norm": 1.328125,
      "learning_rate": 0.000469830569711901,
      "loss": 4.9783,
      "mean_token_accuracy": 0.21873321533203124,
      "num_tokens": 41714235.0,
      "step": 18190
    },
    {
      "entropy": 5.179041004180908,
      "epoch": 1.7478386167146973,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004698134190377999,
      "loss": 4.9606,
      "mean_token_accuracy": 0.22201482951641083,
      "num_tokens": 41725870.0,
      "step": 18195
    },
    {
      "entropy": 5.1740946769714355,
      "epoch": 1.7483189241114312,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046979626384070983,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22523927837610244,
      "num_tokens": 41736500.0,
      "step": 18200
    },
    {
      "entropy": 5.162458038330078,
      "epoch": 1.748799231508165,
      "grad_norm": 1.28125,
      "learning_rate": 0.000469779104121029,
      "loss": 4.8882,
      "mean_token_accuracy": 0.22337753176689149,
      "num_tokens": 41748466.0,
      "step": 18205
    },
    {
      "entropy": 5.255827140808106,
      "epoch": 1.7492795389048992,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046976193987915553,
      "loss": 5.0496,
      "mean_token_accuracy": 0.21399937719106674,
      "num_tokens": 41759728.0,
      "step": 18210
    },
    {
      "entropy": 5.186284017562866,
      "epoch": 1.749759846301633,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004697447711154883,
      "loss": 4.9367,
      "mean_token_accuracy": 0.2159511923789978,
      "num_tokens": 41771628.0,
      "step": 18215
    },
    {
      "entropy": 5.229859972000122,
      "epoch": 1.750240153698367,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046972759783042576,
      "loss": 4.9653,
      "mean_token_accuracy": 0.2190747082233429,
      "num_tokens": 41783224.0,
      "step": 18220
    },
    {
      "entropy": 5.115824270248413,
      "epoch": 1.7507204610951008,
      "grad_norm": 1.125,
      "learning_rate": 0.0004697104200243666,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22871831208467483,
      "num_tokens": 41794899.0,
      "step": 18225
    },
    {
      "entropy": 5.1800790309906,
      "epoch": 1.751200768491835,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004696932376977096,
      "loss": 4.8703,
      "mean_token_accuracy": 0.2250388517975807,
      "num_tokens": 41806336.0,
      "step": 18230
    },
    {
      "entropy": 5.190379953384399,
      "epoch": 1.7516810758885688,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004696760508508538,
      "loss": 4.9505,
      "mean_token_accuracy": 0.2223442539572716,
      "num_tokens": 41817443.0,
      "step": 18235
    },
    {
      "entropy": 5.160542297363281,
      "epoch": 1.7521613832853027,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046965885948419814,
      "loss": 4.9919,
      "mean_token_accuracy": 0.2164471462368965,
      "num_tokens": 41829740.0,
      "step": 18240
    },
    {
      "entropy": 5.125571584701538,
      "epoch": 1.7526416906820366,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004696416635981418,
      "loss": 4.8902,
      "mean_token_accuracy": 0.22456476390361785,
      "num_tokens": 41840519.0,
      "step": 18245
    },
    {
      "entropy": 5.105917882919312,
      "epoch": 1.7531219980787704,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000469624463193084,
      "loss": 4.8481,
      "mean_token_accuracy": 0.22758448570966722,
      "num_tokens": 41850975.0,
      "step": 18250
    },
    {
      "entropy": 5.113386249542236,
      "epoch": 1.7536023054755043,
      "grad_norm": 1.09375,
      "learning_rate": 0.000469607258269424,
      "loss": 4.896,
      "mean_token_accuracy": 0.22474818378686906,
      "num_tokens": 41861854.0,
      "step": 18255
    },
    {
      "entropy": 5.189031267166138,
      "epoch": 1.7540826128722382,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004695900488275614,
      "loss": 5.007,
      "mean_token_accuracy": 0.21779554933309556,
      "num_tokens": 41873279.0,
      "step": 18260
    },
    {
      "entropy": 5.217519235610962,
      "epoch": 1.754562920268972,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004695728348678957,
      "loss": 4.9346,
      "mean_token_accuracy": 0.21305863708257675,
      "num_tokens": 41885545.0,
      "step": 18265
    },
    {
      "entropy": 5.123094701766968,
      "epoch": 1.755043227665706,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004695556163908265,
      "loss": 4.8805,
      "mean_token_accuracy": 0.2211624339222908,
      "num_tokens": 41896168.0,
      "step": 18270
    },
    {
      "entropy": 5.098513746261597,
      "epoch": 1.7555235350624399,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004695383933967536,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22860950231552124,
      "num_tokens": 41907423.0,
      "step": 18275
    },
    {
      "entropy": 5.1227428913116455,
      "epoch": 1.7560038424591737,
      "grad_norm": 1.125,
      "learning_rate": 0.00046952116588607694,
      "loss": 4.8527,
      "mean_token_accuracy": 0.2274855390191078,
      "num_tokens": 41918617.0,
      "step": 18280
    },
    {
      "entropy": 5.146688985824585,
      "epoch": 1.7564841498559076,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004695039338591963,
      "loss": 4.8732,
      "mean_token_accuracy": 0.22586282044649125,
      "num_tokens": 41930235.0,
      "step": 18285
    },
    {
      "entropy": 5.134828805923462,
      "epoch": 1.7569644572526417,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000469486697316512,
      "loss": 4.9634,
      "mean_token_accuracy": 0.21943751126527786,
      "num_tokens": 41941308.0,
      "step": 18290
    },
    {
      "entropy": 5.275082540512085,
      "epoch": 1.7574447646493756,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000469469456258424,
      "loss": 5.0651,
      "mean_token_accuracy": 0.21027645766735076,
      "num_tokens": 41952406.0,
      "step": 18295
    },
    {
      "entropy": 5.245355129241943,
      "epoch": 1.7579250720461095,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004694522106853327,
      "loss": 4.9628,
      "mean_token_accuracy": 0.21077790409326552,
      "num_tokens": 41964642.0,
      "step": 18300
    },
    {
      "entropy": 5.1921234130859375,
      "epoch": 1.7584053794428436,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046943496059763845,
      "loss": 4.9653,
      "mean_token_accuracy": 0.21325814425945283,
      "num_tokens": 41975761.0,
      "step": 18305
    },
    {
      "entropy": 5.124700975418091,
      "epoch": 1.7588856868395775,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046941770599574176,
      "loss": 4.8989,
      "mean_token_accuracy": 0.2256785064935684,
      "num_tokens": 41987596.0,
      "step": 18310
    },
    {
      "entropy": 5.149713659286499,
      "epoch": 1.7593659942363113,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004694004468800433,
      "loss": 4.8859,
      "mean_token_accuracy": 0.21872322410345077,
      "num_tokens": 41999148.0,
      "step": 18315
    },
    {
      "entropy": 5.125066614151001,
      "epoch": 1.7598463016330452,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004693831832509437,
      "loss": 4.8612,
      "mean_token_accuracy": 0.22902813851833342,
      "num_tokens": 42010030.0,
      "step": 18320
    },
    {
      "entropy": 5.19381628036499,
      "epoch": 1.760326609029779,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046936591510884375,
      "loss": 5.0103,
      "mean_token_accuracy": 0.2132745072245598,
      "num_tokens": 42022009.0,
      "step": 18325
    },
    {
      "entropy": 5.233895874023437,
      "epoch": 1.760806916426513,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046934864245414443,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21621835976839066,
      "num_tokens": 42033654.0,
      "step": 18330
    },
    {
      "entropy": 5.123974752426148,
      "epoch": 1.7612872238232469,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046933136528724676,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22414708733558655,
      "num_tokens": 42045521.0,
      "step": 18335
    },
    {
      "entropy": 5.105570316314697,
      "epoch": 1.7617675312199808,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004693140836085518,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22080586552619935,
      "num_tokens": 42057751.0,
      "step": 18340
    },
    {
      "entropy": 5.161794519424438,
      "epoch": 1.7622478386167146,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046929679741846076,
      "loss": 4.9739,
      "mean_token_accuracy": 0.22453261464834212,
      "num_tokens": 42068823.0,
      "step": 18345
    },
    {
      "entropy": 5.1471202850341795,
      "epoch": 1.7627281460134485,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046927950671737505,
      "loss": 4.8587,
      "mean_token_accuracy": 0.22297324985265732,
      "num_tokens": 42079856.0,
      "step": 18350
    },
    {
      "entropy": 5.218841314315796,
      "epoch": 1.7632084534101824,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046926221150569617,
      "loss": 4.9157,
      "mean_token_accuracy": 0.21964731961488723,
      "num_tokens": 42090560.0,
      "step": 18355
    },
    {
      "entropy": 5.2200939655303955,
      "epoch": 1.7636887608069163,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004692449117838255,
      "loss": 4.9988,
      "mean_token_accuracy": 0.21176638156175615,
      "num_tokens": 42102252.0,
      "step": 18360
    },
    {
      "entropy": 5.226511240005493,
      "epoch": 1.7641690682036504,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004692276075521648,
      "loss": 5.0685,
      "mean_token_accuracy": 0.21005309969186783,
      "num_tokens": 42113679.0,
      "step": 18365
    },
    {
      "entropy": 5.167455863952637,
      "epoch": 1.7646493756003843,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004692102988111158,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2179097503423691,
      "num_tokens": 42125919.0,
      "step": 18370
    },
    {
      "entropy": 5.350633573532105,
      "epoch": 1.7651296829971181,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046919298556108023,
      "loss": 5.0943,
      "mean_token_accuracy": 0.2081959679722786,
      "num_tokens": 42138189.0,
      "step": 18375
    },
    {
      "entropy": 5.260043525695801,
      "epoch": 1.765609990393852,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046917566780246036,
      "loss": 4.998,
      "mean_token_accuracy": 0.2203314408659935,
      "num_tokens": 42150232.0,
      "step": 18380
    },
    {
      "entropy": 5.185950994491577,
      "epoch": 1.7660902977905861,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046915834553565793,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21736008822917938,
      "num_tokens": 42160740.0,
      "step": 18385
    },
    {
      "entropy": 5.187232923507691,
      "epoch": 1.76657060518732,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004691410187610753,
      "loss": 4.9342,
      "mean_token_accuracy": 0.22891892045736312,
      "num_tokens": 42171761.0,
      "step": 18390
    },
    {
      "entropy": 5.16988091468811,
      "epoch": 1.767050912584054,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046912368747911465,
      "loss": 4.8846,
      "mean_token_accuracy": 0.22325501441955567,
      "num_tokens": 42183362.0,
      "step": 18395
    },
    {
      "entropy": 5.2661176204681395,
      "epoch": 1.7675312199807878,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046910635169017845,
      "loss": 5.0421,
      "mean_token_accuracy": 0.2092567726969719,
      "num_tokens": 42195610.0,
      "step": 18400
    },
    {
      "entropy": 5.171834993362427,
      "epoch": 1.7680115273775217,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004690890113946691,
      "loss": 4.928,
      "mean_token_accuracy": 0.2105174332857132,
      "num_tokens": 42207489.0,
      "step": 18405
    },
    {
      "entropy": 5.216178894042969,
      "epoch": 1.7684918347742555,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004690716665929893,
      "loss": 4.9508,
      "mean_token_accuracy": 0.21685057282447814,
      "num_tokens": 42217300.0,
      "step": 18410
    },
    {
      "entropy": 5.2149231910705565,
      "epoch": 1.7689721421709894,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046905431728554164,
      "loss": 4.9588,
      "mean_token_accuracy": 0.21947902888059617,
      "num_tokens": 42229031.0,
      "step": 18415
    },
    {
      "entropy": 4.9939906120300295,
      "epoch": 1.7694524495677233,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046903696347272894,
      "loss": 4.6973,
      "mean_token_accuracy": 0.23478817641735078,
      "num_tokens": 42239102.0,
      "step": 18420
    },
    {
      "entropy": 5.175135135650635,
      "epoch": 1.7699327569644572,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046901960515495413,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21834573447704314,
      "num_tokens": 42250436.0,
      "step": 18425
    },
    {
      "entropy": 5.129410266876221,
      "epoch": 1.770413064361191,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004690022423326202,
      "loss": 4.8729,
      "mean_token_accuracy": 0.2321063533425331,
      "num_tokens": 42260544.0,
      "step": 18430
    },
    {
      "entropy": 5.124046754837036,
      "epoch": 1.770893371757925,
      "grad_norm": 1.125,
      "learning_rate": 0.0004689848750061303,
      "loss": 4.8993,
      "mean_token_accuracy": 0.2222402125597,
      "num_tokens": 42272519.0,
      "step": 18435
    },
    {
      "entropy": 5.169499826431275,
      "epoch": 1.7713736791546588,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004689675031758876,
      "loss": 4.8923,
      "mean_token_accuracy": 0.22698742151260376,
      "num_tokens": 42281722.0,
      "step": 18440
    },
    {
      "entropy": 5.151767873764038,
      "epoch": 1.771853986551393,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004689501268422954,
      "loss": 5.0095,
      "mean_token_accuracy": 0.22268947809934617,
      "num_tokens": 42294067.0,
      "step": 18445
    },
    {
      "entropy": 5.188425827026367,
      "epoch": 1.7723342939481268,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046893274600575725,
      "loss": 4.8899,
      "mean_token_accuracy": 0.22142233699560165,
      "num_tokens": 42305547.0,
      "step": 18450
    },
    {
      "entropy": 5.224566888809204,
      "epoch": 1.7728146013448607,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004689153606666765,
      "loss": 4.9586,
      "mean_token_accuracy": 0.20875730067491532,
      "num_tokens": 42315557.0,
      "step": 18455
    },
    {
      "entropy": 5.191870498657226,
      "epoch": 1.7732949087415946,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000468897970825457,
      "loss": 5.0251,
      "mean_token_accuracy": 0.2113093450665474,
      "num_tokens": 42327374.0,
      "step": 18460
    },
    {
      "entropy": 5.1442060470581055,
      "epoch": 1.7737752161383287,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046888057648250233,
      "loss": 4.8807,
      "mean_token_accuracy": 0.22348326742649077,
      "num_tokens": 42338889.0,
      "step": 18465
    },
    {
      "entropy": 5.184911823272705,
      "epoch": 1.7742555235350626,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004688631776382164,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22326247841119767,
      "num_tokens": 42348678.0,
      "step": 18470
    },
    {
      "entropy": 5.147710275650025,
      "epoch": 1.7747358309317964,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046884577429300305,
      "loss": 4.9809,
      "mean_token_accuracy": 0.21520349532365798,
      "num_tokens": 42359090.0,
      "step": 18475
    },
    {
      "entropy": 5.158450937271118,
      "epoch": 1.7752161383285303,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004688283664472665,
      "loss": 4.9148,
      "mean_token_accuracy": 0.21454418003559111,
      "num_tokens": 42370947.0,
      "step": 18480
    },
    {
      "entropy": 5.188743782043457,
      "epoch": 1.7756964457252642,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046881095410141084,
      "loss": 4.9271,
      "mean_token_accuracy": 0.22006626725196837,
      "num_tokens": 42381597.0,
      "step": 18485
    },
    {
      "entropy": 5.194511032104492,
      "epoch": 1.776176753121998,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046879353725584036,
      "loss": 4.8745,
      "mean_token_accuracy": 0.2257644772529602,
      "num_tokens": 42391877.0,
      "step": 18490
    },
    {
      "entropy": 5.0896097183227536,
      "epoch": 1.776657060518732,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046877611591095923,
      "loss": 4.8878,
      "mean_token_accuracy": 0.22146574407815933,
      "num_tokens": 42403916.0,
      "step": 18495
    },
    {
      "entropy": 5.169925928115845,
      "epoch": 1.7771373679154658,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046875869006717224,
      "loss": 4.9619,
      "mean_token_accuracy": 0.21315832883119584,
      "num_tokens": 42415624.0,
      "step": 18500
    },
    {
      "entropy": 5.151780509948731,
      "epoch": 1.7776176753121997,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046874125972488375,
      "loss": 4.9138,
      "mean_token_accuracy": 0.219668011367321,
      "num_tokens": 42427228.0,
      "step": 18505
    },
    {
      "entropy": 5.171441650390625,
      "epoch": 1.7780979827089336,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046872382488449853,
      "loss": 4.9793,
      "mean_token_accuracy": 0.21152068972587584,
      "num_tokens": 42440023.0,
      "step": 18510
    },
    {
      "entropy": 5.184543323516846,
      "epoch": 1.7785782901056675,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046870638554642133,
      "loss": 4.8749,
      "mean_token_accuracy": 0.2245725840330124,
      "num_tokens": 42451126.0,
      "step": 18515
    },
    {
      "entropy": 5.139539527893066,
      "epoch": 1.7790585975024016,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046868894171105704,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21104169487953187,
      "num_tokens": 42462975.0,
      "step": 18520
    },
    {
      "entropy": 5.118284606933594,
      "epoch": 1.7795389048991355,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004686714933788107,
      "loss": 4.855,
      "mean_token_accuracy": 0.22563856095075607,
      "num_tokens": 42474716.0,
      "step": 18525
    },
    {
      "entropy": 5.191065549850464,
      "epoch": 1.7800192122958693,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004686540405500873,
      "loss": 4.9084,
      "mean_token_accuracy": 0.22859731018543245,
      "num_tokens": 42484616.0,
      "step": 18530
    },
    {
      "entropy": 5.1791100025177,
      "epoch": 1.7804995196926032,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004686365832252922,
      "loss": 4.8822,
      "mean_token_accuracy": 0.2233108401298523,
      "num_tokens": 42494840.0,
      "step": 18535
    },
    {
      "entropy": 5.156848001480102,
      "epoch": 1.7809798270893373,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046861912140483056,
      "loss": 4.9511,
      "mean_token_accuracy": 0.21606461107730865,
      "num_tokens": 42505957.0,
      "step": 18540
    },
    {
      "entropy": 5.240969944000244,
      "epoch": 1.7814601344860712,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046860165508910787,
      "loss": 5.0856,
      "mean_token_accuracy": 0.2120821714401245,
      "num_tokens": 42517861.0,
      "step": 18545
    },
    {
      "entropy": 5.137620162963867,
      "epoch": 1.781940441882805,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004685841842785296,
      "loss": 4.8575,
      "mean_token_accuracy": 0.21964601874351503,
      "num_tokens": 42529986.0,
      "step": 18550
    },
    {
      "entropy": 5.111011028289795,
      "epoch": 1.782420749279539,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004685667089735014,
      "loss": 4.8867,
      "mean_token_accuracy": 0.2230542078614235,
      "num_tokens": 42541778.0,
      "step": 18555
    },
    {
      "entropy": 5.212114858627319,
      "epoch": 1.7829010566762729,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046854922917442907,
      "loss": 4.9406,
      "mean_token_accuracy": 0.21493832468986512,
      "num_tokens": 42553776.0,
      "step": 18560
    },
    {
      "entropy": 5.200642395019531,
      "epoch": 1.7833813640730067,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004685317448817182,
      "loss": 4.9693,
      "mean_token_accuracy": 0.21868837922811507,
      "num_tokens": 42566277.0,
      "step": 18565
    },
    {
      "entropy": 5.040183448791504,
      "epoch": 1.7838616714697406,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004685142560957751,
      "loss": 4.7951,
      "mean_token_accuracy": 0.2240560010075569,
      "num_tokens": 42576598.0,
      "step": 18570
    },
    {
      "entropy": 5.133614444732666,
      "epoch": 1.7843419788664745,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004684967628170054,
      "loss": 4.8631,
      "mean_token_accuracy": 0.22733719944953917,
      "num_tokens": 42587850.0,
      "step": 18575
    },
    {
      "entropy": 5.183877897262573,
      "epoch": 1.7848222862632084,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046847926504581553,
      "loss": 4.9112,
      "mean_token_accuracy": 0.21961584389209748,
      "num_tokens": 42598754.0,
      "step": 18580
    },
    {
      "entropy": 5.219208669662476,
      "epoch": 1.7853025936599423,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004684617627826116,
      "loss": 4.9604,
      "mean_token_accuracy": 0.2221649944782257,
      "num_tokens": 42609842.0,
      "step": 18585
    },
    {
      "entropy": 5.107766962051391,
      "epoch": 1.7857829010566761,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004684442560278001,
      "loss": 4.8332,
      "mean_token_accuracy": 0.22912952899932862,
      "num_tokens": 42621641.0,
      "step": 18590
    },
    {
      "entropy": 5.100026845932007,
      "epoch": 1.78626320845341,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046842674478178727,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22217022627592087,
      "num_tokens": 42632523.0,
      "step": 18595
    },
    {
      "entropy": 5.218968868255615,
      "epoch": 1.7867435158501441,
      "grad_norm": 1.125,
      "learning_rate": 0.0004684092290449798,
      "loss": 4.958,
      "mean_token_accuracy": 0.21347346305847167,
      "num_tokens": 42643511.0,
      "step": 18600
    },
    {
      "entropy": 5.223620891571045,
      "epoch": 1.787223823246878,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004683917088177844,
      "loss": 4.8864,
      "mean_token_accuracy": 0.22519985288381578,
      "num_tokens": 42655077.0,
      "step": 18605
    },
    {
      "entropy": 5.23469877243042,
      "epoch": 1.7877041306436119,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004683741841006077,
      "loss": 5.0271,
      "mean_token_accuracy": 0.21238600760698317,
      "num_tokens": 42667454.0,
      "step": 18610
    },
    {
      "entropy": 5.165902709960937,
      "epoch": 1.7881844380403458,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004683566548938567,
      "loss": 4.9077,
      "mean_token_accuracy": 0.22242332696914674,
      "num_tokens": 42678214.0,
      "step": 18615
    },
    {
      "entropy": 5.139394855499267,
      "epoch": 1.7886647454370799,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004683391211979383,
      "loss": 4.8881,
      "mean_token_accuracy": 0.2219072088599205,
      "num_tokens": 42689424.0,
      "step": 18620
    },
    {
      "entropy": 5.31291651725769,
      "epoch": 1.7891450528338138,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004683215830132597,
      "loss": 5.0854,
      "mean_token_accuracy": 0.20756420940160752,
      "num_tokens": 42701679.0,
      "step": 18625
    },
    {
      "entropy": 5.18286771774292,
      "epoch": 1.7896253602305476,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046830404034022786,
      "loss": 4.9174,
      "mean_token_accuracy": 0.2216010123491287,
      "num_tokens": 42711931.0,
      "step": 18630
    },
    {
      "entropy": 5.224659013748169,
      "epoch": 1.7901056676272815,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004682864931792502,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21652059108018876,
      "num_tokens": 42723460.0,
      "step": 18635
    },
    {
      "entropy": 5.211698198318482,
      "epoch": 1.7905859750240154,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004682689415307342,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2227874591946602,
      "num_tokens": 42735389.0,
      "step": 18640
    },
    {
      "entropy": 5.1946195602417,
      "epoch": 1.7910662824207493,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004682513853950872,
      "loss": 4.8943,
      "mean_token_accuracy": 0.2206537440419197,
      "num_tokens": 42747779.0,
      "step": 18645
    },
    {
      "entropy": 5.117613983154297,
      "epoch": 1.7915465898174832,
      "grad_norm": 1.296875,
      "learning_rate": 0.000468233824772717,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21582213938236236,
      "num_tokens": 42758776.0,
      "step": 18650
    },
    {
      "entropy": 5.321246862411499,
      "epoch": 1.792026897214217,
      "grad_norm": 1.1875,
      "learning_rate": 0.000468216259664031,
      "loss": 5.0432,
      "mean_token_accuracy": 0.21038479954004288,
      "num_tokens": 42770234.0,
      "step": 18655
    },
    {
      "entropy": 5.2058374881744385,
      "epoch": 1.792507204610951,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046819869006943727,
      "loss": 4.91,
      "mean_token_accuracy": 0.22692616730928422,
      "num_tokens": 42782076.0,
      "step": 18660
    },
    {
      "entropy": 5.1517956256866455,
      "epoch": 1.7929875120076848,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004681811159893436,
      "loss": 4.8767,
      "mean_token_accuracy": 0.2243885353207588,
      "num_tokens": 42793620.0,
      "step": 18665
    },
    {
      "entropy": 5.204485988616943,
      "epoch": 1.7934678194044187,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046816353742415814,
      "loss": 4.9583,
      "mean_token_accuracy": 0.22410303950309754,
      "num_tokens": 42805125.0,
      "step": 18670
    },
    {
      "entropy": 5.195950365066528,
      "epoch": 1.7939481268011528,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046814595437428885,
      "loss": 4.8574,
      "mean_token_accuracy": 0.22510511726140975,
      "num_tokens": 42815472.0,
      "step": 18675
    },
    {
      "entropy": 5.169199132919312,
      "epoch": 1.7944284341978867,
      "grad_norm": 1.171875,
      "learning_rate": 0.000468128366840144,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2174817755818367,
      "num_tokens": 42826822.0,
      "step": 18680
    },
    {
      "entropy": 5.181671047210694,
      "epoch": 1.7949087415946205,
      "grad_norm": 1.15625,
      "learning_rate": 0.000468110774822132,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22061461210250854,
      "num_tokens": 42839024.0,
      "step": 18685
    },
    {
      "entropy": 5.147944641113281,
      "epoch": 1.7953890489913544,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004680931783206612,
      "loss": 4.8643,
      "mean_token_accuracy": 0.22245632857084274,
      "num_tokens": 42849782.0,
      "step": 18690
    },
    {
      "entropy": 5.1519299983978275,
      "epoch": 1.7958693563880885,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046807557733614014,
      "loss": 4.9194,
      "mean_token_accuracy": 0.22597801983356475,
      "num_tokens": 42859792.0,
      "step": 18695
    },
    {
      "entropy": 5.150907707214356,
      "epoch": 1.7963496637848224,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046805797186897757,
      "loss": 4.9852,
      "mean_token_accuracy": 0.21621638387441636,
      "num_tokens": 42872031.0,
      "step": 18700
    },
    {
      "entropy": 5.2163262367248535,
      "epoch": 1.7968299711815563,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046804036191958206,
      "loss": 4.9228,
      "mean_token_accuracy": 0.22645899951457976,
      "num_tokens": 42882830.0,
      "step": 18705
    },
    {
      "entropy": 5.232586336135864,
      "epoch": 1.7973102785782902,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046802274748836267,
      "loss": 4.9495,
      "mean_token_accuracy": 0.21486039757728576,
      "num_tokens": 42893576.0,
      "step": 18710
    },
    {
      "entropy": 5.089263772964477,
      "epoch": 1.797790585975024,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004680051285757281,
      "loss": 4.8822,
      "mean_token_accuracy": 0.22521999031305312,
      "num_tokens": 42904405.0,
      "step": 18715
    },
    {
      "entropy": 5.192047882080078,
      "epoch": 1.798270893371758,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004679875051820877,
      "loss": 4.9181,
      "mean_token_accuracy": 0.21825072467327117,
      "num_tokens": 42916601.0,
      "step": 18720
    },
    {
      "entropy": 5.214356517791748,
      "epoch": 1.7987512007684918,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004679698773078503,
      "loss": 4.9506,
      "mean_token_accuracy": 0.2163441076874733,
      "num_tokens": 42928977.0,
      "step": 18725
    },
    {
      "entropy": 5.252518033981323,
      "epoch": 1.7992315081652257,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046795224495342554,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22145872861146926,
      "num_tokens": 42940149.0,
      "step": 18730
    },
    {
      "entropy": 5.174649381637574,
      "epoch": 1.7997118155619596,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046793460811922255,
      "loss": 5.0051,
      "mean_token_accuracy": 0.21233069896697998,
      "num_tokens": 42952122.0,
      "step": 18735
    },
    {
      "entropy": 5.157995891571045,
      "epoch": 1.8001921229586935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046791696680565075,
      "loss": 4.8739,
      "mean_token_accuracy": 0.22639018595218657,
      "num_tokens": 42962984.0,
      "step": 18740
    },
    {
      "entropy": 5.237317514419556,
      "epoch": 1.8006724303554273,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046789932101312003,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2110114499926567,
      "num_tokens": 42974409.0,
      "step": 18745
    },
    {
      "entropy": 5.215788173675537,
      "epoch": 1.8011527377521612,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004678816707420397,
      "loss": 4.9835,
      "mean_token_accuracy": 0.215987665951252,
      "num_tokens": 42986600.0,
      "step": 18750
    },
    {
      "entropy": 5.093104648590088,
      "epoch": 1.8016330451488953,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004678640159928198,
      "loss": 4.8671,
      "mean_token_accuracy": 0.22648871839046478,
      "num_tokens": 42998192.0,
      "step": 18755
    },
    {
      "entropy": 5.137369012832641,
      "epoch": 1.8021133525456292,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004678463567658701,
      "loss": 4.8636,
      "mean_token_accuracy": 0.22732312530279158,
      "num_tokens": 43009680.0,
      "step": 18760
    },
    {
      "entropy": 5.129584217071534,
      "epoch": 1.802593659942363,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004678286930616006,
      "loss": 4.8969,
      "mean_token_accuracy": 0.2241714671254158,
      "num_tokens": 43021031.0,
      "step": 18765
    },
    {
      "entropy": 5.157454633712769,
      "epoch": 1.803073967339097,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004678110248804215,
      "loss": 4.9046,
      "mean_token_accuracy": 0.230050627887249,
      "num_tokens": 43031423.0,
      "step": 18770
    },
    {
      "entropy": 5.152748346328735,
      "epoch": 1.803554274735831,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046779335222274293,
      "loss": 4.9453,
      "mean_token_accuracy": 0.2204241082072258,
      "num_tokens": 43042895.0,
      "step": 18775
    },
    {
      "entropy": 5.205005741119384,
      "epoch": 1.804034582132565,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046777567508897515,
      "loss": 5.0363,
      "mean_token_accuracy": 0.21626016795635222,
      "num_tokens": 43053052.0,
      "step": 18780
    },
    {
      "entropy": 5.256972360610962,
      "epoch": 1.8045148895292988,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046775799347952864,
      "loss": 4.9986,
      "mean_token_accuracy": 0.21178092807531357,
      "num_tokens": 43066908.0,
      "step": 18785
    },
    {
      "entropy": 5.213730382919311,
      "epoch": 1.8049951969260327,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004677403073948139,
      "loss": 4.9868,
      "mean_token_accuracy": 0.2203251078724861,
      "num_tokens": 43079736.0,
      "step": 18790
    },
    {
      "entropy": 5.194493389129638,
      "epoch": 1.8054755043227666,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004677226168352416,
      "loss": 5.0094,
      "mean_token_accuracy": 0.21221591532230377,
      "num_tokens": 43091600.0,
      "step": 18795
    },
    {
      "entropy": 5.18105297088623,
      "epoch": 1.8059558117195005,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004677049218012223,
      "loss": 4.9279,
      "mean_token_accuracy": 0.21940095871686935,
      "num_tokens": 43103544.0,
      "step": 18800
    },
    {
      "entropy": 5.18570761680603,
      "epoch": 1.8064361191162344,
      "grad_norm": 1.265625,
      "learning_rate": 0.000467687222293167,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22106406837701797,
      "num_tokens": 43113533.0,
      "step": 18805
    },
    {
      "entropy": 5.097306728363037,
      "epoch": 1.8069164265129682,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004676695183114866,
      "loss": 4.8849,
      "mean_token_accuracy": 0.2233000233769417,
      "num_tokens": 43124917.0,
      "step": 18810
    },
    {
      "entropy": 5.167766952514649,
      "epoch": 1.8073967339097021,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000467651809856592,
      "loss": 4.9251,
      "mean_token_accuracy": 0.2152344301342964,
      "num_tokens": 43136808.0,
      "step": 18815
    },
    {
      "entropy": 5.220812368392944,
      "epoch": 1.807877041306436,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046763409692889446,
      "loss": 4.976,
      "mean_token_accuracy": 0.21131496876478195,
      "num_tokens": 43148567.0,
      "step": 18820
    },
    {
      "entropy": 5.122064304351807,
      "epoch": 1.8083573487031699,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046761637952880516,
      "loss": 4.8313,
      "mean_token_accuracy": 0.22551150918006896,
      "num_tokens": 43160343.0,
      "step": 18825
    },
    {
      "entropy": 5.2211480140686035,
      "epoch": 1.808837656099904,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046759865765673555,
      "loss": 5.0692,
      "mean_token_accuracy": 0.2059829756617546,
      "num_tokens": 43173546.0,
      "step": 18830
    },
    {
      "entropy": 5.246525764465332,
      "epoch": 1.8093179634966379,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000467580931313097,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22781742364168167,
      "num_tokens": 43183846.0,
      "step": 18835
    },
    {
      "entropy": 5.314433813095093,
      "epoch": 1.8097982708933718,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046756320049830106,
      "loss": 5.0323,
      "mean_token_accuracy": 0.21216547340154648,
      "num_tokens": 43195612.0,
      "step": 18840
    },
    {
      "entropy": 5.142131233215332,
      "epoch": 1.8102785782901056,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004675454652127594,
      "loss": 4.9386,
      "mean_token_accuracy": 0.21406775265932082,
      "num_tokens": 43207439.0,
      "step": 18845
    },
    {
      "entropy": 5.20392599105835,
      "epoch": 1.8107588856868397,
      "grad_norm": 1.40625,
      "learning_rate": 0.00046752772545688377,
      "loss": 4.9286,
      "mean_token_accuracy": 0.21485934257507325,
      "num_tokens": 43218066.0,
      "step": 18850
    },
    {
      "entropy": 5.200348949432373,
      "epoch": 1.8112391930835736,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004675099812310861,
      "loss": 4.9504,
      "mean_token_accuracy": 0.21336986869573593,
      "num_tokens": 43230679.0,
      "step": 18855
    },
    {
      "entropy": 5.16449761390686,
      "epoch": 1.8117195004803075,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004674922325357782,
      "loss": 4.9696,
      "mean_token_accuracy": 0.22022107094526291,
      "num_tokens": 43241566.0,
      "step": 18860
    },
    {
      "entropy": 5.277895212173462,
      "epoch": 1.8121998078770414,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046747447937137235,
      "loss": 4.9975,
      "mean_token_accuracy": 0.21173725724220277,
      "num_tokens": 43252630.0,
      "step": 18865
    },
    {
      "entropy": 5.3328132152557375,
      "epoch": 1.8126801152737753,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046745672173828057,
      "loss": 5.1186,
      "mean_token_accuracy": 0.20772823244333266,
      "num_tokens": 43264879.0,
      "step": 18870
    },
    {
      "entropy": 5.169169139862061,
      "epoch": 1.8131604226705091,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004674389596369151,
      "loss": 4.9398,
      "mean_token_accuracy": 0.21993134319782257,
      "num_tokens": 43277321.0,
      "step": 18875
    },
    {
      "entropy": 5.205772972106933,
      "epoch": 1.813640730067243,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046742119306768855,
      "loss": 4.9798,
      "mean_token_accuracy": 0.2221955493092537,
      "num_tokens": 43289711.0,
      "step": 18880
    },
    {
      "entropy": 5.236990308761596,
      "epoch": 1.814121037463977,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004674034220310132,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21330020129680632,
      "num_tokens": 43301372.0,
      "step": 18885
    },
    {
      "entropy": 5.18942666053772,
      "epoch": 1.8146013448607108,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046738564652730176,
      "loss": 4.8155,
      "mean_token_accuracy": 0.22829407155513765,
      "num_tokens": 43312960.0,
      "step": 18890
    },
    {
      "entropy": 5.117136430740357,
      "epoch": 1.8150816522574447,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004673678665569669,
      "loss": 4.9009,
      "mean_token_accuracy": 0.21657486110925675,
      "num_tokens": 43325039.0,
      "step": 18895
    },
    {
      "entropy": 5.232519960403442,
      "epoch": 1.8155619596541785,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004673500821204213,
      "loss": 5.0266,
      "mean_token_accuracy": 0.2168407455086708,
      "num_tokens": 43337064.0,
      "step": 18900
    },
    {
      "entropy": 5.3089766025543215,
      "epoch": 1.8160422670509124,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000467332293218078,
      "loss": 4.9052,
      "mean_token_accuracy": 0.21628721058368683,
      "num_tokens": 43349307.0,
      "step": 18905
    },
    {
      "entropy": 5.125462675094605,
      "epoch": 1.8165225744476465,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046731449985035,
      "loss": 4.9029,
      "mean_token_accuracy": 0.2330257326364517,
      "num_tokens": 43362566.0,
      "step": 18910
    },
    {
      "entropy": 5.1600775718688965,
      "epoch": 1.8170028818443804,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046729670201765036,
      "loss": 5.0265,
      "mean_token_accuracy": 0.21210978478193282,
      "num_tokens": 43374516.0,
      "step": 18915
    },
    {
      "entropy": 5.170306634902954,
      "epoch": 1.8174831892411143,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046727889972039227,
      "loss": 4.8337,
      "mean_token_accuracy": 0.22460660040378572,
      "num_tokens": 43386615.0,
      "step": 18920
    },
    {
      "entropy": 5.214338636398315,
      "epoch": 1.8179634966378482,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046726109295898904,
      "loss": 4.9598,
      "mean_token_accuracy": 0.21729786396026612,
      "num_tokens": 43396582.0,
      "step": 18925
    },
    {
      "entropy": 5.1587584018707275,
      "epoch": 1.8184438040345823,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004672432817338542,
      "loss": 4.8592,
      "mean_token_accuracy": 0.22916062474250792,
      "num_tokens": 43407617.0,
      "step": 18930
    },
    {
      "entropy": 5.191878986358643,
      "epoch": 1.8189241114313162,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046722546604540115,
      "loss": 4.9982,
      "mean_token_accuracy": 0.21417119055986406,
      "num_tokens": 43421048.0,
      "step": 18935
    },
    {
      "entropy": 5.248065567016601,
      "epoch": 1.81940441882805,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004672076458940436,
      "loss": 4.9914,
      "mean_token_accuracy": 0.21493884474039077,
      "num_tokens": 43433325.0,
      "step": 18940
    },
    {
      "entropy": 5.145903205871582,
      "epoch": 1.819884726224784,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046718982128019534,
      "loss": 4.842,
      "mean_token_accuracy": 0.221805602312088,
      "num_tokens": 43444817.0,
      "step": 18945
    },
    {
      "entropy": 5.20827465057373,
      "epoch": 1.8203650336215178,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046717199220427003,
      "loss": 4.9717,
      "mean_token_accuracy": 0.21594235450029373,
      "num_tokens": 43456295.0,
      "step": 18950
    },
    {
      "entropy": 5.203729295730591,
      "epoch": 1.8208453410182517,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046715415866668163,
      "loss": 4.9113,
      "mean_token_accuracy": 0.21596913337707518,
      "num_tokens": 43468218.0,
      "step": 18955
    },
    {
      "entropy": 5.1674182415008545,
      "epoch": 1.8213256484149856,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004671363206678443,
      "loss": 4.9563,
      "mean_token_accuracy": 0.2179381415247917,
      "num_tokens": 43480182.0,
      "step": 18960
    },
    {
      "entropy": 5.20355544090271,
      "epoch": 1.8218059558117194,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046711847820817215,
      "loss": 4.9494,
      "mean_token_accuracy": 0.21507776528596878,
      "num_tokens": 43491108.0,
      "step": 18965
    },
    {
      "entropy": 5.161630582809448,
      "epoch": 1.8222862632084533,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004671006312880794,
      "loss": 4.8922,
      "mean_token_accuracy": 0.22552970051765442,
      "num_tokens": 43503530.0,
      "step": 18970
    },
    {
      "entropy": 5.214934873580932,
      "epoch": 1.8227665706051872,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004670827799079805,
      "loss": 4.9817,
      "mean_token_accuracy": 0.22445722222328185,
      "num_tokens": 43514805.0,
      "step": 18975
    },
    {
      "entropy": 5.2626060962677,
      "epoch": 1.823246878001921,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046706492406828966,
      "loss": 5.079,
      "mean_token_accuracy": 0.21369309723377228,
      "num_tokens": 43526027.0,
      "step": 18980
    },
    {
      "entropy": 5.235404586791992,
      "epoch": 1.8237271853986552,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004670470637694217,
      "loss": 5.0066,
      "mean_token_accuracy": 0.2207424148917198,
      "num_tokens": 43537191.0,
      "step": 18985
    },
    {
      "entropy": 5.215746974945068,
      "epoch": 1.824207492795389,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004670291990117912,
      "loss": 4.9346,
      "mean_token_accuracy": 0.22158151268959045,
      "num_tokens": 43547494.0,
      "step": 18990
    },
    {
      "entropy": 5.17375955581665,
      "epoch": 1.824687800192123,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004670113297958128,
      "loss": 4.8812,
      "mean_token_accuracy": 0.22546596676111222,
      "num_tokens": 43559048.0,
      "step": 18995
    },
    {
      "entropy": 5.240265130996704,
      "epoch": 1.8251681075888568,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046699345612190155,
      "loss": 4.9713,
      "mean_token_accuracy": 0.21515217125415803,
      "num_tokens": 43569571.0,
      "step": 19000
    },
    {
      "entropy": 5.149558448791504,
      "epoch": 1.825648414985591,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046697557799047233,
      "loss": 4.8883,
      "mean_token_accuracy": 0.22440856993198394,
      "num_tokens": 43580092.0,
      "step": 19005
    },
    {
      "entropy": 5.136047983169556,
      "epoch": 1.8261287223823248,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004669576954019403,
      "loss": 4.8636,
      "mean_token_accuracy": 0.21904662400484085,
      "num_tokens": 43592867.0,
      "step": 19010
    },
    {
      "entropy": 5.205252647399902,
      "epoch": 1.8266090297790587,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004669398083567205,
      "loss": 4.9194,
      "mean_token_accuracy": 0.2199709951877594,
      "num_tokens": 43602618.0,
      "step": 19015
    },
    {
      "entropy": 5.230825138092041,
      "epoch": 1.8270893371757926,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004669219168552284,
      "loss": 4.9425,
      "mean_token_accuracy": 0.21868131160736085,
      "num_tokens": 43614425.0,
      "step": 19020
    },
    {
      "entropy": 5.100638437271118,
      "epoch": 1.8275696445725265,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046690402089787916,
      "loss": 4.8063,
      "mean_token_accuracy": 0.22643069177865982,
      "num_tokens": 43627334.0,
      "step": 19025
    },
    {
      "entropy": 5.171082448959351,
      "epoch": 1.8280499519692603,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004668861204850884,
      "loss": 4.9243,
      "mean_token_accuracy": 0.22072537690401078,
      "num_tokens": 43638509.0,
      "step": 19030
    },
    {
      "entropy": 5.135962629318238,
      "epoch": 1.8285302593659942,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046686821561727176,
      "loss": 4.8877,
      "mean_token_accuracy": 0.22678419798612595,
      "num_tokens": 43649633.0,
      "step": 19035
    },
    {
      "entropy": 5.22515549659729,
      "epoch": 1.829010566762728,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004668503062948449,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21895336657762526,
      "num_tokens": 43661175.0,
      "step": 19040
    },
    {
      "entropy": 5.183070468902588,
      "epoch": 1.829490874159462,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004668323925182236,
      "loss": 4.9321,
      "mean_token_accuracy": 0.2172745034098625,
      "num_tokens": 43672000.0,
      "step": 19045
    },
    {
      "entropy": 5.145504426956177,
      "epoch": 1.8299711815561959,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046681447428782377,
      "loss": 4.9025,
      "mean_token_accuracy": 0.22352875620126725,
      "num_tokens": 43682574.0,
      "step": 19050
    },
    {
      "entropy": 5.192563056945801,
      "epoch": 1.8304514889529298,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004667965516040613,
      "loss": 4.9812,
      "mean_token_accuracy": 0.21883598268032073,
      "num_tokens": 43693051.0,
      "step": 19055
    },
    {
      "entropy": 5.2547101974487305,
      "epoch": 1.8309317963496636,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004667786244673526,
      "loss": 4.9311,
      "mean_token_accuracy": 0.21740875989198685,
      "num_tokens": 43705471.0,
      "step": 19060
    },
    {
      "entropy": 5.116177034378052,
      "epoch": 1.8314121037463977,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046676069287811365,
      "loss": 4.8771,
      "mean_token_accuracy": 0.2297815650701523,
      "num_tokens": 43716676.0,
      "step": 19065
    },
    {
      "entropy": 5.252439498901367,
      "epoch": 1.8318924111431316,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004667427568367607,
      "loss": 4.9922,
      "mean_token_accuracy": 0.21458326131105424,
      "num_tokens": 43727997.0,
      "step": 19070
    },
    {
      "entropy": 5.192062854766846,
      "epoch": 1.8323727185398655,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046672481634371047,
      "loss": 4.9706,
      "mean_token_accuracy": 0.21397136598825456,
      "num_tokens": 43739163.0,
      "step": 19075
    },
    {
      "entropy": 5.210564231872558,
      "epoch": 1.8328530259365994,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046670687139937925,
      "loss": 4.9334,
      "mean_token_accuracy": 0.21575036346912385,
      "num_tokens": 43751442.0,
      "step": 19080
    },
    {
      "entropy": 5.176436185836792,
      "epoch": 1.8333333333333335,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004666889220041837,
      "loss": 4.9741,
      "mean_token_accuracy": 0.2195933848619461,
      "num_tokens": 43763430.0,
      "step": 19085
    },
    {
      "entropy": 5.1892101764678955,
      "epoch": 1.8338136407300674,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046667096815854056,
      "loss": 4.857,
      "mean_token_accuracy": 0.23411203026771546,
      "num_tokens": 43774660.0,
      "step": 19090
    },
    {
      "entropy": 5.10307559967041,
      "epoch": 1.8342939481268012,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004666530098628667,
      "loss": 4.8561,
      "mean_token_accuracy": 0.22632951736450196,
      "num_tokens": 43785997.0,
      "step": 19095
    },
    {
      "entropy": 5.191432189941406,
      "epoch": 1.8347742555235351,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004666350471175791,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21191216111183167,
      "num_tokens": 43797744.0,
      "step": 19100
    },
    {
      "entropy": 5.265373373031617,
      "epoch": 1.835254562920269,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004666170799230947,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21302994191646576,
      "num_tokens": 43810415.0,
      "step": 19105
    },
    {
      "entropy": 5.126765298843384,
      "epoch": 1.8357348703170029,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004665991082798307,
      "loss": 4.8877,
      "mean_token_accuracy": 0.22520233392715455,
      "num_tokens": 43822652.0,
      "step": 19110
    },
    {
      "entropy": 5.250832748413086,
      "epoch": 1.8362151777137368,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004665811321882043,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21739140450954436,
      "num_tokens": 43834607.0,
      "step": 19115
    },
    {
      "entropy": 5.138664245605469,
      "epoch": 1.8366954851104706,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046656315164863297,
      "loss": 4.9456,
      "mean_token_accuracy": 0.21869454383850098,
      "num_tokens": 43845124.0,
      "step": 19120
    },
    {
      "entropy": 5.21881742477417,
      "epoch": 1.8371757925072045,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046654516666153403,
      "loss": 5.019,
      "mean_token_accuracy": 0.21503590196371078,
      "num_tokens": 43855964.0,
      "step": 19125
    },
    {
      "entropy": 5.247625541687012,
      "epoch": 1.8376560999039384,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004665271772273251,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22623585760593415,
      "num_tokens": 43866659.0,
      "step": 19130
    },
    {
      "entropy": 5.2445197105407715,
      "epoch": 1.8381364073006723,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004665091833464239,
      "loss": 4.9918,
      "mean_token_accuracy": 0.21184805929660797,
      "num_tokens": 43877410.0,
      "step": 19135
    },
    {
      "entropy": 5.184673881530761,
      "epoch": 1.8386167146974062,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046649118501924805,
      "loss": 5.0154,
      "mean_token_accuracy": 0.2131138727068901,
      "num_tokens": 43889672.0,
      "step": 19140
    },
    {
      "entropy": 5.198420476913452,
      "epoch": 1.8390970220941403,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004664731822462154,
      "loss": 4.9683,
      "mean_token_accuracy": 0.22170667201280594,
      "num_tokens": 43900876.0,
      "step": 19145
    },
    {
      "entropy": 5.187807846069336,
      "epoch": 1.8395773294908742,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046645517502774415,
      "loss": 4.8412,
      "mean_token_accuracy": 0.2253549411892891,
      "num_tokens": 43911460.0,
      "step": 19150
    },
    {
      "entropy": 5.139834547042847,
      "epoch": 1.840057636887608,
      "grad_norm": 1.296875,
      "learning_rate": 0.00046643716336425224,
      "loss": 4.9201,
      "mean_token_accuracy": 0.2207843005657196,
      "num_tokens": 43922344.0,
      "step": 19155
    },
    {
      "entropy": 5.152490186691284,
      "epoch": 1.8405379442843421,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004664191472561578,
      "loss": 4.8553,
      "mean_token_accuracy": 0.22856233716011048,
      "num_tokens": 43933960.0,
      "step": 19160
    },
    {
      "entropy": 5.199237871170044,
      "epoch": 1.841018251681076,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004664011267038792,
      "loss": 5.0515,
      "mean_token_accuracy": 0.212656831741333,
      "num_tokens": 43945805.0,
      "step": 19165
    },
    {
      "entropy": 5.210168075561524,
      "epoch": 1.84149855907781,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046638310170783476,
      "loss": 4.9617,
      "mean_token_accuracy": 0.22276159673929213,
      "num_tokens": 43958318.0,
      "step": 19170
    },
    {
      "entropy": 5.154554843902588,
      "epoch": 1.8419788664745438,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000466365072268443,
      "loss": 4.8359,
      "mean_token_accuracy": 0.22998685240745545,
      "num_tokens": 43969385.0,
      "step": 19175
    },
    {
      "entropy": 5.129644393920898,
      "epoch": 1.8424591738712777,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004663470383861225,
      "loss": 4.9267,
      "mean_token_accuracy": 0.2242673695087433,
      "num_tokens": 43980383.0,
      "step": 19180
    },
    {
      "entropy": 5.21729850769043,
      "epoch": 1.8429394812680115,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000466329000061292,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21995636820793152,
      "num_tokens": 43992900.0,
      "step": 19185
    },
    {
      "entropy": 5.224373435974121,
      "epoch": 1.8434197886647454,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004663109572943702,
      "loss": 4.864,
      "mean_token_accuracy": 0.22331395298242568,
      "num_tokens": 44001823.0,
      "step": 19190
    },
    {
      "entropy": 5.1954779624938965,
      "epoch": 1.8439000960614793,
      "grad_norm": 1.140625,
      "learning_rate": 0.000466292910085776,
      "loss": 5.0213,
      "mean_token_accuracy": 0.21628468930721284,
      "num_tokens": 44014851.0,
      "step": 19195
    },
    {
      "entropy": 5.202737855911255,
      "epoch": 1.8443804034582132,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00046627485843592854,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2179575815796852,
      "num_tokens": 44025191.0,
      "step": 19200
    },
    {
      "entropy": 5.251354360580445,
      "epoch": 1.844860710854947,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046625680234524674,
      "loss": 4.9864,
      "mean_token_accuracy": 0.2152435302734375,
      "num_tokens": 44036319.0,
      "step": 19205
    },
    {
      "entropy": 5.273826169967651,
      "epoch": 1.845341018251681,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046623874181414993,
      "loss": 5.0082,
      "mean_token_accuracy": 0.212562695145607,
      "num_tokens": 44048618.0,
      "step": 19210
    },
    {
      "entropy": 5.179742670059204,
      "epoch": 1.8458213256484148,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004662206768430574,
      "loss": 4.9294,
      "mean_token_accuracy": 0.21890414953231813,
      "num_tokens": 44060754.0,
      "step": 19215
    },
    {
      "entropy": 5.18023829460144,
      "epoch": 1.846301633045149,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004662026074323886,
      "loss": 4.8875,
      "mean_token_accuracy": 0.21874402314424515,
      "num_tokens": 44071239.0,
      "step": 19220
    },
    {
      "entropy": 5.297530174255371,
      "epoch": 1.8467819404418828,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046618453358256303,
      "loss": 4.9805,
      "mean_token_accuracy": 0.2104551762342453,
      "num_tokens": 44082635.0,
      "step": 19225
    },
    {
      "entropy": 5.229009532928467,
      "epoch": 1.8472622478386167,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046616645529400026,
      "loss": 4.9043,
      "mean_token_accuracy": 0.22737446129322053,
      "num_tokens": 44093680.0,
      "step": 19230
    },
    {
      "entropy": 5.169652128219605,
      "epoch": 1.8477425552353506,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004661483725671201,
      "loss": 4.9203,
      "mean_token_accuracy": 0.22150295376777648,
      "num_tokens": 44104490.0,
      "step": 19235
    },
    {
      "entropy": 5.098169231414795,
      "epoch": 1.8482228626320847,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046613028540234226,
      "loss": 4.8587,
      "mean_token_accuracy": 0.22310819625854492,
      "num_tokens": 44116961.0,
      "step": 19240
    },
    {
      "entropy": 5.181032037734985,
      "epoch": 1.8487031700288186,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004661121938000867,
      "loss": 4.883,
      "mean_token_accuracy": 0.22506196647882462,
      "num_tokens": 44127264.0,
      "step": 19245
    },
    {
      "entropy": 5.169287014007568,
      "epoch": 1.8491834774255524,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004660940977607736,
      "loss": 4.898,
      "mean_token_accuracy": 0.2228582575917244,
      "num_tokens": 44139374.0,
      "step": 19250
    },
    {
      "entropy": 5.087090587615966,
      "epoch": 1.8496637848222863,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046607599728482285,
      "loss": 4.8092,
      "mean_token_accuracy": 0.21847768127918243,
      "num_tokens": 44150416.0,
      "step": 19255
    },
    {
      "entropy": 5.212874937057495,
      "epoch": 1.8501440922190202,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046605789237265496,
      "loss": 4.9473,
      "mean_token_accuracy": 0.220275416970253,
      "num_tokens": 44162094.0,
      "step": 19260
    },
    {
      "entropy": 5.234938669204712,
      "epoch": 1.850624399615754,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046603978302469,
      "loss": 4.9691,
      "mean_token_accuracy": 0.21797178238630294,
      "num_tokens": 44174526.0,
      "step": 19265
    },
    {
      "entropy": 5.175544595718383,
      "epoch": 1.851104707012488,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004660216692413486,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2195317953824997,
      "num_tokens": 44185961.0,
      "step": 19270
    },
    {
      "entropy": 5.09603476524353,
      "epoch": 1.8515850144092219,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004660035510230513,
      "loss": 4.8979,
      "mean_token_accuracy": 0.21993931829929353,
      "num_tokens": 44198431.0,
      "step": 19275
    },
    {
      "entropy": 5.218553304672241,
      "epoch": 1.8520653218059557,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004659854283702186,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22929475456476212,
      "num_tokens": 44210518.0,
      "step": 19280
    },
    {
      "entropy": 5.0598798274993895,
      "epoch": 1.8525456292026896,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004659673012832715,
      "loss": 4.814,
      "mean_token_accuracy": 0.2233606457710266,
      "num_tokens": 44222300.0,
      "step": 19285
    },
    {
      "entropy": 5.209354877471924,
      "epoch": 1.8530259365994235,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004659491697626306,
      "loss": 4.9717,
      "mean_token_accuracy": 0.21362395137548446,
      "num_tokens": 44233142.0,
      "step": 19290
    },
    {
      "entropy": 5.169251728057861,
      "epoch": 1.8535062439961574,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046593103380871705,
      "loss": 4.9694,
      "mean_token_accuracy": 0.2204439103603363,
      "num_tokens": 44246507.0,
      "step": 19295
    },
    {
      "entropy": 5.2450323581695555,
      "epoch": 1.8539865513928915,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046591289342195184,
      "loss": 4.9574,
      "mean_token_accuracy": 0.21403302252292633,
      "num_tokens": 44257451.0,
      "step": 19300
    },
    {
      "entropy": 5.148036479949951,
      "epoch": 1.8544668587896254,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004658947486027562,
      "loss": 4.8572,
      "mean_token_accuracy": 0.22469048500061034,
      "num_tokens": 44269157.0,
      "step": 19305
    },
    {
      "entropy": 5.114278507232666,
      "epoch": 1.8549471661863592,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046587659935155124,
      "loss": 4.8653,
      "mean_token_accuracy": 0.22367693334817887,
      "num_tokens": 44279890.0,
      "step": 19310
    },
    {
      "entropy": 5.172156476974488,
      "epoch": 1.8554274735830933,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046585844566875845,
      "loss": 4.8064,
      "mean_token_accuracy": 0.2282515585422516,
      "num_tokens": 44291457.0,
      "step": 19315
    },
    {
      "entropy": 5.132070636749267,
      "epoch": 1.8559077809798272,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004658402875547993,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22107501178979874,
      "num_tokens": 44301692.0,
      "step": 19320
    },
    {
      "entropy": 5.16545615196228,
      "epoch": 1.856388088376561,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046582212501009533,
      "loss": 4.9567,
      "mean_token_accuracy": 0.21611976027488708,
      "num_tokens": 44313540.0,
      "step": 19325
    },
    {
      "entropy": 5.214330959320068,
      "epoch": 1.856868395773295,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046580395803506825,
      "loss": 5.0147,
      "mean_token_accuracy": 0.22329181879758836,
      "num_tokens": 44324970.0,
      "step": 19330
    },
    {
      "entropy": 5.2687489032745365,
      "epoch": 1.8573487031700289,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004657857866301399,
      "loss": 4.9372,
      "mean_token_accuracy": 0.2160044401884079,
      "num_tokens": 44336521.0,
      "step": 19335
    },
    {
      "entropy": 5.099783325195313,
      "epoch": 1.8578290105667628,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00046576761079573204,
      "loss": 4.8327,
      "mean_token_accuracy": 0.2321474239230156,
      "num_tokens": 44347128.0,
      "step": 19340
    },
    {
      "entropy": 5.201531982421875,
      "epoch": 1.8583093179634966,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004657494305322667,
      "loss": 4.9802,
      "mean_token_accuracy": 0.21446518152952193,
      "num_tokens": 44358503.0,
      "step": 19345
    },
    {
      "entropy": 5.229052400588989,
      "epoch": 1.8587896253602305,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004657312458401661,
      "loss": 4.9548,
      "mean_token_accuracy": 0.22427042722702026,
      "num_tokens": 44370849.0,
      "step": 19350
    },
    {
      "entropy": 5.249718284606933,
      "epoch": 1.8592699327569644,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004657130567198522,
      "loss": 4.8671,
      "mean_token_accuracy": 0.2219814956188202,
      "num_tokens": 44383806.0,
      "step": 19355
    },
    {
      "entropy": 5.094815635681153,
      "epoch": 1.8597502401536983,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046569486317174746,
      "loss": 4.7336,
      "mean_token_accuracy": 0.22967343926429748,
      "num_tokens": 44393937.0,
      "step": 19360
    },
    {
      "entropy": 5.154543733596801,
      "epoch": 1.8602305475504322,
      "grad_norm": 1.125,
      "learning_rate": 0.0004656766651962742,
      "loss": 4.9669,
      "mean_token_accuracy": 0.22510907649993897,
      "num_tokens": 44404967.0,
      "step": 19365
    },
    {
      "entropy": 5.166842889785767,
      "epoch": 1.860710854947166,
      "grad_norm": 1.125,
      "learning_rate": 0.0004656584627938551,
      "loss": 4.9372,
      "mean_token_accuracy": 0.21965404003858566,
      "num_tokens": 44416421.0,
      "step": 19370
    },
    {
      "entropy": 5.1779059886932375,
      "epoch": 1.8611911623439001,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046564025596491254,
      "loss": 4.9078,
      "mean_token_accuracy": 0.22394825369119645,
      "num_tokens": 44427145.0,
      "step": 19375
    },
    {
      "entropy": 5.156735181808472,
      "epoch": 1.861671469740634,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004656220447098693,
      "loss": 4.8729,
      "mean_token_accuracy": 0.22575816959142686,
      "num_tokens": 44438456.0,
      "step": 19380
    },
    {
      "entropy": 5.111081790924072,
      "epoch": 1.862151777137368,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004656038290291483,
      "loss": 4.8449,
      "mean_token_accuracy": 0.21852281391620637,
      "num_tokens": 44450366.0,
      "step": 19385
    },
    {
      "entropy": 5.1317895412445065,
      "epoch": 1.8626320845341018,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004655856089231723,
      "loss": 4.7987,
      "mean_token_accuracy": 0.22956424504518508,
      "num_tokens": 44461246.0,
      "step": 19390
    },
    {
      "entropy": 5.189972639083862,
      "epoch": 1.8631123919308359,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004655673843923644,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22044617980718612,
      "num_tokens": 44473099.0,
      "step": 19395
    },
    {
      "entropy": 5.2147908210754395,
      "epoch": 1.8635926993275698,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004655491554371477,
      "loss": 4.9175,
      "mean_token_accuracy": 0.22037432342767715,
      "num_tokens": 44484516.0,
      "step": 19400
    },
    {
      "entropy": 5.094288778305054,
      "epoch": 1.8640730067243036,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046553092205794543,
      "loss": 4.8028,
      "mean_token_accuracy": 0.23056435734033584,
      "num_tokens": 44495500.0,
      "step": 19405
    },
    {
      "entropy": 5.131862831115723,
      "epoch": 1.8645533141210375,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046551268425518096,
      "loss": 4.8639,
      "mean_token_accuracy": 0.22181136161088943,
      "num_tokens": 44506346.0,
      "step": 19410
    },
    {
      "entropy": 5.265206146240234,
      "epoch": 1.8650336215177714,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004654944420292776,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21692022532224656,
      "num_tokens": 44517234.0,
      "step": 19415
    },
    {
      "entropy": 5.139484930038452,
      "epoch": 1.8655139289145053,
      "grad_norm": 1.140625,
      "learning_rate": 0.000465476195380659,
      "loss": 4.817,
      "mean_token_accuracy": 0.22961812913417817,
      "num_tokens": 44528543.0,
      "step": 19420
    },
    {
      "entropy": 5.191559076309204,
      "epoch": 1.8659942363112392,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004654579443097487,
      "loss": 4.9637,
      "mean_token_accuracy": 0.2207016095519066,
      "num_tokens": 44539385.0,
      "step": 19425
    },
    {
      "entropy": 5.093587732315063,
      "epoch": 1.866474543707973,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004654396888169705,
      "loss": 4.7871,
      "mean_token_accuracy": 0.22760196179151534,
      "num_tokens": 44550013.0,
      "step": 19430
    },
    {
      "entropy": 5.181816530227661,
      "epoch": 1.866954851104707,
      "grad_norm": 1.125,
      "learning_rate": 0.00046542142890274816,
      "loss": 4.9414,
      "mean_token_accuracy": 0.21927675604820251,
      "num_tokens": 44561926.0,
      "step": 19435
    },
    {
      "entropy": 5.129086112976074,
      "epoch": 1.8674351585014408,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004654031645675057,
      "loss": 4.8939,
      "mean_token_accuracy": 0.22289102375507355,
      "num_tokens": 44574589.0,
      "step": 19440
    },
    {
      "entropy": 5.185785245895386,
      "epoch": 1.8679154658981747,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004653848958116672,
      "loss": 4.9316,
      "mean_token_accuracy": 0.22065257728099824,
      "num_tokens": 44586531.0,
      "step": 19445
    },
    {
      "entropy": 5.203807020187378,
      "epoch": 1.8683957732949086,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046536662263565667,
      "loss": 4.9,
      "mean_token_accuracy": 0.22689249664545058,
      "num_tokens": 44597432.0,
      "step": 19450
    },
    {
      "entropy": 5.213404798507691,
      "epoch": 1.8688760806916427,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004653483450398985,
      "loss": 4.9666,
      "mean_token_accuracy": 0.21445707529783248,
      "num_tokens": 44609264.0,
      "step": 19455
    },
    {
      "entropy": 5.072569513320923,
      "epoch": 1.8693563880883766,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046533006302481694,
      "loss": 4.8196,
      "mean_token_accuracy": 0.2308601126074791,
      "num_tokens": 44620410.0,
      "step": 19460
    },
    {
      "entropy": 5.157938480377197,
      "epoch": 1.8698366954851104,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004653117765908365,
      "loss": 4.9621,
      "mean_token_accuracy": 0.21627419590950012,
      "num_tokens": 44633123.0,
      "step": 19465
    },
    {
      "entropy": 5.23253173828125,
      "epoch": 1.8703170028818443,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004652934857383816,
      "loss": 4.9575,
      "mean_token_accuracy": 0.21102683544158934,
      "num_tokens": 44644693.0,
      "step": 19470
    },
    {
      "entropy": 5.135721492767334,
      "epoch": 1.8707973102785784,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004652751904678772,
      "loss": 4.865,
      "mean_token_accuracy": 0.23034926503896713,
      "num_tokens": 44656669.0,
      "step": 19475
    },
    {
      "entropy": 5.172334861755371,
      "epoch": 1.8712776176753123,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046525689077974775,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22842198759317398,
      "num_tokens": 44667938.0,
      "step": 19480
    },
    {
      "entropy": 5.202755689620972,
      "epoch": 1.8717579250720462,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046523858667441834,
      "loss": 4.9665,
      "mean_token_accuracy": 0.2158089905977249,
      "num_tokens": 44678967.0,
      "step": 19485
    },
    {
      "entropy": 5.192142200469971,
      "epoch": 1.87223823246878,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004652202781523138,
      "loss": 4.8134,
      "mean_token_accuracy": 0.23218003958463668,
      "num_tokens": 44689583.0,
      "step": 19490
    },
    {
      "entropy": 5.109318447113037,
      "epoch": 1.872718539865514,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004652019652138592,
      "loss": 4.8524,
      "mean_token_accuracy": 0.22568164765834808,
      "num_tokens": 44700942.0,
      "step": 19495
    },
    {
      "entropy": 5.140614080429077,
      "epoch": 1.8731988472622478,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004651836478594798,
      "loss": 4.8629,
      "mean_token_accuracy": 0.22556458860635759,
      "num_tokens": 44712323.0,
      "step": 19500
    },
    {
      "entropy": 5.236519384384155,
      "epoch": 1.8736791546589817,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004651653260896008,
      "loss": 5.0469,
      "mean_token_accuracy": 0.21366228461265563,
      "num_tokens": 44724033.0,
      "step": 19505
    },
    {
      "entropy": 5.19577054977417,
      "epoch": 1.8741594620557156,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046514699990464763,
      "loss": 4.965,
      "mean_token_accuracy": 0.21999471485614777,
      "num_tokens": 44735388.0,
      "step": 19510
    },
    {
      "entropy": 5.16780161857605,
      "epoch": 1.8746397694524495,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004651286693050458,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22349400371313094,
      "num_tokens": 44748165.0,
      "step": 19515
    },
    {
      "entropy": 5.243159341812134,
      "epoch": 1.8751200768491834,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004651103342912207,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2228372275829315,
      "num_tokens": 44758500.0,
      "step": 19520
    },
    {
      "entropy": 5.237855434417725,
      "epoch": 1.8756003842459172,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046509199486359824,
      "loss": 5.0421,
      "mean_token_accuracy": 0.21194336116313933,
      "num_tokens": 44770347.0,
      "step": 19525
    },
    {
      "entropy": 5.206346130371093,
      "epoch": 1.8760806916426513,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046507365102260403,
      "loss": 4.9043,
      "mean_token_accuracy": 0.22151322960853576,
      "num_tokens": 44782365.0,
      "step": 19530
    },
    {
      "entropy": 5.138313627243042,
      "epoch": 1.8765609990393852,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046505530276866417,
      "loss": 4.8138,
      "mean_token_accuracy": 0.22915321737527847,
      "num_tokens": 44793128.0,
      "step": 19535
    },
    {
      "entropy": 5.134373998641967,
      "epoch": 1.877041306436119,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046503695010220443,
      "loss": 4.8921,
      "mean_token_accuracy": 0.2264243707060814,
      "num_tokens": 44804236.0,
      "step": 19540
    },
    {
      "entropy": 5.11343355178833,
      "epoch": 1.877521613832853,
      "grad_norm": 1.09375,
      "learning_rate": 0.000465018593023651,
      "loss": 4.8306,
      "mean_token_accuracy": 0.22449928820133208,
      "num_tokens": 44815983.0,
      "step": 19545
    },
    {
      "entropy": 5.20868182182312,
      "epoch": 1.878001921229587,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004650002315334302,
      "loss": 5.037,
      "mean_token_accuracy": 0.211461578309536,
      "num_tokens": 44828267.0,
      "step": 19550
    },
    {
      "entropy": 5.126822900772095,
      "epoch": 1.878482228626321,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000464981865631968,
      "loss": 4.7859,
      "mean_token_accuracy": 0.22668870091438292,
      "num_tokens": 44838668.0,
      "step": 19555
    },
    {
      "entropy": 5.136554098129272,
      "epoch": 1.8789625360230549,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004649634953196912,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22282899171113968,
      "num_tokens": 44849492.0,
      "step": 19560
    },
    {
      "entropy": 5.297608804702759,
      "epoch": 1.8794428434197887,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046494512059702605,
      "loss": 5.0766,
      "mean_token_accuracy": 0.20706631541252135,
      "num_tokens": 44861388.0,
      "step": 19565
    },
    {
      "entropy": 5.133195209503174,
      "epoch": 1.8799231508165226,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004649267414643992,
      "loss": 4.9056,
      "mean_token_accuracy": 0.2260555610060692,
      "num_tokens": 44873353.0,
      "step": 19570
    },
    {
      "entropy": 5.182441759109497,
      "epoch": 1.8804034582132565,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004649083579222374,
      "loss": 4.9344,
      "mean_token_accuracy": 0.21878827810287477,
      "num_tokens": 44885020.0,
      "step": 19575
    },
    {
      "entropy": 5.263147163391113,
      "epoch": 1.8808837656099904,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046488996997096744,
      "loss": 5.0319,
      "mean_token_accuracy": 0.21065862029790877,
      "num_tokens": 44896743.0,
      "step": 19580
    },
    {
      "entropy": 5.198274898529053,
      "epoch": 1.8813640730067243,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004648715776110162,
      "loss": 4.8848,
      "mean_token_accuracy": 0.22128121703863143,
      "num_tokens": 44909138.0,
      "step": 19585
    },
    {
      "entropy": 5.127352952957153,
      "epoch": 1.8818443804034581,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004648531808428108,
      "loss": 4.9093,
      "mean_token_accuracy": 0.2228813961148262,
      "num_tokens": 44921139.0,
      "step": 19590
    },
    {
      "entropy": 5.144074773788452,
      "epoch": 1.882324687800192,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004648347796667782,
      "loss": 4.792,
      "mean_token_accuracy": 0.2232631877064705,
      "num_tokens": 44933282.0,
      "step": 19595
    },
    {
      "entropy": 5.216866493225098,
      "epoch": 1.882804995196926,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004648163740833458,
      "loss": 4.9514,
      "mean_token_accuracy": 0.21575527638196945,
      "num_tokens": 44946350.0,
      "step": 19600
    },
    {
      "entropy": 5.170894908905029,
      "epoch": 1.8832853025936598,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046479796409294076,
      "loss": 4.8957,
      "mean_token_accuracy": 0.21840279996395112,
      "num_tokens": 44958348.0,
      "step": 19605
    },
    {
      "entropy": 5.15054235458374,
      "epoch": 1.8837656099903939,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004647795496959907,
      "loss": 4.8383,
      "mean_token_accuracy": 0.2310398906469345,
      "num_tokens": 44970206.0,
      "step": 19610
    },
    {
      "entropy": 5.155156135559082,
      "epoch": 1.8842459173871278,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046476113089292286,
      "loss": 4.928,
      "mean_token_accuracy": 0.22069868743419646,
      "num_tokens": 44980814.0,
      "step": 19615
    },
    {
      "entropy": 5.180980205535889,
      "epoch": 1.8847262247838616,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004647427076841651,
      "loss": 4.8616,
      "mean_token_accuracy": 0.2269600361585617,
      "num_tokens": 44991148.0,
      "step": 19620
    },
    {
      "entropy": 5.0984704971313475,
      "epoch": 1.8852065321805955,
      "grad_norm": 1.59375,
      "learning_rate": 0.00046472428007014515,
      "loss": 4.8618,
      "mean_token_accuracy": 0.2320017099380493,
      "num_tokens": 45003070.0,
      "step": 19625
    },
    {
      "entropy": 5.109103441238403,
      "epoch": 1.8856868395773296,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004647058480512907,
      "loss": 4.8849,
      "mean_token_accuracy": 0.21954857259988786,
      "num_tokens": 45015760.0,
      "step": 19630
    },
    {
      "entropy": 5.225310945510865,
      "epoch": 1.8861671469740635,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046468741162802987,
      "loss": 4.9441,
      "mean_token_accuracy": 0.215548600256443,
      "num_tokens": 45026297.0,
      "step": 19635
    },
    {
      "entropy": 5.132082462310791,
      "epoch": 1.8866474543707974,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004646689708007905,
      "loss": 4.8221,
      "mean_token_accuracy": 0.22119199484586716,
      "num_tokens": 45037814.0,
      "step": 19640
    },
    {
      "entropy": 5.133320760726929,
      "epoch": 1.8871277617675313,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046465052557000087,
      "loss": 4.8878,
      "mean_token_accuracy": 0.2298172280192375,
      "num_tokens": 45050632.0,
      "step": 19645
    },
    {
      "entropy": 5.25730938911438,
      "epoch": 1.8876080691642652,
      "grad_norm": 1.5,
      "learning_rate": 0.00046463207593608916,
      "loss": 4.9165,
      "mean_token_accuracy": 0.21724371165037154,
      "num_tokens": 45061866.0,
      "step": 19650
    },
    {
      "entropy": 5.154073858261109,
      "epoch": 1.888088376560999,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004646136218994837,
      "loss": 4.9846,
      "mean_token_accuracy": 0.2128133222460747,
      "num_tokens": 45074550.0,
      "step": 19655
    },
    {
      "entropy": 5.208667135238647,
      "epoch": 1.888568683957733,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046459516346061304,
      "loss": 4.949,
      "mean_token_accuracy": 0.2200430691242218,
      "num_tokens": 45085710.0,
      "step": 19660
    },
    {
      "entropy": 5.195755577087402,
      "epoch": 1.8890489913544668,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046457670061990564,
      "loss": 4.9183,
      "mean_token_accuracy": 0.22321408689022065,
      "num_tokens": 45097917.0,
      "step": 19665
    },
    {
      "entropy": 5.141583776473999,
      "epoch": 1.8895292987512007,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046455823337779024,
      "loss": 4.8609,
      "mean_token_accuracy": 0.2292746901512146,
      "num_tokens": 45108455.0,
      "step": 19670
    },
    {
      "entropy": 5.236583471298218,
      "epoch": 1.8900096061479346,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004645397617346954,
      "loss": 5.0629,
      "mean_token_accuracy": 0.214216972887516,
      "num_tokens": 45120313.0,
      "step": 19675
    },
    {
      "entropy": 5.157782793045044,
      "epoch": 1.8904899135446684,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004645212856910502,
      "loss": 4.8662,
      "mean_token_accuracy": 0.22830743044614793,
      "num_tokens": 45131673.0,
      "step": 19680
    },
    {
      "entropy": 5.140381813049316,
      "epoch": 1.8909702209414025,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004645028052472835,
      "loss": 4.8924,
      "mean_token_accuracy": 0.22009116411209106,
      "num_tokens": 45143589.0,
      "step": 19685
    },
    {
      "entropy": 5.145933818817139,
      "epoch": 1.8914505283381364,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046448432040382444,
      "loss": 4.859,
      "mean_token_accuracy": 0.22494134455919265,
      "num_tokens": 45155904.0,
      "step": 19690
    },
    {
      "entropy": 5.15929012298584,
      "epoch": 1.8919308357348703,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000464465831161102,
      "loss": 4.8706,
      "mean_token_accuracy": 0.2262958660721779,
      "num_tokens": 45166802.0,
      "step": 19695
    },
    {
      "entropy": 5.222515726089478,
      "epoch": 1.8924111431316042,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004644473375195456,
      "loss": 4.9591,
      "mean_token_accuracy": 0.21609064638614656,
      "num_tokens": 45178986.0,
      "step": 19700
    },
    {
      "entropy": 5.1957310199737545,
      "epoch": 1.8928914505283383,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046442883947958466,
      "loss": 4.9095,
      "mean_token_accuracy": 0.2278654932975769,
      "num_tokens": 45190814.0,
      "step": 19705
    },
    {
      "entropy": 5.121941089630127,
      "epoch": 1.8933717579250722,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046441033704164845,
      "loss": 4.8714,
      "mean_token_accuracy": 0.22884267270565034,
      "num_tokens": 45202553.0,
      "step": 19710
    },
    {
      "entropy": 5.203513526916504,
      "epoch": 1.893852065321806,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004643918302061666,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21384691298007966,
      "num_tokens": 45214482.0,
      "step": 19715
    },
    {
      "entropy": 5.1419555187225345,
      "epoch": 1.89433237271854,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004643733189735689,
      "loss": 4.8785,
      "mean_token_accuracy": 0.22470485866069795,
      "num_tokens": 45226675.0,
      "step": 19720
    },
    {
      "entropy": 5.19923415184021,
      "epoch": 1.8948126801152738,
      "grad_norm": 1.109375,
      "learning_rate": 0.000464354803344285,
      "loss": 4.9026,
      "mean_token_accuracy": 0.22384996861219406,
      "num_tokens": 45238995.0,
      "step": 19725
    },
    {
      "entropy": 5.161758708953857,
      "epoch": 1.8952929875120077,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046433628331874496,
      "loss": 4.9428,
      "mean_token_accuracy": 0.22361512780189513,
      "num_tokens": 45250696.0,
      "step": 19730
    },
    {
      "entropy": 5.1652587890625,
      "epoch": 1.8957732949087416,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004643177588973785,
      "loss": 4.8192,
      "mean_token_accuracy": 0.22575814425945281,
      "num_tokens": 45262434.0,
      "step": 19735
    },
    {
      "entropy": 5.197325754165649,
      "epoch": 1.8962536023054755,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004642992300806159,
      "loss": 4.976,
      "mean_token_accuracy": 0.21495808362960817,
      "num_tokens": 45273267.0,
      "step": 19740
    },
    {
      "entropy": 5.153570365905762,
      "epoch": 1.8967339097022093,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004642806968688873,
      "loss": 4.8807,
      "mean_token_accuracy": 0.22358245253562928,
      "num_tokens": 45284601.0,
      "step": 19745
    },
    {
      "entropy": 5.118161153793335,
      "epoch": 1.8972142170989432,
      "grad_norm": 1.359375,
      "learning_rate": 0.00046426215926262295,
      "loss": 4.9069,
      "mean_token_accuracy": 0.2235897809267044,
      "num_tokens": 45296504.0,
      "step": 19750
    },
    {
      "entropy": 5.119107818603515,
      "epoch": 1.897694524495677,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004642436172622532,
      "loss": 4.8281,
      "mean_token_accuracy": 0.22985749393701554,
      "num_tokens": 45307232.0,
      "step": 19755
    },
    {
      "entropy": 5.104911470413208,
      "epoch": 1.898174831892411,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004642250708682086,
      "loss": 4.8001,
      "mean_token_accuracy": 0.23204401284456252,
      "num_tokens": 45317768.0,
      "step": 19760
    },
    {
      "entropy": 5.225300121307373,
      "epoch": 1.898655139289145,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00046420652008091984,
      "loss": 4.9656,
      "mean_token_accuracy": 0.22368326783180237,
      "num_tokens": 45329172.0,
      "step": 19765
    },
    {
      "entropy": 5.244212293624878,
      "epoch": 1.899135446685879,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004641879649008174,
      "loss": 4.9961,
      "mean_token_accuracy": 0.216484871506691,
      "num_tokens": 45339840.0,
      "step": 19770
    },
    {
      "entropy": 5.183950328826905,
      "epoch": 1.8996157540826129,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004641694053283323,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21732640713453294,
      "num_tokens": 45351532.0,
      "step": 19775
    },
    {
      "entropy": 5.18296627998352,
      "epoch": 1.9000960614793467,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046415084136389525,
      "loss": 4.9976,
      "mean_token_accuracy": 0.21421679854393005,
      "num_tokens": 45362872.0,
      "step": 19780
    },
    {
      "entropy": 5.386783075332642,
      "epoch": 1.9005763688760808,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004641322730079374,
      "loss": 5.0501,
      "mean_token_accuracy": 0.21394654661417006,
      "num_tokens": 45374877.0,
      "step": 19785
    },
    {
      "entropy": 5.221222257614135,
      "epoch": 1.9010566762728147,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004641137002608897,
      "loss": 4.9281,
      "mean_token_accuracy": 0.22398762553930282,
      "num_tokens": 45387136.0,
      "step": 19790
    },
    {
      "entropy": 5.06250147819519,
      "epoch": 1.9015369836695486,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046409512312318345,
      "loss": 4.8688,
      "mean_token_accuracy": 0.2260947436094284,
      "num_tokens": 45398310.0,
      "step": 19795
    },
    {
      "entropy": 5.161259508132934,
      "epoch": 1.9020172910662825,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046407654159524994,
      "loss": 4.8839,
      "mean_token_accuracy": 0.22223636656999587,
      "num_tokens": 45410456.0,
      "step": 19800
    },
    {
      "entropy": 5.220488452911377,
      "epoch": 1.9024975984630164,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046405795567752055,
      "loss": 4.8214,
      "mean_token_accuracy": 0.2237703561782837,
      "num_tokens": 45423292.0,
      "step": 19805
    },
    {
      "entropy": 5.0703473567962645,
      "epoch": 1.9029779058597502,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046403936537042686,
      "loss": 4.8884,
      "mean_token_accuracy": 0.2302060157060623,
      "num_tokens": 45434699.0,
      "step": 19810
    },
    {
      "entropy": 5.1167027950286865,
      "epoch": 1.9034582132564841,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046402077067440043,
      "loss": 4.8763,
      "mean_token_accuracy": 0.21490021646022797,
      "num_tokens": 45446849.0,
      "step": 19815
    },
    {
      "entropy": 5.21525297164917,
      "epoch": 1.903938520653218,
      "grad_norm": 1.765625,
      "learning_rate": 0.00046400217158987293,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22067243456840516,
      "num_tokens": 45458070.0,
      "step": 19820
    },
    {
      "entropy": 5.2091490745544435,
      "epoch": 1.9044188280499519,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046398356811727626,
      "loss": 4.9436,
      "mean_token_accuracy": 0.22412577718496324,
      "num_tokens": 45470829.0,
      "step": 19825
    },
    {
      "entropy": 5.151257038116455,
      "epoch": 1.9048991354466858,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004639649602570423,
      "loss": 4.8248,
      "mean_token_accuracy": 0.23086250722408294,
      "num_tokens": 45482354.0,
      "step": 19830
    },
    {
      "entropy": 5.110698318481445,
      "epoch": 1.9053794428434196,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046394634800960314,
      "loss": 4.8383,
      "mean_token_accuracy": 0.2259770855307579,
      "num_tokens": 45493819.0,
      "step": 19835
    },
    {
      "entropy": 5.1807060718536375,
      "epoch": 1.9058597502401537,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046392773137539074,
      "loss": 4.9536,
      "mean_token_accuracy": 0.21950149536132812,
      "num_tokens": 45504877.0,
      "step": 19840
    },
    {
      "entropy": 5.172350072860718,
      "epoch": 1.9063400576368876,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046390911035483744,
      "loss": 4.8858,
      "mean_token_accuracy": 0.22087870985269548,
      "num_tokens": 45516160.0,
      "step": 19845
    },
    {
      "entropy": 5.119910049438476,
      "epoch": 1.9068203650336215,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004638904849483756,
      "loss": 4.866,
      "mean_token_accuracy": 0.22153507322072982,
      "num_tokens": 45529028.0,
      "step": 19850
    },
    {
      "entropy": 5.1236656665802,
      "epoch": 1.9073006724303554,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046387185515643756,
      "loss": 4.8806,
      "mean_token_accuracy": 0.22557639628648757,
      "num_tokens": 45541507.0,
      "step": 19855
    },
    {
      "entropy": 5.2729024410247805,
      "epoch": 1.9077809798270895,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004638532209794559,
      "loss": 5.0152,
      "mean_token_accuracy": 0.21618373990058898,
      "num_tokens": 45553379.0,
      "step": 19860
    },
    {
      "entropy": 5.144784212112427,
      "epoch": 1.9082612872238234,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004638345824178631,
      "loss": 4.8219,
      "mean_token_accuracy": 0.22651127576828003,
      "num_tokens": 45564962.0,
      "step": 19865
    },
    {
      "entropy": 5.21383056640625,
      "epoch": 1.9087415946205573,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046381593947209215,
      "loss": 4.9857,
      "mean_token_accuracy": 0.2193000689148903,
      "num_tokens": 45575270.0,
      "step": 19870
    },
    {
      "entropy": 5.1509575843811035,
      "epoch": 1.9092219020172911,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004637972921425757,
      "loss": 4.849,
      "mean_token_accuracy": 0.22288160473108293,
      "num_tokens": 45587446.0,
      "step": 19875
    },
    {
      "entropy": 5.265188980102539,
      "epoch": 1.909702209414025,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046377864042974675,
      "loss": 5.0362,
      "mean_token_accuracy": 0.22315314263105393,
      "num_tokens": 45599598.0,
      "step": 19880
    },
    {
      "entropy": 5.183281660079956,
      "epoch": 1.910182516810759,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004637599843340384,
      "loss": 4.8263,
      "mean_token_accuracy": 0.2232552722096443,
      "num_tokens": 45610862.0,
      "step": 19885
    },
    {
      "entropy": 5.127306652069092,
      "epoch": 1.9106628242074928,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046374132385588356,
      "loss": 4.8477,
      "mean_token_accuracy": 0.23161111921072006,
      "num_tokens": 45621398.0,
      "step": 19890
    },
    {
      "entropy": 5.139720010757446,
      "epoch": 1.9111431316042267,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046372265899571576,
      "loss": 4.8557,
      "mean_token_accuracy": 0.2234599173069,
      "num_tokens": 45631139.0,
      "step": 19895
    },
    {
      "entropy": 5.185478448867798,
      "epoch": 1.9116234390009605,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046370398975396817,
      "loss": 4.958,
      "mean_token_accuracy": 0.221939554810524,
      "num_tokens": 45642439.0,
      "step": 19900
    },
    {
      "entropy": 5.1586061954498295,
      "epoch": 1.9121037463976944,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004636853161310743,
      "loss": 4.8135,
      "mean_token_accuracy": 0.22794032990932464,
      "num_tokens": 45653701.0,
      "step": 19905
    },
    {
      "entropy": 5.148637151718139,
      "epoch": 1.9125840537944283,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046366663812746764,
      "loss": 4.8789,
      "mean_token_accuracy": 0.23420808762311934,
      "num_tokens": 45663870.0,
      "step": 19910
    },
    {
      "entropy": 5.153133058547974,
      "epoch": 1.9130643611911622,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004636479557435818,
      "loss": 4.8949,
      "mean_token_accuracy": 0.22397466897964477,
      "num_tokens": 45674597.0,
      "step": 19915
    },
    {
      "entropy": 5.190211248397827,
      "epoch": 1.9135446685878963,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046362926897985067,
      "loss": 4.9525,
      "mean_token_accuracy": 0.21714911609888077,
      "num_tokens": 45685871.0,
      "step": 19920
    },
    {
      "entropy": 5.146980476379395,
      "epoch": 1.9140249759846302,
      "grad_norm": 1.21875,
      "learning_rate": 0.000463610577836708,
      "loss": 4.9455,
      "mean_token_accuracy": 0.22072536498308182,
      "num_tokens": 45696742.0,
      "step": 19925
    },
    {
      "entropy": 5.211127662658692,
      "epoch": 1.914505283381364,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046359188231458783,
      "loss": 4.9506,
      "mean_token_accuracy": 0.22291216552257537,
      "num_tokens": 45708398.0,
      "step": 19930
    },
    {
      "entropy": 5.185349225997925,
      "epoch": 1.914985590778098,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046357318241392414,
      "loss": 4.9755,
      "mean_token_accuracy": 0.21860784590244292,
      "num_tokens": 45721525.0,
      "step": 19935
    },
    {
      "entropy": 5.057275485992432,
      "epoch": 1.915465898174832,
      "grad_norm": 1.1875,
      "learning_rate": 0.000463554478135151,
      "loss": 4.8373,
      "mean_token_accuracy": 0.2264331117272377,
      "num_tokens": 45732089.0,
      "step": 19940
    },
    {
      "entropy": 5.173133611679077,
      "epoch": 1.915946205571566,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004635357694787029,
      "loss": 4.8602,
      "mean_token_accuracy": 0.2245445042848587,
      "num_tokens": 45743243.0,
      "step": 19945
    },
    {
      "entropy": 5.179460334777832,
      "epoch": 1.9164265129682998,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000463517056445014,
      "loss": 4.8864,
      "mean_token_accuracy": 0.21932154893875122,
      "num_tokens": 45754113.0,
      "step": 19950
    },
    {
      "entropy": 5.0580676078796385,
      "epoch": 1.9169068203650337,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046349833903451884,
      "loss": 4.8273,
      "mean_token_accuracy": 0.22929075211286545,
      "num_tokens": 45766190.0,
      "step": 19955
    },
    {
      "entropy": 5.1914393424987795,
      "epoch": 1.9173871277617676,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046347961724765196,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2151201084256172,
      "num_tokens": 45777736.0,
      "step": 19960
    },
    {
      "entropy": 5.237403154373169,
      "epoch": 1.9178674351585014,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046346089108484806,
      "loss": 4.9489,
      "mean_token_accuracy": 0.21945378184318542,
      "num_tokens": 45788813.0,
      "step": 19965
    },
    {
      "entropy": 5.175634527206421,
      "epoch": 1.9183477425552353,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046344216054654193,
      "loss": 4.9456,
      "mean_token_accuracy": 0.22117733359336852,
      "num_tokens": 45800671.0,
      "step": 19970
    },
    {
      "entropy": 5.140288257598877,
      "epoch": 1.9188280499519692,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046342342563316833,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22719415724277497,
      "num_tokens": 45811587.0,
      "step": 19975
    },
    {
      "entropy": 5.116430234909058,
      "epoch": 1.919308357348703,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00046340468634516223,
      "loss": 4.7934,
      "mean_token_accuracy": 0.23257973045110703,
      "num_tokens": 45823381.0,
      "step": 19980
    },
    {
      "entropy": 5.216929340362549,
      "epoch": 1.919788664745437,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046338594268295884,
      "loss": 4.9192,
      "mean_token_accuracy": 0.21901257634162902,
      "num_tokens": 45833560.0,
      "step": 19985
    },
    {
      "entropy": 5.173340368270874,
      "epoch": 1.9202689721421708,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004633671946469933,
      "loss": 4.8792,
      "mean_token_accuracy": 0.23102484941482543,
      "num_tokens": 45843823.0,
      "step": 19990
    },
    {
      "entropy": 5.1571966171264645,
      "epoch": 1.920749279538905,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00046334844223770076,
      "loss": 4.9306,
      "mean_token_accuracy": 0.2193652644753456,
      "num_tokens": 45855122.0,
      "step": 19995
    },
    {
      "entropy": 5.153160142898559,
      "epoch": 1.9212295869356388,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046332968545551674,
      "loss": 4.8963,
      "mean_token_accuracy": 0.22554874867200853,
      "num_tokens": 45866638.0,
      "step": 20000
    },
    {
      "entropy": 5.235449171066284,
      "epoch": 1.9217098943323727,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004633109243008765,
      "loss": 4.9931,
      "mean_token_accuracy": 0.20947860330343246,
      "num_tokens": 45878654.0,
      "step": 20005
    },
    {
      "entropy": 5.214382934570312,
      "epoch": 1.9221902017291066,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004632921587742159,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22380865812301637,
      "num_tokens": 45889466.0,
      "step": 20010
    },
    {
      "entropy": 5.163268756866455,
      "epoch": 1.9226705091258407,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046327338887597043,
      "loss": 4.8985,
      "mean_token_accuracy": 0.22524917125701904,
      "num_tokens": 45900750.0,
      "step": 20015
    },
    {
      "entropy": 5.199461793899536,
      "epoch": 1.9231508165225746,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000463254614606576,
      "loss": 4.9438,
      "mean_token_accuracy": 0.21575383096933365,
      "num_tokens": 45912351.0,
      "step": 20020
    },
    {
      "entropy": 5.102425909042358,
      "epoch": 1.9236311239193085,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004632358359664683,
      "loss": 4.8389,
      "mean_token_accuracy": 0.23101456612348556,
      "num_tokens": 45923484.0,
      "step": 20025
    },
    {
      "entropy": 5.131709480285645,
      "epoch": 1.9241114313160423,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046321705295608356,
      "loss": 4.9041,
      "mean_token_accuracy": 0.22721160501241683,
      "num_tokens": 45935512.0,
      "step": 20030
    },
    {
      "entropy": 5.166025066375733,
      "epoch": 1.9245917387127762,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046319826557585764,
      "loss": 4.8426,
      "mean_token_accuracy": 0.22717529833316802,
      "num_tokens": 45946353.0,
      "step": 20035
    },
    {
      "entropy": 5.140226030349732,
      "epoch": 1.92507204610951,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004631794738262269,
      "loss": 4.8539,
      "mean_token_accuracy": 0.22114449143409728,
      "num_tokens": 45958281.0,
      "step": 20040
    },
    {
      "entropy": 5.1322245597839355,
      "epoch": 1.925552353506244,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004631606777076275,
      "loss": 4.8432,
      "mean_token_accuracy": 0.22630993276834488,
      "num_tokens": 45969548.0,
      "step": 20045
    },
    {
      "entropy": 5.137090444564819,
      "epoch": 1.9260326609029779,
      "grad_norm": 1.125,
      "learning_rate": 0.00046314187722049587,
      "loss": 4.8636,
      "mean_token_accuracy": 0.2183900386095047,
      "num_tokens": 45980300.0,
      "step": 20050
    },
    {
      "entropy": 5.128756189346314,
      "epoch": 1.9265129682997117,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046312307236526863,
      "loss": 4.8515,
      "mean_token_accuracy": 0.22304339557886124,
      "num_tokens": 45991280.0,
      "step": 20055
    },
    {
      "entropy": 5.195330810546875,
      "epoch": 1.9269932756964456,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046310426314238217,
      "loss": 4.9558,
      "mean_token_accuracy": 0.218916454911232,
      "num_tokens": 46003074.0,
      "step": 20060
    },
    {
      "entropy": 5.044739055633545,
      "epoch": 1.9274735830931795,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004630854495522733,
      "loss": 4.8395,
      "mean_token_accuracy": 0.2245645359158516,
      "num_tokens": 46014784.0,
      "step": 20065
    },
    {
      "entropy": 5.113285493850708,
      "epoch": 1.9279538904899134,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046306663159537874,
      "loss": 4.8007,
      "mean_token_accuracy": 0.2325538158416748,
      "num_tokens": 46025403.0,
      "step": 20070
    },
    {
      "entropy": 5.167813968658447,
      "epoch": 1.9284341978866475,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046304780927213554,
      "loss": 4.9243,
      "mean_token_accuracy": 0.22493050545454024,
      "num_tokens": 46037171.0,
      "step": 20075
    },
    {
      "entropy": 5.239415979385376,
      "epoch": 1.9289145052833814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046302898258298046,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2217705190181732,
      "num_tokens": 46049219.0,
      "step": 20080
    },
    {
      "entropy": 5.137058401107788,
      "epoch": 1.9293948126801153,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004630101515283509,
      "loss": 4.8214,
      "mean_token_accuracy": 0.23519447296857834,
      "num_tokens": 46061068.0,
      "step": 20085
    },
    {
      "entropy": 5.142185211181641,
      "epoch": 1.9298751200768491,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046299131610868377,
      "loss": 4.8203,
      "mean_token_accuracy": 0.22905095368623735,
      "num_tokens": 46073752.0,
      "step": 20090
    },
    {
      "entropy": 5.197376251220703,
      "epoch": 1.9303554274735832,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004629724763244165,
      "loss": 4.9629,
      "mean_token_accuracy": 0.21947422623634338,
      "num_tokens": 46085908.0,
      "step": 20095
    },
    {
      "entropy": 5.14117922782898,
      "epoch": 1.9308357348703171,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004629536321759866,
      "loss": 4.848,
      "mean_token_accuracy": 0.23375166654586793,
      "num_tokens": 46097467.0,
      "step": 20100
    },
    {
      "entropy": 5.134228515625,
      "epoch": 1.931316042267051,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046293478366383133,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2265674278140068,
      "num_tokens": 46107951.0,
      "step": 20105
    },
    {
      "entropy": 5.2284440994262695,
      "epoch": 1.9317963496637849,
      "grad_norm": 1.125,
      "learning_rate": 0.0004629159307883885,
      "loss": 4.9499,
      "mean_token_accuracy": 0.21272731721401214,
      "num_tokens": 46119862.0,
      "step": 20110
    },
    {
      "entropy": 5.106739711761475,
      "epoch": 1.9322766570605188,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004628970735500958,
      "loss": 4.8569,
      "mean_token_accuracy": 0.2236487865447998,
      "num_tokens": 46130625.0,
      "step": 20115
    },
    {
      "entropy": 5.1711225509643555,
      "epoch": 1.9327569644572526,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046287821194939094,
      "loss": 4.8995,
      "mean_token_accuracy": 0.2217583417892456,
      "num_tokens": 46141819.0,
      "step": 20120
    },
    {
      "entropy": 5.234017610549927,
      "epoch": 1.9332372718539865,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046285934598671186,
      "loss": 4.9922,
      "mean_token_accuracy": 0.21177544742822646,
      "num_tokens": 46154474.0,
      "step": 20125
    },
    {
      "entropy": 5.210927963256836,
      "epoch": 1.9337175792507204,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046284047566249665,
      "loss": 4.9116,
      "mean_token_accuracy": 0.21572579443454742,
      "num_tokens": 46165397.0,
      "step": 20130
    },
    {
      "entropy": 5.114159393310547,
      "epoch": 1.9341978866474543,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046282160097718336,
      "loss": 4.8461,
      "mean_token_accuracy": 0.2252346768975258,
      "num_tokens": 46177202.0,
      "step": 20135
    },
    {
      "entropy": 5.071011543273926,
      "epoch": 1.9346781940441882,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004628027219312102,
      "loss": 4.8961,
      "mean_token_accuracy": 0.2249082639813423,
      "num_tokens": 46189879.0,
      "step": 20140
    },
    {
      "entropy": 5.2582728385925295,
      "epoch": 1.935158501440922,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004627838385250155,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21201104968786239,
      "num_tokens": 46201788.0,
      "step": 20145
    },
    {
      "entropy": 5.235272455215454,
      "epoch": 1.9356388088376562,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046276495075903764,
      "loss": 4.8984,
      "mean_token_accuracy": 0.22626130133867264,
      "num_tokens": 46212316.0,
      "step": 20150
    },
    {
      "entropy": 5.115398025512695,
      "epoch": 1.93611911623439,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046274605863371517,
      "loss": 4.852,
      "mean_token_accuracy": 0.22735101133584976,
      "num_tokens": 46223192.0,
      "step": 20155
    },
    {
      "entropy": 5.1745048522949215,
      "epoch": 1.936599423631124,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004627271621494868,
      "loss": 4.907,
      "mean_token_accuracy": 0.2163966789841652,
      "num_tokens": 46234404.0,
      "step": 20160
    },
    {
      "entropy": 5.199132966995239,
      "epoch": 1.9370797310278578,
      "grad_norm": 1.140625,
      "learning_rate": 0.000462708261306791,
      "loss": 4.8868,
      "mean_token_accuracy": 0.21937177330255508,
      "num_tokens": 46246653.0,
      "step": 20165
    },
    {
      "entropy": 5.202343082427978,
      "epoch": 1.937560038424592,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004626893561060669,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2201458841562271,
      "num_tokens": 46258145.0,
      "step": 20170
    },
    {
      "entropy": 5.160205078125,
      "epoch": 1.9380403458213258,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046267044654775324,
      "loss": 4.8444,
      "mean_token_accuracy": 0.2276844084262848,
      "num_tokens": 46269021.0,
      "step": 20175
    },
    {
      "entropy": 5.058333396911621,
      "epoch": 1.9385206532180597,
      "grad_norm": 1.078125,
      "learning_rate": 0.000462651532632289,
      "loss": 4.8548,
      "mean_token_accuracy": 0.22561895102262497,
      "num_tokens": 46280729.0,
      "step": 20180
    },
    {
      "entropy": 5.13500452041626,
      "epoch": 1.9390009606147935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046263261436011344,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21921314299106598,
      "num_tokens": 46291999.0,
      "step": 20185
    },
    {
      "entropy": 5.192186450958252,
      "epoch": 1.9394812680115274,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004626136917316657,
      "loss": 4.9514,
      "mean_token_accuracy": 0.21912372708320618,
      "num_tokens": 46302740.0,
      "step": 20190
    },
    {
      "entropy": 5.187963008880615,
      "epoch": 1.9399615754082613,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046259476474738514,
      "loss": 4.8634,
      "mean_token_accuracy": 0.2280578851699829,
      "num_tokens": 46313700.0,
      "step": 20195
    },
    {
      "entropy": 5.176633596420288,
      "epoch": 1.9404418828049952,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046257583340771123,
      "loss": 4.9008,
      "mean_token_accuracy": 0.22641313672065735,
      "num_tokens": 46324445.0,
      "step": 20200
    },
    {
      "entropy": 5.0635370254516605,
      "epoch": 1.940922190201729,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004625568977130835,
      "loss": 4.8834,
      "mean_token_accuracy": 0.21839186549186707,
      "num_tokens": 46335192.0,
      "step": 20205
    },
    {
      "entropy": 5.249114465713501,
      "epoch": 1.941402497598463,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004625379576639414,
      "loss": 4.9172,
      "mean_token_accuracy": 0.2253558650612831,
      "num_tokens": 46346067.0,
      "step": 20210
    },
    {
      "entropy": 5.181732940673828,
      "epoch": 1.9418828049951968,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046251901326072487,
      "loss": 4.8515,
      "mean_token_accuracy": 0.2289232060313225,
      "num_tokens": 46356736.0,
      "step": 20215
    },
    {
      "entropy": 5.138217878341675,
      "epoch": 1.9423631123919307,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046250006450387367,
      "loss": 4.922,
      "mean_token_accuracy": 0.22242932319641112,
      "num_tokens": 46368250.0,
      "step": 20220
    },
    {
      "entropy": 5.181773233413696,
      "epoch": 1.9428434197886646,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004624811113938277,
      "loss": 4.998,
      "mean_token_accuracy": 0.21346300542354585,
      "num_tokens": 46379662.0,
      "step": 20225
    },
    {
      "entropy": 5.138122034072876,
      "epoch": 1.9433237271853987,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004624621539310271,
      "loss": 4.8547,
      "mean_token_accuracy": 0.22404130399227143,
      "num_tokens": 46391485.0,
      "step": 20230
    },
    {
      "entropy": 5.207921743392944,
      "epoch": 1.9438040345821326,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046244319211591193,
      "loss": 5.0464,
      "mean_token_accuracy": 0.21266615390777588,
      "num_tokens": 46403608.0,
      "step": 20235
    },
    {
      "entropy": 5.186870908737182,
      "epoch": 1.9442843419788665,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004624242259489223,
      "loss": 4.8498,
      "mean_token_accuracy": 0.2249423399567604,
      "num_tokens": 46416154.0,
      "step": 20240
    },
    {
      "entropy": 5.1637450695037845,
      "epoch": 1.9447646493756003,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046240525543049884,
      "loss": 4.8891,
      "mean_token_accuracy": 0.22535861879587174,
      "num_tokens": 46427654.0,
      "step": 20245
    },
    {
      "entropy": 5.147416496276856,
      "epoch": 1.9452449567723344,
      "grad_norm": 1.3125,
      "learning_rate": 0.00046238628056108176,
      "loss": 4.8509,
      "mean_token_accuracy": 0.22411062568426132,
      "num_tokens": 46438074.0,
      "step": 20250
    },
    {
      "entropy": 5.091103458404541,
      "epoch": 1.9457252641690683,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046236730134111166,
      "loss": 4.8404,
      "mean_token_accuracy": 0.23064495176076888,
      "num_tokens": 46449002.0,
      "step": 20255
    },
    {
      "entropy": 5.144047832489013,
      "epoch": 1.9462055715658022,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004623483177710291,
      "loss": 4.8757,
      "mean_token_accuracy": 0.22446896582841874,
      "num_tokens": 46460576.0,
      "step": 20260
    },
    {
      "entropy": 5.2298510551452635,
      "epoch": 1.946685878962536,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004623293298512751,
      "loss": 4.9389,
      "mean_token_accuracy": 0.22245583683252335,
      "num_tokens": 46471636.0,
      "step": 20265
    },
    {
      "entropy": 5.099475383758545,
      "epoch": 1.94716618635927,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046231033758229026,
      "loss": 4.8055,
      "mean_token_accuracy": 0.22726177871227266,
      "num_tokens": 46482364.0,
      "step": 20270
    },
    {
      "entropy": 5.180945158004761,
      "epoch": 1.9476464937560038,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004622913409645154,
      "loss": 4.9538,
      "mean_token_accuracy": 0.2179654449224472,
      "num_tokens": 46493612.0,
      "step": 20275
    },
    {
      "entropy": 5.122188425064087,
      "epoch": 1.9481268011527377,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004622723399983919,
      "loss": 4.8739,
      "mean_token_accuracy": 0.22594120651483535,
      "num_tokens": 46504825.0,
      "step": 20280
    },
    {
      "entropy": 5.1694494724273685,
      "epoch": 1.9486071085494716,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046225333468436077,
      "loss": 4.8184,
      "mean_token_accuracy": 0.23052066564559937,
      "num_tokens": 46515094.0,
      "step": 20285
    },
    {
      "entropy": 5.2128386974334715,
      "epoch": 1.9490874159462055,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00046223432502286323,
      "loss": 4.9092,
      "mean_token_accuracy": 0.2166774421930313,
      "num_tokens": 46527956.0,
      "step": 20290
    },
    {
      "entropy": 5.097730398178101,
      "epoch": 1.9495677233429394,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046221531101434056,
      "loss": 4.8322,
      "mean_token_accuracy": 0.22769645750522613,
      "num_tokens": 46539446.0,
      "step": 20295
    },
    {
      "entropy": 5.252546691894532,
      "epoch": 1.9500480307396733,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004621962926592343,
      "loss": 4.9484,
      "mean_token_accuracy": 0.22459145933389663,
      "num_tokens": 46551008.0,
      "step": 20300
    },
    {
      "entropy": 5.070315551757813,
      "epoch": 1.9505283381364071,
      "grad_norm": 1.109375,
      "learning_rate": 0.000462177269957986,
      "loss": 4.7654,
      "mean_token_accuracy": 0.23045520782470702,
      "num_tokens": 46563904.0,
      "step": 20305
    },
    {
      "entropy": 5.145900392532349,
      "epoch": 1.9510086455331412,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004621582429110373,
      "loss": 4.9286,
      "mean_token_accuracy": 0.22001587152481078,
      "num_tokens": 46575065.0,
      "step": 20310
    },
    {
      "entropy": 5.1092291355133055,
      "epoch": 1.9514889529298751,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046213921151883,
      "loss": 4.8491,
      "mean_token_accuracy": 0.2251417964696884,
      "num_tokens": 46587683.0,
      "step": 20315
    },
    {
      "entropy": 5.134192895889282,
      "epoch": 1.951969260326609,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004621201757818059,
      "loss": 4.8623,
      "mean_token_accuracy": 0.22971642911434173,
      "num_tokens": 46600060.0,
      "step": 20320
    },
    {
      "entropy": 5.162304019927978,
      "epoch": 1.952449567723343,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046210113570040683,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22297542691230773,
      "num_tokens": 46611619.0,
      "step": 20325
    },
    {
      "entropy": 5.0547631740570065,
      "epoch": 1.952929875120077,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000462082091275075,
      "loss": 4.8069,
      "mean_token_accuracy": 0.22756583243608475,
      "num_tokens": 46622286.0,
      "step": 20330
    },
    {
      "entropy": 5.251196002960205,
      "epoch": 1.9534101825168109,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004620630425062526,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21312947124242781,
      "num_tokens": 46633402.0,
      "step": 20335
    },
    {
      "entropy": 5.063565397262574,
      "epoch": 1.9538904899135447,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004620439893943817,
      "loss": 4.8084,
      "mean_token_accuracy": 0.22850182056427001,
      "num_tokens": 46644889.0,
      "step": 20340
    },
    {
      "entropy": 5.144259786605835,
      "epoch": 1.9543707973102786,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004620249319399049,
      "loss": 4.8987,
      "mean_token_accuracy": 0.23073548525571824,
      "num_tokens": 46656355.0,
      "step": 20345
    },
    {
      "entropy": 5.183876466751099,
      "epoch": 1.9548511047070125,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046200587014326455,
      "loss": 4.9066,
      "mean_token_accuracy": 0.2157456874847412,
      "num_tokens": 46668570.0,
      "step": 20350
    },
    {
      "entropy": 5.1413809299469,
      "epoch": 1.9553314121037464,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004619868040049031,
      "loss": 4.8707,
      "mean_token_accuracy": 0.22115390449762345,
      "num_tokens": 46680015.0,
      "step": 20355
    },
    {
      "entropy": 5.169937181472778,
      "epoch": 1.9558117195004803,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004619677335252633,
      "loss": 4.9262,
      "mean_token_accuracy": 0.219439272582531,
      "num_tokens": 46690872.0,
      "step": 20360
    },
    {
      "entropy": 5.190472650527954,
      "epoch": 1.9562920268972142,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046194865870478793,
      "loss": 4.899,
      "mean_token_accuracy": 0.2188018262386322,
      "num_tokens": 46702760.0,
      "step": 20365
    },
    {
      "entropy": 5.138350152969361,
      "epoch": 1.956772334293948,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046192957954391983,
      "loss": 4.93,
      "mean_token_accuracy": 0.21581437587738037,
      "num_tokens": 46715038.0,
      "step": 20370
    },
    {
      "entropy": 5.193798065185547,
      "epoch": 1.957252641690682,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000461910496043102,
      "loss": 4.916,
      "mean_token_accuracy": 0.22403337061405182,
      "num_tokens": 46725851.0,
      "step": 20375
    },
    {
      "entropy": 5.1309596538543705,
      "epoch": 1.9577329490874158,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004618914082027773,
      "loss": 4.8125,
      "mean_token_accuracy": 0.22524728775024414,
      "num_tokens": 46736380.0,
      "step": 20380
    },
    {
      "entropy": 5.055413293838501,
      "epoch": 1.95821325648415,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046187231602338926,
      "loss": 4.8104,
      "mean_token_accuracy": 0.23012327700853347,
      "num_tokens": 46747872.0,
      "step": 20385
    },
    {
      "entropy": 5.135238122940064,
      "epoch": 1.9586935638808838,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046185321950538086,
      "loss": 4.7783,
      "mean_token_accuracy": 0.2271716147661209,
      "num_tokens": 46759704.0,
      "step": 20390
    },
    {
      "entropy": 5.170301389694214,
      "epoch": 1.9591738712776177,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004618341186491955,
      "loss": 4.9008,
      "mean_token_accuracy": 0.22839677929878235,
      "num_tokens": 46770895.0,
      "step": 20395
    },
    {
      "entropy": 5.154159688949585,
      "epoch": 1.9596541786743515,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004618150134552768,
      "loss": 4.9013,
      "mean_token_accuracy": 0.22131532728672026,
      "num_tokens": 46781190.0,
      "step": 20400
    },
    {
      "entropy": 5.136796283721924,
      "epoch": 1.9601344860710856,
      "grad_norm": 1.1875,
      "learning_rate": 0.000461795903924068,
      "loss": 4.8755,
      "mean_token_accuracy": 0.225905878841877,
      "num_tokens": 46792172.0,
      "step": 20405
    },
    {
      "entropy": 5.200492906570434,
      "epoch": 1.9606147934678195,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046177679005601313,
      "loss": 4.9664,
      "mean_token_accuracy": 0.2152267038822174,
      "num_tokens": 46804189.0,
      "step": 20410
    },
    {
      "entropy": 5.20401029586792,
      "epoch": 1.9610951008645534,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004617576718515558,
      "loss": 4.9006,
      "mean_token_accuracy": 0.21929072886705397,
      "num_tokens": 46815127.0,
      "step": 20415
    },
    {
      "entropy": 5.138180160522461,
      "epoch": 1.9615754082612873,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004617385493111399,
      "loss": 4.9163,
      "mean_token_accuracy": 0.21917709410190583,
      "num_tokens": 46825927.0,
      "step": 20420
    },
    {
      "entropy": 5.229851388931275,
      "epoch": 1.9620557156580212,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004617194224352093,
      "loss": 4.9717,
      "mean_token_accuracy": 0.21480505615472795,
      "num_tokens": 46836689.0,
      "step": 20425
    },
    {
      "entropy": 5.12658839225769,
      "epoch": 1.962536023054755,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004617002912242083,
      "loss": 4.8132,
      "mean_token_accuracy": 0.22959331721067427,
      "num_tokens": 46847650.0,
      "step": 20430
    },
    {
      "entropy": 5.082293748855591,
      "epoch": 1.963016330451489,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046168115567858084,
      "loss": 4.8605,
      "mean_token_accuracy": 0.22965044677257537,
      "num_tokens": 46859469.0,
      "step": 20435
    },
    {
      "entropy": 5.172416067123413,
      "epoch": 1.9634966378482228,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00046166201579877125,
      "loss": 4.8359,
      "mean_token_accuracy": 0.23198784589767457,
      "num_tokens": 46869583.0,
      "step": 20440
    },
    {
      "entropy": 5.221918821334839,
      "epoch": 1.9639769452449567,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004616428715852239,
      "loss": 4.8543,
      "mean_token_accuracy": 0.22763019055128098,
      "num_tokens": 46881357.0,
      "step": 20445
    },
    {
      "entropy": 5.067873334884643,
      "epoch": 1.9644572526416906,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004616237230383833,
      "loss": 4.885,
      "mean_token_accuracy": 0.22169135361909867,
      "num_tokens": 46893627.0,
      "step": 20450
    },
    {
      "entropy": 5.160275411605835,
      "epoch": 1.9649375600384245,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046160457015869414,
      "loss": 4.9423,
      "mean_token_accuracy": 0.21869318336248397,
      "num_tokens": 46904621.0,
      "step": 20455
    },
    {
      "entropy": 5.25049614906311,
      "epoch": 1.9654178674351583,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046158541294660083,
      "loss": 4.9111,
      "mean_token_accuracy": 0.22008464932441713,
      "num_tokens": 46914381.0,
      "step": 20460
    },
    {
      "entropy": 5.155420112609863,
      "epoch": 1.9658981748318924,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004615662514025483,
      "loss": 4.8658,
      "mean_token_accuracy": 0.22083081752061845,
      "num_tokens": 46925139.0,
      "step": 20465
    },
    {
      "entropy": 5.185564994812012,
      "epoch": 1.9663784822286263,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046154708552698147,
      "loss": 4.9042,
      "mean_token_accuracy": 0.21911248415708542,
      "num_tokens": 46937638.0,
      "step": 20470
    },
    {
      "entropy": 5.110273933410644,
      "epoch": 1.9668587896253602,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046152791532034517,
      "loss": 4.903,
      "mean_token_accuracy": 0.22611591964960098,
      "num_tokens": 46949460.0,
      "step": 20475
    },
    {
      "entropy": 5.172116136550903,
      "epoch": 1.967339097022094,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046150874078308463,
      "loss": 4.9454,
      "mean_token_accuracy": 0.21682527363300325,
      "num_tokens": 46961047.0,
      "step": 20480
    },
    {
      "entropy": 5.169101428985596,
      "epoch": 1.9678194044188282,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004614895619156449,
      "loss": 4.875,
      "mean_token_accuracy": 0.22442455142736434,
      "num_tokens": 46972009.0,
      "step": 20485
    },
    {
      "entropy": 5.025904178619385,
      "epoch": 1.968299711815562,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046147037871847125,
      "loss": 4.8168,
      "mean_token_accuracy": 0.22609294056892396,
      "num_tokens": 46982911.0,
      "step": 20490
    },
    {
      "entropy": 5.184283638000489,
      "epoch": 1.968780019212296,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004614511911920092,
      "loss": 4.8865,
      "mean_token_accuracy": 0.2169527530670166,
      "num_tokens": 46995096.0,
      "step": 20495
    },
    {
      "entropy": 5.251112604141236,
      "epoch": 1.9692603266090298,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046143199933670404,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2161327987909317,
      "num_tokens": 47006814.0,
      "step": 20500
    },
    {
      "entropy": 5.104316711425781,
      "epoch": 1.9697406340057637,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004614128031530015,
      "loss": 4.7813,
      "mean_token_accuracy": 0.2299412414431572,
      "num_tokens": 47018204.0,
      "step": 20505
    },
    {
      "entropy": 5.13380994796753,
      "epoch": 1.9702209414024976,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046139360264134724,
      "loss": 4.9024,
      "mean_token_accuracy": 0.22973873764276503,
      "num_tokens": 47029655.0,
      "step": 20510
    },
    {
      "entropy": 5.229656457901001,
      "epoch": 1.9707012487992315,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004613743978021869,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21451540738344194,
      "num_tokens": 47041830.0,
      "step": 20515
    },
    {
      "entropy": 5.224571895599365,
      "epoch": 1.9711815561959654,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046135518863596654,
      "loss": 4.912,
      "mean_token_accuracy": 0.22660491913557052,
      "num_tokens": 47052817.0,
      "step": 20520
    },
    {
      "entropy": 5.1657514572143555,
      "epoch": 1.9716618635926992,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046133597514313204,
      "loss": 4.9155,
      "mean_token_accuracy": 0.21786917001008987,
      "num_tokens": 47064452.0,
      "step": 20525
    },
    {
      "entropy": 5.16720814704895,
      "epoch": 1.9721421709894331,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004613167573241295,
      "loss": 4.9623,
      "mean_token_accuracy": 0.21333869695663452,
      "num_tokens": 47074952.0,
      "step": 20530
    },
    {
      "entropy": 5.198386907577515,
      "epoch": 1.972622478386167,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004612975351794051,
      "loss": 4.898,
      "mean_token_accuracy": 0.22730254977941514,
      "num_tokens": 47085496.0,
      "step": 20535
    },
    {
      "entropy": 5.1445920944213865,
      "epoch": 1.973102785782901,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004612783087094051,
      "loss": 4.9228,
      "mean_token_accuracy": 0.21912720054388046,
      "num_tokens": 47097798.0,
      "step": 20540
    },
    {
      "entropy": 5.226597690582276,
      "epoch": 1.973583093179635,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046125907791457594,
      "loss": 4.9875,
      "mean_token_accuracy": 0.2212499976158142,
      "num_tokens": 47109798.0,
      "step": 20545
    },
    {
      "entropy": 5.223862552642823,
      "epoch": 1.9740634005763689,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046123984279536405,
      "loss": 4.9117,
      "mean_token_accuracy": 0.22669910341501237,
      "num_tokens": 47120073.0,
      "step": 20550
    },
    {
      "entropy": 5.1797607898712155,
      "epoch": 1.9745437079731027,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046122060335221604,
      "loss": 4.8766,
      "mean_token_accuracy": 0.22150231748819352,
      "num_tokens": 47130596.0,
      "step": 20555
    },
    {
      "entropy": 5.087465858459472,
      "epoch": 1.9750240153698368,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046120135958557855,
      "loss": 4.7975,
      "mean_token_accuracy": 0.24019335508346557,
      "num_tokens": 47141660.0,
      "step": 20560
    },
    {
      "entropy": 5.308927440643311,
      "epoch": 1.9755043227665707,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046118211149589843,
      "loss": 5.0685,
      "mean_token_accuracy": 0.20047816038131713,
      "num_tokens": 47153046.0,
      "step": 20565
    },
    {
      "entropy": 5.279877758026123,
      "epoch": 1.9759846301633046,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004611628590836225,
      "loss": 4.9976,
      "mean_token_accuracy": 0.22026793658733368,
      "num_tokens": 47162706.0,
      "step": 20570
    },
    {
      "entropy": 5.137110376358033,
      "epoch": 1.9764649375600385,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004611436023491979,
      "loss": 4.922,
      "mean_token_accuracy": 0.22106708586215973,
      "num_tokens": 47174508.0,
      "step": 20575
    },
    {
      "entropy": 5.129480171203613,
      "epoch": 1.9769452449567724,
      "grad_norm": 1.375,
      "learning_rate": 0.0004611243412930714,
      "loss": 4.8174,
      "mean_token_accuracy": 0.22664519399404526,
      "num_tokens": 47185019.0,
      "step": 20580
    },
    {
      "entropy": 5.127928161621094,
      "epoch": 1.9774255523535063,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046110507591569047,
      "loss": 4.8434,
      "mean_token_accuracy": 0.22695548832416534,
      "num_tokens": 47195538.0,
      "step": 20585
    },
    {
      "entropy": 5.180661725997925,
      "epoch": 1.9779058597502401,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004610858062175023,
      "loss": 4.8864,
      "mean_token_accuracy": 0.2234800562262535,
      "num_tokens": 47206398.0,
      "step": 20590
    },
    {
      "entropy": 5.216695022583008,
      "epoch": 1.978386167146974,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046106653219895417,
      "loss": 4.9707,
      "mean_token_accuracy": 0.21071705967187881,
      "num_tokens": 47217775.0,
      "step": 20595
    },
    {
      "entropy": 5.185318374633789,
      "epoch": 1.978866474543708,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004610472538604938,
      "loss": 4.9245,
      "mean_token_accuracy": 0.2215694174170494,
      "num_tokens": 47230179.0,
      "step": 20600
    },
    {
      "entropy": 5.123435544967651,
      "epoch": 1.9793467819404418,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046102797120256854,
      "loss": 4.7932,
      "mean_token_accuracy": 0.23630284070968627,
      "num_tokens": 47241416.0,
      "step": 20605
    },
    {
      "entropy": 5.122931957244873,
      "epoch": 1.9798270893371757,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004610086842256262,
      "loss": 4.8674,
      "mean_token_accuracy": 0.22120108604431152,
      "num_tokens": 47253916.0,
      "step": 20610
    },
    {
      "entropy": 5.22463092803955,
      "epoch": 1.9803073967339095,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004609893929301146,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21862184703350068,
      "num_tokens": 47265176.0,
      "step": 20615
    },
    {
      "entropy": 5.162397766113282,
      "epoch": 1.9807877041306436,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004609700973164816,
      "loss": 4.8519,
      "mean_token_accuracy": 0.22394065856933593,
      "num_tokens": 47276178.0,
      "step": 20620
    },
    {
      "entropy": 5.052760076522827,
      "epoch": 1.9812680115273775,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004609507973851751,
      "loss": 4.8454,
      "mean_token_accuracy": 0.22459010928869247,
      "num_tokens": 47287607.0,
      "step": 20625
    },
    {
      "entropy": 5.259043216705322,
      "epoch": 1.9817483189241114,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046093149313664316,
      "loss": 5.0117,
      "mean_token_accuracy": 0.21477555334568024,
      "num_tokens": 47300069.0,
      "step": 20630
    },
    {
      "entropy": 5.239381504058838,
      "epoch": 1.9822286263208453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004609121845713342,
      "loss": 4.8609,
      "mean_token_accuracy": 0.2279776617884636,
      "num_tokens": 47312349.0,
      "step": 20635
    },
    {
      "entropy": 5.084532070159912,
      "epoch": 1.9827089337175794,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004608928716896963,
      "loss": 4.8632,
      "mean_token_accuracy": 0.23164936304092407,
      "num_tokens": 47323171.0,
      "step": 20640
    },
    {
      "entropy": 5.191327857971191,
      "epoch": 1.9831892411143133,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004608735544921778,
      "loss": 4.9365,
      "mean_token_accuracy": 0.22291866540908814,
      "num_tokens": 47333990.0,
      "step": 20645
    },
    {
      "entropy": 5.099944257736206,
      "epoch": 1.9836695485110472,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046085423297922745,
      "loss": 4.866,
      "mean_token_accuracy": 0.22436515539884566,
      "num_tokens": 47346081.0,
      "step": 20650
    },
    {
      "entropy": 5.25259804725647,
      "epoch": 1.984149855907781,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046083490715129367,
      "loss": 5.0715,
      "mean_token_accuracy": 0.21124795377254485,
      "num_tokens": 47357979.0,
      "step": 20655
    },
    {
      "entropy": 5.353647613525391,
      "epoch": 1.984630163304515,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004608155770088251,
      "loss": 4.9708,
      "mean_token_accuracy": 0.20855092108249665,
      "num_tokens": 47369066.0,
      "step": 20660
    },
    {
      "entropy": 5.166320753097534,
      "epoch": 1.9851104707012488,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046079624255227066,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21655142456293106,
      "num_tokens": 47380316.0,
      "step": 20665
    },
    {
      "entropy": 5.1111267566680905,
      "epoch": 1.9855907780979827,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004607769037820791,
      "loss": 4.9173,
      "mean_token_accuracy": 0.2168147459626198,
      "num_tokens": 47392040.0,
      "step": 20670
    },
    {
      "entropy": 5.206585454940796,
      "epoch": 1.9860710854947166,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004607575606986995,
      "loss": 4.9096,
      "mean_token_accuracy": 0.2246001571416855,
      "num_tokens": 47404114.0,
      "step": 20675
    },
    {
      "entropy": 5.1461708545684814,
      "epoch": 1.9865513928914504,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004607382133025809,
      "loss": 4.8471,
      "mean_token_accuracy": 0.22355568259954453,
      "num_tokens": 47416581.0,
      "step": 20680
    },
    {
      "entropy": 5.131541156768799,
      "epoch": 1.9870317002881843,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046071886159417257,
      "loss": 4.7705,
      "mean_token_accuracy": 0.23597488850355147,
      "num_tokens": 47427314.0,
      "step": 20685
    },
    {
      "entropy": 5.173101234436035,
      "epoch": 1.9875120076849182,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004606995055739238,
      "loss": 4.9764,
      "mean_token_accuracy": 0.2116803377866745,
      "num_tokens": 47438672.0,
      "step": 20690
    },
    {
      "entropy": 5.255899286270141,
      "epoch": 1.9879923150816523,
      "grad_norm": 1.125,
      "learning_rate": 0.00046068014524228374,
      "loss": 4.9748,
      "mean_token_accuracy": 0.21822184324264526,
      "num_tokens": 47449901.0,
      "step": 20695
    },
    {
      "entropy": 5.1441638469696045,
      "epoch": 1.9884726224783862,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046066078059970217,
      "loss": 4.9486,
      "mean_token_accuracy": 0.2203874945640564,
      "num_tokens": 47462831.0,
      "step": 20700
    },
    {
      "entropy": 5.145195865631104,
      "epoch": 1.98895292987512,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004606414116466286,
      "loss": 4.9132,
      "mean_token_accuracy": 0.22482379972934724,
      "num_tokens": 47474075.0,
      "step": 20705
    },
    {
      "entropy": 5.222511434555054,
      "epoch": 1.989433237271854,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046062203838351267,
      "loss": 5.0088,
      "mean_token_accuracy": 0.2137518495321274,
      "num_tokens": 47486893.0,
      "step": 20710
    },
    {
      "entropy": 5.195823001861572,
      "epoch": 1.989913544668588,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046060266081080414,
      "loss": 4.9107,
      "mean_token_accuracy": 0.2177120178937912,
      "num_tokens": 47497969.0,
      "step": 20715
    },
    {
      "entropy": 5.268104124069214,
      "epoch": 1.990393852065322,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004605832789289531,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21774250715970994,
      "num_tokens": 47509476.0,
      "step": 20720
    },
    {
      "entropy": 5.154200506210327,
      "epoch": 1.9908741594620558,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004605638927384093,
      "loss": 4.8603,
      "mean_token_accuracy": 0.22424486130475998,
      "num_tokens": 47520124.0,
      "step": 20725
    },
    {
      "entropy": 5.130086088180542,
      "epoch": 1.9913544668587897,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046054450223962284,
      "loss": 4.8458,
      "mean_token_accuracy": 0.2233796611428261,
      "num_tokens": 47532079.0,
      "step": 20730
    },
    {
      "entropy": 5.2034660339355465,
      "epoch": 1.9918347742555236,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046052510743304405,
      "loss": 4.88,
      "mean_token_accuracy": 0.22987380027770996,
      "num_tokens": 47544142.0,
      "step": 20735
    },
    {
      "entropy": 5.153724527359008,
      "epoch": 1.9923150816522575,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004605057083191232,
      "loss": 4.885,
      "mean_token_accuracy": 0.22473601698875428,
      "num_tokens": 47555688.0,
      "step": 20740
    },
    {
      "entropy": 5.148333263397217,
      "epoch": 1.9927953890489913,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004604863048983106,
      "loss": 4.8426,
      "mean_token_accuracy": 0.23165808618068695,
      "num_tokens": 47566590.0,
      "step": 20745
    },
    {
      "entropy": 5.1586127281188965,
      "epoch": 1.9932756964457252,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004604668971710568,
      "loss": 4.9177,
      "mean_token_accuracy": 0.22791305780410767,
      "num_tokens": 47577781.0,
      "step": 20750
    },
    {
      "entropy": 5.154198169708252,
      "epoch": 1.993756003842459,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004604474851378124,
      "loss": 4.8383,
      "mean_token_accuracy": 0.23356927186250687,
      "num_tokens": 47589074.0,
      "step": 20755
    },
    {
      "entropy": 5.145243215560913,
      "epoch": 1.994236311239193,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046042806879902803,
      "loss": 4.878,
      "mean_token_accuracy": 0.22673259526491166,
      "num_tokens": 47600841.0,
      "step": 20760
    },
    {
      "entropy": 5.196794843673706,
      "epoch": 1.9947166186359269,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004604086481551546,
      "loss": 4.8994,
      "mean_token_accuracy": 0.22033290117979049,
      "num_tokens": 47612332.0,
      "step": 20765
    },
    {
      "entropy": 5.135573148727417,
      "epoch": 1.9951969260326607,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004603892232066428,
      "loss": 4.8759,
      "mean_token_accuracy": 0.21997257620096206,
      "num_tokens": 47623773.0,
      "step": 20770
    },
    {
      "entropy": 5.122671794891358,
      "epoch": 1.9956772334293948,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046036979395394374,
      "loss": 4.843,
      "mean_token_accuracy": 0.2296233743429184,
      "num_tokens": 47634802.0,
      "step": 20775
    },
    {
      "entropy": 5.265804052352905,
      "epoch": 1.9961575408261287,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004603503603975085,
      "loss": 4.989,
      "mean_token_accuracy": 0.2201553389430046,
      "num_tokens": 47646772.0,
      "step": 20780
    },
    {
      "entropy": 5.203161096572876,
      "epoch": 1.9966378482228626,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004603309225377883,
      "loss": 4.928,
      "mean_token_accuracy": 0.21952929049730302,
      "num_tokens": 47658321.0,
      "step": 20785
    },
    {
      "entropy": 5.056264638900757,
      "epoch": 1.9971181556195965,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004603114803752344,
      "loss": 4.8112,
      "mean_token_accuracy": 0.2377043768763542,
      "num_tokens": 47668996.0,
      "step": 20790
    },
    {
      "entropy": 5.122574663162231,
      "epoch": 1.9975984630163306,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046029203391029813,
      "loss": 4.8665,
      "mean_token_accuracy": 0.2300494521856308,
      "num_tokens": 47679091.0,
      "step": 20795
    },
    {
      "entropy": 5.170337295532226,
      "epoch": 1.9980787704130645,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046027258314343107,
      "loss": 4.8828,
      "mean_token_accuracy": 0.22475721538066865,
      "num_tokens": 47690740.0,
      "step": 20800
    },
    {
      "entropy": 5.135902833938599,
      "epoch": 1.9985590778097984,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046025312807508487,
      "loss": 4.8604,
      "mean_token_accuracy": 0.22518380284309386,
      "num_tokens": 47703514.0,
      "step": 20805
    },
    {
      "entropy": 5.085724973678589,
      "epoch": 1.9990393852065322,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046023366870571097,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22315678000450134,
      "num_tokens": 47715220.0,
      "step": 20810
    },
    {
      "entropy": 5.184278106689453,
      "epoch": 1.9995196926032661,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046021420503576145,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22257687151432037,
      "num_tokens": 47727387.0,
      "step": 20815
    },
    {
      "entropy": 5.251145887374878,
      "epoch": 2.0,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004601947370656879,
      "loss": 4.9691,
      "mean_token_accuracy": 0.21188677251338958,
      "num_tokens": 47737072.0,
      "step": 20820
    },
    {
      "entropy": 5.099743032455445,
      "epoch": 2.000480307396734,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004601752647959426,
      "loss": 4.7696,
      "mean_token_accuracy": 0.2395510956645012,
      "num_tokens": 47748516.0,
      "step": 20825
    },
    {
      "entropy": 5.096894598007202,
      "epoch": 2.0009606147934678,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004601557882269775,
      "loss": 4.7784,
      "mean_token_accuracy": 0.227401065826416,
      "num_tokens": 47760482.0,
      "step": 20830
    },
    {
      "entropy": 5.1745030879974365,
      "epoch": 2.0014409221902016,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004601363073592447,
      "loss": 4.8124,
      "mean_token_accuracy": 0.23437483608722687,
      "num_tokens": 47771204.0,
      "step": 20835
    },
    {
      "entropy": 5.102879381179809,
      "epoch": 2.0019212295869355,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004601168221931967,
      "loss": 4.8028,
      "mean_token_accuracy": 0.2304491639137268,
      "num_tokens": 47782404.0,
      "step": 20840
    },
    {
      "entropy": 5.136948728561402,
      "epoch": 2.0024015369836694,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004600973327292857,
      "loss": 4.8532,
      "mean_token_accuracy": 0.22482473403215408,
      "num_tokens": 47793612.0,
      "step": 20845
    },
    {
      "entropy": 5.151643371582031,
      "epoch": 2.0028818443804033,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046007783896796436,
      "loss": 4.7821,
      "mean_token_accuracy": 0.23377842009067534,
      "num_tokens": 47804467.0,
      "step": 20850
    },
    {
      "entropy": 5.219990921020508,
      "epoch": 2.003362151777137,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004600583409096851,
      "loss": 4.911,
      "mean_token_accuracy": 0.21624902933835982,
      "num_tokens": 47815654.0,
      "step": 20855
    },
    {
      "entropy": 5.169375228881836,
      "epoch": 2.0038424591738715,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046003883855490066,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22423673123121263,
      "num_tokens": 47827630.0,
      "step": 20860
    },
    {
      "entropy": 5.1376800537109375,
      "epoch": 2.0043227665706054,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004600193319040638,
      "loss": 4.7874,
      "mean_token_accuracy": 0.22913518995046617,
      "num_tokens": 47838046.0,
      "step": 20865
    },
    {
      "entropy": 5.210384368896484,
      "epoch": 2.0048030739673393,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045999982095762756,
      "loss": 4.8777,
      "mean_token_accuracy": 0.2252206563949585,
      "num_tokens": 47849506.0,
      "step": 20870
    },
    {
      "entropy": 5.101250123977661,
      "epoch": 2.005283381364073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045998030571604473,
      "loss": 4.7121,
      "mean_token_accuracy": 0.2378280684351921,
      "num_tokens": 47860689.0,
      "step": 20875
    },
    {
      "entropy": 5.084704113006592,
      "epoch": 2.005763688760807,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004599607861797685,
      "loss": 4.7691,
      "mean_token_accuracy": 0.23349439203739167,
      "num_tokens": 47872215.0,
      "step": 20880
    },
    {
      "entropy": 5.172788333892822,
      "epoch": 2.006243996157541,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045994126234925203,
      "loss": 4.9243,
      "mean_token_accuracy": 0.22070244699716568,
      "num_tokens": 47883726.0,
      "step": 20885
    },
    {
      "entropy": 5.2045807361602785,
      "epoch": 2.006724303554275,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045992173422494865,
      "loss": 4.7958,
      "mean_token_accuracy": 0.22505733519792556,
      "num_tokens": 47893915.0,
      "step": 20890
    },
    {
      "entropy": 5.156705713272094,
      "epoch": 2.0072046109510087,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004599022018073117,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22373861372470855,
      "num_tokens": 47905026.0,
      "step": 20895
    },
    {
      "entropy": 5.207870531082153,
      "epoch": 2.0076849183477425,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004598826650967946,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22750947177410125,
      "num_tokens": 47918410.0,
      "step": 20900
    },
    {
      "entropy": 5.2213239669799805,
      "epoch": 2.0081652257444764,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045986312409385105,
      "loss": 4.8676,
      "mean_token_accuracy": 0.22180142253637314,
      "num_tokens": 47929584.0,
      "step": 20905
    },
    {
      "entropy": 5.128156328201294,
      "epoch": 2.0086455331412103,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004598435787989347,
      "loss": 4.841,
      "mean_token_accuracy": 0.21902771890163422,
      "num_tokens": 47940585.0,
      "step": 20910
    },
    {
      "entropy": 5.145664358139038,
      "epoch": 2.009125840537944,
      "grad_norm": 1.125,
      "learning_rate": 0.00045982402921249934,
      "loss": 4.8349,
      "mean_token_accuracy": 0.22053710073232652,
      "num_tokens": 47953951.0,
      "step": 20915
    },
    {
      "entropy": 5.124466276168823,
      "epoch": 2.009606147934678,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004598044753349988,
      "loss": 4.753,
      "mean_token_accuracy": 0.22682977914810182,
      "num_tokens": 47965783.0,
      "step": 20920
    },
    {
      "entropy": 5.107498407363892,
      "epoch": 2.010086455331412,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045978491716688706,
      "loss": 4.7864,
      "mean_token_accuracy": 0.22769156098365784,
      "num_tokens": 47978509.0,
      "step": 20925
    },
    {
      "entropy": 5.222621059417724,
      "epoch": 2.010566762728146,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004597653547086184,
      "loss": 4.8856,
      "mean_token_accuracy": 0.21907687038183213,
      "num_tokens": 47989828.0,
      "step": 20930
    },
    {
      "entropy": 5.151153802871704,
      "epoch": 2.0110470701248797,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004597457879606467,
      "loss": 4.7474,
      "mean_token_accuracy": 0.22689333707094192,
      "num_tokens": 47999683.0,
      "step": 20935
    },
    {
      "entropy": 5.047843933105469,
      "epoch": 2.011527377521614,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045972621692342636,
      "loss": 4.6895,
      "mean_token_accuracy": 0.23427390456199645,
      "num_tokens": 48010526.0,
      "step": 20940
    },
    {
      "entropy": 5.012542295455932,
      "epoch": 2.012007684918348,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045970664159741186,
      "loss": 4.7206,
      "mean_token_accuracy": 0.23594721257686616,
      "num_tokens": 48021383.0,
      "step": 20945
    },
    {
      "entropy": 5.017799663543701,
      "epoch": 2.012487992315082,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045968706198305765,
      "loss": 4.6499,
      "mean_token_accuracy": 0.24006713330745696,
      "num_tokens": 48031459.0,
      "step": 20950
    },
    {
      "entropy": 5.216294479370117,
      "epoch": 2.0129682997118157,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045966747808081824,
      "loss": 4.9134,
      "mean_token_accuracy": 0.2199397414922714,
      "num_tokens": 48043590.0,
      "step": 20955
    },
    {
      "entropy": 5.086314582824707,
      "epoch": 2.0134486071085496,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004596478898911483,
      "loss": 4.7071,
      "mean_token_accuracy": 0.23482942283153535,
      "num_tokens": 48054715.0,
      "step": 20960
    },
    {
      "entropy": 5.115472650527954,
      "epoch": 2.0139289145052834,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045962829741450265,
      "loss": 4.8227,
      "mean_token_accuracy": 0.2256898358464241,
      "num_tokens": 48065606.0,
      "step": 20965
    },
    {
      "entropy": 5.209070014953613,
      "epoch": 2.0144092219020173,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004596087006513361,
      "loss": 4.8688,
      "mean_token_accuracy": 0.22466896325349808,
      "num_tokens": 48076097.0,
      "step": 20970
    },
    {
      "entropy": 5.2381011009216305,
      "epoch": 2.014889529298751,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045958909960210385,
      "loss": 4.8926,
      "mean_token_accuracy": 0.22053786963224412,
      "num_tokens": 48087429.0,
      "step": 20975
    },
    {
      "entropy": 5.070555353164673,
      "epoch": 2.015369836695485,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045956949426726075,
      "loss": 4.7359,
      "mean_token_accuracy": 0.23253799229860306,
      "num_tokens": 48098347.0,
      "step": 20980
    },
    {
      "entropy": 5.00875334739685,
      "epoch": 2.015850144092219,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045954988464726203,
      "loss": 4.6556,
      "mean_token_accuracy": 0.23918845355510712,
      "num_tokens": 48109355.0,
      "step": 20985
    },
    {
      "entropy": 5.152424573898315,
      "epoch": 2.016330451488953,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000459530270742563,
      "loss": 4.872,
      "mean_token_accuracy": 0.21965826004743577,
      "num_tokens": 48121875.0,
      "step": 20990
    },
    {
      "entropy": 5.245805120468139,
      "epoch": 2.0168107588856867,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045951065255361905,
      "loss": 4.8959,
      "mean_token_accuracy": 0.21643051207065583,
      "num_tokens": 48133192.0,
      "step": 20995
    },
    {
      "entropy": 5.115788173675537,
      "epoch": 2.0172910662824206,
      "grad_norm": 1.125,
      "learning_rate": 0.0004594910300808856,
      "loss": 4.7372,
      "mean_token_accuracy": 0.23160056322813033,
      "num_tokens": 48143621.0,
      "step": 21000
    },
    {
      "epoch": 2.0172910662824206,
      "eval_entropy": 4.923085412067765,
      "eval_loss": 4.9685378074646,
      "eval_mean_token_accuracy": 0.22786725892879248,
      "eval_num_tokens": 48143621.0,
      "eval_runtime": 26.8028,
      "eval_samples_per_second": 1224.312,
      "eval_steps_per_second": 153.044,
      "step": 21000
    },
    {
      "entropy": 5.015113162994385,
      "epoch": 2.0177713736791545,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004594714033248183,
      "loss": 4.6898,
      "mean_token_accuracy": 0.23660008758306503,
      "num_tokens": 48153803.0,
      "step": 21005
    },
    {
      "entropy": 5.105776357650757,
      "epoch": 2.0182516810758884,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004594517722858728,
      "loss": 4.7828,
      "mean_token_accuracy": 0.22631603926420213,
      "num_tokens": 48165742.0,
      "step": 21010
    },
    {
      "entropy": 5.064240789413452,
      "epoch": 2.0187319884726227,
      "grad_norm": 1.125,
      "learning_rate": 0.00045943213696450475,
      "loss": 4.7708,
      "mean_token_accuracy": 0.2354864865541458,
      "num_tokens": 48177894.0,
      "step": 21015
    },
    {
      "entropy": 5.154419088363648,
      "epoch": 2.0192122958693566,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045941249736117023,
      "loss": 4.8357,
      "mean_token_accuracy": 0.2230084255337715,
      "num_tokens": 48188536.0,
      "step": 21020
    },
    {
      "entropy": 5.140545272827149,
      "epoch": 2.0196926032660905,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004593928534763251,
      "loss": 4.7601,
      "mean_token_accuracy": 0.23291108906269073,
      "num_tokens": 48199276.0,
      "step": 21025
    },
    {
      "entropy": 5.206594705581665,
      "epoch": 2.0201729106628243,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004593732053104254,
      "loss": 4.9214,
      "mean_token_accuracy": 0.22376627027988433,
      "num_tokens": 48212479.0,
      "step": 21030
    },
    {
      "entropy": 5.135408210754394,
      "epoch": 2.020653218059558,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045935355286392735,
      "loss": 4.7222,
      "mean_token_accuracy": 0.23597335815429688,
      "num_tokens": 48224279.0,
      "step": 21035
    },
    {
      "entropy": 5.134943866729737,
      "epoch": 2.021133525456292,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004593338961372873,
      "loss": 4.8848,
      "mean_token_accuracy": 0.2162734717130661,
      "num_tokens": 48235943.0,
      "step": 21040
    },
    {
      "entropy": 5.034694814682007,
      "epoch": 2.021613832853026,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004593142351309614,
      "loss": 4.6666,
      "mean_token_accuracy": 0.24116964489221573,
      "num_tokens": 48246407.0,
      "step": 21045
    },
    {
      "entropy": 5.1786274909973145,
      "epoch": 2.02209414024976,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004592945698454064,
      "loss": 4.87,
      "mean_token_accuracy": 0.22315399646759032,
      "num_tokens": 48256957.0,
      "step": 21050
    },
    {
      "entropy": 5.049574995040894,
      "epoch": 2.0225744476464937,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045927490028107866,
      "loss": 4.7437,
      "mean_token_accuracy": 0.23032526075839996,
      "num_tokens": 48268253.0,
      "step": 21055
    },
    {
      "entropy": 5.2037682056427,
      "epoch": 2.0230547550432276,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000459255226438435,
      "loss": 4.9299,
      "mean_token_accuracy": 0.22471971809864044,
      "num_tokens": 48279980.0,
      "step": 21060
    },
    {
      "entropy": 5.2044741153717045,
      "epoch": 2.0235350624399615,
      "grad_norm": 1.0625,
      "learning_rate": 0.000459235548317932,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2234876736998558,
      "num_tokens": 48291199.0,
      "step": 21065
    },
    {
      "entropy": 5.175772571563721,
      "epoch": 2.0240153698366954,
      "grad_norm": 1.125,
      "learning_rate": 0.00045921586592002667,
      "loss": 4.8603,
      "mean_token_accuracy": 0.22201622128486634,
      "num_tokens": 48301610.0,
      "step": 21070
    },
    {
      "entropy": 5.1607225894927975,
      "epoch": 2.0244956772334293,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000459196179245176,
      "loss": 4.8428,
      "mean_token_accuracy": 0.2213815599679947,
      "num_tokens": 48313655.0,
      "step": 21075
    },
    {
      "entropy": 5.167400932312011,
      "epoch": 2.024975984630163,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004591764882938369,
      "loss": 4.7569,
      "mean_token_accuracy": 0.2336047410964966,
      "num_tokens": 48325385.0,
      "step": 21080
    },
    {
      "entropy": 5.127931785583496,
      "epoch": 2.025456292026897,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004591567930664667,
      "loss": 4.7367,
      "mean_token_accuracy": 0.23095783293247224,
      "num_tokens": 48335986.0,
      "step": 21085
    },
    {
      "entropy": 5.131039142608643,
      "epoch": 2.025936599423631,
      "grad_norm": 1.125,
      "learning_rate": 0.0004591370935635226,
      "loss": 4.8681,
      "mean_token_accuracy": 0.22674043476581573,
      "num_tokens": 48347196.0,
      "step": 21090
    },
    {
      "entropy": 5.204093074798584,
      "epoch": 2.0264169068203652,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004591173897854619,
      "loss": 4.9035,
      "mean_token_accuracy": 0.22117702066898345,
      "num_tokens": 48357944.0,
      "step": 21095
    },
    {
      "entropy": 5.2365165710449215,
      "epoch": 2.026897214217099,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004590976817327422,
      "loss": 4.9949,
      "mean_token_accuracy": 0.21333225816488266,
      "num_tokens": 48369838.0,
      "step": 21100
    },
    {
      "entropy": 5.149486351013183,
      "epoch": 2.027377521613833,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004590779694058209,
      "loss": 4.7849,
      "mean_token_accuracy": 0.2293430209159851,
      "num_tokens": 48381107.0,
      "step": 21105
    },
    {
      "entropy": 5.078858470916748,
      "epoch": 2.027857829010567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045905825280515586,
      "loss": 4.8353,
      "mean_token_accuracy": 0.22689204514026642,
      "num_tokens": 48392699.0,
      "step": 21110
    },
    {
      "entropy": 5.183833312988281,
      "epoch": 2.0283381364073008,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045903853193120464,
      "loss": 4.8773,
      "mean_token_accuracy": 0.2245073661208153,
      "num_tokens": 48403836.0,
      "step": 21115
    },
    {
      "entropy": 5.1207905292510985,
      "epoch": 2.0288184438040346,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045901880678442524,
      "loss": 4.7651,
      "mean_token_accuracy": 0.23643454909324646,
      "num_tokens": 48417373.0,
      "step": 21120
    },
    {
      "entropy": 5.15496597290039,
      "epoch": 2.0292987512007685,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045899907736527556,
      "loss": 4.8159,
      "mean_token_accuracy": 0.23182263523340224,
      "num_tokens": 48428033.0,
      "step": 21125
    },
    {
      "entropy": 5.082381772994995,
      "epoch": 2.0297790585975024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045897934367421364,
      "loss": 4.7571,
      "mean_token_accuracy": 0.23105890452861785,
      "num_tokens": 48440147.0,
      "step": 21130
    },
    {
      "entropy": 5.045969390869141,
      "epoch": 2.0302593659942363,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004589596057116977,
      "loss": 4.7898,
      "mean_token_accuracy": 0.2291718080639839,
      "num_tokens": 48451406.0,
      "step": 21135
    },
    {
      "entropy": 5.1496696949005125,
      "epoch": 2.03073967339097,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045893986347818593,
      "loss": 4.7756,
      "mean_token_accuracy": 0.23712951987981795,
      "num_tokens": 48464393.0,
      "step": 21140
    },
    {
      "entropy": 5.1563661098480225,
      "epoch": 2.031219980787704,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004589201169741368,
      "loss": 4.8639,
      "mean_token_accuracy": 0.2257084757089615,
      "num_tokens": 48476433.0,
      "step": 21145
    },
    {
      "entropy": 5.139056634902954,
      "epoch": 2.031700288184438,
      "grad_norm": 1.125,
      "learning_rate": 0.00045890036620000856,
      "loss": 4.7879,
      "mean_token_accuracy": 0.22651064693927764,
      "num_tokens": 48488372.0,
      "step": 21150
    },
    {
      "entropy": 5.063483810424804,
      "epoch": 2.032180595581172,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045888061115626,
      "loss": 4.7508,
      "mean_token_accuracy": 0.2360047772526741,
      "num_tokens": 48498834.0,
      "step": 21155
    },
    {
      "entropy": 5.1776988983154295,
      "epoch": 2.0326609029779057,
      "grad_norm": 1.125,
      "learning_rate": 0.0004588608518433496,
      "loss": 4.7881,
      "mean_token_accuracy": 0.22805428504943848,
      "num_tokens": 48509160.0,
      "step": 21160
    },
    {
      "entropy": 5.1091419696807865,
      "epoch": 2.0331412103746396,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004588410882617362,
      "loss": 4.8285,
      "mean_token_accuracy": 0.2278285652399063,
      "num_tokens": 48522005.0,
      "step": 21165
    },
    {
      "entropy": 5.117874002456665,
      "epoch": 2.0336215177713735,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004588213204118786,
      "loss": 4.7971,
      "mean_token_accuracy": 0.2283453032374382,
      "num_tokens": 48534238.0,
      "step": 21170
    },
    {
      "entropy": 5.144563627243042,
      "epoch": 2.034101825168108,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045880154829423586,
      "loss": 4.8192,
      "mean_token_accuracy": 0.2281987711787224,
      "num_tokens": 48546440.0,
      "step": 21175
    },
    {
      "entropy": 5.198111820220947,
      "epoch": 2.0345821325648417,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004587817719092668,
      "loss": 4.8057,
      "mean_token_accuracy": 0.235726235806942,
      "num_tokens": 48556893.0,
      "step": 21180
    },
    {
      "entropy": 5.118719482421875,
      "epoch": 2.0350624399615755,
      "grad_norm": 1.125,
      "learning_rate": 0.00045876199125743087,
      "loss": 4.7945,
      "mean_token_accuracy": 0.2280433163046837,
      "num_tokens": 48568483.0,
      "step": 21185
    },
    {
      "entropy": 5.168294191360474,
      "epoch": 2.0355427473583094,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004587422063391871,
      "loss": 4.8414,
      "mean_token_accuracy": 0.21844590455293655,
      "num_tokens": 48579352.0,
      "step": 21190
    },
    {
      "entropy": 5.184268808364868,
      "epoch": 2.0360230547550433,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004587224171549949,
      "loss": 4.8693,
      "mean_token_accuracy": 0.21767326891422273,
      "num_tokens": 48591447.0,
      "step": 21195
    },
    {
      "entropy": 5.149103450775146,
      "epoch": 2.036503362151777,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045870262370531376,
      "loss": 4.8517,
      "mean_token_accuracy": 0.2161777585744858,
      "num_tokens": 48603724.0,
      "step": 21200
    },
    {
      "entropy": 5.235502767562866,
      "epoch": 2.036983669548511,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045868282599060314,
      "loss": 4.8791,
      "mean_token_accuracy": 0.21950249671936034,
      "num_tokens": 48616319.0,
      "step": 21205
    },
    {
      "entropy": 5.130740213394165,
      "epoch": 2.037463976945245,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004586630240113227,
      "loss": 4.8146,
      "mean_token_accuracy": 0.23369720429182053,
      "num_tokens": 48627121.0,
      "step": 21210
    },
    {
      "entropy": 5.064772653579712,
      "epoch": 2.037944284341979,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004586432177679322,
      "loss": 4.7323,
      "mean_token_accuracy": 0.23671976178884507,
      "num_tokens": 48638687.0,
      "step": 21215
    },
    {
      "entropy": 5.1230486869812015,
      "epoch": 2.0384245917387127,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045862340726089153,
      "loss": 4.895,
      "mean_token_accuracy": 0.2212316170334816,
      "num_tokens": 48649480.0,
      "step": 21220
    },
    {
      "entropy": 5.252587985992432,
      "epoch": 2.0389048991354466,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004586035924906606,
      "loss": 4.9526,
      "mean_token_accuracy": 0.21867176592350007,
      "num_tokens": 48660197.0,
      "step": 21225
    },
    {
      "entropy": 5.074550771713257,
      "epoch": 2.0393852065321805,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045858377345769946,
      "loss": 4.7241,
      "mean_token_accuracy": 0.23025636076927186,
      "num_tokens": 48671711.0,
      "step": 21230
    },
    {
      "entropy": 5.064471673965454,
      "epoch": 2.0398655139289144,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004585639501624682,
      "loss": 4.7357,
      "mean_token_accuracy": 0.23403332829475404,
      "num_tokens": 48683285.0,
      "step": 21235
    },
    {
      "entropy": 5.174516201019287,
      "epoch": 2.0403458213256482,
      "grad_norm": 1.046875,
      "learning_rate": 0.000458544122605427,
      "loss": 4.8062,
      "mean_token_accuracy": 0.23280756324529647,
      "num_tokens": 48696308.0,
      "step": 21240
    },
    {
      "entropy": 5.062208223342895,
      "epoch": 2.040826128722382,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045852429078703646,
      "loss": 4.6483,
      "mean_token_accuracy": 0.23936488777399062,
      "num_tokens": 48706999.0,
      "step": 21245
    },
    {
      "entropy": 5.096370887756348,
      "epoch": 2.0413064361191164,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045850445470775673,
      "loss": 4.7985,
      "mean_token_accuracy": 0.23230074793100358,
      "num_tokens": 48717248.0,
      "step": 21250
    },
    {
      "entropy": 5.136594343185425,
      "epoch": 2.0417867435158503,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004584846143680485,
      "loss": 4.8802,
      "mean_token_accuracy": 0.22621916979551315,
      "num_tokens": 48727948.0,
      "step": 21255
    },
    {
      "entropy": 5.183815860748291,
      "epoch": 2.042267050912584,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004584647697683723,
      "loss": 4.8246,
      "mean_token_accuracy": 0.2285064160823822,
      "num_tokens": 48740474.0,
      "step": 21260
    },
    {
      "entropy": 5.188474464416504,
      "epoch": 2.042747358309318,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045844492090918904,
      "loss": 4.8638,
      "mean_token_accuracy": 0.22091308683156968,
      "num_tokens": 48753415.0,
      "step": 21265
    },
    {
      "entropy": 5.100908422470093,
      "epoch": 2.043227665706052,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045842506779095936,
      "loss": 4.7444,
      "mean_token_accuracy": 0.23560899794101714,
      "num_tokens": 48764466.0,
      "step": 21270
    },
    {
      "entropy": 5.0673106670379635,
      "epoch": 2.043707973102786,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004584052104141443,
      "loss": 4.7563,
      "mean_token_accuracy": 0.23317614793777466,
      "num_tokens": 48777513.0,
      "step": 21275
    },
    {
      "entropy": 5.2186503410339355,
      "epoch": 2.0441882804995197,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004583853487792048,
      "loss": 4.8721,
      "mean_token_accuracy": 0.2298312559723854,
      "num_tokens": 48788732.0,
      "step": 21280
    },
    {
      "entropy": 5.115804719924927,
      "epoch": 2.0446685878962536,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004583654828866021,
      "loss": 4.7416,
      "mean_token_accuracy": 0.2294040337204933,
      "num_tokens": 48799620.0,
      "step": 21285
    },
    {
      "entropy": 5.027311134338379,
      "epoch": 2.0451488952929875,
      "grad_norm": 1.125,
      "learning_rate": 0.0004583456127367973,
      "loss": 4.7914,
      "mean_token_accuracy": 0.22863316982984544,
      "num_tokens": 48811914.0,
      "step": 21290
    },
    {
      "entropy": 5.083560800552368,
      "epoch": 2.0456292026897214,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004583257383302519,
      "loss": 4.7815,
      "mean_token_accuracy": 0.23334577828645706,
      "num_tokens": 48822610.0,
      "step": 21295
    },
    {
      "entropy": 5.160345602035522,
      "epoch": 2.0461095100864553,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004583058596674271,
      "loss": 4.8886,
      "mean_token_accuracy": 0.21999238133430482,
      "num_tokens": 48834296.0,
      "step": 21300
    },
    {
      "entropy": 5.072899580001831,
      "epoch": 2.046589817483189,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004582859767487846,
      "loss": 4.7002,
      "mean_token_accuracy": 0.23565882295370102,
      "num_tokens": 48846455.0,
      "step": 21305
    },
    {
      "entropy": 5.162654113769531,
      "epoch": 2.047070124879923,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045826608957478604,
      "loss": 4.9115,
      "mean_token_accuracy": 0.22564785182476044,
      "num_tokens": 48858364.0,
      "step": 21310
    },
    {
      "entropy": 5.173195457458496,
      "epoch": 2.047550432276657,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045824619814589297,
      "loss": 4.8373,
      "mean_token_accuracy": 0.22557061463594436,
      "num_tokens": 48869829.0,
      "step": 21315
    },
    {
      "entropy": 5.135656547546387,
      "epoch": 2.0480307396733908,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004582263024625674,
      "loss": 4.8091,
      "mean_token_accuracy": 0.22933683842420577,
      "num_tokens": 48880951.0,
      "step": 21320
    },
    {
      "entropy": 5.109381484985351,
      "epoch": 2.048511047070125,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004582064025252711,
      "loss": 4.7947,
      "mean_token_accuracy": 0.23121156841516494,
      "num_tokens": 48893367.0,
      "step": 21325
    },
    {
      "entropy": 5.180786848068237,
      "epoch": 2.048991354466859,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004581864983344661,
      "loss": 4.8484,
      "mean_token_accuracy": 0.22610592544078828,
      "num_tokens": 48904400.0,
      "step": 21330
    },
    {
      "entropy": 5.140589189529419,
      "epoch": 2.049471661863593,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004581665898906147,
      "loss": 4.8455,
      "mean_token_accuracy": 0.23180918395519257,
      "num_tokens": 48916305.0,
      "step": 21335
    },
    {
      "entropy": 5.123523283004761,
      "epoch": 2.0499519692603267,
      "grad_norm": 1.125,
      "learning_rate": 0.00045814667719417887,
      "loss": 4.828,
      "mean_token_accuracy": 0.21703954190015792,
      "num_tokens": 48927160.0,
      "step": 21340
    },
    {
      "entropy": 5.148010683059693,
      "epoch": 2.0504322766570606,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000458126760245621,
      "loss": 4.8548,
      "mean_token_accuracy": 0.2234545037150383,
      "num_tokens": 48940776.0,
      "step": 21345
    },
    {
      "entropy": 5.198182439804077,
      "epoch": 2.0509125840537945,
      "grad_norm": 1.125,
      "learning_rate": 0.0004581068390454036,
      "loss": 4.846,
      "mean_token_accuracy": 0.22127244025468826,
      "num_tokens": 48952375.0,
      "step": 21350
    },
    {
      "entropy": 5.13310399055481,
      "epoch": 2.0513928914505284,
      "grad_norm": 1.125,
      "learning_rate": 0.00045808691359398905,
      "loss": 4.8262,
      "mean_token_accuracy": 0.23141152113676072,
      "num_tokens": 48964684.0,
      "step": 21355
    },
    {
      "entropy": 5.14719820022583,
      "epoch": 2.0518731988472623,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004580669838918401,
      "loss": 4.8424,
      "mean_token_accuracy": 0.22869929522275925,
      "num_tokens": 48975654.0,
      "step": 21360
    },
    {
      "entropy": 5.097676849365234,
      "epoch": 2.052353506243996,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045804704993941935,
      "loss": 4.8204,
      "mean_token_accuracy": 0.22690718322992326,
      "num_tokens": 48987983.0,
      "step": 21365
    },
    {
      "entropy": 5.038017749786377,
      "epoch": 2.05283381364073,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045802711173718966,
      "loss": 4.7269,
      "mean_token_accuracy": 0.23645653426647187,
      "num_tokens": 48998728.0,
      "step": 21370
    },
    {
      "entropy": 5.150280141830445,
      "epoch": 2.053314121037464,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004580071692856138,
      "loss": 4.8491,
      "mean_token_accuracy": 0.22369205355644226,
      "num_tokens": 49009177.0,
      "step": 21375
    },
    {
      "entropy": 5.126007318496704,
      "epoch": 2.053794428434198,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045798722258515504,
      "loss": 4.7816,
      "mean_token_accuracy": 0.23385382890701295,
      "num_tokens": 49021550.0,
      "step": 21380
    },
    {
      "entropy": 5.186445140838623,
      "epoch": 2.0542747358309317,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045796727163627623,
      "loss": 4.7917,
      "mean_token_accuracy": 0.22648132890462874,
      "num_tokens": 49032349.0,
      "step": 21385
    },
    {
      "entropy": 5.0493019104003904,
      "epoch": 2.0547550432276656,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004579473164394408,
      "loss": 4.7481,
      "mean_token_accuracy": 0.23788137137889862,
      "num_tokens": 49044829.0,
      "step": 21390
    },
    {
      "entropy": 5.122381973266601,
      "epoch": 2.0552353506243994,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045792735699511176,
      "loss": 4.876,
      "mean_token_accuracy": 0.22797955572605133,
      "num_tokens": 49057319.0,
      "step": 21395
    },
    {
      "entropy": 5.159272003173828,
      "epoch": 2.0557156580211333,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045790739330375276,
      "loss": 4.789,
      "mean_token_accuracy": 0.232901431620121,
      "num_tokens": 49068872.0,
      "step": 21400
    },
    {
      "entropy": 5.103936004638672,
      "epoch": 2.0561959654178676,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045788742536582717,
      "loss": 4.7415,
      "mean_token_accuracy": 0.2300373926758766,
      "num_tokens": 49081080.0,
      "step": 21405
    },
    {
      "entropy": 5.02402548789978,
      "epoch": 2.0566762728146015,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045786745318179866,
      "loss": 4.6789,
      "mean_token_accuracy": 0.23625958114862441,
      "num_tokens": 49092535.0,
      "step": 21410
    },
    {
      "entropy": 5.151445436477661,
      "epoch": 2.0571565802113354,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004578474767521309,
      "loss": 4.8904,
      "mean_token_accuracy": 0.21825755387544632,
      "num_tokens": 49103743.0,
      "step": 21415
    },
    {
      "entropy": 5.182199907302857,
      "epoch": 2.0576368876080693,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045782749607728765,
      "loss": 4.8715,
      "mean_token_accuracy": 0.22956449687480926,
      "num_tokens": 49115346.0,
      "step": 21420
    },
    {
      "entropy": 5.106340026855468,
      "epoch": 2.058117195004803,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045780751115773286,
      "loss": 4.792,
      "mean_token_accuracy": 0.23183710277080535,
      "num_tokens": 49126000.0,
      "step": 21425
    },
    {
      "entropy": 5.0790119647979735,
      "epoch": 2.058597502401537,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004577875219939304,
      "loss": 4.7453,
      "mean_token_accuracy": 0.23397108316421508,
      "num_tokens": 49136617.0,
      "step": 21430
    },
    {
      "entropy": 5.136586046218872,
      "epoch": 2.059077809798271,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004577675285863446,
      "loss": 4.8292,
      "mean_token_accuracy": 0.22454218715429305,
      "num_tokens": 49147804.0,
      "step": 21435
    },
    {
      "entropy": 5.143794250488281,
      "epoch": 2.059558117195005,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045774753093543943,
      "loss": 4.7313,
      "mean_token_accuracy": 0.23686419427394867,
      "num_tokens": 49158367.0,
      "step": 21440
    },
    {
      "entropy": 5.116992473602295,
      "epoch": 2.0600384245917387,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004577275290416791,
      "loss": 4.8338,
      "mean_token_accuracy": 0.22878028005361556,
      "num_tokens": 49170236.0,
      "step": 21445
    },
    {
      "entropy": 5.108990573883057,
      "epoch": 2.0605187319884726,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004577075229055283,
      "loss": 4.8028,
      "mean_token_accuracy": 0.23080692440271378,
      "num_tokens": 49183379.0,
      "step": 21450
    },
    {
      "entropy": 5.121269464492798,
      "epoch": 2.0609990393852065,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045768751252745133,
      "loss": 4.8492,
      "mean_token_accuracy": 0.22216108590364456,
      "num_tokens": 49194649.0,
      "step": 21455
    },
    {
      "entropy": 5.154722833633423,
      "epoch": 2.0614793467819403,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045766749790791274,
      "loss": 4.8359,
      "mean_token_accuracy": 0.22991353571414946,
      "num_tokens": 49206523.0,
      "step": 21460
    },
    {
      "entropy": 5.114092922210693,
      "epoch": 2.061959654178674,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004576474790473773,
      "loss": 4.8598,
      "mean_token_accuracy": 0.22149042338132857,
      "num_tokens": 49219020.0,
      "step": 21465
    },
    {
      "entropy": 5.128277397155761,
      "epoch": 2.062439961575408,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045762745594630973,
      "loss": 4.7971,
      "mean_token_accuracy": 0.22919617146253585,
      "num_tokens": 49230751.0,
      "step": 21470
    },
    {
      "entropy": 5.230525827407837,
      "epoch": 2.062920268972142,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004576074286051749,
      "loss": 4.8536,
      "mean_token_accuracy": 0.22659927159547805,
      "num_tokens": 49241758.0,
      "step": 21475
    },
    {
      "entropy": 5.07979531288147,
      "epoch": 2.063400576368876,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045758739702443787,
      "loss": 4.795,
      "mean_token_accuracy": 0.23413648456335068,
      "num_tokens": 49253404.0,
      "step": 21480
    },
    {
      "entropy": 5.197508811950684,
      "epoch": 2.06388088376561,
      "grad_norm": 1.125,
      "learning_rate": 0.0004575673612045636,
      "loss": 4.9034,
      "mean_token_accuracy": 0.2174158439040184,
      "num_tokens": 49264000.0,
      "step": 21485
    },
    {
      "entropy": 5.183927869796753,
      "epoch": 2.064361191162344,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004575473211460173,
      "loss": 4.8606,
      "mean_token_accuracy": 0.22664424628019333,
      "num_tokens": 49275657.0,
      "step": 21490
    },
    {
      "entropy": 5.149008083343506,
      "epoch": 2.064841498559078,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004575272768492644,
      "loss": 4.9448,
      "mean_token_accuracy": 0.21756215393543243,
      "num_tokens": 49288570.0,
      "step": 21495
    },
    {
      "entropy": 5.153464269638062,
      "epoch": 2.065321805955812,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045750722831476993,
      "loss": 4.8367,
      "mean_token_accuracy": 0.21927328258752823,
      "num_tokens": 49300982.0,
      "step": 21500
    },
    {
      "entropy": 5.165737724304199,
      "epoch": 2.0658021133525457,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045748717554299964,
      "loss": 4.8712,
      "mean_token_accuracy": 0.2252198040485382,
      "num_tokens": 49313141.0,
      "step": 21505
    },
    {
      "entropy": 5.120110702514649,
      "epoch": 2.0662824207492796,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000457467118534419,
      "loss": 4.8146,
      "mean_token_accuracy": 0.22986338287591934,
      "num_tokens": 49324726.0,
      "step": 21510
    },
    {
      "entropy": 5.234684085845947,
      "epoch": 2.0667627281460135,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004574470572894938,
      "loss": 4.9056,
      "mean_token_accuracy": 0.21907796263694762,
      "num_tokens": 49336187.0,
      "step": 21515
    },
    {
      "entropy": 5.215353393554688,
      "epoch": 2.0672430355427474,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004574269918086895,
      "loss": 4.8389,
      "mean_token_accuracy": 0.227256141602993,
      "num_tokens": 49348075.0,
      "step": 21520
    },
    {
      "entropy": 5.111878395080566,
      "epoch": 2.0677233429394812,
      "grad_norm": 2.390625,
      "learning_rate": 0.0004574069220924722,
      "loss": 4.7939,
      "mean_token_accuracy": 0.2338130995631218,
      "num_tokens": 49358339.0,
      "step": 21525
    },
    {
      "entropy": 5.150820732116699,
      "epoch": 2.068203650336215,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004573868481413079,
      "loss": 4.839,
      "mean_token_accuracy": 0.2302357167005539,
      "num_tokens": 49369650.0,
      "step": 21530
    },
    {
      "entropy": 5.12345175743103,
      "epoch": 2.068683957732949,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045736676995566244,
      "loss": 4.8562,
      "mean_token_accuracy": 0.23156656622886657,
      "num_tokens": 49381982.0,
      "step": 21535
    },
    {
      "entropy": 5.136140203475952,
      "epoch": 2.069164265129683,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045734668753600217,
      "loss": 4.8161,
      "mean_token_accuracy": 0.2273861736059189,
      "num_tokens": 49393442.0,
      "step": 21540
    },
    {
      "entropy": 5.164246320724487,
      "epoch": 2.0696445725264168,
      "grad_norm": 1.125,
      "learning_rate": 0.00045732660088279326,
      "loss": 4.8079,
      "mean_token_accuracy": 0.2364514708518982,
      "num_tokens": 49404311.0,
      "step": 21545
    },
    {
      "entropy": 5.210529518127442,
      "epoch": 2.0701248799231506,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045730650999650216,
      "loss": 4.8938,
      "mean_token_accuracy": 0.22465836703777314,
      "num_tokens": 49417587.0,
      "step": 21550
    },
    {
      "entropy": 5.207915782928467,
      "epoch": 2.0706051873198845,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045728641487759506,
      "loss": 4.8689,
      "mean_token_accuracy": 0.22325333058834076,
      "num_tokens": 49429119.0,
      "step": 21555
    },
    {
      "entropy": 5.130094766616821,
      "epoch": 2.071085494716619,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004572663155265388,
      "loss": 4.7982,
      "mean_token_accuracy": 0.22695179730653764,
      "num_tokens": 49440734.0,
      "step": 21560
    },
    {
      "entropy": 5.076013803482056,
      "epoch": 2.0715658021133527,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004572462119437999,
      "loss": 4.7601,
      "mean_token_accuracy": 0.22727279216051102,
      "num_tokens": 49452830.0,
      "step": 21565
    },
    {
      "entropy": 5.12640414237976,
      "epoch": 2.0720461095100866,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045722610412984513,
      "loss": 4.8035,
      "mean_token_accuracy": 0.22549251914024354,
      "num_tokens": 49464380.0,
      "step": 21570
    },
    {
      "entropy": 5.047779226303101,
      "epoch": 2.0725264169068205,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004572059920851412,
      "loss": 4.7619,
      "mean_token_accuracy": 0.2369871973991394,
      "num_tokens": 49475436.0,
      "step": 21575
    },
    {
      "entropy": 5.107104349136352,
      "epoch": 2.0730067243035544,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045718587581015534,
      "loss": 4.8181,
      "mean_token_accuracy": 0.22511211186647415,
      "num_tokens": 49487015.0,
      "step": 21580
    },
    {
      "entropy": 5.130113887786865,
      "epoch": 2.0734870317002883,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004571657553053544,
      "loss": 4.7118,
      "mean_token_accuracy": 0.2317424312233925,
      "num_tokens": 49498485.0,
      "step": 21585
    },
    {
      "entropy": 5.213483142852783,
      "epoch": 2.073967339097022,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004571456305712055,
      "loss": 4.8636,
      "mean_token_accuracy": 0.2245493561029434,
      "num_tokens": 49508899.0,
      "step": 21590
    },
    {
      "entropy": 5.082722806930542,
      "epoch": 2.074447646493756,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004571255016081759,
      "loss": 4.7467,
      "mean_token_accuracy": 0.22882274836301802,
      "num_tokens": 49520956.0,
      "step": 21595
    },
    {
      "entropy": 5.188198852539062,
      "epoch": 2.07492795389049,
      "grad_norm": 1.203125,
      "learning_rate": 0.000457105368416733,
      "loss": 4.8977,
      "mean_token_accuracy": 0.21681189835071563,
      "num_tokens": 49532726.0,
      "step": 21600
    },
    {
      "entropy": 5.1883338451385494,
      "epoch": 2.0754082612872238,
      "grad_norm": 1.125,
      "learning_rate": 0.00045708523099734417,
      "loss": 4.8692,
      "mean_token_accuracy": 0.2222958728671074,
      "num_tokens": 49545156.0,
      "step": 21605
    },
    {
      "entropy": 5.117680597305298,
      "epoch": 2.0758885686839577,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045706508935047693,
      "loss": 4.8237,
      "mean_token_accuracy": 0.2316017270088196,
      "num_tokens": 49557552.0,
      "step": 21610
    },
    {
      "entropy": 5.095915412902832,
      "epoch": 2.0763688760806915,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000457044943476599,
      "loss": 4.8217,
      "mean_token_accuracy": 0.2226182371377945,
      "num_tokens": 49567906.0,
      "step": 21615
    },
    {
      "entropy": 5.105267906188965,
      "epoch": 2.0768491834774254,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045702479337617795,
      "loss": 4.746,
      "mean_token_accuracy": 0.2362363889813423,
      "num_tokens": 49578641.0,
      "step": 21620
    },
    {
      "entropy": 5.1341650009155275,
      "epoch": 2.0773294908741593,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004570046390496818,
      "loss": 4.8151,
      "mean_token_accuracy": 0.23247731626033782,
      "num_tokens": 49590449.0,
      "step": 21625
    },
    {
      "entropy": 5.147923231124878,
      "epoch": 2.077809798270893,
      "grad_norm": 1.125,
      "learning_rate": 0.0004569844804975783,
      "loss": 4.8498,
      "mean_token_accuracy": 0.22743574380874634,
      "num_tokens": 49600877.0,
      "step": 21630
    },
    {
      "entropy": 5.108972883224487,
      "epoch": 2.0782901056676275,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004569643177203356,
      "loss": 4.8407,
      "mean_token_accuracy": 0.2232835277915001,
      "num_tokens": 49612310.0,
      "step": 21635
    },
    {
      "entropy": 5.123304748535157,
      "epoch": 2.0787704130643614,
      "grad_norm": 1.125,
      "learning_rate": 0.0004569441507184218,
      "loss": 4.8152,
      "mean_token_accuracy": 0.2188327953219414,
      "num_tokens": 49625347.0,
      "step": 21640
    },
    {
      "entropy": 5.191255760192871,
      "epoch": 2.0792507204610953,
      "grad_norm": 1.125,
      "learning_rate": 0.00045692397949230495,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22124880999326707,
      "num_tokens": 49637943.0,
      "step": 21645
    },
    {
      "entropy": 5.106200408935547,
      "epoch": 2.079731027857829,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045690380404245364,
      "loss": 4.7223,
      "mean_token_accuracy": 0.23784611523151397,
      "num_tokens": 49649318.0,
      "step": 21650
    },
    {
      "entropy": 5.084951639175415,
      "epoch": 2.080211335254563,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045688362436933607,
      "loss": 4.8006,
      "mean_token_accuracy": 0.23124517798423766,
      "num_tokens": 49661368.0,
      "step": 21655
    },
    {
      "entropy": 5.12960786819458,
      "epoch": 2.080691642651297,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004568634404734208,
      "loss": 4.7419,
      "mean_token_accuracy": 0.23737003058195114,
      "num_tokens": 49673566.0,
      "step": 21660
    },
    {
      "entropy": 5.069616985321045,
      "epoch": 2.081171950048031,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004568432523551765,
      "loss": 4.7761,
      "mean_token_accuracy": 0.23235907554626464,
      "num_tokens": 49685151.0,
      "step": 21665
    },
    {
      "entropy": 5.163156604766845,
      "epoch": 2.0816522574447647,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004568230600150718,
      "loss": 4.8037,
      "mean_token_accuracy": 0.233746762573719,
      "num_tokens": 49696409.0,
      "step": 21670
    },
    {
      "entropy": 5.113603353500366,
      "epoch": 2.0821325648414986,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004568028634535757,
      "loss": 4.7791,
      "mean_token_accuracy": 0.23244920819997789,
      "num_tokens": 49709031.0,
      "step": 21675
    },
    {
      "entropy": 4.996547985076904,
      "epoch": 2.0826128722382324,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004567826626711568,
      "loss": 4.6992,
      "mean_token_accuracy": 0.23367761969566345,
      "num_tokens": 49719646.0,
      "step": 21680
    },
    {
      "entropy": 5.03615436553955,
      "epoch": 2.0830931796349663,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004567624576682843,
      "loss": 4.7517,
      "mean_token_accuracy": 0.2370401442050934,
      "num_tokens": 49730384.0,
      "step": 21685
    },
    {
      "entropy": 5.163003444671631,
      "epoch": 2.0835734870317,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004567422484454273,
      "loss": 4.8899,
      "mean_token_accuracy": 0.22283249199390412,
      "num_tokens": 49742256.0,
      "step": 21690
    },
    {
      "entropy": 5.153074789047241,
      "epoch": 2.084053794428434,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045672203500305493,
      "loss": 4.7538,
      "mean_token_accuracy": 0.23399983644485473,
      "num_tokens": 49753547.0,
      "step": 21695
    },
    {
      "entropy": 5.122174215316773,
      "epoch": 2.084534101825168,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00045670181734163654,
      "loss": 4.8947,
      "mean_token_accuracy": 0.22197608947753905,
      "num_tokens": 49765472.0,
      "step": 21700
    },
    {
      "entropy": 5.055600786209107,
      "epoch": 2.085014409221902,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004566815954616414,
      "loss": 4.6997,
      "mean_token_accuracy": 0.23437589704990386,
      "num_tokens": 49777031.0,
      "step": 21705
    },
    {
      "entropy": 5.181737947463989,
      "epoch": 2.0854947166186357,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045666136936353913,
      "loss": 4.7996,
      "mean_token_accuracy": 0.2343893676996231,
      "num_tokens": 49787923.0,
      "step": 21710
    },
    {
      "entropy": 5.080281352996826,
      "epoch": 2.08597502401537,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004566411390477993,
      "loss": 4.7825,
      "mean_token_accuracy": 0.22684745192527772,
      "num_tokens": 49799652.0,
      "step": 21715
    },
    {
      "entropy": 5.01281704902649,
      "epoch": 2.086455331412104,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045662090451489156,
      "loss": 4.7923,
      "mean_token_accuracy": 0.22947124838829042,
      "num_tokens": 49811278.0,
      "step": 21720
    },
    {
      "entropy": 5.167512130737305,
      "epoch": 2.086935638808838,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045660066576528577,
      "loss": 4.8186,
      "mean_token_accuracy": 0.23003543317317962,
      "num_tokens": 49823560.0,
      "step": 21725
    },
    {
      "entropy": 5.101685810089111,
      "epoch": 2.0874159462055717,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004565804227994518,
      "loss": 4.7053,
      "mean_token_accuracy": 0.2399987295269966,
      "num_tokens": 49834662.0,
      "step": 21730
    },
    {
      "entropy": 5.066118478775024,
      "epoch": 2.0878962536023056,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004565601756178595,
      "loss": 4.7793,
      "mean_token_accuracy": 0.2338222235441208,
      "num_tokens": 49845832.0,
      "step": 21735
    },
    {
      "entropy": 5.115154123306274,
      "epoch": 2.0883765609990395,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004565399242209791,
      "loss": 4.809,
      "mean_token_accuracy": 0.22988557070493698,
      "num_tokens": 49856420.0,
      "step": 21740
    },
    {
      "entropy": 5.086192941665649,
      "epoch": 2.0888568683957733,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004565196686092807,
      "loss": 4.6802,
      "mean_token_accuracy": 0.24067697823047637,
      "num_tokens": 49867680.0,
      "step": 21745
    },
    {
      "entropy": 5.086302614212036,
      "epoch": 2.089337175792507,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004564994087832346,
      "loss": 4.7763,
      "mean_token_accuracy": 0.2241254925727844,
      "num_tokens": 49879246.0,
      "step": 21750
    },
    {
      "entropy": 5.116818952560425,
      "epoch": 2.089817483189241,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045647914474331123,
      "loss": 4.7559,
      "mean_token_accuracy": 0.23548246771097184,
      "num_tokens": 49890536.0,
      "step": 21755
    },
    {
      "entropy": 5.201633787155151,
      "epoch": 2.090297790585975,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045645887648998094,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22182239592075348,
      "num_tokens": 49902331.0,
      "step": 21760
    },
    {
      "entropy": 5.127918243408203,
      "epoch": 2.090778097982709,
      "grad_norm": 1.125,
      "learning_rate": 0.00045643860402371433,
      "loss": 4.7433,
      "mean_token_accuracy": 0.24435743540525437,
      "num_tokens": 49913814.0,
      "step": 21765
    },
    {
      "entropy": 5.069564914703369,
      "epoch": 2.0912584053794427,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004564183273449821,
      "loss": 4.7097,
      "mean_token_accuracy": 0.2350170284509659,
      "num_tokens": 49924592.0,
      "step": 21770
    },
    {
      "entropy": 5.08256254196167,
      "epoch": 2.0917387127761766,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004563980464542551,
      "loss": 4.8422,
      "mean_token_accuracy": 0.2219977781176567,
      "num_tokens": 49936526.0,
      "step": 21775
    },
    {
      "entropy": 5.166136121749878,
      "epoch": 2.0922190201729105,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045637776135200406,
      "loss": 4.8195,
      "mean_token_accuracy": 0.2308672323822975,
      "num_tokens": 49947655.0,
      "step": 21780
    },
    {
      "entropy": 5.110087442398071,
      "epoch": 2.0926993275696444,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004563574720386999,
      "loss": 4.8235,
      "mean_token_accuracy": 0.22972238808870316,
      "num_tokens": 49959251.0,
      "step": 21785
    },
    {
      "entropy": 5.124665832519531,
      "epoch": 2.0931796349663783,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004563371785148139,
      "loss": 4.8598,
      "mean_token_accuracy": 0.22206338346004487,
      "num_tokens": 49970411.0,
      "step": 21790
    },
    {
      "entropy": 5.166639375686645,
      "epoch": 2.0936599423631126,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045631688078081695,
      "loss": 4.9344,
      "mean_token_accuracy": 0.22105189114809037,
      "num_tokens": 49983935.0,
      "step": 21795
    },
    {
      "entropy": 5.14625358581543,
      "epoch": 2.0941402497598465,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004562965788371805,
      "loss": 4.755,
      "mean_token_accuracy": 0.23398739397525786,
      "num_tokens": 49995278.0,
      "step": 21800
    },
    {
      "entropy": 5.076023530960083,
      "epoch": 2.0946205571565804,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004562762726843758,
      "loss": 4.7763,
      "mean_token_accuracy": 0.2299480363726616,
      "num_tokens": 50005534.0,
      "step": 21805
    },
    {
      "entropy": 5.100980281829834,
      "epoch": 2.0951008645533142,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045625596232287436,
      "loss": 4.8081,
      "mean_token_accuracy": 0.22836074084043503,
      "num_tokens": 50016583.0,
      "step": 21810
    },
    {
      "entropy": 5.110647821426392,
      "epoch": 2.095581171950048,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004562356477531477,
      "loss": 4.7573,
      "mean_token_accuracy": 0.22743655741214752,
      "num_tokens": 50027290.0,
      "step": 21815
    },
    {
      "entropy": 5.181838750839233,
      "epoch": 2.096061479346782,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004562153289756674,
      "loss": 4.8686,
      "mean_token_accuracy": 0.22955266386270523,
      "num_tokens": 50039671.0,
      "step": 21820
    },
    {
      "entropy": 5.133626890182495,
      "epoch": 2.096541786743516,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004561950059909053,
      "loss": 4.7852,
      "mean_token_accuracy": 0.23353853076696396,
      "num_tokens": 50051639.0,
      "step": 21825
    },
    {
      "entropy": 5.120704650878906,
      "epoch": 2.0970220941402498,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004561746787993332,
      "loss": 4.8649,
      "mean_token_accuracy": 0.2157152697443962,
      "num_tokens": 50064475.0,
      "step": 21830
    },
    {
      "entropy": 5.136310338973999,
      "epoch": 2.0975024015369836,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045615434740142307,
      "loss": 4.7858,
      "mean_token_accuracy": 0.22716640681028366,
      "num_tokens": 50076038.0,
      "step": 21835
    },
    {
      "entropy": 5.143194580078125,
      "epoch": 2.0979827089337175,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045613401179764686,
      "loss": 4.7991,
      "mean_token_accuracy": 0.22955496311187745,
      "num_tokens": 50086425.0,
      "step": 21840
    },
    {
      "entropy": 5.17017502784729,
      "epoch": 2.0984630163304514,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045611367198847676,
      "loss": 4.864,
      "mean_token_accuracy": 0.22988806962966918,
      "num_tokens": 50098320.0,
      "step": 21845
    },
    {
      "entropy": 5.075951623916626,
      "epoch": 2.0989433237271853,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000456093327974385,
      "loss": 4.7757,
      "mean_token_accuracy": 0.23008698225021362,
      "num_tokens": 50109619.0,
      "step": 21850
    },
    {
      "entropy": 5.108696126937867,
      "epoch": 2.099423631123919,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004560729797558438,
      "loss": 4.7704,
      "mean_token_accuracy": 0.23491482436656952,
      "num_tokens": 50121043.0,
      "step": 21855
    },
    {
      "entropy": 5.145803070068359,
      "epoch": 2.099903938520653,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004560526273333259,
      "loss": 4.8781,
      "mean_token_accuracy": 0.22382663637399675,
      "num_tokens": 50132673.0,
      "step": 21860
    },
    {
      "entropy": 5.069375038146973,
      "epoch": 2.100384245917387,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045603227070730346,
      "loss": 4.7205,
      "mean_token_accuracy": 0.22840524166822435,
      "num_tokens": 50144393.0,
      "step": 21865
    },
    {
      "entropy": 5.105737161636353,
      "epoch": 2.1008645533141213,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045601190987824933,
      "loss": 4.7681,
      "mean_token_accuracy": 0.22679009586572646,
      "num_tokens": 50156278.0,
      "step": 21870
    },
    {
      "entropy": 5.075602293014526,
      "epoch": 2.101344860710855,
      "grad_norm": 1.125,
      "learning_rate": 0.00045599154484663606,
      "loss": 4.8119,
      "mean_token_accuracy": 0.22936656177043915,
      "num_tokens": 50167652.0,
      "step": 21875
    },
    {
      "entropy": 5.122927856445313,
      "epoch": 2.101825168107589,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045597117561293663,
      "loss": 4.8876,
      "mean_token_accuracy": 0.2306849718093872,
      "num_tokens": 50179482.0,
      "step": 21880
    },
    {
      "entropy": 5.150292110443115,
      "epoch": 2.102305475504323,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004559508021776238,
      "loss": 4.8103,
      "mean_token_accuracy": 0.23201755583286285,
      "num_tokens": 50190137.0,
      "step": 21885
    },
    {
      "entropy": 5.337715530395508,
      "epoch": 2.1027857829010568,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004559304245411707,
      "loss": 4.9838,
      "mean_token_accuracy": 0.2196114778518677,
      "num_tokens": 50201853.0,
      "step": 21890
    },
    {
      "entropy": 5.141672086715698,
      "epoch": 2.1032660902977907,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045591004270405044,
      "loss": 4.7773,
      "mean_token_accuracy": 0.23600849509239197,
      "num_tokens": 50212635.0,
      "step": 21895
    },
    {
      "entropy": 5.004324674606323,
      "epoch": 2.1037463976945245,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004558896566667361,
      "loss": 4.7148,
      "mean_token_accuracy": 0.22909828275442123,
      "num_tokens": 50224842.0,
      "step": 21900
    },
    {
      "entropy": 5.153818798065186,
      "epoch": 2.1042267050912584,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045586926642970113,
      "loss": 4.8736,
      "mean_token_accuracy": 0.2258002758026123,
      "num_tokens": 50236741.0,
      "step": 21905
    },
    {
      "entropy": 5.190854787826538,
      "epoch": 2.1047070124879923,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004558488719934188,
      "loss": 4.7842,
      "mean_token_accuracy": 0.2248753637075424,
      "num_tokens": 50247572.0,
      "step": 21910
    },
    {
      "entropy": 5.192377376556396,
      "epoch": 2.105187319884726,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004558284733583627,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22839969247579575,
      "num_tokens": 50258566.0,
      "step": 21915
    },
    {
      "entropy": 5.117787456512451,
      "epoch": 2.10566762728146,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045580807052500645,
      "loss": 4.8143,
      "mean_token_accuracy": 0.2317997172474861,
      "num_tokens": 50270577.0,
      "step": 21920
    },
    {
      "entropy": 5.11265926361084,
      "epoch": 2.106147934678194,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004557876634938236,
      "loss": 4.8146,
      "mean_token_accuracy": 0.23246464282274246,
      "num_tokens": 50281124.0,
      "step": 21925
    },
    {
      "entropy": 5.138473463058472,
      "epoch": 2.106628242074928,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004557672522652881,
      "loss": 4.7291,
      "mean_token_accuracy": 0.23490789383649827,
      "num_tokens": 50291945.0,
      "step": 21930
    },
    {
      "entropy": 5.102387142181397,
      "epoch": 2.1071085494716617,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004557468368398738,
      "loss": 4.7787,
      "mean_token_accuracy": 0.2338991716504097,
      "num_tokens": 50303312.0,
      "step": 21935
    },
    {
      "entropy": 5.091919422149658,
      "epoch": 2.1075888568683956,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004557264172180546,
      "loss": 4.8084,
      "mean_token_accuracy": 0.2282189294695854,
      "num_tokens": 50314791.0,
      "step": 21940
    },
    {
      "entropy": 5.206293821334839,
      "epoch": 2.1080691642651295,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004557059934003046,
      "loss": 4.8797,
      "mean_token_accuracy": 0.224047289788723,
      "num_tokens": 50326438.0,
      "step": 21945
    },
    {
      "entropy": 5.030334520339966,
      "epoch": 2.108549471661864,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004556855653870981,
      "loss": 4.6294,
      "mean_token_accuracy": 0.24084686040878295,
      "num_tokens": 50338091.0,
      "step": 21950
    },
    {
      "entropy": 5.028448581695557,
      "epoch": 2.1090297790585977,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004556651331789092,
      "loss": 4.7274,
      "mean_token_accuracy": 0.23792562782764434,
      "num_tokens": 50350387.0,
      "step": 21955
    },
    {
      "entropy": 5.127129220962525,
      "epoch": 2.1095100864553316,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004556446967762125,
      "loss": 4.7905,
      "mean_token_accuracy": 0.23103787750005722,
      "num_tokens": 50362689.0,
      "step": 21960
    },
    {
      "entropy": 5.1273902416229244,
      "epoch": 2.1099903938520654,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045562425617948226,
      "loss": 4.7442,
      "mean_token_accuracy": 0.2344693049788475,
      "num_tokens": 50373290.0,
      "step": 21965
    },
    {
      "entropy": 5.082629776000976,
      "epoch": 2.1104707012487993,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045560381138919315,
      "loss": 4.814,
      "mean_token_accuracy": 0.22500716149806976,
      "num_tokens": 50385267.0,
      "step": 21970
    },
    {
      "entropy": 5.133916139602661,
      "epoch": 2.110951008645533,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045558336240581984,
      "loss": 4.8339,
      "mean_token_accuracy": 0.22592054754495622,
      "num_tokens": 50398014.0,
      "step": 21975
    },
    {
      "entropy": 5.165552616119385,
      "epoch": 2.111431316042267,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045556290922983705,
      "loss": 4.8073,
      "mean_token_accuracy": 0.23242446035146713,
      "num_tokens": 50408661.0,
      "step": 21980
    },
    {
      "entropy": 5.098717164993286,
      "epoch": 2.111911623439001,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004555424518617197,
      "loss": 4.7906,
      "mean_token_accuracy": 0.23299820870161056,
      "num_tokens": 50419231.0,
      "step": 21985
    },
    {
      "entropy": 5.145823335647583,
      "epoch": 2.112391930835735,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045552199030194274,
      "loss": 4.7756,
      "mean_token_accuracy": 0.23035596311092377,
      "num_tokens": 50430412.0,
      "step": 21990
    },
    {
      "entropy": 5.233340120315551,
      "epoch": 2.1128722382324687,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045550152455098113,
      "loss": 4.9363,
      "mean_token_accuracy": 0.22196089178323747,
      "num_tokens": 50442431.0,
      "step": 21995
    },
    {
      "entropy": 5.1133557796478275,
      "epoch": 2.1133525456292026,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004554810546093102,
      "loss": 4.8202,
      "mean_token_accuracy": 0.2278466984629631,
      "num_tokens": 50454417.0,
      "step": 22000
    },
    {
      "entropy": 5.088508892059326,
      "epoch": 2.1138328530259365,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000455460580477405,
      "loss": 4.8151,
      "mean_token_accuracy": 0.229460608959198,
      "num_tokens": 50466447.0,
      "step": 22005
    },
    {
      "entropy": 5.105055141448974,
      "epoch": 2.1143131604226704,
      "grad_norm": 1.140625,
      "learning_rate": 0.000455440102155741,
      "loss": 4.7766,
      "mean_token_accuracy": 0.2265194684267044,
      "num_tokens": 50477843.0,
      "step": 22010
    },
    {
      "entropy": 5.117995738983154,
      "epoch": 2.1147934678194042,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004554196196447937,
      "loss": 4.8433,
      "mean_token_accuracy": 0.22670915722846985,
      "num_tokens": 50489407.0,
      "step": 22015
    },
    {
      "entropy": 5.145983600616455,
      "epoch": 2.115273775216138,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004553991329450385,
      "loss": 4.7833,
      "mean_token_accuracy": 0.2258385479450226,
      "num_tokens": 50499492.0,
      "step": 22020
    },
    {
      "entropy": 5.064445972442627,
      "epoch": 2.115754082612872,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045537864205695116,
      "loss": 4.8366,
      "mean_token_accuracy": 0.22338130623102187,
      "num_tokens": 50511741.0,
      "step": 22025
    },
    {
      "entropy": 5.050553178787231,
      "epoch": 2.1162343900096063,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004553581469810073,
      "loss": 4.7426,
      "mean_token_accuracy": 0.2319988414645195,
      "num_tokens": 50523238.0,
      "step": 22030
    },
    {
      "entropy": 5.08756160736084,
      "epoch": 2.11671469740634,
      "grad_norm": 1.125,
      "learning_rate": 0.00045533764771768287,
      "loss": 4.7762,
      "mean_token_accuracy": 0.23490686118602752,
      "num_tokens": 50534420.0,
      "step": 22035
    },
    {
      "entropy": 5.039917039871216,
      "epoch": 2.117195004803074,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00045531714426745373,
      "loss": 4.7357,
      "mean_token_accuracy": 0.2376674309372902,
      "num_tokens": 50545785.0,
      "step": 22040
    },
    {
      "entropy": 5.131248044967651,
      "epoch": 2.117675312199808,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004552966366307959,
      "loss": 4.809,
      "mean_token_accuracy": 0.2245887726545334,
      "num_tokens": 50556441.0,
      "step": 22045
    },
    {
      "entropy": 5.049736785888672,
      "epoch": 2.118155619596542,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004552761248081856,
      "loss": 4.7397,
      "mean_token_accuracy": 0.23258329182863235,
      "num_tokens": 50568519.0,
      "step": 22050
    },
    {
      "entropy": 5.1407571792602536,
      "epoch": 2.1186359269932757,
      "grad_norm": 1.09375,
      "learning_rate": 0.000455255608800099,
      "loss": 4.8227,
      "mean_token_accuracy": 0.22938024252653122,
      "num_tokens": 50579469.0,
      "step": 22055
    },
    {
      "entropy": 4.97040319442749,
      "epoch": 2.1191162343900096,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045523508860701237,
      "loss": 4.6165,
      "mean_token_accuracy": 0.24262254387140275,
      "num_tokens": 50591405.0,
      "step": 22060
    },
    {
      "entropy": 5.094077730178833,
      "epoch": 2.1195965417867435,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004552145642294021,
      "loss": 4.8152,
      "mean_token_accuracy": 0.23994069695472717,
      "num_tokens": 50603742.0,
      "step": 22065
    },
    {
      "entropy": 5.103747606277466,
      "epoch": 2.1200768491834774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045519403566774493,
      "loss": 4.7944,
      "mean_token_accuracy": 0.23053058832883835,
      "num_tokens": 50615753.0,
      "step": 22070
    },
    {
      "entropy": 5.094579029083252,
      "epoch": 2.1205571565802113,
      "grad_norm": 1.125,
      "learning_rate": 0.0004551735029225172,
      "loss": 4.712,
      "mean_token_accuracy": 0.23386039286851884,
      "num_tokens": 50626715.0,
      "step": 22075
    },
    {
      "entropy": 5.058100938796997,
      "epoch": 2.121037463976945,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045515296599419583,
      "loss": 4.7379,
      "mean_token_accuracy": 0.2351568579673767,
      "num_tokens": 50637552.0,
      "step": 22080
    },
    {
      "entropy": 5.105273103713989,
      "epoch": 2.121517771373679,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004551324248832574,
      "loss": 4.8782,
      "mean_token_accuracy": 0.22308872640132904,
      "num_tokens": 50648143.0,
      "step": 22085
    },
    {
      "entropy": 5.113940763473511,
      "epoch": 2.121998078770413,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004551118795901791,
      "loss": 4.8527,
      "mean_token_accuracy": 0.21738509833812714,
      "num_tokens": 50658878.0,
      "step": 22090
    },
    {
      "entropy": 5.0933012008667,
      "epoch": 2.122478386167147,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004550913301154376,
      "loss": 4.8041,
      "mean_token_accuracy": 0.23284170031547546,
      "num_tokens": 50670912.0,
      "step": 22095
    },
    {
      "entropy": 5.127492475509643,
      "epoch": 2.1229586935638807,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004550707764595103,
      "loss": 4.8382,
      "mean_token_accuracy": 0.22272917330265046,
      "num_tokens": 50683030.0,
      "step": 22100
    },
    {
      "entropy": 5.171814107894898,
      "epoch": 2.123439000960615,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045505021862287434,
      "loss": 4.8513,
      "mean_token_accuracy": 0.21623581051826476,
      "num_tokens": 50694654.0,
      "step": 22105
    },
    {
      "entropy": 5.180832195281982,
      "epoch": 2.123919308357349,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045502965660600684,
      "loss": 4.876,
      "mean_token_accuracy": 0.22136667668819426,
      "num_tokens": 50706277.0,
      "step": 22110
    },
    {
      "entropy": 5.099314785003662,
      "epoch": 2.1243996157540828,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004550090904093853,
      "loss": 4.7757,
      "mean_token_accuracy": 0.23358724564313887,
      "num_tokens": 50717610.0,
      "step": 22115
    },
    {
      "entropy": 5.178705215454102,
      "epoch": 2.1248799231508166,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004549885200334872,
      "loss": 4.955,
      "mean_token_accuracy": 0.21713918596506118,
      "num_tokens": 50728827.0,
      "step": 22120
    },
    {
      "entropy": 5.073518371582031,
      "epoch": 2.1253602305475505,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004549679454787901,
      "loss": 4.7486,
      "mean_token_accuracy": 0.23730372339487077,
      "num_tokens": 50740098.0,
      "step": 22125
    },
    {
      "entropy": 5.130878973007202,
      "epoch": 2.1258405379442844,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045494736674577175,
      "loss": 4.8146,
      "mean_token_accuracy": 0.22786290049552918,
      "num_tokens": 50752166.0,
      "step": 22130
    },
    {
      "entropy": 5.119168710708618,
      "epoch": 2.1263208453410183,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004549267838349099,
      "loss": 4.8083,
      "mean_token_accuracy": 0.23251488357782363,
      "num_tokens": 50763561.0,
      "step": 22135
    },
    {
      "entropy": 5.166716051101685,
      "epoch": 2.126801152737752,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004549061967466823,
      "loss": 4.8634,
      "mean_token_accuracy": 0.23351675868034363,
      "num_tokens": 50774194.0,
      "step": 22140
    },
    {
      "entropy": 5.0733240127563475,
      "epoch": 2.127281460134486,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004548856054815671,
      "loss": 4.8076,
      "mean_token_accuracy": 0.2318489745259285,
      "num_tokens": 50784873.0,
      "step": 22145
    },
    {
      "entropy": 5.175985860824585,
      "epoch": 2.12776176753122,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045486501004004225,
      "loss": 4.9079,
      "mean_token_accuracy": 0.22240222841501237,
      "num_tokens": 50795687.0,
      "step": 22150
    },
    {
      "entropy": 5.188103723526001,
      "epoch": 2.128242074927954,
      "grad_norm": 1.125,
      "learning_rate": 0.000454844410422586,
      "loss": 4.8538,
      "mean_token_accuracy": 0.22479133307933807,
      "num_tokens": 50806304.0,
      "step": 22155
    },
    {
      "entropy": 5.002657890319824,
      "epoch": 2.1287223823246877,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045482380662967655,
      "loss": 4.6923,
      "mean_token_accuracy": 0.23126361817121505,
      "num_tokens": 50816775.0,
      "step": 22160
    },
    {
      "entropy": 5.181390762329102,
      "epoch": 2.1292026897214216,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004548031986617923,
      "loss": 4.862,
      "mean_token_accuracy": 0.2216108813881874,
      "num_tokens": 50828266.0,
      "step": 22165
    },
    {
      "entropy": 5.157224512100219,
      "epoch": 2.1296829971181555,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004547825865194117,
      "loss": 4.8079,
      "mean_token_accuracy": 0.2268590971827507,
      "num_tokens": 50840140.0,
      "step": 22170
    },
    {
      "entropy": 5.16111273765564,
      "epoch": 2.1301633045148893,
      "grad_norm": 1.125,
      "learning_rate": 0.00045476197020301323,
      "loss": 4.878,
      "mean_token_accuracy": 0.22430993765592575,
      "num_tokens": 50850415.0,
      "step": 22175
    },
    {
      "entropy": 5.072700500488281,
      "epoch": 2.1306436119116237,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045474134971307554,
      "loss": 4.7971,
      "mean_token_accuracy": 0.22537578642368317,
      "num_tokens": 50861396.0,
      "step": 22180
    },
    {
      "entropy": 5.128022480010986,
      "epoch": 2.1311239193083575,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004547207250500775,
      "loss": 4.8071,
      "mean_token_accuracy": 0.22539043724536895,
      "num_tokens": 50873931.0,
      "step": 22185
    },
    {
      "entropy": 5.128237771987915,
      "epoch": 2.1316042267050914,
      "grad_norm": 1.15625,
      "learning_rate": 0.000454700096214498,
      "loss": 4.7768,
      "mean_token_accuracy": 0.22611454874277115,
      "num_tokens": 50884557.0,
      "step": 22190
    },
    {
      "entropy": 5.071543216705322,
      "epoch": 2.1320845341018253,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045467946320681567,
      "loss": 4.7127,
      "mean_token_accuracy": 0.23287661820650102,
      "num_tokens": 50895757.0,
      "step": 22195
    },
    {
      "entropy": 5.105447483062744,
      "epoch": 2.132564841498559,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004546588260275098,
      "loss": 4.8146,
      "mean_token_accuracy": 0.2302413299679756,
      "num_tokens": 50906437.0,
      "step": 22200
    },
    {
      "entropy": 5.04798526763916,
      "epoch": 2.133045148895293,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045463818467705955,
      "loss": 4.7699,
      "mean_token_accuracy": 0.23630340844392778,
      "num_tokens": 50916689.0,
      "step": 22205
    },
    {
      "entropy": 5.192591381072998,
      "epoch": 2.133525456292027,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000454617539155944,
      "loss": 4.8639,
      "mean_token_accuracy": 0.2207561105489731,
      "num_tokens": 50928266.0,
      "step": 22210
    },
    {
      "entropy": 5.198041343688965,
      "epoch": 2.134005763688761,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045459688946464255,
      "loss": 4.8549,
      "mean_token_accuracy": 0.22911514043807985,
      "num_tokens": 50940220.0,
      "step": 22215
    },
    {
      "entropy": 5.055037641525269,
      "epoch": 2.1344860710854947,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004545762356036346,
      "loss": 4.7703,
      "mean_token_accuracy": 0.2360699877142906,
      "num_tokens": 50952668.0,
      "step": 22220
    },
    {
      "entropy": 5.05750675201416,
      "epoch": 2.1349663784822286,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004545555775733998,
      "loss": 4.7923,
      "mean_token_accuracy": 0.2279730573296547,
      "num_tokens": 50964624.0,
      "step": 22225
    },
    {
      "entropy": 5.165986347198486,
      "epoch": 2.1354466858789625,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045453491537441747,
      "loss": 4.8747,
      "mean_token_accuracy": 0.22170280516147614,
      "num_tokens": 50976704.0,
      "step": 22230
    },
    {
      "entropy": 5.080795097351074,
      "epoch": 2.1359269932756964,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045451424900716763,
      "loss": 4.7234,
      "mean_token_accuracy": 0.23465065211057662,
      "num_tokens": 50987986.0,
      "step": 22235
    },
    {
      "entropy": 5.078517484664917,
      "epoch": 2.1364073006724302,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045449357847212994,
      "loss": 4.7585,
      "mean_token_accuracy": 0.23064401000738144,
      "num_tokens": 50999624.0,
      "step": 22240
    },
    {
      "entropy": 5.125305318832398,
      "epoch": 2.136887608069164,
      "grad_norm": 1.0,
      "learning_rate": 0.0004544729037697844,
      "loss": 4.7775,
      "mean_token_accuracy": 0.23089563399553298,
      "num_tokens": 51011800.0,
      "step": 22245
    },
    {
      "entropy": 5.074768590927124,
      "epoch": 2.137367915465898,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045445222490061093,
      "loss": 4.7263,
      "mean_token_accuracy": 0.23572536259889604,
      "num_tokens": 51023004.0,
      "step": 22250
    },
    {
      "entropy": 4.963903903961182,
      "epoch": 2.1378482228626323,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004544315418650897,
      "loss": 4.6727,
      "mean_token_accuracy": 0.23558634966611863,
      "num_tokens": 51033993.0,
      "step": 22255
    },
    {
      "entropy": 4.997238636016846,
      "epoch": 2.138328530259366,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004544108546637008,
      "loss": 4.7637,
      "mean_token_accuracy": 0.23835617005825044,
      "num_tokens": 51045269.0,
      "step": 22260
    },
    {
      "entropy": 5.070559024810791,
      "epoch": 2.1388088376561,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004543901632969247,
      "loss": 4.6955,
      "mean_token_accuracy": 0.23294834941625595,
      "num_tokens": 51056044.0,
      "step": 22265
    },
    {
      "entropy": 5.112168169021606,
      "epoch": 2.139289145052834,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045436946776524157,
      "loss": 4.7775,
      "mean_token_accuracy": 0.23480267077684402,
      "num_tokens": 51067096.0,
      "step": 22270
    },
    {
      "entropy": 5.075560855865478,
      "epoch": 2.139769452449568,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045434876806913204,
      "loss": 4.7912,
      "mean_token_accuracy": 0.23536744415760041,
      "num_tokens": 51080050.0,
      "step": 22275
    },
    {
      "entropy": 5.147746467590332,
      "epoch": 2.1402497598463017,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004543280642090767,
      "loss": 4.8485,
      "mean_token_accuracy": 0.2215731456875801,
      "num_tokens": 51091333.0,
      "step": 22280
    },
    {
      "entropy": 5.156608772277832,
      "epoch": 2.1407300672430356,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004543073561855562,
      "loss": 4.8549,
      "mean_token_accuracy": 0.2260780692100525,
      "num_tokens": 51102508.0,
      "step": 22285
    },
    {
      "entropy": 5.062401819229126,
      "epoch": 2.1412103746397695,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004542866439990513,
      "loss": 4.7948,
      "mean_token_accuracy": 0.22423228174448012,
      "num_tokens": 51114427.0,
      "step": 22290
    },
    {
      "entropy": 5.166135263442993,
      "epoch": 2.1416906820365034,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004542659276500429,
      "loss": 4.8005,
      "mean_token_accuracy": 0.23201976716518402,
      "num_tokens": 51126486.0,
      "step": 22295
    },
    {
      "entropy": 5.096059131622314,
      "epoch": 2.1421709894332372,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045424520713901204,
      "loss": 4.7386,
      "mean_token_accuracy": 0.23788861334323883,
      "num_tokens": 51136813.0,
      "step": 22300
    },
    {
      "entropy": 5.000286674499511,
      "epoch": 2.142651296829971,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004542244824664396,
      "loss": 4.6695,
      "mean_token_accuracy": 0.23362847417593002,
      "num_tokens": 51148257.0,
      "step": 22305
    },
    {
      "entropy": 5.134123373031616,
      "epoch": 2.143131604226705,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045420375363280696,
      "loss": 4.8815,
      "mean_token_accuracy": 0.21876666247844695,
      "num_tokens": 51159159.0,
      "step": 22310
    },
    {
      "entropy": 5.07529993057251,
      "epoch": 2.143611911623439,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045418302063859526,
      "loss": 4.7429,
      "mean_token_accuracy": 0.23444928526878356,
      "num_tokens": 51170900.0,
      "step": 22315
    },
    {
      "entropy": 5.203510046005249,
      "epoch": 2.1440922190201728,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045416228348428583,
      "loss": 4.9142,
      "mean_token_accuracy": 0.21714494824409486,
      "num_tokens": 51183707.0,
      "step": 22320
    },
    {
      "entropy": 5.076108455657959,
      "epoch": 2.1445725264169067,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045414154217036023,
      "loss": 4.7185,
      "mean_token_accuracy": 0.2371895879507065,
      "num_tokens": 51195114.0,
      "step": 22325
    },
    {
      "entropy": 5.046933889389038,
      "epoch": 2.1450528338136405,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045412079669730006,
      "loss": 4.7258,
      "mean_token_accuracy": 0.23803662955760957,
      "num_tokens": 51206210.0,
      "step": 22330
    },
    {
      "entropy": 5.066112375259399,
      "epoch": 2.1455331412103744,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004541000470655867,
      "loss": 4.7923,
      "mean_token_accuracy": 0.23359201550483705,
      "num_tokens": 51217657.0,
      "step": 22335
    },
    {
      "entropy": 5.176227664947509,
      "epoch": 2.1460134486071087,
      "grad_norm": 1.125,
      "learning_rate": 0.00045407929327570215,
      "loss": 4.864,
      "mean_token_accuracy": 0.22981271296739578,
      "num_tokens": 51227836.0,
      "step": 22340
    },
    {
      "entropy": 5.112975072860718,
      "epoch": 2.1464937560038426,
      "grad_norm": 1.125,
      "learning_rate": 0.0004540585353281282,
      "loss": 4.771,
      "mean_token_accuracy": 0.24184567630290985,
      "num_tokens": 51238936.0,
      "step": 22345
    },
    {
      "entropy": 5.146573734283447,
      "epoch": 2.1469740634005765,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004540377732233467,
      "loss": 4.9238,
      "mean_token_accuracy": 0.22959763556718826,
      "num_tokens": 51250696.0,
      "step": 22350
    },
    {
      "entropy": 5.17834963798523,
      "epoch": 2.1474543707973104,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004540170069618397,
      "loss": 4.8281,
      "mean_token_accuracy": 0.2297007694840431,
      "num_tokens": 51262414.0,
      "step": 22355
    },
    {
      "entropy": 5.042438554763794,
      "epoch": 2.1479346781940443,
      "grad_norm": 1.125,
      "learning_rate": 0.00045399623654408946,
      "loss": 4.7296,
      "mean_token_accuracy": 0.22824445217847825,
      "num_tokens": 51273267.0,
      "step": 22360
    },
    {
      "entropy": 5.052595233917236,
      "epoch": 2.148414985590778,
      "grad_norm": 1.046875,
      "learning_rate": 0.000453975461970578,
      "loss": 4.8204,
      "mean_token_accuracy": 0.23316656649112702,
      "num_tokens": 51284783.0,
      "step": 22365
    },
    {
      "entropy": 5.105958318710327,
      "epoch": 2.148895292987512,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004539546832417879,
      "loss": 4.7546,
      "mean_token_accuracy": 0.23036372065544128,
      "num_tokens": 51296048.0,
      "step": 22370
    },
    {
      "entropy": 5.079136371612549,
      "epoch": 2.149375600384246,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045393390035820136,
      "loss": 4.7764,
      "mean_token_accuracy": 0.2301976889371872,
      "num_tokens": 51306731.0,
      "step": 22375
    },
    {
      "entropy": 5.127407360076904,
      "epoch": 2.14985590778098,
      "grad_norm": 1.09375,
      "learning_rate": 0.000453913113320301,
      "loss": 4.8839,
      "mean_token_accuracy": 0.22248636037111283,
      "num_tokens": 51318431.0,
      "step": 22380
    },
    {
      "entropy": 5.158274555206299,
      "epoch": 2.1503362151777137,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004538923221285694,
      "loss": 4.8474,
      "mean_token_accuracy": 0.2337651014328003,
      "num_tokens": 51330585.0,
      "step": 22385
    },
    {
      "entropy": 5.073586702346802,
      "epoch": 2.1508165225744476,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004538715267834893,
      "loss": 4.7778,
      "mean_token_accuracy": 0.229562845826149,
      "num_tokens": 51341936.0,
      "step": 22390
    },
    {
      "entropy": 5.109851503372193,
      "epoch": 2.1512968299711814,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004538507272855434,
      "loss": 4.8308,
      "mean_token_accuracy": 0.23133473843336105,
      "num_tokens": 51353270.0,
      "step": 22395
    },
    {
      "entropy": 5.216899585723877,
      "epoch": 2.1517771373679153,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045382992363521486,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22154793441295623,
      "num_tokens": 51365470.0,
      "step": 22400
    },
    {
      "entropy": 5.095748043060302,
      "epoch": 2.152257444764649,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045380911583298633,
      "loss": 4.808,
      "mean_token_accuracy": 0.22889665514230728,
      "num_tokens": 51375911.0,
      "step": 22405
    },
    {
      "entropy": 5.103784036636353,
      "epoch": 2.152737752161383,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045378830387934123,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22170654386281968,
      "num_tokens": 51387082.0,
      "step": 22410
    },
    {
      "entropy": 5.111545658111572,
      "epoch": 2.1532180595581174,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004537674877747626,
      "loss": 4.7968,
      "mean_token_accuracy": 0.2291550636291504,
      "num_tokens": 51398826.0,
      "step": 22415
    },
    {
      "entropy": 5.127094268798828,
      "epoch": 2.1536983669548513,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045374666751973365,
      "loss": 4.7857,
      "mean_token_accuracy": 0.22865325957536697,
      "num_tokens": 51410848.0,
      "step": 22420
    },
    {
      "entropy": 5.0831092357635494,
      "epoch": 2.154178674351585,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045372584311473784,
      "loss": 4.7181,
      "mean_token_accuracy": 0.2314053237438202,
      "num_tokens": 51422272.0,
      "step": 22425
    },
    {
      "entropy": 5.023440980911255,
      "epoch": 2.154658981748319,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004537050145602587,
      "loss": 4.7411,
      "mean_token_accuracy": 0.2306130051612854,
      "num_tokens": 51433139.0,
      "step": 22430
    },
    {
      "entropy": 5.098038148880005,
      "epoch": 2.155139289145053,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004536841818567798,
      "loss": 4.735,
      "mean_token_accuracy": 0.22860486358404158,
      "num_tokens": 51445582.0,
      "step": 22435
    },
    {
      "entropy": 5.095311784744263,
      "epoch": 2.155619596541787,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004536633450047847,
      "loss": 4.6825,
      "mean_token_accuracy": 0.23743011355400084,
      "num_tokens": 51456859.0,
      "step": 22440
    },
    {
      "entropy": 5.064573001861572,
      "epoch": 2.1560999039385207,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045364250400475734,
      "loss": 4.8519,
      "mean_token_accuracy": 0.2304494008421898,
      "num_tokens": 51468939.0,
      "step": 22445
    },
    {
      "entropy": 5.072771167755127,
      "epoch": 2.1565802113352546,
      "grad_norm": 1.125,
      "learning_rate": 0.0004536216588571814,
      "loss": 4.8047,
      "mean_token_accuracy": 0.22680765092372895,
      "num_tokens": 51480390.0,
      "step": 22450
    },
    {
      "entropy": 5.101582479476929,
      "epoch": 2.1570605187319885,
      "grad_norm": 1.0625,
      "learning_rate": 0.000453600809562541,
      "loss": 4.7832,
      "mean_token_accuracy": 0.23274296522140503,
      "num_tokens": 51493194.0,
      "step": 22455
    },
    {
      "entropy": 5.0897070407867435,
      "epoch": 2.1575408261287223,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004535799561213202,
      "loss": 4.7822,
      "mean_token_accuracy": 0.22871144860982895,
      "num_tokens": 51505084.0,
      "step": 22460
    },
    {
      "entropy": 5.155414867401123,
      "epoch": 2.158021133525456,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000453559098534003,
      "loss": 4.815,
      "mean_token_accuracy": 0.23038263916969298,
      "num_tokens": 51516474.0,
      "step": 22465
    },
    {
      "entropy": 5.107330179214477,
      "epoch": 2.15850144092219,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004535382368010738,
      "loss": 4.8048,
      "mean_token_accuracy": 0.23444980531930923,
      "num_tokens": 51526734.0,
      "step": 22470
    },
    {
      "entropy": 5.058229780197143,
      "epoch": 2.158981748318924,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045351737092301676,
      "loss": 4.7758,
      "mean_token_accuracy": 0.23163246661424636,
      "num_tokens": 51538230.0,
      "step": 22475
    },
    {
      "entropy": 5.108958292007446,
      "epoch": 2.159462055715658,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004534965009003165,
      "loss": 4.7933,
      "mean_token_accuracy": 0.23016343861818314,
      "num_tokens": 51551671.0,
      "step": 22480
    },
    {
      "entropy": 5.082572364807129,
      "epoch": 2.1599423631123917,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004534756267334576,
      "loss": 4.7926,
      "mean_token_accuracy": 0.2300082892179489,
      "num_tokens": 51563034.0,
      "step": 22485
    },
    {
      "entropy": 5.035448694229126,
      "epoch": 2.160422670509126,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045345474842292455,
      "loss": 4.6836,
      "mean_token_accuracy": 0.23904351890087128,
      "num_tokens": 51575210.0,
      "step": 22490
    },
    {
      "entropy": 5.0996935844421385,
      "epoch": 2.16090297790586,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004534338659692022,
      "loss": 4.8502,
      "mean_token_accuracy": 0.2275824710726738,
      "num_tokens": 51585586.0,
      "step": 22495
    },
    {
      "entropy": 5.113663530349731,
      "epoch": 2.161383285302594,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004534129793727753,
      "loss": 4.8281,
      "mean_token_accuracy": 0.22628759890794753,
      "num_tokens": 51596519.0,
      "step": 22500
    },
    {
      "entropy": 5.059601306915283,
      "epoch": 2.1618635926993277,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004533920886341288,
      "loss": 4.7744,
      "mean_token_accuracy": 0.23476321548223494,
      "num_tokens": 51608421.0,
      "step": 22505
    },
    {
      "entropy": 5.142944574356079,
      "epoch": 2.1623439000960616,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004533711937537477,
      "loss": 4.8457,
      "mean_token_accuracy": 0.22144338190555574,
      "num_tokens": 51620615.0,
      "step": 22510
    },
    {
      "entropy": 5.174992227554322,
      "epoch": 2.1628242074927955,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004533502947321171,
      "loss": 4.8289,
      "mean_token_accuracy": 0.22758959978818893,
      "num_tokens": 51632797.0,
      "step": 22515
    },
    {
      "entropy": 5.113536071777344,
      "epoch": 2.1633045148895294,
      "grad_norm": 1.125,
      "learning_rate": 0.0004533293915697223,
      "loss": 4.8181,
      "mean_token_accuracy": 0.2225467935204506,
      "num_tokens": 51644735.0,
      "step": 22520
    },
    {
      "entropy": 5.071394777297973,
      "epoch": 2.1637848222862632,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045330848426704853,
      "loss": 4.7999,
      "mean_token_accuracy": 0.2258875384926796,
      "num_tokens": 51656270.0,
      "step": 22525
    },
    {
      "entropy": 5.085580444335937,
      "epoch": 2.164265129682997,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004532875728245813,
      "loss": 4.7477,
      "mean_token_accuracy": 0.23762866258621215,
      "num_tokens": 51668223.0,
      "step": 22530
    },
    {
      "entropy": 5.080254364013672,
      "epoch": 2.164745437079731,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045326665724280594,
      "loss": 4.7154,
      "mean_token_accuracy": 0.23662257045507432,
      "num_tokens": 51680298.0,
      "step": 22535
    },
    {
      "entropy": 5.155526971817016,
      "epoch": 2.165225744476465,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045324573752220814,
      "loss": 4.896,
      "mean_token_accuracy": 0.22130007445812225,
      "num_tokens": 51692803.0,
      "step": 22540
    },
    {
      "entropy": 5.0807374000549315,
      "epoch": 2.1657060518731988,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045322481366327365,
      "loss": 4.7772,
      "mean_token_accuracy": 0.22820665538311005,
      "num_tokens": 51704478.0,
      "step": 22545
    },
    {
      "entropy": 5.127686500549316,
      "epoch": 2.1661863592699326,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004532038856664882,
      "loss": 4.7686,
      "mean_token_accuracy": 0.23359857946634294,
      "num_tokens": 51714627.0,
      "step": 22550
    },
    {
      "entropy": 5.0890522480010985,
      "epoch": 2.1666666666666665,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004531829535323376,
      "loss": 4.7395,
      "mean_token_accuracy": 0.24152956902980804,
      "num_tokens": 51726297.0,
      "step": 22555
    },
    {
      "entropy": 5.025962829589844,
      "epoch": 2.1671469740634004,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000453162017261308,
      "loss": 4.7669,
      "mean_token_accuracy": 0.2363984242081642,
      "num_tokens": 51737676.0,
      "step": 22560
    },
    {
      "entropy": 5.09314193725586,
      "epoch": 2.1676272814601343,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004531410768538854,
      "loss": 4.8034,
      "mean_token_accuracy": 0.23226175010204314,
      "num_tokens": 51749532.0,
      "step": 22565
    },
    {
      "entropy": 5.11493935585022,
      "epoch": 2.168107588856868,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045312013231055596,
      "loss": 4.8083,
      "mean_token_accuracy": 0.23494400084018707,
      "num_tokens": 51761047.0,
      "step": 22570
    },
    {
      "entropy": 5.156091403961182,
      "epoch": 2.1685878962536025,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045309918363180593,
      "loss": 4.7785,
      "mean_token_accuracy": 0.233763624727726,
      "num_tokens": 51772134.0,
      "step": 22575
    },
    {
      "entropy": 5.20871753692627,
      "epoch": 2.1690682036503364,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045307823081812166,
      "loss": 4.909,
      "mean_token_accuracy": 0.21739418357610701,
      "num_tokens": 51785775.0,
      "step": 22580
    },
    {
      "entropy": 5.043304014205932,
      "epoch": 2.1695485110470702,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045305727386998977,
      "loss": 4.7484,
      "mean_token_accuracy": 0.23272988349199294,
      "num_tokens": 51797306.0,
      "step": 22585
    },
    {
      "entropy": 5.095614671707153,
      "epoch": 2.170028818443804,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004530363127878966,
      "loss": 4.8506,
      "mean_token_accuracy": 0.22515084445476533,
      "num_tokens": 51809312.0,
      "step": 22590
    },
    {
      "entropy": 5.108261251449585,
      "epoch": 2.170509125840538,
      "grad_norm": 1.390625,
      "learning_rate": 0.00045301534757232885,
      "loss": 4.8171,
      "mean_token_accuracy": 0.23263057619333266,
      "num_tokens": 51821646.0,
      "step": 22595
    },
    {
      "entropy": 5.064534521102905,
      "epoch": 2.170989433237272,
      "grad_norm": 1.125,
      "learning_rate": 0.0004529943782237735,
      "loss": 4.7056,
      "mean_token_accuracy": 0.23698605895042418,
      "num_tokens": 51831546.0,
      "step": 22600
    },
    {
      "entropy": 5.100519037246704,
      "epoch": 2.1714697406340058,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045297340474271717,
      "loss": 4.8346,
      "mean_token_accuracy": 0.22837483435869216,
      "num_tokens": 51842273.0,
      "step": 22605
    },
    {
      "entropy": 5.14804482460022,
      "epoch": 2.1719500480307397,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004529524271296468,
      "loss": 4.8493,
      "mean_token_accuracy": 0.22853796780109406,
      "num_tokens": 51854715.0,
      "step": 22610
    },
    {
      "entropy": 5.0822783470153805,
      "epoch": 2.1724303554274735,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045293144538504943,
      "loss": 4.7774,
      "mean_token_accuracy": 0.22889146357774734,
      "num_tokens": 51866819.0,
      "step": 22615
    },
    {
      "entropy": 5.08549747467041,
      "epoch": 2.1729106628242074,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004529104595094124,
      "loss": 4.7661,
      "mean_token_accuracy": 0.2411769896745682,
      "num_tokens": 51878651.0,
      "step": 22620
    },
    {
      "entropy": 5.1355376720428465,
      "epoch": 2.1733909702209413,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045288946950322264,
      "loss": 4.8141,
      "mean_token_accuracy": 0.2291857123374939,
      "num_tokens": 51888640.0,
      "step": 22625
    },
    {
      "entropy": 5.054444742202759,
      "epoch": 2.173871277617675,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004528684753669677,
      "loss": 4.7401,
      "mean_token_accuracy": 0.2393981397151947,
      "num_tokens": 51899810.0,
      "step": 22630
    },
    {
      "entropy": 5.095856428146362,
      "epoch": 2.174351585014409,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004528474771011349,
      "loss": 4.8171,
      "mean_token_accuracy": 0.22483566254377366,
      "num_tokens": 51910985.0,
      "step": 22635
    },
    {
      "entropy": 5.112012672424316,
      "epoch": 2.174831892411143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045282647470621176,
      "loss": 4.7468,
      "mean_token_accuracy": 0.2374114587903023,
      "num_tokens": 51922550.0,
      "step": 22640
    },
    {
      "entropy": 5.067946195602417,
      "epoch": 2.175312199807877,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045280546818268595,
      "loss": 4.6714,
      "mean_token_accuracy": 0.2336398482322693,
      "num_tokens": 51934319.0,
      "step": 22645
    },
    {
      "entropy": 5.096906900405884,
      "epoch": 2.175792507204611,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004527844575310452,
      "loss": 4.9065,
      "mean_token_accuracy": 0.21944654881954193,
      "num_tokens": 51945516.0,
      "step": 22650
    },
    {
      "entropy": 5.201075220108033,
      "epoch": 2.176272814601345,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045276344275177715,
      "loss": 4.8103,
      "mean_token_accuracy": 0.2266676679253578,
      "num_tokens": 51957033.0,
      "step": 22655
    },
    {
      "entropy": 5.087824535369873,
      "epoch": 2.176753121998079,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045274242384536984,
      "loss": 4.8073,
      "mean_token_accuracy": 0.23012621700763702,
      "num_tokens": 51968211.0,
      "step": 22660
    },
    {
      "entropy": 5.028482866287232,
      "epoch": 2.177233429394813,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004527214008123113,
      "loss": 4.7134,
      "mean_token_accuracy": 0.23513562679290773,
      "num_tokens": 51978709.0,
      "step": 22665
    },
    {
      "entropy": 5.10414686203003,
      "epoch": 2.1777137367915467,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004527003736530895,
      "loss": 4.8509,
      "mean_token_accuracy": 0.22367929965257644,
      "num_tokens": 51989997.0,
      "step": 22670
    },
    {
      "entropy": 5.045032691955567,
      "epoch": 2.1781940441882806,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045267934236819265,
      "loss": 4.7516,
      "mean_token_accuracy": 0.23588275611400605,
      "num_tokens": 52001730.0,
      "step": 22675
    },
    {
      "entropy": 5.154171705245972,
      "epoch": 2.1786743515850144,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004526583069581091,
      "loss": 4.8311,
      "mean_token_accuracy": 0.2206468716263771,
      "num_tokens": 52012267.0,
      "step": 22680
    },
    {
      "entropy": 5.210177230834961,
      "epoch": 2.1791546589817483,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004526372674233272,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22339679598808287,
      "num_tokens": 52024700.0,
      "step": 22685
    },
    {
      "entropy": 5.1391921043396,
      "epoch": 2.179634966378482,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045261622376433543,
      "loss": 4.7769,
      "mean_token_accuracy": 0.23246812373399733,
      "num_tokens": 52036685.0,
      "step": 22690
    },
    {
      "entropy": 5.103837347030639,
      "epoch": 2.180115273775216,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045259517598162237,
      "loss": 4.7837,
      "mean_token_accuracy": 0.23136062175035477,
      "num_tokens": 52048911.0,
      "step": 22695
    },
    {
      "entropy": 5.0935615539550785,
      "epoch": 2.18059558117195,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004525741240756766,
      "loss": 4.7867,
      "mean_token_accuracy": 0.23264979422092438,
      "num_tokens": 52060039.0,
      "step": 22700
    },
    {
      "entropy": 5.080454635620117,
      "epoch": 2.181075888568684,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004525530680469871,
      "loss": 4.7756,
      "mean_token_accuracy": 0.22993822395801544,
      "num_tokens": 52070549.0,
      "step": 22705
    },
    {
      "entropy": 5.16889967918396,
      "epoch": 2.1815561959654177,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045253200789604245,
      "loss": 4.8675,
      "mean_token_accuracy": 0.22030745893716813,
      "num_tokens": 52081573.0,
      "step": 22710
    },
    {
      "entropy": 5.036343908309936,
      "epoch": 2.1820365033621516,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045251094362333186,
      "loss": 4.7773,
      "mean_token_accuracy": 0.23250143080949784,
      "num_tokens": 52093694.0,
      "step": 22715
    },
    {
      "entropy": 5.12799186706543,
      "epoch": 2.1825168107588855,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004524898752293441,
      "loss": 4.8818,
      "mean_token_accuracy": 0.2292369395494461,
      "num_tokens": 52104784.0,
      "step": 22720
    },
    {
      "entropy": 5.173839807510376,
      "epoch": 2.18299711815562,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045246880271456857,
      "loss": 4.849,
      "mean_token_accuracy": 0.22246391475200653,
      "num_tokens": 52116444.0,
      "step": 22725
    },
    {
      "entropy": 5.004988813400269,
      "epoch": 2.1834774255523537,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004524477260794944,
      "loss": 4.7411,
      "mean_token_accuracy": 0.23672550916671753,
      "num_tokens": 52130264.0,
      "step": 22730
    },
    {
      "entropy": 5.097243165969848,
      "epoch": 2.1839577329490876,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045242664532461094,
      "loss": 4.7738,
      "mean_token_accuracy": 0.2250171884894371,
      "num_tokens": 52142688.0,
      "step": 22735
    },
    {
      "entropy": 5.227225208282471,
      "epoch": 2.1844380403458215,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045240556045040767,
      "loss": 4.8788,
      "mean_token_accuracy": 0.2186577081680298,
      "num_tokens": 52154823.0,
      "step": 22740
    },
    {
      "entropy": 5.112895202636719,
      "epoch": 2.1849183477425553,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045238447145737397,
      "loss": 4.8219,
      "mean_token_accuracy": 0.23602428585290908,
      "num_tokens": 52165542.0,
      "step": 22745
    },
    {
      "entropy": 5.185267162322998,
      "epoch": 2.185398655139289,
      "grad_norm": 1.0,
      "learning_rate": 0.00045236337834599966,
      "loss": 4.8165,
      "mean_token_accuracy": 0.2334161326289177,
      "num_tokens": 52176475.0,
      "step": 22750
    },
    {
      "entropy": 5.073167705535889,
      "epoch": 2.185878962536023,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045234228111677434,
      "loss": 4.7711,
      "mean_token_accuracy": 0.22861799001693725,
      "num_tokens": 52188467.0,
      "step": 22755
    },
    {
      "entropy": 5.077366733551026,
      "epoch": 2.186359269932757,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004523211797701878,
      "loss": 4.7308,
      "mean_token_accuracy": 0.23736548274755478,
      "num_tokens": 52199518.0,
      "step": 22760
    },
    {
      "entropy": 5.05582971572876,
      "epoch": 2.186839577329491,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045230007430673014,
      "loss": 4.767,
      "mean_token_accuracy": 0.23630660325288771,
      "num_tokens": 52211429.0,
      "step": 22765
    },
    {
      "entropy": 5.198104524612427,
      "epoch": 2.1873198847262247,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004522789647268911,
      "loss": 4.8429,
      "mean_token_accuracy": 0.22684505581855774,
      "num_tokens": 52222258.0,
      "step": 22770
    },
    {
      "entropy": 5.053358125686645,
      "epoch": 2.1878001921229586,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004522578510311609,
      "loss": 4.7359,
      "mean_token_accuracy": 0.23054501861333848,
      "num_tokens": 52233660.0,
      "step": 22775
    },
    {
      "entropy": 5.053745412826538,
      "epoch": 2.1882804995196925,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045223673322002984,
      "loss": 4.763,
      "mean_token_accuracy": 0.23531855642795563,
      "num_tokens": 52244721.0,
      "step": 22780
    },
    {
      "entropy": 5.1854479789733885,
      "epoch": 2.1887608069164264,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045221561129398804,
      "loss": 4.8576,
      "mean_token_accuracy": 0.22568602561950685,
      "num_tokens": 52256099.0,
      "step": 22785
    },
    {
      "entropy": 5.12065896987915,
      "epoch": 2.1892411143131603,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000452194485253526,
      "loss": 4.7945,
      "mean_token_accuracy": 0.2370864599943161,
      "num_tokens": 52266943.0,
      "step": 22790
    },
    {
      "entropy": 5.027578735351563,
      "epoch": 2.189721421709894,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004521733550991342,
      "loss": 4.7864,
      "mean_token_accuracy": 0.23362623006105424,
      "num_tokens": 52279074.0,
      "step": 22795
    },
    {
      "entropy": 5.134048843383789,
      "epoch": 2.1902017291066285,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045215222083130316,
      "loss": 4.8125,
      "mean_token_accuracy": 0.22818947285413743,
      "num_tokens": 52290613.0,
      "step": 22800
    },
    {
      "entropy": 5.150738286972046,
      "epoch": 2.1906820365033624,
      "grad_norm": 1.125,
      "learning_rate": 0.0004521310824505236,
      "loss": 4.8143,
      "mean_token_accuracy": 0.2255667641758919,
      "num_tokens": 52301903.0,
      "step": 22805
    },
    {
      "entropy": 5.052944993972778,
      "epoch": 2.1911623439000962,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045210993995728623,
      "loss": 4.7903,
      "mean_token_accuracy": 0.23248525112867355,
      "num_tokens": 52312738.0,
      "step": 22810
    },
    {
      "entropy": 5.033452320098877,
      "epoch": 2.19164265129683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000452088793352082,
      "loss": 4.7435,
      "mean_token_accuracy": 0.23512278348207474,
      "num_tokens": 52325315.0,
      "step": 22815
    },
    {
      "entropy": 5.211318302154541,
      "epoch": 2.192122958693564,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004520676426354018,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22204600125551224,
      "num_tokens": 52338330.0,
      "step": 22820
    },
    {
      "entropy": 5.187323522567749,
      "epoch": 2.192603266090298,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004520464878077368,
      "loss": 4.9222,
      "mean_token_accuracy": 0.21692733615636825,
      "num_tokens": 52349758.0,
      "step": 22825
    },
    {
      "entropy": 5.069988107681274,
      "epoch": 2.1930835734870318,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045202532886957805,
      "loss": 4.7386,
      "mean_token_accuracy": 0.23619951456785201,
      "num_tokens": 52361901.0,
      "step": 22830
    },
    {
      "entropy": 5.115238952636719,
      "epoch": 2.1935638808837656,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045200416582141676,
      "loss": 4.8079,
      "mean_token_accuracy": 0.2353666439652443,
      "num_tokens": 52372912.0,
      "step": 22835
    },
    {
      "entropy": 5.0756243705749515,
      "epoch": 2.1940441882804995,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004519829986637444,
      "loss": 4.797,
      "mean_token_accuracy": 0.22749230414628982,
      "num_tokens": 52384270.0,
      "step": 22840
    },
    {
      "entropy": 5.145582103729248,
      "epoch": 2.1945244956772334,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004519618273970523,
      "loss": 4.8638,
      "mean_token_accuracy": 0.22478014528751372,
      "num_tokens": 52396555.0,
      "step": 22845
    },
    {
      "entropy": 5.126102590560913,
      "epoch": 2.1950048030739673,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045194065202183205,
      "loss": 4.7687,
      "mean_token_accuracy": 0.22973854541778566,
      "num_tokens": 52408183.0,
      "step": 22850
    },
    {
      "entropy": 5.14082670211792,
      "epoch": 2.195485110470701,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004519194725385752,
      "loss": 4.9301,
      "mean_token_accuracy": 0.22707783579826354,
      "num_tokens": 52420285.0,
      "step": 22855
    },
    {
      "entropy": 5.141588163375855,
      "epoch": 2.195965417867435,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045189828894777364,
      "loss": 4.8055,
      "mean_token_accuracy": 0.226527664065361,
      "num_tokens": 52431558.0,
      "step": 22860
    },
    {
      "entropy": 5.202821207046509,
      "epoch": 2.196445725264169,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045187710124991904,
      "loss": 4.891,
      "mean_token_accuracy": 0.22294100224971772,
      "num_tokens": 52443189.0,
      "step": 22865
    },
    {
      "entropy": 5.114332294464111,
      "epoch": 2.196926032660903,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004518559094455034,
      "loss": 4.8133,
      "mean_token_accuracy": 0.23092198222875596,
      "num_tokens": 52454781.0,
      "step": 22870
    },
    {
      "entropy": 5.047515487670898,
      "epoch": 2.1974063400576367,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004518347135350187,
      "loss": 4.7438,
      "mean_token_accuracy": 0.24109113067388535,
      "num_tokens": 52466424.0,
      "step": 22875
    },
    {
      "entropy": 5.1014703750610355,
      "epoch": 2.1978866474543706,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045181351351895703,
      "loss": 4.7203,
      "mean_token_accuracy": 0.23490461111068725,
      "num_tokens": 52477043.0,
      "step": 22880
    },
    {
      "entropy": 5.01074595451355,
      "epoch": 2.198366954851105,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004517923093978106,
      "loss": 4.6898,
      "mean_token_accuracy": 0.2375525489449501,
      "num_tokens": 52488321.0,
      "step": 22885
    },
    {
      "entropy": 5.103656768798828,
      "epoch": 2.1988472622478388,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004517711011720717,
      "loss": 4.7988,
      "mean_token_accuracy": 0.22859315276145936,
      "num_tokens": 52499071.0,
      "step": 22890
    },
    {
      "entropy": 5.112888193130493,
      "epoch": 2.1993275696445727,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004517498888422327,
      "loss": 4.7125,
      "mean_token_accuracy": 0.24221968203783034,
      "num_tokens": 52509846.0,
      "step": 22895
    },
    {
      "entropy": 5.175854682922363,
      "epoch": 2.1998078770413065,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004517286724087862,
      "loss": 4.9208,
      "mean_token_accuracy": 0.21725360602140426,
      "num_tokens": 52521450.0,
      "step": 22900
    },
    {
      "entropy": 5.0794912815094,
      "epoch": 2.2002881844380404,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004517074518722247,
      "loss": 4.7845,
      "mean_token_accuracy": 0.23274537175893784,
      "num_tokens": 52533115.0,
      "step": 22905
    },
    {
      "entropy": 5.030126285552979,
      "epoch": 2.2007684918347743,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045168622723304084,
      "loss": 4.7025,
      "mean_token_accuracy": 0.24132169634103776,
      "num_tokens": 52543921.0,
      "step": 22910
    },
    {
      "entropy": 5.097835826873779,
      "epoch": 2.201248799231508,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004516649984917274,
      "loss": 4.7688,
      "mean_token_accuracy": 0.23708610981702805,
      "num_tokens": 52554956.0,
      "step": 22915
    },
    {
      "entropy": 5.0629435062408445,
      "epoch": 2.201729106628242,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045164376564877734,
      "loss": 4.7772,
      "mean_token_accuracy": 0.2271813690662384,
      "num_tokens": 52567386.0,
      "step": 22920
    },
    {
      "entropy": 5.117129993438721,
      "epoch": 2.202209414024976,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045162252870468354,
      "loss": 4.8159,
      "mean_token_accuracy": 0.23417202085256578,
      "num_tokens": 52578533.0,
      "step": 22925
    },
    {
      "entropy": 5.222954607009887,
      "epoch": 2.20268972142171,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004516012876599391,
      "loss": 4.9414,
      "mean_token_accuracy": 0.21920910477638245,
      "num_tokens": 52590283.0,
      "step": 22930
    },
    {
      "entropy": 5.063795375823974,
      "epoch": 2.2031700288184437,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045158004251503715,
      "loss": 4.7391,
      "mean_token_accuracy": 0.23735689371824265,
      "num_tokens": 52602362.0,
      "step": 22935
    },
    {
      "entropy": 4.971229219436646,
      "epoch": 2.2036503362151776,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045155879327047087,
      "loss": 4.7026,
      "mean_token_accuracy": 0.23837947696447373,
      "num_tokens": 52614057.0,
      "step": 22940
    },
    {
      "entropy": 5.032861566543579,
      "epoch": 2.2041306436119115,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004515375399267338,
      "loss": 4.7661,
      "mean_token_accuracy": 0.22884205132722854,
      "num_tokens": 52624298.0,
      "step": 22945
    },
    {
      "entropy": 5.1293806552886965,
      "epoch": 2.2046109510086453,
      "grad_norm": 1.0,
      "learning_rate": 0.00045151628248431925,
      "loss": 4.8148,
      "mean_token_accuracy": 0.2307080715894699,
      "num_tokens": 52637043.0,
      "step": 22950
    },
    {
      "entropy": 5.188603496551513,
      "epoch": 2.2050912584053792,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045149502094372077,
      "loss": 4.8855,
      "mean_token_accuracy": 0.2264219269156456,
      "num_tokens": 52649326.0,
      "step": 22955
    },
    {
      "entropy": 5.132386541366577,
      "epoch": 2.2055715658021136,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045147375530543195,
      "loss": 4.7854,
      "mean_token_accuracy": 0.2312495529651642,
      "num_tokens": 52660973.0,
      "step": 22960
    },
    {
      "entropy": 5.058101749420166,
      "epoch": 2.2060518731988474,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045145248556994653,
      "loss": 4.722,
      "mean_token_accuracy": 0.23750037848949432,
      "num_tokens": 52671943.0,
      "step": 22965
    },
    {
      "entropy": 5.080811214447022,
      "epoch": 2.2065321805955813,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004514312117377584,
      "loss": 4.816,
      "mean_token_accuracy": 0.22809068411588668,
      "num_tokens": 52684634.0,
      "step": 22970
    },
    {
      "entropy": 5.074667024612427,
      "epoch": 2.207012487992315,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004514099338093613,
      "loss": 4.7811,
      "mean_token_accuracy": 0.23760328739881514,
      "num_tokens": 52695681.0,
      "step": 22975
    },
    {
      "entropy": 5.054156684875489,
      "epoch": 2.207492795389049,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004513886517852496,
      "loss": 4.7218,
      "mean_token_accuracy": 0.24085892885923385,
      "num_tokens": 52706967.0,
      "step": 22980
    },
    {
      "entropy": 5.143637561798096,
      "epoch": 2.207973102785783,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000451367365665917,
      "loss": 4.7982,
      "mean_token_accuracy": 0.22624808102846145,
      "num_tokens": 52719226.0,
      "step": 22985
    },
    {
      "entropy": 4.97782530784607,
      "epoch": 2.208453410182517,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045134607545185785,
      "loss": 4.6548,
      "mean_token_accuracy": 0.24088680744171143,
      "num_tokens": 52730141.0,
      "step": 22990
    },
    {
      "entropy": 5.107984972000122,
      "epoch": 2.2089337175792507,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004513247811435666,
      "loss": 4.8594,
      "mean_token_accuracy": 0.2248896077275276,
      "num_tokens": 52741059.0,
      "step": 22995
    },
    {
      "entropy": 5.146727514266968,
      "epoch": 2.2094140249759846,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045130348274153735,
      "loss": 4.8387,
      "mean_token_accuracy": 0.2260206162929535,
      "num_tokens": 52752598.0,
      "step": 23000
    },
    {
      "entropy": 4.998793983459473,
      "epoch": 2.2098943323727185,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045128218024626486,
      "loss": 4.7221,
      "mean_token_accuracy": 0.23424574881792068,
      "num_tokens": 52764620.0,
      "step": 23005
    },
    {
      "entropy": 5.073661470413208,
      "epoch": 2.2103746397694524,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004512608736582436,
      "loss": 4.7226,
      "mean_token_accuracy": 0.235737182199955,
      "num_tokens": 52774756.0,
      "step": 23010
    },
    {
      "entropy": 5.220461416244507,
      "epoch": 2.2108549471661862,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004512395629779682,
      "loss": 4.8805,
      "mean_token_accuracy": 0.2228524148464203,
      "num_tokens": 52786742.0,
      "step": 23015
    },
    {
      "entropy": 5.186991024017334,
      "epoch": 2.21133525456292,
      "grad_norm": 1.125,
      "learning_rate": 0.0004512182482059335,
      "loss": 4.8364,
      "mean_token_accuracy": 0.22229242771863938,
      "num_tokens": 52798303.0,
      "step": 23020
    },
    {
      "entropy": 5.055900907516479,
      "epoch": 2.211815561959654,
      "grad_norm": 1.125,
      "learning_rate": 0.0004511969293426343,
      "loss": 4.7849,
      "mean_token_accuracy": 0.2319771021604538,
      "num_tokens": 52808930.0,
      "step": 23025
    },
    {
      "entropy": 5.035862398147583,
      "epoch": 2.212295869356388,
      "grad_norm": 1.125,
      "learning_rate": 0.00045117560638856567,
      "loss": 4.747,
      "mean_token_accuracy": 0.22902755439281464,
      "num_tokens": 52820066.0,
      "step": 23030
    },
    {
      "entropy": 5.130739736557007,
      "epoch": 2.212776176753122,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004511542793442225,
      "loss": 4.7913,
      "mean_token_accuracy": 0.2316420778632164,
      "num_tokens": 52830839.0,
      "step": 23035
    },
    {
      "entropy": 5.128929901123047,
      "epoch": 2.213256484149856,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004511329482101001,
      "loss": 4.7937,
      "mean_token_accuracy": 0.22558759152889252,
      "num_tokens": 52843112.0,
      "step": 23040
    },
    {
      "entropy": 5.116424512863159,
      "epoch": 2.21373679154659,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004511116129866936,
      "loss": 4.879,
      "mean_token_accuracy": 0.22888087779283522,
      "num_tokens": 52854737.0,
      "step": 23045
    },
    {
      "entropy": 5.220840406417847,
      "epoch": 2.214217098943324,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045109027367449845,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22465359270572663,
      "num_tokens": 52866876.0,
      "step": 23050
    },
    {
      "entropy": 5.166212320327759,
      "epoch": 2.2146974063400577,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045106893027400995,
      "loss": 4.8219,
      "mean_token_accuracy": 0.22468414455652236,
      "num_tokens": 52878731.0,
      "step": 23055
    },
    {
      "entropy": 5.169036626815796,
      "epoch": 2.2151777137367916,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045104758278572375,
      "loss": 4.8222,
      "mean_token_accuracy": 0.22637639194726944,
      "num_tokens": 52891011.0,
      "step": 23060
    },
    {
      "entropy": 5.031224584579467,
      "epoch": 2.2156580211335255,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004510262312101355,
      "loss": 4.7625,
      "mean_token_accuracy": 0.23082911670207978,
      "num_tokens": 52902334.0,
      "step": 23065
    },
    {
      "entropy": 5.2142016887664795,
      "epoch": 2.2161383285302594,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004510048755477407,
      "loss": 4.8947,
      "mean_token_accuracy": 0.2183098629117012,
      "num_tokens": 52913839.0,
      "step": 23070
    },
    {
      "entropy": 5.168850946426391,
      "epoch": 2.2166186359269933,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004509835157990354,
      "loss": 4.8161,
      "mean_token_accuracy": 0.22825908809900283,
      "num_tokens": 52925588.0,
      "step": 23075
    },
    {
      "entropy": 5.229594802856445,
      "epoch": 2.217098943323727,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00045096215196451547,
      "loss": 4.9438,
      "mean_token_accuracy": 0.2200305461883545,
      "num_tokens": 52938474.0,
      "step": 23080
    },
    {
      "entropy": 5.022704410552978,
      "epoch": 2.217579250720461,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045094078404467683,
      "loss": 4.7193,
      "mean_token_accuracy": 0.23483040779829026,
      "num_tokens": 52950157.0,
      "step": 23085
    },
    {
      "entropy": 5.093522310256958,
      "epoch": 2.218059558117195,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045091941204001564,
      "loss": 4.8468,
      "mean_token_accuracy": 0.2259957730770111,
      "num_tokens": 52960334.0,
      "step": 23090
    },
    {
      "entropy": 5.139597654342651,
      "epoch": 2.218539865513929,
      "grad_norm": 1.046875,
      "learning_rate": 0.000450898035951028,
      "loss": 4.8861,
      "mean_token_accuracy": 0.2294859379529953,
      "num_tokens": 52972251.0,
      "step": 23095
    },
    {
      "entropy": 5.145685434341431,
      "epoch": 2.2190201729106627,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045087665577821034,
      "loss": 4.763,
      "mean_token_accuracy": 0.23695262223482133,
      "num_tokens": 52983897.0,
      "step": 23100
    },
    {
      "entropy": 5.176052904129028,
      "epoch": 2.2195004803073966,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000450855271522059,
      "loss": 4.8595,
      "mean_token_accuracy": 0.21844857484102248,
      "num_tokens": 52995821.0,
      "step": 23105
    },
    {
      "entropy": 5.098303699493409,
      "epoch": 2.219980787704131,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045083388318307044,
      "loss": 4.801,
      "mean_token_accuracy": 0.22858060747385026,
      "num_tokens": 53007890.0,
      "step": 23110
    },
    {
      "entropy": 5.099214839935303,
      "epoch": 2.2204610951008648,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004508124907617411,
      "loss": 4.7962,
      "mean_token_accuracy": 0.22958508431911467,
      "num_tokens": 53019390.0,
      "step": 23115
    },
    {
      "entropy": 5.116458892822266,
      "epoch": 2.2209414024975986,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004507910942585679,
      "loss": 4.7903,
      "mean_token_accuracy": 0.2310373529791832,
      "num_tokens": 53031318.0,
      "step": 23120
    },
    {
      "entropy": 5.086399698257447,
      "epoch": 2.2214217098943325,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004507696936740475,
      "loss": 4.6746,
      "mean_token_accuracy": 0.24217380881309508,
      "num_tokens": 53043121.0,
      "step": 23125
    },
    {
      "entropy": 5.049490356445313,
      "epoch": 2.2219020172910664,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004507482890086767,
      "loss": 4.7959,
      "mean_token_accuracy": 0.229643514752388,
      "num_tokens": 53055600.0,
      "step": 23130
    },
    {
      "entropy": 5.028835916519165,
      "epoch": 2.2223823246878003,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004507268802629525,
      "loss": 4.6989,
      "mean_token_accuracy": 0.24149151295423507,
      "num_tokens": 53067075.0,
      "step": 23135
    },
    {
      "entropy": 5.002907419204712,
      "epoch": 2.222862632084534,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004507054674373719,
      "loss": 4.7684,
      "mean_token_accuracy": 0.23617046773433686,
      "num_tokens": 53078374.0,
      "step": 23140
    },
    {
      "entropy": 5.094421100616455,
      "epoch": 2.223342939481268,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045068405053243216,
      "loss": 4.7611,
      "mean_token_accuracy": 0.2335207685828209,
      "num_tokens": 53089177.0,
      "step": 23145
    },
    {
      "entropy": 4.9939733982086185,
      "epoch": 2.223823246878002,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004506626295486304,
      "loss": 4.6337,
      "mean_token_accuracy": 0.24132752269506455,
      "num_tokens": 53099957.0,
      "step": 23150
    },
    {
      "entropy": 5.119339799880981,
      "epoch": 2.224303554274736,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045064120448646405,
      "loss": 4.8129,
      "mean_token_accuracy": 0.22773957699537278,
      "num_tokens": 53113369.0,
      "step": 23155
    },
    {
      "entropy": 5.110342264175415,
      "epoch": 2.2247838616714697,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004506197753464304,
      "loss": 4.7515,
      "mean_token_accuracy": 0.2369021952152252,
      "num_tokens": 53124647.0,
      "step": 23160
    },
    {
      "entropy": 5.099501848220825,
      "epoch": 2.2252641690682036,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045059834212902707,
      "loss": 4.8286,
      "mean_token_accuracy": 0.22955333292484284,
      "num_tokens": 53135218.0,
      "step": 23165
    },
    {
      "entropy": 5.0327675342559814,
      "epoch": 2.2257444764649374,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00045057690483475167,
      "loss": 4.7814,
      "mean_token_accuracy": 0.2319254606962204,
      "num_tokens": 53148077.0,
      "step": 23170
    },
    {
      "entropy": 5.1574948787689205,
      "epoch": 2.2262247838616713,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004505554634641019,
      "loss": 4.8341,
      "mean_token_accuracy": 0.22938669919967652,
      "num_tokens": 53159866.0,
      "step": 23175
    },
    {
      "entropy": 5.098292684555053,
      "epoch": 2.226705091258405,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045053401801757554,
      "loss": 4.774,
      "mean_token_accuracy": 0.2299387827515602,
      "num_tokens": 53171151.0,
      "step": 23180
    },
    {
      "entropy": 5.162169742584228,
      "epoch": 2.227185398655139,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045051256849567054,
      "loss": 4.8068,
      "mean_token_accuracy": 0.22838256508111954,
      "num_tokens": 53182400.0,
      "step": 23185
    },
    {
      "entropy": 5.1758131980896,
      "epoch": 2.227665706051873,
      "grad_norm": 1.0,
      "learning_rate": 0.00045049111489888486,
      "loss": 4.9087,
      "mean_token_accuracy": 0.22053113877773284,
      "num_tokens": 53194248.0,
      "step": 23190
    },
    {
      "entropy": 5.121346282958984,
      "epoch": 2.2281460134486073,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004504696572277165,
      "loss": 4.8092,
      "mean_token_accuracy": 0.23302749395370484,
      "num_tokens": 53205484.0,
      "step": 23195
    },
    {
      "entropy": 5.093762302398682,
      "epoch": 2.228626320845341,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045044819548266385,
      "loss": 4.7702,
      "mean_token_accuracy": 0.23627068996429443,
      "num_tokens": 53217070.0,
      "step": 23200
    },
    {
      "entropy": 5.036428689956665,
      "epoch": 2.229106628242075,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045042672966422506,
      "loss": 4.7415,
      "mean_token_accuracy": 0.2350686311721802,
      "num_tokens": 53228431.0,
      "step": 23205
    },
    {
      "entropy": 5.162332201004029,
      "epoch": 2.229586935638809,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00045040525977289847,
      "loss": 4.8874,
      "mean_token_accuracy": 0.22714349031448364,
      "num_tokens": 53240718.0,
      "step": 23210
    },
    {
      "entropy": 5.2310747623443605,
      "epoch": 2.230067243035543,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004503837858091826,
      "loss": 4.8437,
      "mean_token_accuracy": 0.22913905531167983,
      "num_tokens": 53251542.0,
      "step": 23215
    },
    {
      "entropy": 5.126812505722046,
      "epoch": 2.2305475504322767,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045036230777357604,
      "loss": 4.7678,
      "mean_token_accuracy": 0.23932116329669953,
      "num_tokens": 53261915.0,
      "step": 23220
    },
    {
      "entropy": 5.087382841110229,
      "epoch": 2.2310278578290106,
      "grad_norm": 1.0,
      "learning_rate": 0.0004503408256665774,
      "loss": 4.8261,
      "mean_token_accuracy": 0.23212223947048188,
      "num_tokens": 53273287.0,
      "step": 23225
    },
    {
      "entropy": 5.121749830245972,
      "epoch": 2.2315081652257445,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045031933948868545,
      "loss": 4.8667,
      "mean_token_accuracy": 0.22219200134277345,
      "num_tokens": 53284482.0,
      "step": 23230
    },
    {
      "entropy": 5.101685905456543,
      "epoch": 2.2319884726224783,
      "grad_norm": 0.96875,
      "learning_rate": 0.00045029784924039903,
      "loss": 4.735,
      "mean_token_accuracy": 0.23122657239437103,
      "num_tokens": 53296235.0,
      "step": 23235
    },
    {
      "entropy": 5.208931922912598,
      "epoch": 2.2324687800192122,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00045027635492221716,
      "loss": 4.8903,
      "mean_token_accuracy": 0.22234119176864625,
      "num_tokens": 53306784.0,
      "step": 23240
    },
    {
      "entropy": 5.066908836364746,
      "epoch": 2.232949087415946,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045025485653463866,
      "loss": 4.7926,
      "mean_token_accuracy": 0.2368413805961609,
      "num_tokens": 53318303.0,
      "step": 23245
    },
    {
      "entropy": 5.133825874328613,
      "epoch": 2.23342939481268,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004502333540781628,
      "loss": 4.7665,
      "mean_token_accuracy": 0.2348608300089836,
      "num_tokens": 53330196.0,
      "step": 23250
    },
    {
      "entropy": 5.127715635299682,
      "epoch": 2.233909702209414,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004502118475532888,
      "loss": 4.7862,
      "mean_token_accuracy": 0.23026148825883866,
      "num_tokens": 53342250.0,
      "step": 23255
    },
    {
      "entropy": 5.00491189956665,
      "epoch": 2.2343900096061478,
      "grad_norm": 1.078125,
      "learning_rate": 0.000450190336960516,
      "loss": 4.7336,
      "mean_token_accuracy": 0.22646015882492065,
      "num_tokens": 53353050.0,
      "step": 23260
    },
    {
      "entropy": 5.059219408035278,
      "epoch": 2.2348703170028816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004501688223003438,
      "loss": 4.7281,
      "mean_token_accuracy": 0.2415449783205986,
      "num_tokens": 53364702.0,
      "step": 23265
    },
    {
      "entropy": 5.079537868499756,
      "epoch": 2.235350624399616,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004501473035732717,
      "loss": 4.7602,
      "mean_token_accuracy": 0.22756219357252122,
      "num_tokens": 53375790.0,
      "step": 23270
    },
    {
      "entropy": 5.128358221054077,
      "epoch": 2.23583093179635,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004501257807797993,
      "loss": 4.7982,
      "mean_token_accuracy": 0.22635925710201263,
      "num_tokens": 53387206.0,
      "step": 23275
    },
    {
      "entropy": 5.080106401443482,
      "epoch": 2.2363112391930837,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045010425392042624,
      "loss": 4.8076,
      "mean_token_accuracy": 0.22965409904718398,
      "num_tokens": 53399519.0,
      "step": 23280
    },
    {
      "entropy": 5.119130277633667,
      "epoch": 2.2367915465898176,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004500827229956524,
      "loss": 4.7493,
      "mean_token_accuracy": 0.23283777236938477,
      "num_tokens": 53411187.0,
      "step": 23285
    },
    {
      "entropy": 5.162726736068725,
      "epoch": 2.2372718539865515,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045006118800597757,
      "loss": 4.843,
      "mean_token_accuracy": 0.22757963091135025,
      "num_tokens": 53421328.0,
      "step": 23290
    },
    {
      "entropy": 5.085423135757447,
      "epoch": 2.2377521613832854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045003964895190177,
      "loss": 4.847,
      "mean_token_accuracy": 0.22638614177703859,
      "num_tokens": 53433155.0,
      "step": 23295
    },
    {
      "entropy": 5.122412776947021,
      "epoch": 2.2382324687800192,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004500181058339251,
      "loss": 4.8034,
      "mean_token_accuracy": 0.2230328142642975,
      "num_tokens": 53445498.0,
      "step": 23300
    },
    {
      "entropy": 5.120661640167237,
      "epoch": 2.238712776176753,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004499965586525478,
      "loss": 4.7672,
      "mean_token_accuracy": 0.23427130579948424,
      "num_tokens": 53458005.0,
      "step": 23305
    },
    {
      "entropy": 5.0678764343261715,
      "epoch": 2.239193083573487,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044997500740826993,
      "loss": 4.8472,
      "mean_token_accuracy": 0.2360531583428383,
      "num_tokens": 53470052.0,
      "step": 23310
    },
    {
      "entropy": 5.053494787216186,
      "epoch": 2.239673390970221,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000449953452101592,
      "loss": 4.795,
      "mean_token_accuracy": 0.23531701415777206,
      "num_tokens": 53480665.0,
      "step": 23315
    },
    {
      "entropy": 5.1003560543060305,
      "epoch": 2.2401536983669548,
      "grad_norm": 1.0,
      "learning_rate": 0.00044993189273301445,
      "loss": 4.7554,
      "mean_token_accuracy": 0.22907655835151672,
      "num_tokens": 53491903.0,
      "step": 23320
    },
    {
      "entropy": 5.069397449493408,
      "epoch": 2.2406340057636887,
      "grad_norm": 1.125,
      "learning_rate": 0.0004499103293030377,
      "loss": 4.7856,
      "mean_token_accuracy": 0.2328623726963997,
      "num_tokens": 53502806.0,
      "step": 23325
    },
    {
      "entropy": 5.08409366607666,
      "epoch": 2.2411143131604225,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004498887618121625,
      "loss": 4.773,
      "mean_token_accuracy": 0.23178012520074845,
      "num_tokens": 53513738.0,
      "step": 23330
    },
    {
      "entropy": 5.065647745132447,
      "epoch": 2.2415946205571564,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004498671902608897,
      "loss": 4.7561,
      "mean_token_accuracy": 0.2316155642271042,
      "num_tokens": 53525546.0,
      "step": 23335
    },
    {
      "entropy": 5.093524217605591,
      "epoch": 2.2420749279538903,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004498456146497198,
      "loss": 4.7598,
      "mean_token_accuracy": 0.23062681555747985,
      "num_tokens": 53537162.0,
      "step": 23340
    },
    {
      "entropy": 5.090532970428467,
      "epoch": 2.2425552353506246,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044982403497915405,
      "loss": 4.8371,
      "mean_token_accuracy": 0.2295659750699997,
      "num_tokens": 53549165.0,
      "step": 23345
    },
    {
      "entropy": 5.124613332748413,
      "epoch": 2.2430355427473585,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044980245124969333,
      "loss": 4.7956,
      "mean_token_accuracy": 0.23347476571798326,
      "num_tokens": 53561641.0,
      "step": 23350
    },
    {
      "entropy": 5.077760457992554,
      "epoch": 2.2435158501440924,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004497808634618386,
      "loss": 4.7582,
      "mean_token_accuracy": 0.24206406325101854,
      "num_tokens": 53572350.0,
      "step": 23355
    },
    {
      "entropy": 5.10000319480896,
      "epoch": 2.2439961575408263,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004497592716160913,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22264520078897476,
      "num_tokens": 53584191.0,
      "step": 23360
    },
    {
      "entropy": 5.193553733825683,
      "epoch": 2.24447646493756,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044973767571295273,
      "loss": 4.848,
      "mean_token_accuracy": 0.22456169873476028,
      "num_tokens": 53595021.0,
      "step": 23365
    },
    {
      "entropy": 5.198762607574463,
      "epoch": 2.244956772334294,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004497160757529241,
      "loss": 4.839,
      "mean_token_accuracy": 0.23482302576303482,
      "num_tokens": 53606992.0,
      "step": 23370
    },
    {
      "entropy": 5.072748804092408,
      "epoch": 2.245437079731028,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044969447173650695,
      "loss": 4.7388,
      "mean_token_accuracy": 0.24042182117700578,
      "num_tokens": 53616548.0,
      "step": 23375
    },
    {
      "entropy": 5.0779595375061035,
      "epoch": 2.245917387127762,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000449672863664203,
      "loss": 4.7667,
      "mean_token_accuracy": 0.23548574596643448,
      "num_tokens": 53629078.0,
      "step": 23380
    },
    {
      "entropy": 5.082695341110229,
      "epoch": 2.2463976945244957,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044965125153651375,
      "loss": 4.7015,
      "mean_token_accuracy": 0.24380909502506257,
      "num_tokens": 53640059.0,
      "step": 23385
    },
    {
      "entropy": 5.119830846786499,
      "epoch": 2.2468780019212296,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004496296353539411,
      "loss": 4.907,
      "mean_token_accuracy": 0.22078739255666732,
      "num_tokens": 53651806.0,
      "step": 23390
    },
    {
      "entropy": 5.150289726257324,
      "epoch": 2.2473583093179634,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004496080151169869,
      "loss": 4.8732,
      "mean_token_accuracy": 0.22324226945638656,
      "num_tokens": 53662180.0,
      "step": 23395
    },
    {
      "entropy": 5.14017014503479,
      "epoch": 2.2478386167146973,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044958639082615294,
      "loss": 4.8781,
      "mean_token_accuracy": 0.22525733709335327,
      "num_tokens": 53674520.0,
      "step": 23400
    },
    {
      "entropy": 5.247305679321289,
      "epoch": 2.248318924111431,
      "grad_norm": 1.125,
      "learning_rate": 0.0004495647624819415,
      "loss": 4.9322,
      "mean_token_accuracy": 0.2178337574005127,
      "num_tokens": 53686051.0,
      "step": 23405
    },
    {
      "entropy": 5.1200151443481445,
      "epoch": 2.248799231508165,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044954313008485457,
      "loss": 4.6867,
      "mean_token_accuracy": 0.24275511950254441,
      "num_tokens": 53697171.0,
      "step": 23410
    },
    {
      "entropy": 5.126438045501709,
      "epoch": 2.249279538904899,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044952149363539453,
      "loss": 4.8015,
      "mean_token_accuracy": 0.23062219768762587,
      "num_tokens": 53707776.0,
      "step": 23415
    },
    {
      "entropy": 5.140814208984375,
      "epoch": 2.249759846301633,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004494998531340635,
      "loss": 4.8412,
      "mean_token_accuracy": 0.2232404589653015,
      "num_tokens": 53719430.0,
      "step": 23420
    },
    {
      "entropy": 5.048863315582276,
      "epoch": 2.2502401536983667,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004494782085813641,
      "loss": 4.6884,
      "mean_token_accuracy": 0.24226518124341964,
      "num_tokens": 53730708.0,
      "step": 23425
    },
    {
      "entropy": 5.077973699569702,
      "epoch": 2.250720461095101,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004494565599777988,
      "loss": 4.7936,
      "mean_token_accuracy": 0.22824919372797012,
      "num_tokens": 53741623.0,
      "step": 23430
    },
    {
      "entropy": 5.093221998214721,
      "epoch": 2.251200768491835,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044943490732387025,
      "loss": 4.7263,
      "mean_token_accuracy": 0.2359860673546791,
      "num_tokens": 53752348.0,
      "step": 23435
    },
    {
      "entropy": 5.097160863876343,
      "epoch": 2.251681075888569,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004494132506200811,
      "loss": 4.7496,
      "mean_token_accuracy": 0.23401957005262375,
      "num_tokens": 53763973.0,
      "step": 23440
    },
    {
      "entropy": 4.999422025680542,
      "epoch": 2.2521613832853027,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004493915898669341,
      "loss": 4.6354,
      "mean_token_accuracy": 0.2394458457827568,
      "num_tokens": 53775139.0,
      "step": 23445
    },
    {
      "entropy": 5.075583124160767,
      "epoch": 2.2526416906820366,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004493699250649323,
      "loss": 4.8434,
      "mean_token_accuracy": 0.23190349340438843,
      "num_tokens": 53788324.0,
      "step": 23450
    },
    {
      "entropy": 5.129440879821777,
      "epoch": 2.2531219980787704,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004493482562145786,
      "loss": 4.7334,
      "mean_token_accuracy": 0.23334760814905167,
      "num_tokens": 53800019.0,
      "step": 23455
    },
    {
      "entropy": 5.137174606323242,
      "epoch": 2.2536023054755043,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044932658331637605,
      "loss": 4.8535,
      "mean_token_accuracy": 0.22261528968811034,
      "num_tokens": 53812155.0,
      "step": 23460
    },
    {
      "entropy": 5.058337640762329,
      "epoch": 2.254082612872238,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004493049063708279,
      "loss": 4.7759,
      "mean_token_accuracy": 0.2384394034743309,
      "num_tokens": 53823770.0,
      "step": 23465
    },
    {
      "entropy": 5.203836679458618,
      "epoch": 2.254562920268972,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044928322537843746,
      "loss": 4.9612,
      "mean_token_accuracy": 0.2205788418650627,
      "num_tokens": 53836037.0,
      "step": 23470
    },
    {
      "entropy": 5.098265600204468,
      "epoch": 2.255043227665706,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044926154033970793,
      "loss": 4.6896,
      "mean_token_accuracy": 0.23346782773733138,
      "num_tokens": 53848517.0,
      "step": 23475
    },
    {
      "entropy": 5.093183469772339,
      "epoch": 2.25552353506244,
      "grad_norm": 1.109375,
      "learning_rate": 0.000449239851255143,
      "loss": 4.7459,
      "mean_token_accuracy": 0.24112479388713837,
      "num_tokens": 53860143.0,
      "step": 23480
    },
    {
      "entropy": 5.178523349761963,
      "epoch": 2.2560038424591737,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044921815812524606,
      "loss": 4.8499,
      "mean_token_accuracy": 0.22999966889619827,
      "num_tokens": 53871173.0,
      "step": 23485
    },
    {
      "entropy": 5.091372728347778,
      "epoch": 2.2564841498559076,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044919646095052077,
      "loss": 4.7738,
      "mean_token_accuracy": 0.23395557701587677,
      "num_tokens": 53882791.0,
      "step": 23490
    },
    {
      "entropy": 5.099833393096924,
      "epoch": 2.2569644572526415,
      "grad_norm": 1.109375,
      "learning_rate": 0.000449174759731471,
      "loss": 4.8118,
      "mean_token_accuracy": 0.23298768252134322,
      "num_tokens": 53894785.0,
      "step": 23495
    },
    {
      "entropy": 5.14121036529541,
      "epoch": 2.2574447646493754,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044915305446860046,
      "loss": 4.8025,
      "mean_token_accuracy": 0.2299412190914154,
      "num_tokens": 53905989.0,
      "step": 23500
    },
    {
      "entropy": 5.044490909576416,
      "epoch": 2.2579250720461097,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044913134516241305,
      "loss": 4.7092,
      "mean_token_accuracy": 0.23634413331747056,
      "num_tokens": 53918233.0,
      "step": 23505
    },
    {
      "entropy": 5.05875973701477,
      "epoch": 2.2584053794428436,
      "grad_norm": 1.125,
      "learning_rate": 0.0004491096318134129,
      "loss": 4.7998,
      "mean_token_accuracy": 0.2311341106891632,
      "num_tokens": 53929886.0,
      "step": 23510
    },
    {
      "entropy": 5.20423436164856,
      "epoch": 2.2588856868395775,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004490879144221042,
      "loss": 4.9491,
      "mean_token_accuracy": 0.21492367386817932,
      "num_tokens": 53941788.0,
      "step": 23515
    },
    {
      "entropy": 5.135285997390747,
      "epoch": 2.2593659942363113,
      "grad_norm": 1.0,
      "learning_rate": 0.00044906619298899097,
      "loss": 4.7497,
      "mean_token_accuracy": 0.2349345088005066,
      "num_tokens": 53952753.0,
      "step": 23520
    },
    {
      "entropy": 5.049922132492066,
      "epoch": 2.2598463016330452,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004490444675145777,
      "loss": 4.7854,
      "mean_token_accuracy": 0.23332255631685256,
      "num_tokens": 53964435.0,
      "step": 23525
    },
    {
      "entropy": 5.134873867034912,
      "epoch": 2.260326609029779,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004490227379993686,
      "loss": 4.8334,
      "mean_token_accuracy": 0.23176417797803878,
      "num_tokens": 53976906.0,
      "step": 23530
    },
    {
      "entropy": 5.125746440887451,
      "epoch": 2.260806916426513,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004490010044438683,
      "loss": 4.7615,
      "mean_token_accuracy": 0.23967888206243515,
      "num_tokens": 53988076.0,
      "step": 23535
    },
    {
      "entropy": 4.983516693115234,
      "epoch": 2.261287223823247,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044897926684858133,
      "loss": 4.6705,
      "mean_token_accuracy": 0.2390793889760971,
      "num_tokens": 53998896.0,
      "step": 23540
    },
    {
      "entropy": 5.018841123580932,
      "epoch": 2.2617675312199808,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044895752521401246,
      "loss": 4.7663,
      "mean_token_accuracy": 0.23176145255565644,
      "num_tokens": 54010855.0,
      "step": 23545
    },
    {
      "entropy": 5.13550124168396,
      "epoch": 2.2622478386167146,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004489357795406663,
      "loss": 4.801,
      "mean_token_accuracy": 0.23310249596834182,
      "num_tokens": 54021630.0,
      "step": 23550
    },
    {
      "entropy": 5.072286319732666,
      "epoch": 2.2627281460134485,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004489140298290479,
      "loss": 4.7335,
      "mean_token_accuracy": 0.23013039380311967,
      "num_tokens": 54033715.0,
      "step": 23555
    },
    {
      "entropy": 5.028193426132202,
      "epoch": 2.2632084534101824,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044889227607966217,
      "loss": 4.81,
      "mean_token_accuracy": 0.22755009084939956,
      "num_tokens": 54046633.0,
      "step": 23560
    },
    {
      "entropy": 5.121305847167969,
      "epoch": 2.2636887608069163,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044887051829301406,
      "loss": 4.7926,
      "mean_token_accuracy": 0.2326028436422348,
      "num_tokens": 54058389.0,
      "step": 23565
    },
    {
      "entropy": 5.1233641624450685,
      "epoch": 2.26416906820365,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044884875646960886,
      "loss": 4.8624,
      "mean_token_accuracy": 0.22554973661899566,
      "num_tokens": 54070338.0,
      "step": 23570
    },
    {
      "entropy": 5.067175817489624,
      "epoch": 2.264649375600384,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044882699060995175,
      "loss": 4.7398,
      "mean_token_accuracy": 0.23018296360969542,
      "num_tokens": 54082511.0,
      "step": 23575
    },
    {
      "entropy": 5.071782445907592,
      "epoch": 2.2651296829971184,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004488052207145481,
      "loss": 4.7382,
      "mean_token_accuracy": 0.23689772486686705,
      "num_tokens": 54094141.0,
      "step": 23580
    },
    {
      "entropy": 5.1482759475708,
      "epoch": 2.2656099903938522,
      "grad_norm": 1.375,
      "learning_rate": 0.00044878344678390324,
      "loss": 4.8669,
      "mean_token_accuracy": 0.2273583933711052,
      "num_tokens": 54104837.0,
      "step": 23585
    },
    {
      "entropy": 5.209051465988159,
      "epoch": 2.266090297790586,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044876166881852286,
      "loss": 4.847,
      "mean_token_accuracy": 0.22356017231941222,
      "num_tokens": 54116924.0,
      "step": 23590
    },
    {
      "entropy": 5.120183658599854,
      "epoch": 2.26657060518732,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004487398868189125,
      "loss": 4.772,
      "mean_token_accuracy": 0.23515297919511796,
      "num_tokens": 54127187.0,
      "step": 23595
    },
    {
      "entropy": 5.0530791759490965,
      "epoch": 2.267050912584054,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044871810078557777,
      "loss": 4.7756,
      "mean_token_accuracy": 0.23454860150814055,
      "num_tokens": 54139024.0,
      "step": 23600
    },
    {
      "entropy": 5.102918815612793,
      "epoch": 2.2675312199807878,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004486963107190247,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22554150819778443,
      "num_tokens": 54150643.0,
      "step": 23605
    },
    {
      "entropy": 5.1482460498809814,
      "epoch": 2.2680115273775217,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044867451661975894,
      "loss": 4.7758,
      "mean_token_accuracy": 0.23260141611099244,
      "num_tokens": 54162171.0,
      "step": 23610
    },
    {
      "entropy": 5.10244574546814,
      "epoch": 2.2684918347742555,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044865271848828673,
      "loss": 4.7821,
      "mean_token_accuracy": 0.2298636943101883,
      "num_tokens": 54173104.0,
      "step": 23615
    },
    {
      "entropy": 5.069700860977173,
      "epoch": 2.2689721421709894,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000448630916325114,
      "loss": 4.7631,
      "mean_token_accuracy": 0.23645809143781663,
      "num_tokens": 54183956.0,
      "step": 23620
    },
    {
      "entropy": 5.140821409225464,
      "epoch": 2.2694524495677233,
      "grad_norm": 1.0625,
      "learning_rate": 0.000448609110130747,
      "loss": 4.8338,
      "mean_token_accuracy": 0.23555280715227128,
      "num_tokens": 54195087.0,
      "step": 23625
    },
    {
      "entropy": 5.083088970184326,
      "epoch": 2.269932756964457,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044858729990569193,
      "loss": 4.735,
      "mean_token_accuracy": 0.23396464586257934,
      "num_tokens": 54206684.0,
      "step": 23630
    },
    {
      "entropy": 5.041690158843994,
      "epoch": 2.270413064361191,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044856548565045523,
      "loss": 4.7491,
      "mean_token_accuracy": 0.23365791141986847,
      "num_tokens": 54218155.0,
      "step": 23635
    },
    {
      "entropy": 5.08332462310791,
      "epoch": 2.270893371757925,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044854366736554323,
      "loss": 4.8267,
      "mean_token_accuracy": 0.22983042001724244,
      "num_tokens": 54229124.0,
      "step": 23640
    },
    {
      "entropy": 5.074865198135376,
      "epoch": 2.271373679154659,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044852184505146274,
      "loss": 4.7626,
      "mean_token_accuracy": 0.23736238479614258,
      "num_tokens": 54240889.0,
      "step": 23645
    },
    {
      "entropy": 5.084311008453369,
      "epoch": 2.2718539865513927,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004485000187087202,
      "loss": 4.7305,
      "mean_token_accuracy": 0.23949097394943236,
      "num_tokens": 54252388.0,
      "step": 23650
    },
    {
      "entropy": 5.09113073348999,
      "epoch": 2.272334293948127,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004484781883378224,
      "loss": 4.7686,
      "mean_token_accuracy": 0.23424773961305617,
      "num_tokens": 54264702.0,
      "step": 23655
    },
    {
      "entropy": 5.098014211654663,
      "epoch": 2.2728146013448605,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044845635393927623,
      "loss": 4.8439,
      "mean_token_accuracy": 0.22995774894952775,
      "num_tokens": 54277064.0,
      "step": 23660
    },
    {
      "entropy": 5.058008480072021,
      "epoch": 2.273294908741595,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004484345155135886,
      "loss": 4.7282,
      "mean_token_accuracy": 0.24007131457328795,
      "num_tokens": 54287958.0,
      "step": 23665
    },
    {
      "entropy": 5.068008184432983,
      "epoch": 2.2737752161383287,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004484126730612664,
      "loss": 4.7604,
      "mean_token_accuracy": 0.2369404539465904,
      "num_tokens": 54299428.0,
      "step": 23670
    },
    {
      "entropy": 5.150881052017212,
      "epoch": 2.2742555235350626,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000448390826582817,
      "loss": 4.8343,
      "mean_token_accuracy": 0.23052153140306472,
      "num_tokens": 54312249.0,
      "step": 23675
    },
    {
      "entropy": 5.209943580627441,
      "epoch": 2.2747358309317964,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044836897607874744,
      "loss": 4.8855,
      "mean_token_accuracy": 0.2259225904941559,
      "num_tokens": 54324408.0,
      "step": 23680
    },
    {
      "entropy": 5.146414136886596,
      "epoch": 2.2752161383285303,
      "grad_norm": 1.015625,
      "learning_rate": 0.000448347121549565,
      "loss": 4.8226,
      "mean_token_accuracy": 0.23238304108381272,
      "num_tokens": 54336092.0,
      "step": 23685
    },
    {
      "entropy": 5.028502130508423,
      "epoch": 2.275696445725264,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004483252629957771,
      "loss": 4.7043,
      "mean_token_accuracy": 0.24290446192026138,
      "num_tokens": 54347231.0,
      "step": 23690
    },
    {
      "entropy": 5.082117938995362,
      "epoch": 2.276176753121998,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044830340041789133,
      "loss": 4.7639,
      "mean_token_accuracy": 0.23573375046253203,
      "num_tokens": 54359096.0,
      "step": 23695
    },
    {
      "entropy": 5.064883136749268,
      "epoch": 2.276657060518732,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004482815338164152,
      "loss": 4.683,
      "mean_token_accuracy": 0.2430448964238167,
      "num_tokens": 54370433.0,
      "step": 23700
    },
    {
      "entropy": 5.170016717910767,
      "epoch": 2.277137367915466,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004482596631918564,
      "loss": 4.8201,
      "mean_token_accuracy": 0.2244497910141945,
      "num_tokens": 54381430.0,
      "step": 23705
    },
    {
      "entropy": 5.095471620559692,
      "epoch": 2.2776176753121997,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044823778854472267,
      "loss": 4.7863,
      "mean_token_accuracy": 0.2219826892018318,
      "num_tokens": 54392840.0,
      "step": 23710
    },
    {
      "entropy": 5.1966516971588135,
      "epoch": 2.2780979827089336,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004482159098755219,
      "loss": 4.9034,
      "mean_token_accuracy": 0.22104487419128419,
      "num_tokens": 54405776.0,
      "step": 23715
    },
    {
      "entropy": 5.07906494140625,
      "epoch": 2.2785782901056675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004481940271847621,
      "loss": 4.7357,
      "mean_token_accuracy": 0.23345702588558198,
      "num_tokens": 54417392.0,
      "step": 23720
    },
    {
      "entropy": 5.144957399368286,
      "epoch": 2.2790585975024014,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004481721404729513,
      "loss": 4.8974,
      "mean_token_accuracy": 0.22101654410362243,
      "num_tokens": 54428407.0,
      "step": 23725
    },
    {
      "entropy": 5.184346866607666,
      "epoch": 2.2795389048991357,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004481502497405975,
      "loss": 4.795,
      "mean_token_accuracy": 0.2342268005013466,
      "num_tokens": 54439667.0,
      "step": 23730
    },
    {
      "entropy": 5.078746032714844,
      "epoch": 2.280019212295869,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004481283549882091,
      "loss": 4.7445,
      "mean_token_accuracy": 0.2313960686326027,
      "num_tokens": 54451966.0,
      "step": 23735
    },
    {
      "entropy": 5.189362049102783,
      "epoch": 2.2804995196926034,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044810645621629443,
      "loss": 4.9105,
      "mean_token_accuracy": 0.22013920098543166,
      "num_tokens": 54462896.0,
      "step": 23740
    },
    {
      "entropy": 5.166595029830932,
      "epoch": 2.2809798270893373,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044808455342536176,
      "loss": 4.8287,
      "mean_token_accuracy": 0.23614353984594344,
      "num_tokens": 54474056.0,
      "step": 23745
    },
    {
      "entropy": 5.140611600875855,
      "epoch": 2.281460134486071,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044806264661591976,
      "loss": 4.8053,
      "mean_token_accuracy": 0.2257085010409355,
      "num_tokens": 54485846.0,
      "step": 23750
    },
    {
      "entropy": 5.0847326755523685,
      "epoch": 2.281940441882805,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004480407357884771,
      "loss": 4.7254,
      "mean_token_accuracy": 0.2342715159058571,
      "num_tokens": 54496020.0,
      "step": 23755
    },
    {
      "entropy": 5.187861394882202,
      "epoch": 2.282420749279539,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044801882094354226,
      "loss": 4.8918,
      "mean_token_accuracy": 0.2177417740225792,
      "num_tokens": 54507388.0,
      "step": 23760
    },
    {
      "entropy": 5.144430255889892,
      "epoch": 2.282901056676273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004479969020816242,
      "loss": 4.8023,
      "mean_token_accuracy": 0.23507119864225387,
      "num_tokens": 54519159.0,
      "step": 23765
    },
    {
      "entropy": 5.178164958953857,
      "epoch": 2.2833813640730067,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044797497920323175,
      "loss": 4.9633,
      "mean_token_accuracy": 0.22422045469284058,
      "num_tokens": 54530258.0,
      "step": 23770
    },
    {
      "entropy": 5.182456684112549,
      "epoch": 2.2838616714697406,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004479530523088739,
      "loss": 4.9191,
      "mean_token_accuracy": 0.2203189730644226,
      "num_tokens": 54542626.0,
      "step": 23775
    },
    {
      "entropy": 5.146267366409302,
      "epoch": 2.2843419788664745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004479311213990598,
      "loss": 4.7922,
      "mean_token_accuracy": 0.23174499869346618,
      "num_tokens": 54553762.0,
      "step": 23780
    },
    {
      "entropy": 5.117496299743652,
      "epoch": 2.2848222862632084,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044790918647429854,
      "loss": 4.7303,
      "mean_token_accuracy": 0.23345881700515747,
      "num_tokens": 54565484.0,
      "step": 23785
    },
    {
      "entropy": 5.066101884841919,
      "epoch": 2.2853025936599423,
      "grad_norm": 1.0,
      "learning_rate": 0.00044788724753509935,
      "loss": 4.7615,
      "mean_token_accuracy": 0.23681660592556,
      "num_tokens": 54577321.0,
      "step": 23790
    },
    {
      "entropy": 5.018876600265503,
      "epoch": 2.285782901056676,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004478653045819717,
      "loss": 4.7326,
      "mean_token_accuracy": 0.24045118242502211,
      "num_tokens": 54587796.0,
      "step": 23795
    },
    {
      "entropy": 5.170510292053223,
      "epoch": 2.28626320845341,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004478433576154249,
      "loss": 4.8523,
      "mean_token_accuracy": 0.2238484725356102,
      "num_tokens": 54600637.0,
      "step": 23800
    },
    {
      "entropy": 5.167998456954956,
      "epoch": 2.286743515850144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004478214066359687,
      "loss": 4.8291,
      "mean_token_accuracy": 0.23148442059755325,
      "num_tokens": 54610690.0,
      "step": 23805
    },
    {
      "entropy": 5.114355564117432,
      "epoch": 2.287223823246878,
      "grad_norm": 1.125,
      "learning_rate": 0.00044779945164411254,
      "loss": 4.8322,
      "mean_token_accuracy": 0.2297376960515976,
      "num_tokens": 54622417.0,
      "step": 23810
    },
    {
      "entropy": 5.108760070800781,
      "epoch": 2.287704130643612,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004477774926403662,
      "loss": 4.7535,
      "mean_token_accuracy": 0.23886296898126602,
      "num_tokens": 54633190.0,
      "step": 23815
    },
    {
      "entropy": 5.1435870170593265,
      "epoch": 2.288184438040346,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004477555296252396,
      "loss": 4.7547,
      "mean_token_accuracy": 0.23206374943256378,
      "num_tokens": 54643507.0,
      "step": 23820
    },
    {
      "entropy": 4.985277462005615,
      "epoch": 2.28866474543708,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044773356259924255,
      "loss": 4.6994,
      "mean_token_accuracy": 0.23732422292232513,
      "num_tokens": 54653835.0,
      "step": 23825
    },
    {
      "entropy": 5.113461065292358,
      "epoch": 2.2891450528338138,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044771159156288505,
      "loss": 4.7947,
      "mean_token_accuracy": 0.23351410031318665,
      "num_tokens": 54664908.0,
      "step": 23830
    },
    {
      "entropy": 5.066265344619751,
      "epoch": 2.2896253602305476,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004476896165166772,
      "loss": 4.779,
      "mean_token_accuracy": 0.23599461913108827,
      "num_tokens": 54677714.0,
      "step": 23835
    },
    {
      "entropy": 5.0937357425689695,
      "epoch": 2.2901056676272815,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044766763746112936,
      "loss": 4.8139,
      "mean_token_accuracy": 0.22718480229377747,
      "num_tokens": 54688919.0,
      "step": 23840
    },
    {
      "entropy": 5.111744832992554,
      "epoch": 2.2905859750240154,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004476456543967517,
      "loss": 4.804,
      "mean_token_accuracy": 0.2341846615076065,
      "num_tokens": 54700089.0,
      "step": 23845
    },
    {
      "entropy": 5.1170495510101315,
      "epoch": 2.2910662824207493,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044762366732405454,
      "loss": 4.7453,
      "mean_token_accuracy": 0.2395706906914711,
      "num_tokens": 54711233.0,
      "step": 23850
    },
    {
      "entropy": 5.131341409683228,
      "epoch": 2.291546589817483,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004476016762435484,
      "loss": 4.8805,
      "mean_token_accuracy": 0.22512992918491365,
      "num_tokens": 54723140.0,
      "step": 23855
    },
    {
      "entropy": 5.149951314926147,
      "epoch": 2.292026897214217,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004475796811557439,
      "loss": 4.8023,
      "mean_token_accuracy": 0.23573793321847916,
      "num_tokens": 54734196.0,
      "step": 23860
    },
    {
      "entropy": 5.1301428318023685,
      "epoch": 2.292507204610951,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044755768206115155,
      "loss": 4.8724,
      "mean_token_accuracy": 0.21495721191167833,
      "num_tokens": 54747395.0,
      "step": 23865
    },
    {
      "entropy": 5.162541103363037,
      "epoch": 2.292987512007685,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004475356789602823,
      "loss": 4.8432,
      "mean_token_accuracy": 0.22741710245609284,
      "num_tokens": 54761137.0,
      "step": 23870
    },
    {
      "entropy": 5.1252992153167725,
      "epoch": 2.2934678194044187,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044751367185364696,
      "loss": 4.7368,
      "mean_token_accuracy": 0.23414889425039292,
      "num_tokens": 54772181.0,
      "step": 23875
    },
    {
      "entropy": 5.065300750732422,
      "epoch": 2.2939481268011526,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044749166074175634,
      "loss": 4.7756,
      "mean_token_accuracy": 0.23413179069757462,
      "num_tokens": 54783759.0,
      "step": 23880
    },
    {
      "entropy": 5.067009830474854,
      "epoch": 2.2944284341978864,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044746964562512154,
      "loss": 4.6988,
      "mean_token_accuracy": 0.23515940308570862,
      "num_tokens": 54796265.0,
      "step": 23885
    },
    {
      "entropy": 5.0589280128479,
      "epoch": 2.2949087415946208,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044744762650425376,
      "loss": 4.7884,
      "mean_token_accuracy": 0.23344510197639465,
      "num_tokens": 54808159.0,
      "step": 23890
    },
    {
      "entropy": 5.168301582336426,
      "epoch": 2.2953890489913547,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044742560337966415,
      "loss": 4.8777,
      "mean_token_accuracy": 0.23123983293771744,
      "num_tokens": 54819070.0,
      "step": 23895
    },
    {
      "entropy": 5.052987957000733,
      "epoch": 2.2958693563880885,
      "grad_norm": 1.15625,
      "learning_rate": 0.000447403576251864,
      "loss": 4.7658,
      "mean_token_accuracy": 0.2241657167673111,
      "num_tokens": 54829210.0,
      "step": 23900
    },
    {
      "entropy": 5.043314409255982,
      "epoch": 2.2963496637848224,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004473815451213648,
      "loss": 4.7475,
      "mean_token_accuracy": 0.23271748870611192,
      "num_tokens": 54841238.0,
      "step": 23905
    },
    {
      "entropy": 5.126592111587525,
      "epoch": 2.2968299711815563,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004473595099886779,
      "loss": 4.8145,
      "mean_token_accuracy": 0.23608639538288118,
      "num_tokens": 54852531.0,
      "step": 23910
    },
    {
      "entropy": 5.13033881187439,
      "epoch": 2.29731027857829,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004473374708543151,
      "loss": 4.7827,
      "mean_token_accuracy": 0.23330678641796113,
      "num_tokens": 54864089.0,
      "step": 23915
    },
    {
      "entropy": 5.083416271209717,
      "epoch": 2.297790585975024,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004473154277187878,
      "loss": 4.748,
      "mean_token_accuracy": 0.2305147111415863,
      "num_tokens": 54874956.0,
      "step": 23920
    },
    {
      "entropy": 5.0938615798950195,
      "epoch": 2.298270893371758,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044729338058260805,
      "loss": 4.7896,
      "mean_token_accuracy": 0.23220585733652116,
      "num_tokens": 54885608.0,
      "step": 23925
    },
    {
      "entropy": 5.093524932861328,
      "epoch": 2.298751200768492,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004472713294462876,
      "loss": 4.7282,
      "mean_token_accuracy": 0.23458857089281082,
      "num_tokens": 54896782.0,
      "step": 23930
    },
    {
      "entropy": 5.115837240219117,
      "epoch": 2.2992315081652257,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044724927431033843,
      "loss": 4.8146,
      "mean_token_accuracy": 0.2266061559319496,
      "num_tokens": 54908179.0,
      "step": 23935
    },
    {
      "entropy": 5.131691360473633,
      "epoch": 2.2997118155619596,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004472272151752725,
      "loss": 4.8047,
      "mean_token_accuracy": 0.2324026569724083,
      "num_tokens": 54921114.0,
      "step": 23940
    },
    {
      "entropy": 5.133564090728759,
      "epoch": 2.3001921229586935,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004472051520416022,
      "loss": 4.7736,
      "mean_token_accuracy": 0.2303609609603882,
      "num_tokens": 54931398.0,
      "step": 23945
    },
    {
      "entropy": 5.152296495437622,
      "epoch": 2.3006724303554273,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004471830849098395,
      "loss": 4.842,
      "mean_token_accuracy": 0.22772619873285294,
      "num_tokens": 54942354.0,
      "step": 23950
    },
    {
      "entropy": 5.081222152709961,
      "epoch": 2.3011527377521612,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044716101378049683,
      "loss": 4.7591,
      "mean_token_accuracy": 0.2396457850933075,
      "num_tokens": 54953700.0,
      "step": 23955
    },
    {
      "entropy": 4.967060565948486,
      "epoch": 2.301633045148895,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044713893865408667,
      "loss": 4.7054,
      "mean_token_accuracy": 0.23864023983478547,
      "num_tokens": 54965544.0,
      "step": 23960
    },
    {
      "entropy": 5.07809624671936,
      "epoch": 2.3021133525456294,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004471168595311215,
      "loss": 4.7412,
      "mean_token_accuracy": 0.2382942870259285,
      "num_tokens": 54976811.0,
      "step": 23965
    },
    {
      "entropy": 5.0738249778747555,
      "epoch": 2.302593659942363,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044709477641211395,
      "loss": 4.7335,
      "mean_token_accuracy": 0.24010078608989716,
      "num_tokens": 54987508.0,
      "step": 23970
    },
    {
      "entropy": 5.187147521972657,
      "epoch": 2.303073967339097,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004470726892975767,
      "loss": 4.9559,
      "mean_token_accuracy": 0.21804401129484177,
      "num_tokens": 55000276.0,
      "step": 23975
    },
    {
      "entropy": 5.112174463272095,
      "epoch": 2.303554274735831,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044705059818802255,
      "loss": 4.776,
      "mean_token_accuracy": 0.23761829882860183,
      "num_tokens": 55012397.0,
      "step": 23980
    },
    {
      "entropy": 5.0756970882415775,
      "epoch": 2.304034582132565,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004470285030839644,
      "loss": 4.7383,
      "mean_token_accuracy": 0.23044277727603912,
      "num_tokens": 55024358.0,
      "step": 23985
    },
    {
      "entropy": 4.970513105392456,
      "epoch": 2.304514889529299,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044700640398591526,
      "loss": 4.7784,
      "mean_token_accuracy": 0.23175662159919738,
      "num_tokens": 55035887.0,
      "step": 23990
    },
    {
      "entropy": 5.150806427001953,
      "epoch": 2.3049951969260327,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004469843008943881,
      "loss": 4.8553,
      "mean_token_accuracy": 0.2252778336405754,
      "num_tokens": 55048775.0,
      "step": 23995
    },
    {
      "entropy": 5.194459104537964,
      "epoch": 2.3054755043227666,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004469621938098962,
      "loss": 4.8122,
      "mean_token_accuracy": 0.23306870609521865,
      "num_tokens": 55059994.0,
      "step": 24000
    },
    {
      "epoch": 2.3054755043227666,
      "eval_entropy": 4.967041379820714,
      "eval_loss": 4.919524669647217,
      "eval_mean_token_accuracy": 0.2330568053385925,
      "eval_num_tokens": 55059994.0,
      "eval_runtime": 26.6089,
      "eval_samples_per_second": 1233.232,
      "eval_steps_per_second": 154.159,
      "step": 24000
    },
    {
      "entropy": 5.110465574264526,
      "epoch": 2.3059558117195005,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004469400827329528,
      "loss": 4.83,
      "mean_token_accuracy": 0.22905360609292985,
      "num_tokens": 55071950.0,
      "step": 24005
    },
    {
      "entropy": 4.9932629585266115,
      "epoch": 2.3064361191162344,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004469179676640712,
      "loss": 4.6853,
      "mean_token_accuracy": 0.24508388787508012,
      "num_tokens": 55085260.0,
      "step": 24010
    },
    {
      "entropy": 5.12148380279541,
      "epoch": 2.3069164265129682,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004468958486037649,
      "loss": 4.7001,
      "mean_token_accuracy": 0.22994801104068757,
      "num_tokens": 55095794.0,
      "step": 24015
    },
    {
      "entropy": 5.105727100372315,
      "epoch": 2.307396733909702,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004468737255525474,
      "loss": 4.7812,
      "mean_token_accuracy": 0.2325472801923752,
      "num_tokens": 55106476.0,
      "step": 24020
    },
    {
      "entropy": 5.104316663742066,
      "epoch": 2.307877041306436,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004468515985109324,
      "loss": 4.8296,
      "mean_token_accuracy": 0.23250334858894348,
      "num_tokens": 55118277.0,
      "step": 24025
    },
    {
      "entropy": 5.10371356010437,
      "epoch": 2.30835734870317,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004468294674794335,
      "loss": 4.7946,
      "mean_token_accuracy": 0.2293125197291374,
      "num_tokens": 55129902.0,
      "step": 24030
    },
    {
      "entropy": 5.040475130081177,
      "epoch": 2.3088376560999038,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004468073324585647,
      "loss": 4.7398,
      "mean_token_accuracy": 0.22803521454334258,
      "num_tokens": 55143342.0,
      "step": 24035
    },
    {
      "entropy": 5.146928834915161,
      "epoch": 2.309317963496638,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004467851934488398,
      "loss": 4.8156,
      "mean_token_accuracy": 0.23593219518661498,
      "num_tokens": 55154116.0,
      "step": 24040
    },
    {
      "entropy": 5.130366611480713,
      "epoch": 2.3097982708933715,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004467630504507727,
      "loss": 4.7838,
      "mean_token_accuracy": 0.23294282406568528,
      "num_tokens": 55165548.0,
      "step": 24045
    },
    {
      "entropy": 5.137736463546753,
      "epoch": 2.310278578290106,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004467409034648776,
      "loss": 4.8423,
      "mean_token_accuracy": 0.22782959789037704,
      "num_tokens": 55176047.0,
      "step": 24050
    },
    {
      "entropy": 5.072274827957154,
      "epoch": 2.3107588856868397,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004467187524916688,
      "loss": 4.7862,
      "mean_token_accuracy": 0.231942056119442,
      "num_tokens": 55188715.0,
      "step": 24055
    },
    {
      "entropy": 4.995149564743042,
      "epoch": 2.3112391930835736,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004466965975316604,
      "loss": 4.6738,
      "mean_token_accuracy": 0.2425564184784889,
      "num_tokens": 55198988.0,
      "step": 24060
    },
    {
      "entropy": 4.973041296005249,
      "epoch": 2.3117195004803075,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044667443858536685,
      "loss": 4.6585,
      "mean_token_accuracy": 0.24218691885471344,
      "num_tokens": 55210000.0,
      "step": 24065
    },
    {
      "entropy": 5.112592983245849,
      "epoch": 2.3121998078770414,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004466522756533026,
      "loss": 4.7967,
      "mean_token_accuracy": 0.22654343843460084,
      "num_tokens": 55221794.0,
      "step": 24070
    },
    {
      "entropy": 5.148695468902588,
      "epoch": 2.3126801152737753,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004466301087359822,
      "loss": 4.845,
      "mean_token_accuracy": 0.22544953525066375,
      "num_tokens": 55233166.0,
      "step": 24075
    },
    {
      "entropy": 5.057459306716919,
      "epoch": 2.313160422670509,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044660793783392035,
      "loss": 4.7273,
      "mean_token_accuracy": 0.23736212253570557,
      "num_tokens": 55244179.0,
      "step": 24080
    },
    {
      "entropy": 5.044998645782471,
      "epoch": 2.313640730067243,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004465857629476317,
      "loss": 4.7435,
      "mean_token_accuracy": 0.23495712876319885,
      "num_tokens": 55255556.0,
      "step": 24085
    },
    {
      "entropy": 5.214394664764404,
      "epoch": 2.314121037463977,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004465635840776312,
      "loss": 4.8918,
      "mean_token_accuracy": 0.22667918652296065,
      "num_tokens": 55267001.0,
      "step": 24090
    },
    {
      "entropy": 5.204922914505005,
      "epoch": 2.314601344860711,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044654140122443373,
      "loss": 4.8224,
      "mean_token_accuracy": 0.231950643658638,
      "num_tokens": 55278253.0,
      "step": 24095
    },
    {
      "entropy": 5.086462736129761,
      "epoch": 2.3150816522574447,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004465192143885542,
      "loss": 4.7839,
      "mean_token_accuracy": 0.23172909617424012,
      "num_tokens": 55289601.0,
      "step": 24100
    },
    {
      "entropy": 5.051853370666504,
      "epoch": 2.3155619596541785,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044649702357050787,
      "loss": 4.7841,
      "mean_token_accuracy": 0.2329975113272667,
      "num_tokens": 55300661.0,
      "step": 24105
    },
    {
      "entropy": 5.052013397216797,
      "epoch": 2.3160422670509124,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004464748287708099,
      "loss": 4.7612,
      "mean_token_accuracy": 0.23609595000743866,
      "num_tokens": 55311890.0,
      "step": 24110
    },
    {
      "entropy": 5.023010396957398,
      "epoch": 2.3165225744476463,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044645262998997557,
      "loss": 4.7117,
      "mean_token_accuracy": 0.24315921515226363,
      "num_tokens": 55322241.0,
      "step": 24115
    },
    {
      "entropy": 5.094870615005493,
      "epoch": 2.31700288184438,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044643042722852024,
      "loss": 4.6749,
      "mean_token_accuracy": 0.24713147431612015,
      "num_tokens": 55334064.0,
      "step": 24120
    },
    {
      "entropy": 5.073682975769043,
      "epoch": 2.3174831892411145,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004464082204869595,
      "loss": 4.8142,
      "mean_token_accuracy": 0.2292729988694191,
      "num_tokens": 55346266.0,
      "step": 24125
    },
    {
      "entropy": 5.032088136672973,
      "epoch": 2.3179634966378484,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004463860097658088,
      "loss": 4.7359,
      "mean_token_accuracy": 0.2398496761918068,
      "num_tokens": 55357395.0,
      "step": 24130
    },
    {
      "entropy": 5.0814313888549805,
      "epoch": 2.3184438040345823,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004463637950655839,
      "loss": 4.7318,
      "mean_token_accuracy": 0.23123213201761245,
      "num_tokens": 55367982.0,
      "step": 24135
    },
    {
      "entropy": 5.091415643692017,
      "epoch": 2.318924111431316,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044634157638680054,
      "loss": 4.8127,
      "mean_token_accuracy": 0.2341153532266617,
      "num_tokens": 55379822.0,
      "step": 24140
    },
    {
      "entropy": 5.111624240875244,
      "epoch": 2.31940441882805,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044631935372997455,
      "loss": 4.7744,
      "mean_token_accuracy": 0.2243826448917389,
      "num_tokens": 55391271.0,
      "step": 24145
    },
    {
      "entropy": 5.079347038269043,
      "epoch": 2.319884726224784,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004462971270956219,
      "loss": 4.7062,
      "mean_token_accuracy": 0.2315313309431076,
      "num_tokens": 55401514.0,
      "step": 24150
    },
    {
      "entropy": 5.109730434417725,
      "epoch": 2.320365033621518,
      "grad_norm": 1.125,
      "learning_rate": 0.0004462748964842586,
      "loss": 4.8104,
      "mean_token_accuracy": 0.23129841983318328,
      "num_tokens": 55412979.0,
      "step": 24155
    },
    {
      "entropy": 5.065424060821533,
      "epoch": 2.3208453410182517,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004462526618964008,
      "loss": 4.7386,
      "mean_token_accuracy": 0.23363394290208817,
      "num_tokens": 55423985.0,
      "step": 24160
    },
    {
      "entropy": 5.094624328613281,
      "epoch": 2.3213256484149856,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004462304233325647,
      "loss": 4.7721,
      "mean_token_accuracy": 0.22987855821847916,
      "num_tokens": 55434638.0,
      "step": 24165
    },
    {
      "entropy": 5.119213485717774,
      "epoch": 2.3218059558117194,
      "grad_norm": 1.0,
      "learning_rate": 0.0004462081807932666,
      "loss": 4.8175,
      "mean_token_accuracy": 0.22581660747528076,
      "num_tokens": 55445402.0,
      "step": 24170
    },
    {
      "entropy": 5.139634704589843,
      "epoch": 2.3222862632084533,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004461859342790229,
      "loss": 4.7422,
      "mean_token_accuracy": 0.23479513078927994,
      "num_tokens": 55456707.0,
      "step": 24175
    },
    {
      "entropy": 5.054404354095459,
      "epoch": 2.322766570605187,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004461636837903501,
      "loss": 4.7255,
      "mean_token_accuracy": 0.23432399034500123,
      "num_tokens": 55468473.0,
      "step": 24180
    },
    {
      "entropy": 4.977806520462036,
      "epoch": 2.323246878001921,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004461414293277649,
      "loss": 4.6302,
      "mean_token_accuracy": 0.24134107083082199,
      "num_tokens": 55479623.0,
      "step": 24185
    },
    {
      "entropy": 5.132324647903443,
      "epoch": 2.323727185398655,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004461191708917839,
      "loss": 4.8362,
      "mean_token_accuracy": 0.23134353756904602,
      "num_tokens": 55490036.0,
      "step": 24190
    },
    {
      "entropy": 5.0639444351196286,
      "epoch": 2.324207492795389,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044609690848292376,
      "loss": 4.7767,
      "mean_token_accuracy": 0.23550268709659578,
      "num_tokens": 55501605.0,
      "step": 24195
    },
    {
      "entropy": 5.084042739868164,
      "epoch": 2.324687800192123,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004460746421017016,
      "loss": 4.6949,
      "mean_token_accuracy": 0.2416678711771965,
      "num_tokens": 55512029.0,
      "step": 24200
    },
    {
      "entropy": 5.1158007144927975,
      "epoch": 2.325168107588857,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044605237174863405,
      "loss": 4.7984,
      "mean_token_accuracy": 0.23438495993614197,
      "num_tokens": 55524260.0,
      "step": 24205
    },
    {
      "entropy": 5.029112339019775,
      "epoch": 2.325648414985591,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004460300974242386,
      "loss": 4.7849,
      "mean_token_accuracy": 0.23255055099725724,
      "num_tokens": 55536520.0,
      "step": 24210
    },
    {
      "entropy": 5.176107120513916,
      "epoch": 2.326128722382325,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004460078191290319,
      "loss": 4.8401,
      "mean_token_accuracy": 0.22214179188013078,
      "num_tokens": 55547896.0,
      "step": 24215
    },
    {
      "entropy": 5.163871955871582,
      "epoch": 2.3266090297790587,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044598553686353153,
      "loss": 4.7646,
      "mean_token_accuracy": 0.2297682896256447,
      "num_tokens": 55559229.0,
      "step": 24220
    },
    {
      "entropy": 5.1030967235565186,
      "epoch": 2.3270893371757926,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044596325062825476,
      "loss": 4.7846,
      "mean_token_accuracy": 0.23466531932353973,
      "num_tokens": 55569531.0,
      "step": 24225
    },
    {
      "entropy": 5.054196405410766,
      "epoch": 2.3275696445725265,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004459409604237189,
      "loss": 4.7343,
      "mean_token_accuracy": 0.23412058502435684,
      "num_tokens": 55580454.0,
      "step": 24230
    },
    {
      "entropy": 5.062302732467652,
      "epoch": 2.3280499519692603,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044591866625044154,
      "loss": 4.7522,
      "mean_token_accuracy": 0.24423030763864517,
      "num_tokens": 55591336.0,
      "step": 24235
    },
    {
      "entropy": 5.151030397415161,
      "epoch": 2.3285302593659942,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004458963681089403,
      "loss": 4.818,
      "mean_token_accuracy": 0.23043319582939148,
      "num_tokens": 55601955.0,
      "step": 24240
    },
    {
      "entropy": 5.177875661849976,
      "epoch": 2.329010566762728,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004458740659997328,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22355221956968307,
      "num_tokens": 55614130.0,
      "step": 24245
    },
    {
      "entropy": 5.109447050094604,
      "epoch": 2.329490874159462,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004458517599233369,
      "loss": 4.7966,
      "mean_token_accuracy": 0.23202638179063798,
      "num_tokens": 55625447.0,
      "step": 24250
    },
    {
      "entropy": 5.0980446338653564,
      "epoch": 2.329971181556196,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004458294498802706,
      "loss": 4.7612,
      "mean_token_accuracy": 0.23456941694021224,
      "num_tokens": 55636426.0,
      "step": 24255
    },
    {
      "entropy": 5.003779411315918,
      "epoch": 2.3304514889529298,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004458071358710516,
      "loss": 4.7177,
      "mean_token_accuracy": 0.2293532133102417,
      "num_tokens": 55647095.0,
      "step": 24260
    },
    {
      "entropy": 5.00334734916687,
      "epoch": 2.3309317963496636,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004457848178961981,
      "loss": 4.7024,
      "mean_token_accuracy": 0.23526566177606584,
      "num_tokens": 55660362.0,
      "step": 24265
    },
    {
      "entropy": 5.088176631927491,
      "epoch": 2.3314121037463975,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044576249595622833,
      "loss": 4.7266,
      "mean_token_accuracy": 0.23541588485240936,
      "num_tokens": 55671760.0,
      "step": 24270
    },
    {
      "entropy": 5.054433393478393,
      "epoch": 2.331892411143132,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004457401700516603,
      "loss": 4.7773,
      "mean_token_accuracy": 0.22757106572389602,
      "num_tokens": 55683826.0,
      "step": 24275
    },
    {
      "entropy": 5.147644805908203,
      "epoch": 2.3323727185398653,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044571784018301267,
      "loss": 4.7813,
      "mean_token_accuracy": 0.23552829772233963,
      "num_tokens": 55694939.0,
      "step": 24280
    },
    {
      "entropy": 5.092920541763306,
      "epoch": 2.3328530259365996,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044569550635080365,
      "loss": 4.7842,
      "mean_token_accuracy": 0.23183344304561615,
      "num_tokens": 55707101.0,
      "step": 24285
    },
    {
      "entropy": 5.052872800827027,
      "epoch": 2.3333333333333335,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044567316855555184,
      "loss": 4.6999,
      "mean_token_accuracy": 0.23395263403654099,
      "num_tokens": 55718418.0,
      "step": 24290
    },
    {
      "entropy": 5.125878429412841,
      "epoch": 2.3338136407300674,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004456508267977759,
      "loss": 4.9315,
      "mean_token_accuracy": 0.21754334568977357,
      "num_tokens": 55729229.0,
      "step": 24295
    },
    {
      "entropy": 5.159682989120483,
      "epoch": 2.3342939481268012,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044562848107799444,
      "loss": 4.8071,
      "mean_token_accuracy": 0.22653307020664215,
      "num_tokens": 55740719.0,
      "step": 24300
    },
    {
      "entropy": 5.091968202590943,
      "epoch": 2.334774255523535,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00044560613139672627,
      "loss": 4.6953,
      "mean_token_accuracy": 0.23848137259483337,
      "num_tokens": 55752363.0,
      "step": 24305
    },
    {
      "entropy": 5.0234363079071045,
      "epoch": 2.335254562920269,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044558377775449036,
      "loss": 4.7057,
      "mean_token_accuracy": 0.23967838287353516,
      "num_tokens": 55763851.0,
      "step": 24310
    },
    {
      "entropy": 5.069035482406616,
      "epoch": 2.335734870317003,
      "grad_norm": 1.0,
      "learning_rate": 0.00044556142015180573,
      "loss": 4.756,
      "mean_token_accuracy": 0.2410675033926964,
      "num_tokens": 55776118.0,
      "step": 24315
    },
    {
      "entropy": 4.985824871063232,
      "epoch": 2.3362151777137368,
      "grad_norm": 1.0,
      "learning_rate": 0.00044553905858919134,
      "loss": 4.6798,
      "mean_token_accuracy": 0.2417192742228508,
      "num_tokens": 55788106.0,
      "step": 24320
    },
    {
      "entropy": 5.103778219223022,
      "epoch": 2.3366954851104706,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004455166930671664,
      "loss": 4.7768,
      "mean_token_accuracy": 0.22946872860193251,
      "num_tokens": 55798608.0,
      "step": 24325
    },
    {
      "entropy": 5.192025279998779,
      "epoch": 2.3371757925072045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044549432358625014,
      "loss": 4.8065,
      "mean_token_accuracy": 0.2321943536400795,
      "num_tokens": 55810963.0,
      "step": 24330
    },
    {
      "entropy": 5.046109247207641,
      "epoch": 2.3376560999039384,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000445471950146962,
      "loss": 4.6758,
      "mean_token_accuracy": 0.23509302586317063,
      "num_tokens": 55821593.0,
      "step": 24335
    },
    {
      "entropy": 5.127198696136475,
      "epoch": 2.3381364073006723,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004454495727498214,
      "loss": 4.8641,
      "mean_token_accuracy": 0.2251267686486244,
      "num_tokens": 55833602.0,
      "step": 24340
    },
    {
      "entropy": 5.235980987548828,
      "epoch": 2.338616714697406,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004454271913953478,
      "loss": 4.8838,
      "mean_token_accuracy": 0.22301825433969497,
      "num_tokens": 55844745.0,
      "step": 24345
    },
    {
      "entropy": 5.204687786102295,
      "epoch": 2.3390970220941405,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044540480608406093,
      "loss": 4.872,
      "mean_token_accuracy": 0.22576149851083754,
      "num_tokens": 55857193.0,
      "step": 24350
    },
    {
      "entropy": 5.048415470123291,
      "epoch": 2.339577329490874,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004453824168164804,
      "loss": 4.7647,
      "mean_token_accuracy": 0.23150048702955245,
      "num_tokens": 55868954.0,
      "step": 24355
    },
    {
      "entropy": 5.058977222442627,
      "epoch": 2.3400576368876083,
      "grad_norm": 1.25,
      "learning_rate": 0.0004453600235931261,
      "loss": 4.7596,
      "mean_token_accuracy": 0.22982732504606246,
      "num_tokens": 55880485.0,
      "step": 24360
    },
    {
      "entropy": 5.182627391815186,
      "epoch": 2.340537944284342,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000445337626414518,
      "loss": 4.826,
      "mean_token_accuracy": 0.22760045081377028,
      "num_tokens": 55893043.0,
      "step": 24365
    },
    {
      "entropy": 5.072005701065064,
      "epoch": 2.341018251681076,
      "grad_norm": 1.125,
      "learning_rate": 0.00044531522528117593,
      "loss": 4.6813,
      "mean_token_accuracy": 0.2374671146273613,
      "num_tokens": 55905776.0,
      "step": 24370
    },
    {
      "entropy": 5.070609378814697,
      "epoch": 2.34149855907781,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044529282019362007,
      "loss": 4.7834,
      "mean_token_accuracy": 0.23412936329841613,
      "num_tokens": 55916424.0,
      "step": 24375
    },
    {
      "entropy": 5.188404750823975,
      "epoch": 2.341978866474544,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044527041115237056,
      "loss": 4.9269,
      "mean_token_accuracy": 0.22076896727085113,
      "num_tokens": 55928510.0,
      "step": 24380
    },
    {
      "entropy": 5.21982889175415,
      "epoch": 2.3424591738712777,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004452479981579477,
      "loss": 4.8799,
      "mean_token_accuracy": 0.22602816224098204,
      "num_tokens": 55940536.0,
      "step": 24385
    },
    {
      "entropy": 5.130209970474243,
      "epoch": 2.3429394812680115,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004452255812108719,
      "loss": 4.8134,
      "mean_token_accuracy": 0.23122189193964005,
      "num_tokens": 55951837.0,
      "step": 24390
    },
    {
      "entropy": 4.954881381988526,
      "epoch": 2.3434197886647454,
      "grad_norm": 1.0,
      "learning_rate": 0.0004452031603116636,
      "loss": 4.6495,
      "mean_token_accuracy": 0.24795227795839309,
      "num_tokens": 55963641.0,
      "step": 24395
    },
    {
      "entropy": 5.13444938659668,
      "epoch": 2.3439000960614793,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044518073546084326,
      "loss": 4.8219,
      "mean_token_accuracy": 0.22782014608383178,
      "num_tokens": 55975205.0,
      "step": 24400
    },
    {
      "entropy": 5.183595085144043,
      "epoch": 2.344380403458213,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004451583066589316,
      "loss": 4.8326,
      "mean_token_accuracy": 0.22484428137540818,
      "num_tokens": 55985456.0,
      "step": 24405
    },
    {
      "entropy": 5.135547637939453,
      "epoch": 2.344860710854947,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044513587390644925,
      "loss": 4.8282,
      "mean_token_accuracy": 0.2278609275817871,
      "num_tokens": 55997946.0,
      "step": 24410
    },
    {
      "entropy": 5.15059461593628,
      "epoch": 2.345341018251681,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044511343720391724,
      "loss": 4.7992,
      "mean_token_accuracy": 0.22993275821208953,
      "num_tokens": 56008907.0,
      "step": 24415
    },
    {
      "entropy": 5.036718797683716,
      "epoch": 2.345821325648415,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004450909965518563,
      "loss": 4.7515,
      "mean_token_accuracy": 0.23488381803035735,
      "num_tokens": 56019153.0,
      "step": 24420
    },
    {
      "entropy": 4.976629447937012,
      "epoch": 2.3463016330451487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044506855195078755,
      "loss": 4.6379,
      "mean_token_accuracy": 0.23901573121547698,
      "num_tokens": 56029944.0,
      "step": 24425
    },
    {
      "entropy": 5.142466259002686,
      "epoch": 2.3467819404418826,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044504610340123185,
      "loss": 4.7459,
      "mean_token_accuracy": 0.23296163678169252,
      "num_tokens": 56041782.0,
      "step": 24430
    },
    {
      "entropy": 5.202021598815918,
      "epoch": 2.347262247838617,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044502365090371066,
      "loss": 4.8643,
      "mean_token_accuracy": 0.23029387146234512,
      "num_tokens": 56052803.0,
      "step": 24435
    },
    {
      "entropy": 5.126140832901001,
      "epoch": 2.347742555235351,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004450011944587452,
      "loss": 4.7911,
      "mean_token_accuracy": 0.2339429348707199,
      "num_tokens": 56062503.0,
      "step": 24440
    },
    {
      "entropy": 5.022159719467163,
      "epoch": 2.3482228626320847,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044497873406685673,
      "loss": 4.752,
      "mean_token_accuracy": 0.23670969009399415,
      "num_tokens": 56073147.0,
      "step": 24445
    },
    {
      "entropy": 5.094202136993408,
      "epoch": 2.3487031700288186,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004449562697285669,
      "loss": 4.83,
      "mean_token_accuracy": 0.22617395371198654,
      "num_tokens": 56084450.0,
      "step": 24450
    },
    {
      "entropy": 5.052973461151123,
      "epoch": 2.3491834774255524,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044493380144439707,
      "loss": 4.6725,
      "mean_token_accuracy": 0.24197966903448104,
      "num_tokens": 56094869.0,
      "step": 24455
    },
    {
      "entropy": 5.0543232440948485,
      "epoch": 2.3496637848222863,
      "grad_norm": 1.0,
      "learning_rate": 0.000444911329214869,
      "loss": 4.728,
      "mean_token_accuracy": 0.24060946851968765,
      "num_tokens": 56105640.0,
      "step": 24460
    },
    {
      "entropy": 5.139142560958862,
      "epoch": 2.35014409221902,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044488885304050434,
      "loss": 4.856,
      "mean_token_accuracy": 0.222569477558136,
      "num_tokens": 56116756.0,
      "step": 24465
    },
    {
      "entropy": 5.110070466995239,
      "epoch": 2.350624399615754,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000444866372921825,
      "loss": 4.7867,
      "mean_token_accuracy": 0.23002996891736985,
      "num_tokens": 56129674.0,
      "step": 24470
    },
    {
      "entropy": 5.110382890701294,
      "epoch": 2.351104707012488,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044484388885935287,
      "loss": 4.8448,
      "mean_token_accuracy": 0.2263624057173729,
      "num_tokens": 56141360.0,
      "step": 24475
    },
    {
      "entropy": 5.064094591140747,
      "epoch": 2.351585014409222,
      "grad_norm": 1.0,
      "learning_rate": 0.00044482140085361005,
      "loss": 4.6735,
      "mean_token_accuracy": 0.24753724187612533,
      "num_tokens": 56152995.0,
      "step": 24480
    },
    {
      "entropy": 5.137344264984131,
      "epoch": 2.3520653218059557,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044479890890511853,
      "loss": 4.8573,
      "mean_token_accuracy": 0.23367035537958145,
      "num_tokens": 56164515.0,
      "step": 24485
    },
    {
      "entropy": 5.095425701141357,
      "epoch": 2.3525456292026896,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044477641301440054,
      "loss": 4.805,
      "mean_token_accuracy": 0.23389451056718827,
      "num_tokens": 56175983.0,
      "step": 24490
    },
    {
      "entropy": 5.056402587890625,
      "epoch": 2.3530259365994235,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004447539131819784,
      "loss": 4.6958,
      "mean_token_accuracy": 0.23665628731250762,
      "num_tokens": 56187176.0,
      "step": 24495
    },
    {
      "entropy": 5.114434146881104,
      "epoch": 2.3535062439961574,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044473140940837436,
      "loss": 4.7612,
      "mean_token_accuracy": 0.23370923548936845,
      "num_tokens": 56198713.0,
      "step": 24500
    },
    {
      "entropy": 5.025687170028687,
      "epoch": 2.3539865513928913,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044470890169411107,
      "loss": 4.7364,
      "mean_token_accuracy": 0.23166906833648682,
      "num_tokens": 56210473.0,
      "step": 24505
    },
    {
      "entropy": 5.094594287872314,
      "epoch": 2.3544668587896256,
      "grad_norm": 1.03125,
      "learning_rate": 0.000444686390039711,
      "loss": 4.8532,
      "mean_token_accuracy": 0.223676198720932,
      "num_tokens": 56221850.0,
      "step": 24510
    },
    {
      "entropy": 5.097413873672485,
      "epoch": 2.3549471661863595,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004446638744456968,
      "loss": 4.7579,
      "mean_token_accuracy": 0.2332083359360695,
      "num_tokens": 56232359.0,
      "step": 24515
    },
    {
      "entropy": 5.059114217758179,
      "epoch": 2.3554274735830933,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044464135491259135,
      "loss": 4.732,
      "mean_token_accuracy": 0.2376992419362068,
      "num_tokens": 56243385.0,
      "step": 24520
    },
    {
      "entropy": 5.077806663513184,
      "epoch": 2.3559077809798272,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004446188314409172,
      "loss": 4.7401,
      "mean_token_accuracy": 0.24371936917304993,
      "num_tokens": 56254713.0,
      "step": 24525
    },
    {
      "entropy": 5.016921186447144,
      "epoch": 2.356388088376561,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004445963040311975,
      "loss": 4.6979,
      "mean_token_accuracy": 0.24295390397310257,
      "num_tokens": 56266227.0,
      "step": 24530
    },
    {
      "entropy": 5.190720653533935,
      "epoch": 2.356868395773295,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044457377268395526,
      "loss": 4.9352,
      "mean_token_accuracy": 0.22228452563285828,
      "num_tokens": 56276258.0,
      "step": 24535
    },
    {
      "entropy": 5.166210460662842,
      "epoch": 2.357348703170029,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044455123739971355,
      "loss": 4.8458,
      "mean_token_accuracy": 0.2260269895195961,
      "num_tokens": 56287909.0,
      "step": 24540
    },
    {
      "entropy": 5.135364103317261,
      "epoch": 2.3578290105667628,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044452869817899554,
      "loss": 4.8355,
      "mean_token_accuracy": 0.23661180436611176,
      "num_tokens": 56299161.0,
      "step": 24545
    },
    {
      "entropy": 5.1210136890411375,
      "epoch": 2.3583093179634966,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004445061550223246,
      "loss": 4.8061,
      "mean_token_accuracy": 0.23572715371847153,
      "num_tokens": 56311333.0,
      "step": 24550
    },
    {
      "entropy": 5.077708578109741,
      "epoch": 2.3587896253602305,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044448360793022403,
      "loss": 4.6932,
      "mean_token_accuracy": 0.24269042015075684,
      "num_tokens": 56322349.0,
      "step": 24555
    },
    {
      "entropy": 5.114160633087158,
      "epoch": 2.3592699327569644,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004444610569032174,
      "loss": 4.8314,
      "mean_token_accuracy": 0.23055192828178406,
      "num_tokens": 56333169.0,
      "step": 24560
    },
    {
      "entropy": 5.023568534851075,
      "epoch": 2.3597502401536983,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004444385019418281,
      "loss": 4.6821,
      "mean_token_accuracy": 0.2420537903904915,
      "num_tokens": 56345419.0,
      "step": 24565
    },
    {
      "entropy": 5.074182796478271,
      "epoch": 2.360230547550432,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044441594304658004,
      "loss": 4.7138,
      "mean_token_accuracy": 0.24078634679317473,
      "num_tokens": 56355848.0,
      "step": 24570
    },
    {
      "entropy": 5.157995653152466,
      "epoch": 2.360710854947166,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004443933802179968,
      "loss": 4.9146,
      "mean_token_accuracy": 0.2232978403568268,
      "num_tokens": 56368369.0,
      "step": 24575
    },
    {
      "entropy": 5.055591249465943,
      "epoch": 2.3611911623439,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044437081345660224,
      "loss": 4.7259,
      "mean_token_accuracy": 0.23656278401613234,
      "num_tokens": 56379069.0,
      "step": 24580
    },
    {
      "entropy": 5.105720567703247,
      "epoch": 2.3616714697406342,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004443482427629204,
      "loss": 4.7799,
      "mean_token_accuracy": 0.2349407911300659,
      "num_tokens": 56390325.0,
      "step": 24585
    },
    {
      "entropy": 5.133857011795044,
      "epoch": 2.3621517771373677,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004443256681374751,
      "loss": 4.8251,
      "mean_token_accuracy": 0.23864502757787703,
      "num_tokens": 56401336.0,
      "step": 24590
    },
    {
      "entropy": 5.12314076423645,
      "epoch": 2.362632084534102,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004443030895807907,
      "loss": 4.8631,
      "mean_token_accuracy": 0.23199960142374038,
      "num_tokens": 56412547.0,
      "step": 24595
    },
    {
      "entropy": 5.225598573684692,
      "epoch": 2.363112391930836,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044428050709339117,
      "loss": 4.8552,
      "mean_token_accuracy": 0.23228445947170256,
      "num_tokens": 56423757.0,
      "step": 24600
    },
    {
      "entropy": 5.057822799682617,
      "epoch": 2.3635926993275698,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000444257920675801,
      "loss": 4.7866,
      "mean_token_accuracy": 0.2332039773464203,
      "num_tokens": 56433936.0,
      "step": 24605
    },
    {
      "entropy": 5.018553400039673,
      "epoch": 2.3640730067243036,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044423533032854454,
      "loss": 4.7258,
      "mean_token_accuracy": 0.2384037643671036,
      "num_tokens": 56446504.0,
      "step": 24610
    },
    {
      "entropy": 5.172363567352295,
      "epoch": 2.3645533141210375,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004442127360521462,
      "loss": 4.8708,
      "mean_token_accuracy": 0.22684390246868133,
      "num_tokens": 56459019.0,
      "step": 24615
    },
    {
      "entropy": 5.129667568206787,
      "epoch": 2.3650336215177714,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004441901378471306,
      "loss": 4.7813,
      "mean_token_accuracy": 0.22786300182342528,
      "num_tokens": 56470208.0,
      "step": 24620
    },
    {
      "entropy": 5.070704936981201,
      "epoch": 2.3655139289145053,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044416753571402233,
      "loss": 4.7834,
      "mean_token_accuracy": 0.22928062975406646,
      "num_tokens": 56482188.0,
      "step": 24625
    },
    {
      "entropy": 5.043357467651367,
      "epoch": 2.365994236311239,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004441449296533462,
      "loss": 4.7312,
      "mean_token_accuracy": 0.23165766149759293,
      "num_tokens": 56493599.0,
      "step": 24630
    },
    {
      "entropy": 5.1696032047271725,
      "epoch": 2.366474543707973,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044412231966562717,
      "loss": 4.8783,
      "mean_token_accuracy": 0.22342826724052428,
      "num_tokens": 56504448.0,
      "step": 24635
    },
    {
      "entropy": 5.141905307769775,
      "epoch": 2.366954851104707,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044409970575139,
      "loss": 4.8044,
      "mean_token_accuracy": 0.22882331907749176,
      "num_tokens": 56517361.0,
      "step": 24640
    },
    {
      "entropy": 5.0394429683685305,
      "epoch": 2.367435158501441,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004440770879111598,
      "loss": 4.6549,
      "mean_token_accuracy": 0.2430874899029732,
      "num_tokens": 56530152.0,
      "step": 24645
    },
    {
      "entropy": 5.178493881225586,
      "epoch": 2.3679154658981747,
      "grad_norm": 1.125,
      "learning_rate": 0.00044405446614546163,
      "loss": 4.8724,
      "mean_token_accuracy": 0.22889245748519899,
      "num_tokens": 56541121.0,
      "step": 24650
    },
    {
      "entropy": 5.037426376342774,
      "epoch": 2.3683957732949086,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004440318404548208,
      "loss": 4.6863,
      "mean_token_accuracy": 0.24107389599084855,
      "num_tokens": 56551697.0,
      "step": 24655
    },
    {
      "entropy": 5.222705268859864,
      "epoch": 2.3688760806916425,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044400921083976246,
      "loss": 4.98,
      "mean_token_accuracy": 0.2186206564307213,
      "num_tokens": 56564250.0,
      "step": 24660
    },
    {
      "entropy": 5.175839900970459,
      "epoch": 2.3693563880883763,
      "grad_norm": 1.125,
      "learning_rate": 0.0004439865773008122,
      "loss": 4.8002,
      "mean_token_accuracy": 0.22791826725006104,
      "num_tokens": 56574963.0,
      "step": 24665
    },
    {
      "entropy": 5.055000305175781,
      "epoch": 2.3698366954851107,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004439639398384953,
      "loss": 4.7499,
      "mean_token_accuracy": 0.2359452486038208,
      "num_tokens": 56585516.0,
      "step": 24670
    },
    {
      "entropy": 5.022207880020142,
      "epoch": 2.3703170028818445,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044394129845333756,
      "loss": 4.7022,
      "mean_token_accuracy": 0.23406234681606292,
      "num_tokens": 56597459.0,
      "step": 24675
    },
    {
      "entropy": 5.122415208816529,
      "epoch": 2.3707973102785784,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004439186531458645,
      "loss": 4.7958,
      "mean_token_accuracy": 0.2331722319126129,
      "num_tokens": 56608942.0,
      "step": 24680
    },
    {
      "entropy": 5.176864385604858,
      "epoch": 2.3712776176753123,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044389600391660185,
      "loss": 4.8479,
      "mean_token_accuracy": 0.22404264360666276,
      "num_tokens": 56621444.0,
      "step": 24685
    },
    {
      "entropy": 5.135802316665649,
      "epoch": 2.371757925072046,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044387335076607554,
      "loss": 4.7055,
      "mean_token_accuracy": 0.24162033945322037,
      "num_tokens": 56631734.0,
      "step": 24690
    },
    {
      "entropy": 5.019165468215943,
      "epoch": 2.37223823246878,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004438506936948115,
      "loss": 4.7538,
      "mean_token_accuracy": 0.23724590688943864,
      "num_tokens": 56644140.0,
      "step": 24695
    },
    {
      "entropy": 5.037785911560059,
      "epoch": 2.372718539865514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044382803270333565,
      "loss": 4.7063,
      "mean_token_accuracy": 0.24110988825559615,
      "num_tokens": 56655071.0,
      "step": 24700
    },
    {
      "entropy": 5.086750078201294,
      "epoch": 2.373198847262248,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004438053677921743,
      "loss": 4.7457,
      "mean_token_accuracy": 0.23549226820468902,
      "num_tokens": 56667434.0,
      "step": 24705
    },
    {
      "entropy": 5.143817186355591,
      "epoch": 2.3736791546589817,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044378269896185344,
      "loss": 4.8317,
      "mean_token_accuracy": 0.22771190255880355,
      "num_tokens": 56677675.0,
      "step": 24710
    },
    {
      "entropy": 5.088802671432495,
      "epoch": 2.3741594620557156,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004437600262128996,
      "loss": 4.7404,
      "mean_token_accuracy": 0.23774995505809784,
      "num_tokens": 56687849.0,
      "step": 24715
    },
    {
      "entropy": 5.093300676345825,
      "epoch": 2.3746397694524495,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000443737349545839,
      "loss": 4.7578,
      "mean_token_accuracy": 0.23018800616264343,
      "num_tokens": 56699637.0,
      "step": 24720
    },
    {
      "entropy": 5.202822685241699,
      "epoch": 2.3751200768491834,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044371466896119823,
      "loss": 4.8476,
      "mean_token_accuracy": 0.23083342164754866,
      "num_tokens": 56711412.0,
      "step": 24725
    },
    {
      "entropy": 5.218271684646607,
      "epoch": 2.3756003842459172,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044369198445950384,
      "loss": 4.8623,
      "mean_token_accuracy": 0.2346072092652321,
      "num_tokens": 56723379.0,
      "step": 24730
    },
    {
      "entropy": 5.134843635559082,
      "epoch": 2.376080691642651,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004436692960412824,
      "loss": 4.8833,
      "mean_token_accuracy": 0.22743797153234482,
      "num_tokens": 56734936.0,
      "step": 24735
    },
    {
      "entropy": 5.069939804077149,
      "epoch": 2.376560999039385,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004436466037070608,
      "loss": 4.7242,
      "mean_token_accuracy": 0.2421581819653511,
      "num_tokens": 56746925.0,
      "step": 24740
    },
    {
      "entropy": 5.1908531665802,
      "epoch": 2.3770413064361193,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044362390745736585,
      "loss": 4.8798,
      "mean_token_accuracy": 0.23411790728569032,
      "num_tokens": 56759034.0,
      "step": 24745
    },
    {
      "entropy": 5.138485527038574,
      "epoch": 2.377521613832853,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004436012072927245,
      "loss": 4.7858,
      "mean_token_accuracy": 0.2295556589961052,
      "num_tokens": 56770882.0,
      "step": 24750
    },
    {
      "entropy": 5.0240624904632565,
      "epoch": 2.378001921229587,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044357850321366375,
      "loss": 4.721,
      "mean_token_accuracy": 0.24274455159902572,
      "num_tokens": 56782110.0,
      "step": 24755
    },
    {
      "entropy": 5.088750267028809,
      "epoch": 2.378482228626321,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004435557952207107,
      "loss": 4.7559,
      "mean_token_accuracy": 0.23276188522577285,
      "num_tokens": 56793689.0,
      "step": 24760
    },
    {
      "entropy": 5.046028327941895,
      "epoch": 2.378962536023055,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044353308331439257,
      "loss": 4.7654,
      "mean_token_accuracy": 0.23359298408031465,
      "num_tokens": 56805315.0,
      "step": 24765
    },
    {
      "entropy": 5.117844438552856,
      "epoch": 2.3794428434197887,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004435103674952367,
      "loss": 4.8409,
      "mean_token_accuracy": 0.22817204147577286,
      "num_tokens": 56816738.0,
      "step": 24770
    },
    {
      "entropy": 5.224308109283447,
      "epoch": 2.3799231508165226,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044348764776377047,
      "loss": 4.9673,
      "mean_token_accuracy": 0.2240326777100563,
      "num_tokens": 56828613.0,
      "step": 24775
    },
    {
      "entropy": 5.171450853347778,
      "epoch": 2.3804034582132565,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004434649241205214,
      "loss": 4.8086,
      "mean_token_accuracy": 0.2302727147936821,
      "num_tokens": 56840373.0,
      "step": 24780
    },
    {
      "entropy": 5.134899139404297,
      "epoch": 2.3808837656099904,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044344219656601704,
      "loss": 4.8483,
      "mean_token_accuracy": 0.22254785299301147,
      "num_tokens": 56850581.0,
      "step": 24785
    },
    {
      "entropy": 5.109383678436279,
      "epoch": 2.3813640730067243,
      "grad_norm": 1.015625,
      "learning_rate": 0.000443419465100785,
      "loss": 4.7683,
      "mean_token_accuracy": 0.2374254807829857,
      "num_tokens": 56861576.0,
      "step": 24790
    },
    {
      "entropy": 5.1563536643981935,
      "epoch": 2.381844380403458,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004433967297253531,
      "loss": 4.8766,
      "mean_token_accuracy": 0.23381539881229402,
      "num_tokens": 56872833.0,
      "step": 24795
    },
    {
      "entropy": 5.130254220962525,
      "epoch": 2.382324687800192,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044337399044024924,
      "loss": 4.8211,
      "mean_token_accuracy": 0.23404806703329087,
      "num_tokens": 56884367.0,
      "step": 24800
    },
    {
      "entropy": 4.977699089050293,
      "epoch": 2.382804995196926,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004433512472460012,
      "loss": 4.6737,
      "mean_token_accuracy": 0.2421378791332245,
      "num_tokens": 56895233.0,
      "step": 24805
    },
    {
      "entropy": 5.033953428268433,
      "epoch": 2.38328530259366,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044332850014313713,
      "loss": 4.7974,
      "mean_token_accuracy": 0.2327971488237381,
      "num_tokens": 56907133.0,
      "step": 24810
    },
    {
      "entropy": 5.189722967147827,
      "epoch": 2.3837656099903937,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004433057491321851,
      "loss": 4.8536,
      "mean_token_accuracy": 0.23125628381967545,
      "num_tokens": 56919372.0,
      "step": 24815
    },
    {
      "entropy": 5.107855606079101,
      "epoch": 2.384245917387128,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044328299421367333,
      "loss": 4.8227,
      "mean_token_accuracy": 0.22587259560823442,
      "num_tokens": 56930523.0,
      "step": 24820
    },
    {
      "entropy": 5.007207632064819,
      "epoch": 2.3847262247838614,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004432602353881302,
      "loss": 4.6635,
      "mean_token_accuracy": 0.2368766114115715,
      "num_tokens": 56941324.0,
      "step": 24825
    },
    {
      "entropy": 4.9703474044799805,
      "epoch": 2.3852065321805958,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044323747265608395,
      "loss": 4.6579,
      "mean_token_accuracy": 0.24127011448144914,
      "num_tokens": 56952359.0,
      "step": 24830
    },
    {
      "entropy": 5.107290506362915,
      "epoch": 2.3856868395773296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004432147060180632,
      "loss": 4.7367,
      "mean_token_accuracy": 0.24027741849422454,
      "num_tokens": 56963456.0,
      "step": 24835
    },
    {
      "entropy": 5.024862623214721,
      "epoch": 2.3861671469740635,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044319193547459645,
      "loss": 4.6859,
      "mean_token_accuracy": 0.23604112565517427,
      "num_tokens": 56975029.0,
      "step": 24840
    },
    {
      "entropy": 5.04381365776062,
      "epoch": 2.3866474543707974,
      "grad_norm": 1.125,
      "learning_rate": 0.0004431691610262124,
      "loss": 4.7753,
      "mean_token_accuracy": 0.2338072270154953,
      "num_tokens": 56987159.0,
      "step": 24845
    },
    {
      "entropy": 5.108178901672363,
      "epoch": 2.3871277617675313,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044314638267343976,
      "loss": 4.8502,
      "mean_token_accuracy": 0.23129929155111312,
      "num_tokens": 56998359.0,
      "step": 24850
    },
    {
      "entropy": 5.104422950744629,
      "epoch": 2.387608069164265,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004431236004168075,
      "loss": 4.8214,
      "mean_token_accuracy": 0.22530155181884765,
      "num_tokens": 57009980.0,
      "step": 24855
    },
    {
      "entropy": 5.116705894470215,
      "epoch": 2.388088376560999,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004431008142568444,
      "loss": 4.7968,
      "mean_token_accuracy": 0.23201425969600678,
      "num_tokens": 57022531.0,
      "step": 24860
    },
    {
      "entropy": 5.16227011680603,
      "epoch": 2.388568683957733,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044307802419407954,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22191356271505355,
      "num_tokens": 57034746.0,
      "step": 24865
    },
    {
      "entropy": 5.094029808044434,
      "epoch": 2.389048991354467,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004430552302290421,
      "loss": 4.8161,
      "mean_token_accuracy": 0.22856799364089966,
      "num_tokens": 57046496.0,
      "step": 24870
    },
    {
      "entropy": 5.025378847122193,
      "epoch": 2.3895292987512007,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004430324323622611,
      "loss": 4.7517,
      "mean_token_accuracy": 0.22921628057956694,
      "num_tokens": 57059083.0,
      "step": 24875
    },
    {
      "entropy": 5.1534104347229,
      "epoch": 2.3900096061479346,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044300963059426605,
      "loss": 4.7437,
      "mean_token_accuracy": 0.23937105238437653,
      "num_tokens": 57070077.0,
      "step": 24880
    },
    {
      "entropy": 5.1272345066070555,
      "epoch": 2.3904899135446684,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044298682492558637,
      "loss": 4.7175,
      "mean_token_accuracy": 0.23633715212345124,
      "num_tokens": 57081397.0,
      "step": 24885
    },
    {
      "entropy": 5.094593715667725,
      "epoch": 2.3909702209414023,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044296401535675136,
      "loss": 4.8129,
      "mean_token_accuracy": 0.23150604814291,
      "num_tokens": 57093882.0,
      "step": 24890
    },
    {
      "entropy": 5.129273223876953,
      "epoch": 2.3914505283381366,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044294120188829056,
      "loss": 4.8818,
      "mean_token_accuracy": 0.22558727115392685,
      "num_tokens": 57104603.0,
      "step": 24895
    },
    {
      "entropy": 5.177860975265503,
      "epoch": 2.39193083573487,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004429183845207339,
      "loss": 4.8338,
      "mean_token_accuracy": 0.2269103616476059,
      "num_tokens": 57115512.0,
      "step": 24900
    },
    {
      "entropy": 5.111091756820679,
      "epoch": 2.3924111431316044,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004428955632546108,
      "loss": 4.7797,
      "mean_token_accuracy": 0.232811863720417,
      "num_tokens": 57125934.0,
      "step": 24905
    },
    {
      "entropy": 5.018808031082154,
      "epoch": 2.3928914505283383,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004428727380904514,
      "loss": 4.7047,
      "mean_token_accuracy": 0.24079825729131699,
      "num_tokens": 57136437.0,
      "step": 24910
    },
    {
      "entropy": 5.067674064636231,
      "epoch": 2.393371757925072,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044284990902878545,
      "loss": 4.7632,
      "mean_token_accuracy": 0.22171027660369874,
      "num_tokens": 57147542.0,
      "step": 24915
    },
    {
      "entropy": 5.085659599304199,
      "epoch": 2.393852065321806,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044282707607014304,
      "loss": 4.7489,
      "mean_token_accuracy": 0.23067063093185425,
      "num_tokens": 57159248.0,
      "step": 24920
    },
    {
      "entropy": 5.076380443572998,
      "epoch": 2.39433237271854,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044280423921505427,
      "loss": 4.7694,
      "mean_token_accuracy": 0.23429475724697113,
      "num_tokens": 57171627.0,
      "step": 24925
    },
    {
      "entropy": 5.1194539070129395,
      "epoch": 2.394812680115274,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004427813984640493,
      "loss": 4.761,
      "mean_token_accuracy": 0.24031212329864501,
      "num_tokens": 57181604.0,
      "step": 24930
    },
    {
      "entropy": 5.0122722625732425,
      "epoch": 2.3952929875120077,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004427585538176585,
      "loss": 4.7023,
      "mean_token_accuracy": 0.23775070905685425,
      "num_tokens": 57191409.0,
      "step": 24935
    },
    {
      "entropy": 5.041350984573365,
      "epoch": 2.3957732949087416,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044273570527641223,
      "loss": 4.7247,
      "mean_token_accuracy": 0.23627171218395232,
      "num_tokens": 57204606.0,
      "step": 24940
    },
    {
      "entropy": 5.1117020606994625,
      "epoch": 2.3962536023054755,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044271285284084097,
      "loss": 4.8377,
      "mean_token_accuracy": 0.22930252850055693,
      "num_tokens": 57215641.0,
      "step": 24945
    },
    {
      "entropy": 5.115767240524292,
      "epoch": 2.3967339097022093,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004426899965114752,
      "loss": 4.8178,
      "mean_token_accuracy": 0.22832341641187667,
      "num_tokens": 57226520.0,
      "step": 24950
    },
    {
      "entropy": 5.158056592941284,
      "epoch": 2.3972142170989432,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044266713628884566,
      "loss": 4.8457,
      "mean_token_accuracy": 0.21970202773809433,
      "num_tokens": 57237859.0,
      "step": 24955
    },
    {
      "entropy": 5.129123497009277,
      "epoch": 2.397694524495677,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044264427217348315,
      "loss": 4.717,
      "mean_token_accuracy": 0.24113675355911254,
      "num_tokens": 57248727.0,
      "step": 24960
    },
    {
      "entropy": 5.050986337661743,
      "epoch": 2.398174831892411,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004426214041659184,
      "loss": 4.7535,
      "mean_token_accuracy": 0.23984145373106003,
      "num_tokens": 57259644.0,
      "step": 24965
    },
    {
      "entropy": 5.112276887893676,
      "epoch": 2.398655139289145,
      "grad_norm": 1.0,
      "learning_rate": 0.0004425985322666824,
      "loss": 4.8,
      "mean_token_accuracy": 0.2295980304479599,
      "num_tokens": 57271140.0,
      "step": 24970
    },
    {
      "entropy": 5.1218297481536865,
      "epoch": 2.3991354466858787,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004425756564763061,
      "loss": 4.7815,
      "mean_token_accuracy": 0.23568786084651946,
      "num_tokens": 57281205.0,
      "step": 24975
    },
    {
      "entropy": 5.080163383483887,
      "epoch": 2.399615754082613,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044255277679532075,
      "loss": 4.7421,
      "mean_token_accuracy": 0.23275472223758698,
      "num_tokens": 57292056.0,
      "step": 24980
    },
    {
      "entropy": 5.104554891586304,
      "epoch": 2.400096061479347,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044252989322425735,
      "loss": 4.7461,
      "mean_token_accuracy": 0.236335951089859,
      "num_tokens": 57303444.0,
      "step": 24985
    },
    {
      "entropy": 5.114625644683838,
      "epoch": 2.400576368876081,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044250700576364734,
      "loss": 4.7744,
      "mean_token_accuracy": 0.22687787860631942,
      "num_tokens": 57314784.0,
      "step": 24990
    },
    {
      "entropy": 5.000076580047607,
      "epoch": 2.4010566762728147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000442484114414022,
      "loss": 4.6938,
      "mean_token_accuracy": 0.24125445783138275,
      "num_tokens": 57325896.0,
      "step": 24995
    },
    {
      "entropy": 5.074965858459473,
      "epoch": 2.4015369836695486,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004424612191759129,
      "loss": 4.8204,
      "mean_token_accuracy": 0.2237904578447342,
      "num_tokens": 57338338.0,
      "step": 25000
    },
    {
      "entropy": 5.126633548736573,
      "epoch": 2.4020172910662825,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004424383200498515,
      "loss": 4.8145,
      "mean_token_accuracy": 0.23302128165960312,
      "num_tokens": 57350038.0,
      "step": 25005
    },
    {
      "entropy": 5.159433364868164,
      "epoch": 2.4024975984630164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004424154170363696,
      "loss": 4.8461,
      "mean_token_accuracy": 0.22654461115598679,
      "num_tokens": 57362371.0,
      "step": 25010
    },
    {
      "entropy": 5.1346272945404055,
      "epoch": 2.4029779058597502,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004423925101359987,
      "loss": 4.8127,
      "mean_token_accuracy": 0.2227134719491005,
      "num_tokens": 57374032.0,
      "step": 25015
    },
    {
      "entropy": 5.081220531463623,
      "epoch": 2.403458213256484,
      "grad_norm": 1.0,
      "learning_rate": 0.0004423695993492709,
      "loss": 4.7342,
      "mean_token_accuracy": 0.23666941672563552,
      "num_tokens": 57385705.0,
      "step": 25020
    },
    {
      "entropy": 5.04365029335022,
      "epoch": 2.403938520653218,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004423466846767179,
      "loss": 4.7166,
      "mean_token_accuracy": 0.24139521569013594,
      "num_tokens": 57396863.0,
      "step": 25025
    },
    {
      "entropy": 5.234294462203979,
      "epoch": 2.404418828049952,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044232376611887185,
      "loss": 4.9756,
      "mean_token_accuracy": 0.22222196608781813,
      "num_tokens": 57408486.0,
      "step": 25030
    },
    {
      "entropy": 5.1494776725769045,
      "epoch": 2.4048991354466858,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044230084367626477,
      "loss": 4.7951,
      "mean_token_accuracy": 0.23207901418209076,
      "num_tokens": 57419899.0,
      "step": 25035
    },
    {
      "entropy": 5.1677226543426515,
      "epoch": 2.4053794428434196,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004422779173494288,
      "loss": 4.7765,
      "mean_token_accuracy": 0.23284651190042496,
      "num_tokens": 57431620.0,
      "step": 25040
    },
    {
      "entropy": 5.063530826568604,
      "epoch": 2.4058597502401535,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004422549871388965,
      "loss": 4.7909,
      "mean_token_accuracy": 0.22949365079402922,
      "num_tokens": 57443985.0,
      "step": 25045
    },
    {
      "entropy": 5.104589939117432,
      "epoch": 2.4063400576368874,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044223205304519994,
      "loss": 4.7798,
      "mean_token_accuracy": 0.2340580180287361,
      "num_tokens": 57455160.0,
      "step": 25050
    },
    {
      "entropy": 5.141811752319336,
      "epoch": 2.4068203650336217,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004422091150688717,
      "loss": 4.8076,
      "mean_token_accuracy": 0.23721980303525925,
      "num_tokens": 57467557.0,
      "step": 25055
    },
    {
      "entropy": 5.1415762424469,
      "epoch": 2.4073006724303556,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004421861732104443,
      "loss": 4.818,
      "mean_token_accuracy": 0.22998912930488585,
      "num_tokens": 57478749.0,
      "step": 25060
    },
    {
      "entropy": 5.04126558303833,
      "epoch": 2.4077809798270895,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004421632274704504,
      "loss": 4.7407,
      "mean_token_accuracy": 0.23503637909889222,
      "num_tokens": 57489899.0,
      "step": 25065
    },
    {
      "entropy": 5.186473798751831,
      "epoch": 2.4082612872238234,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004421402778494227,
      "loss": 4.7906,
      "mean_token_accuracy": 0.22827454507350922,
      "num_tokens": 57502086.0,
      "step": 25070
    },
    {
      "entropy": 5.081528043746948,
      "epoch": 2.4087415946205573,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004421173243478941,
      "loss": 4.7015,
      "mean_token_accuracy": 0.23745288401842118,
      "num_tokens": 57513555.0,
      "step": 25075
    },
    {
      "entropy": 5.166746711730957,
      "epoch": 2.409221902017291,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044209436696639745,
      "loss": 4.8111,
      "mean_token_accuracy": 0.22953476160764694,
      "num_tokens": 57525100.0,
      "step": 25080
    },
    {
      "entropy": 5.095049428939819,
      "epoch": 2.409702209414025,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044207140570546574,
      "loss": 4.7677,
      "mean_token_accuracy": 0.23114321380853653,
      "num_tokens": 57536355.0,
      "step": 25085
    },
    {
      "entropy": 5.042273664474488,
      "epoch": 2.410182516810759,
      "grad_norm": 1.0,
      "learning_rate": 0.00044204844056563216,
      "loss": 4.7189,
      "mean_token_accuracy": 0.23847521096467972,
      "num_tokens": 57547260.0,
      "step": 25090
    },
    {
      "entropy": 5.089078617095947,
      "epoch": 2.410662824207493,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004420254715474297,
      "loss": 4.7621,
      "mean_token_accuracy": 0.23298065960407258,
      "num_tokens": 57558299.0,
      "step": 25095
    },
    {
      "entropy": 5.050247859954834,
      "epoch": 2.4111431316042267,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044200249865139187,
      "loss": 4.7568,
      "mean_token_accuracy": 0.23560206294059755,
      "num_tokens": 57569393.0,
      "step": 25100
    },
    {
      "entropy": 5.162860631942749,
      "epoch": 2.4116234390009605,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044197952187805185,
      "loss": 4.8044,
      "mean_token_accuracy": 0.2286287248134613,
      "num_tokens": 57580585.0,
      "step": 25105
    },
    {
      "entropy": 5.21138744354248,
      "epoch": 2.4121037463976944,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044195654122794324,
      "loss": 4.8592,
      "mean_token_accuracy": 0.23075156211853026,
      "num_tokens": 57592887.0,
      "step": 25110
    },
    {
      "entropy": 5.177557182312012,
      "epoch": 2.4125840537944283,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004419335567015994,
      "loss": 4.8423,
      "mean_token_accuracy": 0.22508063912391663,
      "num_tokens": 57605129.0,
      "step": 25115
    },
    {
      "entropy": 4.9699546813964846,
      "epoch": 2.413064361191162,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004419105682995542,
      "loss": 4.6689,
      "mean_token_accuracy": 0.23967505842447281,
      "num_tokens": 57616852.0,
      "step": 25120
    },
    {
      "entropy": 5.087364530563354,
      "epoch": 2.413544668587896,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004418875760223411,
      "loss": 4.8098,
      "mean_token_accuracy": 0.2304730862379074,
      "num_tokens": 57629808.0,
      "step": 25125
    },
    {
      "entropy": 5.207426071166992,
      "epoch": 2.4140249759846304,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044186457987049405,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22527200728654861,
      "num_tokens": 57641025.0,
      "step": 25130
    },
    {
      "entropy": 5.0586817264556885,
      "epoch": 2.414505283381364,
      "grad_norm": 1.109375,
      "learning_rate": 0.000441841579844547,
      "loss": 4.7455,
      "mean_token_accuracy": 0.23246509730815887,
      "num_tokens": 57652240.0,
      "step": 25135
    },
    {
      "entropy": 5.0275397300720215,
      "epoch": 2.414985590778098,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004418185759450338,
      "loss": 4.696,
      "mean_token_accuracy": 0.24545542299747466,
      "num_tokens": 57663842.0,
      "step": 25140
    },
    {
      "entropy": 5.09443507194519,
      "epoch": 2.415465898174832,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004417955681724887,
      "loss": 4.7122,
      "mean_token_accuracy": 0.2399942934513092,
      "num_tokens": 57674299.0,
      "step": 25145
    },
    {
      "entropy": 5.028635120391845,
      "epoch": 2.415946205571566,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044177255652744576,
      "loss": 4.7215,
      "mean_token_accuracy": 0.24242948591709138,
      "num_tokens": 57684858.0,
      "step": 25150
    },
    {
      "entropy": 5.173546934127808,
      "epoch": 2.4164265129683,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044174954101043926,
      "loss": 4.8115,
      "mean_token_accuracy": 0.22744420170783997,
      "num_tokens": 57696542.0,
      "step": 25155
    },
    {
      "entropy": 5.076199007034302,
      "epoch": 2.4169068203650337,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044172652162200354,
      "loss": 4.7042,
      "mean_token_accuracy": 0.24288443624973297,
      "num_tokens": 57708603.0,
      "step": 25160
    },
    {
      "entropy": 5.055661869049072,
      "epoch": 2.4173871277617676,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004417034983626731,
      "loss": 4.779,
      "mean_token_accuracy": 0.23503173589706422,
      "num_tokens": 57720273.0,
      "step": 25165
    },
    {
      "entropy": 5.114255332946778,
      "epoch": 2.4178674351585014,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004416804712329825,
      "loss": 4.7123,
      "mean_token_accuracy": 0.2370807036757469,
      "num_tokens": 57730915.0,
      "step": 25170
    },
    {
      "entropy": 5.005442762374878,
      "epoch": 2.4183477425552353,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044165744023346614,
      "loss": 4.7425,
      "mean_token_accuracy": 0.24020758867263795,
      "num_tokens": 57743379.0,
      "step": 25175
    },
    {
      "entropy": 5.034661054611206,
      "epoch": 2.418828049951969,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044163440536465904,
      "loss": 4.7769,
      "mean_token_accuracy": 0.2337392747402191,
      "num_tokens": 57754441.0,
      "step": 25180
    },
    {
      "entropy": 5.147348403930664,
      "epoch": 2.419308357348703,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044161136662709577,
      "loss": 4.7877,
      "mean_token_accuracy": 0.23125338256359101,
      "num_tokens": 57765701.0,
      "step": 25185
    },
    {
      "entropy": 5.143546724319458,
      "epoch": 2.419788664745437,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044158832402131133,
      "loss": 4.8756,
      "mean_token_accuracy": 0.2272426813840866,
      "num_tokens": 57776774.0,
      "step": 25190
    },
    {
      "entropy": 5.157384920120239,
      "epoch": 2.420268972142171,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044156527754784066,
      "loss": 4.8274,
      "mean_token_accuracy": 0.2321350619196892,
      "num_tokens": 57787618.0,
      "step": 25195
    },
    {
      "entropy": 5.208212661743164,
      "epoch": 2.4207492795389047,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044154222720721887,
      "loss": 4.8736,
      "mean_token_accuracy": 0.22514759451150895,
      "num_tokens": 57800133.0,
      "step": 25200
    },
    {
      "entropy": 5.1266951084136965,
      "epoch": 2.421229586935639,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000441519172999981,
      "loss": 4.7415,
      "mean_token_accuracy": 0.23400021344423294,
      "num_tokens": 57812736.0,
      "step": 25205
    },
    {
      "entropy": 5.088744974136352,
      "epoch": 2.4217098943323725,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004414961149266625,
      "loss": 4.7399,
      "mean_token_accuracy": 0.23914081901311873,
      "num_tokens": 57824459.0,
      "step": 25210
    },
    {
      "entropy": 5.097274208068848,
      "epoch": 2.422190201729107,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044147305298779856,
      "loss": 4.805,
      "mean_token_accuracy": 0.23188676089048385,
      "num_tokens": 57836292.0,
      "step": 25215
    },
    {
      "entropy": 5.097728157043457,
      "epoch": 2.4226705091258407,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004414499871839247,
      "loss": 4.7726,
      "mean_token_accuracy": 0.2336946427822113,
      "num_tokens": 57848197.0,
      "step": 25220
    },
    {
      "entropy": 5.146145582199097,
      "epoch": 2.4231508165225746,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004414269175155763,
      "loss": 4.7666,
      "mean_token_accuracy": 0.2287415772676468,
      "num_tokens": 57859491.0,
      "step": 25225
    },
    {
      "entropy": 5.058282279968262,
      "epoch": 2.4236311239193085,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004414038439832891,
      "loss": 4.6724,
      "mean_token_accuracy": 0.248495015501976,
      "num_tokens": 57870199.0,
      "step": 25230
    },
    {
      "entropy": 5.134005403518676,
      "epoch": 2.4241114313160423,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004413807665875988,
      "loss": 4.8957,
      "mean_token_accuracy": 0.22890284806489944,
      "num_tokens": 57882059.0,
      "step": 25235
    },
    {
      "entropy": 5.104201030731201,
      "epoch": 2.4245917387127762,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044135768532904104,
      "loss": 4.7715,
      "mean_token_accuracy": 0.23680859953165054,
      "num_tokens": 57892316.0,
      "step": 25240
    },
    {
      "entropy": 5.066174459457398,
      "epoch": 2.42507204610951,
      "grad_norm": 1.078125,
      "learning_rate": 0.000441334600208152,
      "loss": 4.7231,
      "mean_token_accuracy": 0.23976736068725585,
      "num_tokens": 57903273.0,
      "step": 25245
    },
    {
      "entropy": 5.111947822570801,
      "epoch": 2.425552353506244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044131151122546724,
      "loss": 4.8659,
      "mean_token_accuracy": 0.22582025974988937,
      "num_tokens": 57914484.0,
      "step": 25250
    },
    {
      "entropy": 5.079412984848022,
      "epoch": 2.426032660902978,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044128841838152313,
      "loss": 4.6668,
      "mean_token_accuracy": 0.23892130851745605,
      "num_tokens": 57925213.0,
      "step": 25255
    },
    {
      "entropy": 5.1167895793914795,
      "epoch": 2.4265129682997117,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004412653216768558,
      "loss": 4.764,
      "mean_token_accuracy": 0.22619060426950455,
      "num_tokens": 57937422.0,
      "step": 25260
    },
    {
      "entropy": 5.1064495086669925,
      "epoch": 2.4269932756964456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004412422211120013,
      "loss": 4.7581,
      "mean_token_accuracy": 0.23461629301309586,
      "num_tokens": 57947576.0,
      "step": 25265
    },
    {
      "entropy": 5.038341283798218,
      "epoch": 2.4274735830931795,
      "grad_norm": 1.0,
      "learning_rate": 0.0004412191166874961,
      "loss": 4.7407,
      "mean_token_accuracy": 0.23312466591596603,
      "num_tokens": 57958998.0,
      "step": 25270
    },
    {
      "entropy": 5.144538736343383,
      "epoch": 2.4279538904899134,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004411960084038766,
      "loss": 4.8164,
      "mean_token_accuracy": 0.23059480488300324,
      "num_tokens": 57970126.0,
      "step": 25275
    },
    {
      "entropy": 5.129548263549805,
      "epoch": 2.4284341978866473,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044117289626167917,
      "loss": 4.7571,
      "mean_token_accuracy": 0.23756093829870223,
      "num_tokens": 57981523.0,
      "step": 25280
    },
    {
      "entropy": 5.0713263034820555,
      "epoch": 2.428914505283381,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004411497802614406,
      "loss": 4.7131,
      "mean_token_accuracy": 0.24007366597652435,
      "num_tokens": 57992113.0,
      "step": 25285
    },
    {
      "entropy": 5.130498838424683,
      "epoch": 2.4293948126801155,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004411266604036975,
      "loss": 4.8208,
      "mean_token_accuracy": 0.22891694605350493,
      "num_tokens": 58005163.0,
      "step": 25290
    },
    {
      "entropy": 5.153032684326172,
      "epoch": 2.4298751200768494,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044110353668898674,
      "loss": 4.7865,
      "mean_token_accuracy": 0.22758774161338807,
      "num_tokens": 58015095.0,
      "step": 25295
    },
    {
      "entropy": 5.06425895690918,
      "epoch": 2.4303554274735832,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004410804091178449,
      "loss": 4.7131,
      "mean_token_accuracy": 0.23937252312898635,
      "num_tokens": 58025522.0,
      "step": 25300
    },
    {
      "entropy": 5.087374210357666,
      "epoch": 2.430835734870317,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004410572776908092,
      "loss": 4.7734,
      "mean_token_accuracy": 0.23314732909202576,
      "num_tokens": 58036567.0,
      "step": 25305
    },
    {
      "entropy": 5.093733644485473,
      "epoch": 2.431316042267051,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044103414240841664,
      "loss": 4.7251,
      "mean_token_accuracy": 0.23997088968753816,
      "num_tokens": 58048321.0,
      "step": 25310
    },
    {
      "entropy": 5.094263315200806,
      "epoch": 2.431796349663785,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004410110032712043,
      "loss": 4.8109,
      "mean_token_accuracy": 0.22674921303987502,
      "num_tokens": 58060505.0,
      "step": 25315
    },
    {
      "entropy": 5.085807132720947,
      "epoch": 2.4322766570605188,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004409878602797094,
      "loss": 4.7318,
      "mean_token_accuracy": 0.24517549574375153,
      "num_tokens": 58071926.0,
      "step": 25320
    },
    {
      "entropy": 5.12570972442627,
      "epoch": 2.4327569644572526,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044096471343446923,
      "loss": 4.8038,
      "mean_token_accuracy": 0.23613174855709076,
      "num_tokens": 58082963.0,
      "step": 25325
    },
    {
      "entropy": 5.1096268653869625,
      "epoch": 2.4332372718539865,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004409415627360213,
      "loss": 4.7841,
      "mean_token_accuracy": 0.2344191551208496,
      "num_tokens": 58095343.0,
      "step": 25330
    },
    {
      "entropy": 5.002459621429443,
      "epoch": 2.4337175792507204,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044091840818490303,
      "loss": 4.6934,
      "mean_token_accuracy": 0.24174196869134904,
      "num_tokens": 58106763.0,
      "step": 25335
    },
    {
      "entropy": 5.098709583282471,
      "epoch": 2.4341978866474543,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044089524978165197,
      "loss": 4.7138,
      "mean_token_accuracy": 0.24017773121595382,
      "num_tokens": 58116942.0,
      "step": 25340
    },
    {
      "entropy": 5.051480770111084,
      "epoch": 2.434678194044188,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044087208752680577,
      "loss": 4.6693,
      "mean_token_accuracy": 0.24185501784086227,
      "num_tokens": 58127365.0,
      "step": 25345
    },
    {
      "entropy": 5.079211044311523,
      "epoch": 2.435158501440922,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004408489214209023,
      "loss": 4.8167,
      "mean_token_accuracy": 0.2244767278432846,
      "num_tokens": 58137572.0,
      "step": 25350
    },
    {
      "entropy": 5.17046365737915,
      "epoch": 2.435638808837656,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004408257514644793,
      "loss": 4.8894,
      "mean_token_accuracy": 0.23011724799871444,
      "num_tokens": 58150012.0,
      "step": 25355
    },
    {
      "entropy": 5.0342365264892575,
      "epoch": 2.43611911623439,
      "grad_norm": 0.96875,
      "learning_rate": 0.00044080257765807476,
      "loss": 4.6133,
      "mean_token_accuracy": 0.24943419247865678,
      "num_tokens": 58160611.0,
      "step": 25360
    },
    {
      "entropy": 5.133284950256348,
      "epoch": 2.436599423631124,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004407794000022267,
      "loss": 4.8995,
      "mean_token_accuracy": 0.22139323949813844,
      "num_tokens": 58172113.0,
      "step": 25365
    },
    {
      "entropy": 5.139159250259399,
      "epoch": 2.437079731027858,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004407562184974732,
      "loss": 4.7878,
      "mean_token_accuracy": 0.23307765573263167,
      "num_tokens": 58183118.0,
      "step": 25370
    },
    {
      "entropy": 5.091373300552368,
      "epoch": 2.437560038424592,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004407330331443526,
      "loss": 4.7594,
      "mean_token_accuracy": 0.2336428537964821,
      "num_tokens": 58194534.0,
      "step": 25375
    },
    {
      "entropy": 5.125935316085815,
      "epoch": 2.438040345821326,
      "grad_norm": 0.921875,
      "learning_rate": 0.000440709843943403,
      "loss": 4.8108,
      "mean_token_accuracy": 0.2305499017238617,
      "num_tokens": 58205952.0,
      "step": 25380
    },
    {
      "entropy": 5.120977115631104,
      "epoch": 2.4385206532180597,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000440686650895163,
      "loss": 4.7468,
      "mean_token_accuracy": 0.2368416279554367,
      "num_tokens": 58217811.0,
      "step": 25385
    },
    {
      "entropy": 5.184613800048828,
      "epoch": 2.4390009606147935,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044066345400017084,
      "loss": 4.8804,
      "mean_token_accuracy": 0.2276732936501503,
      "num_tokens": 58228658.0,
      "step": 25390
    },
    {
      "entropy": 5.112476444244384,
      "epoch": 2.4394812680115274,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044064025325896524,
      "loss": 4.8298,
      "mean_token_accuracy": 0.227210533618927,
      "num_tokens": 58242130.0,
      "step": 25395
    },
    {
      "entropy": 5.129372644424438,
      "epoch": 2.4399615754082613,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044061704867208484,
      "loss": 4.7553,
      "mean_token_accuracy": 0.23601716458797456,
      "num_tokens": 58253533.0,
      "step": 25400
    },
    {
      "entropy": 5.109399747848511,
      "epoch": 2.440441882804995,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044059384024006825,
      "loss": 4.7625,
      "mean_token_accuracy": 0.2311705693602562,
      "num_tokens": 58264381.0,
      "step": 25405
    },
    {
      "entropy": 5.082081604003906,
      "epoch": 2.440922190201729,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004405706279634545,
      "loss": 4.8401,
      "mean_token_accuracy": 0.2272565856575966,
      "num_tokens": 58275114.0,
      "step": 25410
    },
    {
      "entropy": 5.079714679718018,
      "epoch": 2.441402497598463,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00044054741184278243,
      "loss": 4.7217,
      "mean_token_accuracy": 0.2352374330163002,
      "num_tokens": 58287349.0,
      "step": 25415
    },
    {
      "entropy": 5.066584300994873,
      "epoch": 2.441882804995197,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044052419187859095,
      "loss": 4.7234,
      "mean_token_accuracy": 0.24268799871206284,
      "num_tokens": 58298787.0,
      "step": 25420
    },
    {
      "entropy": 5.07541675567627,
      "epoch": 2.4423631123919307,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004405009680714193,
      "loss": 4.8325,
      "mean_token_accuracy": 0.2259284421801567,
      "num_tokens": 58311019.0,
      "step": 25425
    },
    {
      "entropy": 5.065648937225342,
      "epoch": 2.4428434197886646,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004404777404218065,
      "loss": 4.7427,
      "mean_token_accuracy": 0.23616894334554672,
      "num_tokens": 58322692.0,
      "step": 25430
    },
    {
      "entropy": 5.1116297245025635,
      "epoch": 2.4433237271853985,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000440454508930292,
      "loss": 4.8031,
      "mean_token_accuracy": 0.23099347352981567,
      "num_tokens": 58333638.0,
      "step": 25435
    },
    {
      "entropy": 5.099006319046021,
      "epoch": 2.443804034582133,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004404312735974152,
      "loss": 4.7625,
      "mean_token_accuracy": 0.23393814414739608,
      "num_tokens": 58345291.0,
      "step": 25440
    },
    {
      "entropy": 5.183764505386352,
      "epoch": 2.4442843419788662,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00044040803442371533,
      "loss": 4.8394,
      "mean_token_accuracy": 0.23117247819900513,
      "num_tokens": 58357303.0,
      "step": 25445
    },
    {
      "entropy": 5.064045238494873,
      "epoch": 2.4447646493756006,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004403847914097321,
      "loss": 4.7676,
      "mean_token_accuracy": 0.23392631709575654,
      "num_tokens": 58369804.0,
      "step": 25450
    },
    {
      "entropy": 5.061349296569825,
      "epoch": 2.4452449567723344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044036154455600517,
      "loss": 4.7961,
      "mean_token_accuracy": 0.2300865888595581,
      "num_tokens": 58381027.0,
      "step": 25455
    },
    {
      "entropy": 5.079055309295654,
      "epoch": 2.4457252641690683,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004403382938630741,
      "loss": 4.7284,
      "mean_token_accuracy": 0.2324882924556732,
      "num_tokens": 58392165.0,
      "step": 25460
    },
    {
      "entropy": 5.137262344360352,
      "epoch": 2.446205571565802,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044031503933147887,
      "loss": 4.7635,
      "mean_token_accuracy": 0.23302173167467116,
      "num_tokens": 58402494.0,
      "step": 25465
    },
    {
      "entropy": 5.030036067962646,
      "epoch": 2.446685878962536,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044029178096175934,
      "loss": 4.7084,
      "mean_token_accuracy": 0.2400970160961151,
      "num_tokens": 58413735.0,
      "step": 25470
    },
    {
      "entropy": 5.0946910858154295,
      "epoch": 2.44716618635927,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004402685187544554,
      "loss": 4.7603,
      "mean_token_accuracy": 0.22815975546836853,
      "num_tokens": 58425977.0,
      "step": 25475
    },
    {
      "entropy": 5.093181991577149,
      "epoch": 2.447646493756004,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004402452527101072,
      "loss": 4.7635,
      "mean_token_accuracy": 0.23100874722003936,
      "num_tokens": 58436845.0,
      "step": 25480
    },
    {
      "entropy": 5.067014122009278,
      "epoch": 2.4481268011527377,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004402219828292549,
      "loss": 4.7133,
      "mean_token_accuracy": 0.23937568068504333,
      "num_tokens": 58448113.0,
      "step": 25485
    },
    {
      "entropy": 5.128396129608154,
      "epoch": 2.4486071085494716,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004401987091124388,
      "loss": 4.7809,
      "mean_token_accuracy": 0.2275865614414215,
      "num_tokens": 58459226.0,
      "step": 25490
    },
    {
      "entropy": 5.124465370178223,
      "epoch": 2.4490874159462055,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004401754315601992,
      "loss": 4.7844,
      "mean_token_accuracy": 0.23099374771118164,
      "num_tokens": 58471354.0,
      "step": 25495
    },
    {
      "entropy": 5.110172700881958,
      "epoch": 2.4495677233429394,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004401521501730765,
      "loss": 4.8022,
      "mean_token_accuracy": 0.23507220894098282,
      "num_tokens": 58482859.0,
      "step": 25500
    },
    {
      "entropy": 5.126191759109497,
      "epoch": 2.4500480307396733,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044012886495161144,
      "loss": 4.8171,
      "mean_token_accuracy": 0.22465186864137648,
      "num_tokens": 58496351.0,
      "step": 25505
    },
    {
      "entropy": 5.119567966461181,
      "epoch": 2.450528338136407,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004401055758963443,
      "loss": 4.8106,
      "mean_token_accuracy": 0.23310638219118118,
      "num_tokens": 58508055.0,
      "step": 25510
    },
    {
      "entropy": 5.045698881149292,
      "epoch": 2.451008645533141,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000440082283007816,
      "loss": 4.7657,
      "mean_token_accuracy": 0.2347000703215599,
      "num_tokens": 58519344.0,
      "step": 25515
    },
    {
      "entropy": 5.196641635894776,
      "epoch": 2.451488952929875,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044005898628656734,
      "loss": 4.8507,
      "mean_token_accuracy": 0.22363806813955306,
      "num_tokens": 58530141.0,
      "step": 25520
    },
    {
      "entropy": 5.1229606628417965,
      "epoch": 2.4519692603266092,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000440035685733139,
      "loss": 4.7427,
      "mean_token_accuracy": 0.2366889864206314,
      "num_tokens": 58541527.0,
      "step": 25525
    },
    {
      "entropy": 5.121675777435303,
      "epoch": 2.452449567723343,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004400123813480722,
      "loss": 4.8646,
      "mean_token_accuracy": 0.2331099271774292,
      "num_tokens": 58552205.0,
      "step": 25530
    },
    {
      "entropy": 5.122437763214111,
      "epoch": 2.452929875120077,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043998907313190787,
      "loss": 4.7131,
      "mean_token_accuracy": 0.23764855712652205,
      "num_tokens": 58563563.0,
      "step": 25535
    },
    {
      "entropy": 5.093289422988891,
      "epoch": 2.453410182516811,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004399657610851873,
      "loss": 4.775,
      "mean_token_accuracy": 0.22591578215360641,
      "num_tokens": 58575605.0,
      "step": 25540
    },
    {
      "entropy": 5.131378602981568,
      "epoch": 2.4538904899135447,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043994244520845146,
      "loss": 4.8766,
      "mean_token_accuracy": 0.22445763796567916,
      "num_tokens": 58587761.0,
      "step": 25545
    },
    {
      "entropy": 5.1941510200500485,
      "epoch": 2.4543707973102786,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004399191255022418,
      "loss": 4.7331,
      "mean_token_accuracy": 0.2340010166168213,
      "num_tokens": 58598541.0,
      "step": 25550
    },
    {
      "entropy": 5.11373348236084,
      "epoch": 2.4548511047070125,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004398958019670998,
      "loss": 4.7696,
      "mean_token_accuracy": 0.23594623804092407,
      "num_tokens": 58610833.0,
      "step": 25555
    },
    {
      "entropy": 5.138592100143432,
      "epoch": 2.4553314121037464,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043987247460356696,
      "loss": 4.8042,
      "mean_token_accuracy": 0.2289872944355011,
      "num_tokens": 58620857.0,
      "step": 25560
    },
    {
      "entropy": 5.0673120498657225,
      "epoch": 2.4558117195004803,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004398491434121847,
      "loss": 4.7785,
      "mean_token_accuracy": 0.24006226807832717,
      "num_tokens": 58631245.0,
      "step": 25565
    },
    {
      "entropy": 5.10088152885437,
      "epoch": 2.456292026897214,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000439825808393495,
      "loss": 4.7671,
      "mean_token_accuracy": 0.2364627242088318,
      "num_tokens": 58642995.0,
      "step": 25570
    },
    {
      "entropy": 5.009462118148804,
      "epoch": 2.456772334293948,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004398024695480394,
      "loss": 4.7187,
      "mean_token_accuracy": 0.24038867354393006,
      "num_tokens": 58654817.0,
      "step": 25575
    },
    {
      "entropy": 5.071001863479614,
      "epoch": 2.457252641690682,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004397791268763598,
      "loss": 4.7674,
      "mean_token_accuracy": 0.23346448838710784,
      "num_tokens": 58665912.0,
      "step": 25580
    },
    {
      "entropy": 5.119217348098755,
      "epoch": 2.457732949087416,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043975578037899814,
      "loss": 4.7671,
      "mean_token_accuracy": 0.23359060436487197,
      "num_tokens": 58678080.0,
      "step": 25585
    },
    {
      "entropy": 5.130401945114135,
      "epoch": 2.4582132564841497,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004397324300564966,
      "loss": 4.6968,
      "mean_token_accuracy": 0.23818675726652144,
      "num_tokens": 58689902.0,
      "step": 25590
    },
    {
      "entropy": 5.080589342117309,
      "epoch": 2.4586935638808836,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004397090759093971,
      "loss": 4.7228,
      "mean_token_accuracy": 0.2348733052611351,
      "num_tokens": 58701857.0,
      "step": 25595
    },
    {
      "entropy": 5.079823589324951,
      "epoch": 2.459173871277618,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043968571793824194,
      "loss": 4.7707,
      "mean_token_accuracy": 0.2306866392493248,
      "num_tokens": 58712775.0,
      "step": 25600
    },
    {
      "entropy": 5.053331470489502,
      "epoch": 2.4596541786743518,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004396623561435734,
      "loss": 4.7238,
      "mean_token_accuracy": 0.23509093821048738,
      "num_tokens": 58723583.0,
      "step": 25605
    },
    {
      "entropy": 5.140745401382446,
      "epoch": 2.4601344860710856,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004396389905259339,
      "loss": 4.8359,
      "mean_token_accuracy": 0.23397689014673234,
      "num_tokens": 58734187.0,
      "step": 25610
    },
    {
      "entropy": 5.111422061920166,
      "epoch": 2.4606147934678195,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043961562108586603,
      "loss": 4.7716,
      "mean_token_accuracy": 0.2273463174700737,
      "num_tokens": 58746138.0,
      "step": 25615
    },
    {
      "entropy": 5.133679008483886,
      "epoch": 2.4610951008645534,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043959224782391215,
      "loss": 4.8978,
      "mean_token_accuracy": 0.22781252562999726,
      "num_tokens": 58757765.0,
      "step": 25620
    },
    {
      "entropy": 5.267997694015503,
      "epoch": 2.4615754082612873,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000439568870740615,
      "loss": 4.9573,
      "mean_token_accuracy": 0.209496933221817,
      "num_tokens": 58769957.0,
      "step": 25625
    },
    {
      "entropy": 5.268808460235595,
      "epoch": 2.462055715658021,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004395454898365174,
      "loss": 4.8817,
      "mean_token_accuracy": 0.2285192534327507,
      "num_tokens": 58781019.0,
      "step": 25630
    },
    {
      "entropy": 5.0845808506011965,
      "epoch": 2.462536023054755,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043952210511216205,
      "loss": 4.7955,
      "mean_token_accuracy": 0.23109103888273239,
      "num_tokens": 58793427.0,
      "step": 25635
    },
    {
      "entropy": 5.149553012847901,
      "epoch": 2.463016330451489,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043949871656809205,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22437111139297486,
      "num_tokens": 58804406.0,
      "step": 25640
    },
    {
      "entropy": 5.1259218692779545,
      "epoch": 2.463496637848223,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043947532420485024,
      "loss": 4.7825,
      "mean_token_accuracy": 0.23405844271183013,
      "num_tokens": 58815783.0,
      "step": 25645
    },
    {
      "entropy": 5.114241647720337,
      "epoch": 2.4639769452449567,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004394519280229798,
      "loss": 4.8766,
      "mean_token_accuracy": 0.23240289092063904,
      "num_tokens": 58826968.0,
      "step": 25650
    },
    {
      "entropy": 5.139990282058716,
      "epoch": 2.4644572526416906,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043942852802302397,
      "loss": 4.73,
      "mean_token_accuracy": 0.2382521227002144,
      "num_tokens": 58839090.0,
      "step": 25655
    },
    {
      "entropy": 5.055558395385742,
      "epoch": 2.4649375600384245,
      "grad_norm": 1.0,
      "learning_rate": 0.0004394051242055259,
      "loss": 4.7152,
      "mean_token_accuracy": 0.2340341255068779,
      "num_tokens": 58851636.0,
      "step": 25660
    },
    {
      "entropy": 5.105636835098267,
      "epoch": 2.4654178674351583,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004393817165710291,
      "loss": 4.8077,
      "mean_token_accuracy": 0.2381417855620384,
      "num_tokens": 58863777.0,
      "step": 25665
    },
    {
      "entropy": 5.09263014793396,
      "epoch": 2.465898174831892,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043935830512007687,
      "loss": 4.7916,
      "mean_token_accuracy": 0.2294021874666214,
      "num_tokens": 58875454.0,
      "step": 25670
    },
    {
      "entropy": 5.1145477294921875,
      "epoch": 2.4663784822286265,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00043933488985321286,
      "loss": 4.7419,
      "mean_token_accuracy": 0.23965275287628174,
      "num_tokens": 58887932.0,
      "step": 25675
    },
    {
      "entropy": 5.092795276641846,
      "epoch": 2.46685878962536,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004393114707709807,
      "loss": 4.7877,
      "mean_token_accuracy": 0.22680242508649825,
      "num_tokens": 58900167.0,
      "step": 25680
    },
    {
      "entropy": 5.03729395866394,
      "epoch": 2.4673390970220943,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004392880478739241,
      "loss": 4.708,
      "mean_token_accuracy": 0.24039701074361802,
      "num_tokens": 58911769.0,
      "step": 25685
    },
    {
      "entropy": 5.055427932739258,
      "epoch": 2.467819404418828,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004392646211625869,
      "loss": 4.7798,
      "mean_token_accuracy": 0.23929691910743714,
      "num_tokens": 58923747.0,
      "step": 25690
    },
    {
      "entropy": 5.170050048828125,
      "epoch": 2.468299711815562,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004392411906375129,
      "loss": 4.8043,
      "mean_token_accuracy": 0.23504370003938674,
      "num_tokens": 58934801.0,
      "step": 25695
    },
    {
      "entropy": 5.130199575424195,
      "epoch": 2.468780019212296,
      "grad_norm": 1.0,
      "learning_rate": 0.00043921775629924615,
      "loss": 4.8008,
      "mean_token_accuracy": 0.23775578588247298,
      "num_tokens": 58946566.0,
      "step": 25700
    },
    {
      "entropy": 5.021524667739868,
      "epoch": 2.46926032660903,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043919431814833077,
      "loss": 4.7243,
      "mean_token_accuracy": 0.24153468608856202,
      "num_tokens": 58957066.0,
      "step": 25705
    },
    {
      "entropy": 5.045628786087036,
      "epoch": 2.4697406340057637,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043917087618531084,
      "loss": 4.6876,
      "mean_token_accuracy": 0.24304801374673843,
      "num_tokens": 58967553.0,
      "step": 25710
    },
    {
      "entropy": 5.099869251251221,
      "epoch": 2.4702209414024976,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004391474304107307,
      "loss": 4.7994,
      "mean_token_accuracy": 0.2361670970916748,
      "num_tokens": 58979038.0,
      "step": 25715
    },
    {
      "entropy": 5.138790607452393,
      "epoch": 2.4707012487992315,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043912398082513463,
      "loss": 4.7517,
      "mean_token_accuracy": 0.23147303313016893,
      "num_tokens": 58989564.0,
      "step": 25720
    },
    {
      "entropy": 5.068999195098877,
      "epoch": 2.4711815561959654,
      "grad_norm": 1.0,
      "learning_rate": 0.000439100527429067,
      "loss": 4.7867,
      "mean_token_accuracy": 0.2377003848552704,
      "num_tokens": 59001789.0,
      "step": 25725
    },
    {
      "entropy": 5.082029342651367,
      "epoch": 2.4716618635926992,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043907707022307243,
      "loss": 4.797,
      "mean_token_accuracy": 0.23598384261131286,
      "num_tokens": 59013518.0,
      "step": 25730
    },
    {
      "entropy": 5.082039022445679,
      "epoch": 2.472142170989433,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043905360920769553,
      "loss": 4.6945,
      "mean_token_accuracy": 0.2389338195323944,
      "num_tokens": 59024739.0,
      "step": 25735
    },
    {
      "entropy": 5.082338571548462,
      "epoch": 2.472622478386167,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000439030144383481,
      "loss": 4.7462,
      "mean_token_accuracy": 0.2376638799905777,
      "num_tokens": 59036142.0,
      "step": 25740
    },
    {
      "entropy": 5.119328022003174,
      "epoch": 2.473102785782901,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043900667575097355,
      "loss": 4.8147,
      "mean_token_accuracy": 0.23234879225492477,
      "num_tokens": 59046675.0,
      "step": 25745
    },
    {
      "entropy": 5.082459926605225,
      "epoch": 2.473583093179635,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004389832033107181,
      "loss": 4.7279,
      "mean_token_accuracy": 0.23376445472240448,
      "num_tokens": 59057193.0,
      "step": 25750
    },
    {
      "entropy": 5.053757762908935,
      "epoch": 2.4740634005763686,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043895972706325953,
      "loss": 4.7537,
      "mean_token_accuracy": 0.2340223655104637,
      "num_tokens": 59069501.0,
      "step": 25755
    },
    {
      "entropy": 5.164645671844482,
      "epoch": 2.474543707973103,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000438936247009143,
      "loss": 4.882,
      "mean_token_accuracy": 0.22205574661493302,
      "num_tokens": 59080973.0,
      "step": 25760
    },
    {
      "entropy": 5.071099948883057,
      "epoch": 2.475024015369837,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043891276314891365,
      "loss": 4.7492,
      "mean_token_accuracy": 0.2319207340478897,
      "num_tokens": 59092541.0,
      "step": 25765
    },
    {
      "entropy": 5.041993093490601,
      "epoch": 2.4755043227665707,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004388892754831166,
      "loss": 4.7377,
      "mean_token_accuracy": 0.23146856874227523,
      "num_tokens": 59103936.0,
      "step": 25770
    },
    {
      "entropy": 5.033396053314209,
      "epoch": 2.4759846301633046,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004388657840122973,
      "loss": 4.761,
      "mean_token_accuracy": 0.23518873900175094,
      "num_tokens": 59114714.0,
      "step": 25775
    },
    {
      "entropy": 5.129553651809692,
      "epoch": 2.4764649375600385,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000438842288737001,
      "loss": 4.7664,
      "mean_token_accuracy": 0.23006606847047806,
      "num_tokens": 59126678.0,
      "step": 25780
    },
    {
      "entropy": 5.076951408386231,
      "epoch": 2.4769452449567724,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043881878965777325,
      "loss": 4.7837,
      "mean_token_accuracy": 0.23669403344392775,
      "num_tokens": 59137732.0,
      "step": 25785
    },
    {
      "entropy": 5.065786743164063,
      "epoch": 2.4774255523535063,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043879528677515973,
      "loss": 4.7345,
      "mean_token_accuracy": 0.2324807032942772,
      "num_tokens": 59149231.0,
      "step": 25790
    },
    {
      "entropy": 5.190707445144653,
      "epoch": 2.47790585975024,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043877178008970596,
      "loss": 4.8788,
      "mean_token_accuracy": 0.22177693694829942,
      "num_tokens": 59161033.0,
      "step": 25795
    },
    {
      "entropy": 5.041792583465576,
      "epoch": 2.478386167146974,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004387482696019578,
      "loss": 4.6916,
      "mean_token_accuracy": 0.24278185814619063,
      "num_tokens": 59171906.0,
      "step": 25800
    },
    {
      "entropy": 5.1255138397216795,
      "epoch": 2.478866474543708,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043872475531246105,
      "loss": 4.8704,
      "mean_token_accuracy": 0.22886819541454315,
      "num_tokens": 59185036.0,
      "step": 25805
    },
    {
      "entropy": 5.146501731872559,
      "epoch": 2.479346781940442,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043870123722176166,
      "loss": 4.8392,
      "mean_token_accuracy": 0.22884862273931503,
      "num_tokens": 59196397.0,
      "step": 25810
    },
    {
      "entropy": 5.132103204727173,
      "epoch": 2.4798270893371757,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004386777153304056,
      "loss": 4.8334,
      "mean_token_accuracy": 0.23319937735795976,
      "num_tokens": 59208180.0,
      "step": 25815
    },
    {
      "entropy": 5.235309267044068,
      "epoch": 2.4803073967339095,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043865418963893896,
      "loss": 4.9267,
      "mean_token_accuracy": 0.22410188466310502,
      "num_tokens": 59219651.0,
      "step": 25820
    },
    {
      "entropy": 5.020664930343628,
      "epoch": 2.4807877041306434,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004386306601479081,
      "loss": 4.6647,
      "mean_token_accuracy": 0.2449666291475296,
      "num_tokens": 59231327.0,
      "step": 25825
    },
    {
      "entropy": 5.048181962966919,
      "epoch": 2.4812680115273773,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004386071268578591,
      "loss": 4.7819,
      "mean_token_accuracy": 0.22921958714723586,
      "num_tokens": 59243606.0,
      "step": 25830
    },
    {
      "entropy": 5.169221353530884,
      "epoch": 2.4817483189241116,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043858358976933844,
      "loss": 4.8137,
      "mean_token_accuracy": 0.22859895527362822,
      "num_tokens": 59255824.0,
      "step": 25835
    },
    {
      "entropy": 5.061475563049316,
      "epoch": 2.4822286263208455,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043856004888289264,
      "loss": 4.7093,
      "mean_token_accuracy": 0.2365119904279709,
      "num_tokens": 59267405.0,
      "step": 25840
    },
    {
      "entropy": 5.053994512557983,
      "epoch": 2.4827089337175794,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004385365041990681,
      "loss": 4.7906,
      "mean_token_accuracy": 0.23043682873249055,
      "num_tokens": 59279063.0,
      "step": 25845
    },
    {
      "entropy": 4.931280326843262,
      "epoch": 2.4831892411143133,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004385129557184116,
      "loss": 4.5641,
      "mean_token_accuracy": 0.251629002392292,
      "num_tokens": 59289838.0,
      "step": 25850
    },
    {
      "entropy": 5.035507297515869,
      "epoch": 2.483669548511047,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043848940344146976,
      "loss": 4.7003,
      "mean_token_accuracy": 0.23514868468046188,
      "num_tokens": 59301505.0,
      "step": 25855
    },
    {
      "entropy": 5.005031633377075,
      "epoch": 2.484149855907781,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004384658473687894,
      "loss": 4.6741,
      "mean_token_accuracy": 0.24170637279748916,
      "num_tokens": 59312934.0,
      "step": 25860
    },
    {
      "entropy": 5.103130245208741,
      "epoch": 2.484630163304515,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004384422875009176,
      "loss": 4.7591,
      "mean_token_accuracy": 0.23573997467756272,
      "num_tokens": 59324104.0,
      "step": 25865
    },
    {
      "entropy": 5.1372294425964355,
      "epoch": 2.485110470701249,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004384187238384011,
      "loss": 4.7873,
      "mean_token_accuracy": 0.23286372125148774,
      "num_tokens": 59334978.0,
      "step": 25870
    },
    {
      "entropy": 5.038167238235474,
      "epoch": 2.4855907780979827,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004383951563817871,
      "loss": 4.7125,
      "mean_token_accuracy": 0.23766422569751738,
      "num_tokens": 59345779.0,
      "step": 25875
    },
    {
      "entropy": 5.168970394134521,
      "epoch": 2.4860710854947166,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004383715851316227,
      "loss": 4.8653,
      "mean_token_accuracy": 0.22742612957954406,
      "num_tokens": 59358870.0,
      "step": 25880
    },
    {
      "entropy": 5.22148699760437,
      "epoch": 2.4865513928914504,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043834801008845527,
      "loss": 4.8362,
      "mean_token_accuracy": 0.2313341200351715,
      "num_tokens": 59369831.0,
      "step": 25885
    },
    {
      "entropy": 5.027270889282226,
      "epoch": 2.4870317002881843,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004383244312528321,
      "loss": 4.6773,
      "mean_token_accuracy": 0.24435664266347884,
      "num_tokens": 59381105.0,
      "step": 25890
    },
    {
      "entropy": 5.091629123687744,
      "epoch": 2.487512007684918,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004383008486253006,
      "loss": 4.7751,
      "mean_token_accuracy": 0.23087029606103898,
      "num_tokens": 59391923.0,
      "step": 25895
    },
    {
      "entropy": 5.094354104995728,
      "epoch": 2.487992315081652,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043827726220640827,
      "loss": 4.7352,
      "mean_token_accuracy": 0.23938317447900773,
      "num_tokens": 59402986.0,
      "step": 25900
    },
    {
      "entropy": 5.059929990768433,
      "epoch": 2.488472622478386,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043825367199670274,
      "loss": 4.7572,
      "mean_token_accuracy": 0.2399807319045067,
      "num_tokens": 59413656.0,
      "step": 25905
    },
    {
      "entropy": 5.000602769851684,
      "epoch": 2.4889529298751203,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004382300779967318,
      "loss": 4.6915,
      "mean_token_accuracy": 0.2420386478304863,
      "num_tokens": 59425396.0,
      "step": 25910
    },
    {
      "entropy": 5.1699141502380375,
      "epoch": 2.489433237271854,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043820648020704303,
      "loss": 4.8146,
      "mean_token_accuracy": 0.23585128039121628,
      "num_tokens": 59436859.0,
      "step": 25915
    },
    {
      "entropy": 5.1679497241973875,
      "epoch": 2.489913544668588,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043818287862818444,
      "loss": 4.857,
      "mean_token_accuracy": 0.23326811343431472,
      "num_tokens": 59448415.0,
      "step": 25920
    },
    {
      "entropy": 5.025774717330933,
      "epoch": 2.490393852065322,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000438159273260704,
      "loss": 4.6585,
      "mean_token_accuracy": 0.2442651003599167,
      "num_tokens": 59459477.0,
      "step": 25925
    },
    {
      "entropy": 4.988399934768677,
      "epoch": 2.490874159462056,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004381356641051497,
      "loss": 4.7085,
      "mean_token_accuracy": 0.24467783272266388,
      "num_tokens": 59471184.0,
      "step": 25930
    },
    {
      "entropy": 5.068174219131469,
      "epoch": 2.4913544668587897,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004381120511620697,
      "loss": 4.728,
      "mean_token_accuracy": 0.2401551067829132,
      "num_tokens": 59482706.0,
      "step": 25935
    },
    {
      "entropy": 5.083991813659668,
      "epoch": 2.4918347742555236,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043808843443201217,
      "loss": 4.7716,
      "mean_token_accuracy": 0.2327693909406662,
      "num_tokens": 59493822.0,
      "step": 25940
    },
    {
      "entropy": 5.088975811004639,
      "epoch": 2.4923150816522575,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004380648139155255,
      "loss": 4.7039,
      "mean_token_accuracy": 0.2357386976480484,
      "num_tokens": 59504577.0,
      "step": 25945
    },
    {
      "entropy": 5.08472785949707,
      "epoch": 2.4927953890489913,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004380411896131581,
      "loss": 4.7703,
      "mean_token_accuracy": 0.22917639315128327,
      "num_tokens": 59517395.0,
      "step": 25950
    },
    {
      "entropy": 5.148755693435669,
      "epoch": 2.493275696445725,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043801756152545836,
      "loss": 4.7769,
      "mean_token_accuracy": 0.23621094822883607,
      "num_tokens": 59529270.0,
      "step": 25955
    },
    {
      "entropy": 5.1277303218841555,
      "epoch": 2.493756003842459,
      "grad_norm": 1.0,
      "learning_rate": 0.00043799392965297496,
      "loss": 4.766,
      "mean_token_accuracy": 0.23299470394849778,
      "num_tokens": 59540866.0,
      "step": 25960
    },
    {
      "entropy": 5.1604410171508786,
      "epoch": 2.494236311239193,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004379702939962564,
      "loss": 4.8704,
      "mean_token_accuracy": 0.22758614271879196,
      "num_tokens": 59553201.0,
      "step": 25965
    },
    {
      "entropy": 5.060476064682007,
      "epoch": 2.494716618635927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004379466545558516,
      "loss": 4.7294,
      "mean_token_accuracy": 0.24059762954711914,
      "num_tokens": 59563054.0,
      "step": 25970
    },
    {
      "entropy": 5.0093017578125,
      "epoch": 2.4951969260326607,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043792301133230933,
      "loss": 4.6962,
      "mean_token_accuracy": 0.233958438038826,
      "num_tokens": 59574849.0,
      "step": 25975
    },
    {
      "entropy": 5.000476741790772,
      "epoch": 2.4956772334293946,
      "grad_norm": 1.125,
      "learning_rate": 0.0004378993643261785,
      "loss": 4.66,
      "mean_token_accuracy": 0.2410842090845108,
      "num_tokens": 59585567.0,
      "step": 25980
    },
    {
      "entropy": 5.10311598777771,
      "epoch": 2.496157540826129,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043787571353800814,
      "loss": 4.7531,
      "mean_token_accuracy": 0.24289715439081191,
      "num_tokens": 59596810.0,
      "step": 25985
    },
    {
      "entropy": 5.0336995124816895,
      "epoch": 2.4966378482228624,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004378520589683475,
      "loss": 4.7654,
      "mean_token_accuracy": 0.23552702069282533,
      "num_tokens": 59608704.0,
      "step": 25990
    },
    {
      "entropy": 5.047534036636352,
      "epoch": 2.4971181556195967,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043782840061774544,
      "loss": 4.675,
      "mean_token_accuracy": 0.24850208163261414,
      "num_tokens": 59619325.0,
      "step": 25995
    },
    {
      "entropy": 5.156967449188232,
      "epoch": 2.4975984630163306,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043780473848675143,
      "loss": 4.8622,
      "mean_token_accuracy": 0.23040475100278854,
      "num_tokens": 59630181.0,
      "step": 26000
    },
    {
      "entropy": 5.06505970954895,
      "epoch": 2.4980787704130645,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004377810725759149,
      "loss": 4.7784,
      "mean_token_accuracy": 0.22555986195802688,
      "num_tokens": 59642722.0,
      "step": 26005
    },
    {
      "entropy": 5.130134201049804,
      "epoch": 2.4985590778097984,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043775740288578516,
      "loss": 4.7798,
      "mean_token_accuracy": 0.23086844086647035,
      "num_tokens": 59653675.0,
      "step": 26010
    },
    {
      "entropy": 5.184849739074707,
      "epoch": 2.4990393852065322,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004377337294169118,
      "loss": 4.8745,
      "mean_token_accuracy": 0.22333396971225739,
      "num_tokens": 59664897.0,
      "step": 26015
    },
    {
      "entropy": 5.0594089984893795,
      "epoch": 2.499519692603266,
      "grad_norm": 1.0,
      "learning_rate": 0.00043771005216984457,
      "loss": 4.7231,
      "mean_token_accuracy": 0.23794975578784944,
      "num_tokens": 59676554.0,
      "step": 26020
    },
    {
      "entropy": 5.122758388519287,
      "epoch": 2.5,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000437686371145133,
      "loss": 4.8438,
      "mean_token_accuracy": 0.2225934237241745,
      "num_tokens": 59688030.0,
      "step": 26025
    },
    {
      "entropy": 5.095273494720459,
      "epoch": 2.500480307396734,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000437662686343327,
      "loss": 4.7405,
      "mean_token_accuracy": 0.2351490467786789,
      "num_tokens": 59699721.0,
      "step": 26030
    },
    {
      "entropy": 5.066417169570923,
      "epoch": 2.5009606147934678,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004376389977649764,
      "loss": 4.7134,
      "mean_token_accuracy": 0.23639643490314483,
      "num_tokens": 59710426.0,
      "step": 26035
    },
    {
      "entropy": 5.0053812026977536,
      "epoch": 2.5014409221902016,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004376153054106313,
      "loss": 4.6974,
      "mean_token_accuracy": 0.2386977642774582,
      "num_tokens": 59720837.0,
      "step": 26040
    },
    {
      "entropy": 5.090166473388672,
      "epoch": 2.5019212295869355,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004375916092808416,
      "loss": 4.7937,
      "mean_token_accuracy": 0.23464784622192383,
      "num_tokens": 59732045.0,
      "step": 26045
    },
    {
      "entropy": 5.134308004379273,
      "epoch": 2.5024015369836694,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004375679093761575,
      "loss": 4.7736,
      "mean_token_accuracy": 0.22946055084466935,
      "num_tokens": 59744735.0,
      "step": 26050
    },
    {
      "entropy": 5.093188619613647,
      "epoch": 2.5028818443804033,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043754420569712925,
      "loss": 4.7075,
      "mean_token_accuracy": 0.23653424978256227,
      "num_tokens": 59755604.0,
      "step": 26055
    },
    {
      "entropy": 5.051211786270142,
      "epoch": 2.5033621517771376,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043752049824430736,
      "loss": 4.7414,
      "mean_token_accuracy": 0.23855720311403275,
      "num_tokens": 59765892.0,
      "step": 26060
    },
    {
      "entropy": 5.053389978408814,
      "epoch": 2.503842459173871,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043749678701824197,
      "loss": 4.7621,
      "mean_token_accuracy": 0.22930600345134736,
      "num_tokens": 59779068.0,
      "step": 26065
    },
    {
      "entropy": 5.166787576675415,
      "epoch": 2.5043227665706054,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004374730720194837,
      "loss": 4.8104,
      "mean_token_accuracy": 0.2403150275349617,
      "num_tokens": 59789894.0,
      "step": 26070
    },
    {
      "entropy": 5.017841100692749,
      "epoch": 2.5048030739673393,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004374493532485832,
      "loss": 4.7195,
      "mean_token_accuracy": 0.24235603958368301,
      "num_tokens": 59801384.0,
      "step": 26075
    },
    {
      "entropy": 5.0752510070800785,
      "epoch": 2.505283381364073,
      "grad_norm": 1.125,
      "learning_rate": 0.000437425630706091,
      "loss": 4.7673,
      "mean_token_accuracy": 0.23451309353113176,
      "num_tokens": 59813471.0,
      "step": 26080
    },
    {
      "entropy": 5.179547119140625,
      "epoch": 2.505763688760807,
      "grad_norm": 1.015625,
      "learning_rate": 0.000437401904392558,
      "loss": 4.8244,
      "mean_token_accuracy": 0.2328708916902542,
      "num_tokens": 59824495.0,
      "step": 26085
    },
    {
      "entropy": 5.188300752639771,
      "epoch": 2.506243996157541,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043737817430853504,
      "loss": 4.8365,
      "mean_token_accuracy": 0.2258144125342369,
      "num_tokens": 59835588.0,
      "step": 26090
    },
    {
      "entropy": 5.120651006698608,
      "epoch": 2.506724303554275,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043735444045457303,
      "loss": 4.7973,
      "mean_token_accuracy": 0.22862798869609832,
      "num_tokens": 59847093.0,
      "step": 26095
    },
    {
      "entropy": 5.074661064147949,
      "epoch": 2.5072046109510087,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043733070283122306,
      "loss": 4.6699,
      "mean_token_accuracy": 0.2424640476703644,
      "num_tokens": 59858109.0,
      "step": 26100
    },
    {
      "entropy": 5.024330472946167,
      "epoch": 2.5076849183477425,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043730696143903607,
      "loss": 4.7177,
      "mean_token_accuracy": 0.23278080970048903,
      "num_tokens": 59870161.0,
      "step": 26105
    },
    {
      "entropy": 5.1049049377441404,
      "epoch": 2.5081652257444764,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004372832162785635,
      "loss": 4.7693,
      "mean_token_accuracy": 0.23215897530317306,
      "num_tokens": 59882068.0,
      "step": 26110
    },
    {
      "entropy": 5.156821823120117,
      "epoch": 2.5086455331412103,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004372594673503565,
      "loss": 4.7267,
      "mean_token_accuracy": 0.2407372400164604,
      "num_tokens": 59892418.0,
      "step": 26115
    },
    {
      "entropy": 5.0893913269042965,
      "epoch": 2.509125840537944,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004372357146549664,
      "loss": 4.7845,
      "mean_token_accuracy": 0.2369270518422127,
      "num_tokens": 59902790.0,
      "step": 26120
    },
    {
      "entropy": 5.031317043304443,
      "epoch": 2.509606147934678,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043721195819294487,
      "loss": 4.744,
      "mean_token_accuracy": 0.23270730376243592,
      "num_tokens": 59914982.0,
      "step": 26125
    },
    {
      "entropy": 5.118412017822266,
      "epoch": 2.510086455331412,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004371881979648433,
      "loss": 4.8686,
      "mean_token_accuracy": 0.22457899600267411,
      "num_tokens": 59925961.0,
      "step": 26130
    },
    {
      "entropy": 5.1303483009338375,
      "epoch": 2.5105667627281463,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004371644339712133,
      "loss": 4.8142,
      "mean_token_accuracy": 0.23506382256746292,
      "num_tokens": 59938709.0,
      "step": 26135
    },
    {
      "entropy": 5.08832688331604,
      "epoch": 2.5110470701248797,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004371406662126067,
      "loss": 4.7244,
      "mean_token_accuracy": 0.23736357539892197,
      "num_tokens": 59949981.0,
      "step": 26140
    },
    {
      "entropy": 5.035450983047485,
      "epoch": 2.511527377521614,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043711689468957534,
      "loss": 4.6767,
      "mean_token_accuracy": 0.2443459615111351,
      "num_tokens": 59961741.0,
      "step": 26145
    },
    {
      "entropy": 5.008074808120727,
      "epoch": 2.5120076849183475,
      "grad_norm": 1.1875,
      "learning_rate": 0.00043709311940267107,
      "loss": 4.7293,
      "mean_token_accuracy": 0.23666201829910277,
      "num_tokens": 59973399.0,
      "step": 26150
    },
    {
      "entropy": 5.072469329833984,
      "epoch": 2.512487992315082,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004370693403524458,
      "loss": 4.7237,
      "mean_token_accuracy": 0.24049130082130432,
      "num_tokens": 59985180.0,
      "step": 26155
    },
    {
      "entropy": 5.134912586212158,
      "epoch": 2.5129682997118157,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004370455575394518,
      "loss": 4.7524,
      "mean_token_accuracy": 0.23687663972377776,
      "num_tokens": 59995391.0,
      "step": 26160
    },
    {
      "entropy": 5.030692100524902,
      "epoch": 2.5134486071085496,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004370217709642411,
      "loss": 4.7794,
      "mean_token_accuracy": 0.23255509734153748,
      "num_tokens": 60006092.0,
      "step": 26165
    },
    {
      "entropy": 5.114044237136841,
      "epoch": 2.5139289145052834,
      "grad_norm": 0.9375,
      "learning_rate": 0.000436997980627366,
      "loss": 4.8468,
      "mean_token_accuracy": 0.23392362147569656,
      "num_tokens": 60017615.0,
      "step": 26170
    },
    {
      "entropy": 5.144384717941284,
      "epoch": 2.5144092219020173,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043697418652937877,
      "loss": 4.7631,
      "mean_token_accuracy": 0.22936453223228453,
      "num_tokens": 60030062.0,
      "step": 26175
    },
    {
      "entropy": 5.066312265396118,
      "epoch": 2.514889529298751,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004369503886708319,
      "loss": 4.7402,
      "mean_token_accuracy": 0.23545787185430528,
      "num_tokens": 60042188.0,
      "step": 26180
    },
    {
      "entropy": 5.107606315612793,
      "epoch": 2.515369836695485,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043692658705227796,
      "loss": 4.802,
      "mean_token_accuracy": 0.2373108446598053,
      "num_tokens": 60054532.0,
      "step": 26185
    },
    {
      "entropy": 5.199697017669678,
      "epoch": 2.515850144092219,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043690278167426945,
      "loss": 4.8954,
      "mean_token_accuracy": 0.22837022095918655,
      "num_tokens": 60066026.0,
      "step": 26190
    },
    {
      "entropy": 5.087201070785523,
      "epoch": 2.516330451488953,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004368789725373591,
      "loss": 4.6893,
      "mean_token_accuracy": 0.24281687289476395,
      "num_tokens": 60079679.0,
      "step": 26195
    },
    {
      "entropy": 5.092996644973755,
      "epoch": 2.5168107588856867,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043685515964209977,
      "loss": 4.7854,
      "mean_token_accuracy": 0.23427806347608565,
      "num_tokens": 60089971.0,
      "step": 26200
    },
    {
      "entropy": 5.037321472167969,
      "epoch": 2.5172910662824206,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004368313429890441,
      "loss": 4.6966,
      "mean_token_accuracy": 0.23830353915691377,
      "num_tokens": 60100200.0,
      "step": 26205
    },
    {
      "entropy": 5.109868717193604,
      "epoch": 2.517771373679155,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004368075225787453,
      "loss": 4.7532,
      "mean_token_accuracy": 0.23272859007120134,
      "num_tokens": 60112031.0,
      "step": 26210
    },
    {
      "entropy": 5.0705207824707035,
      "epoch": 2.5182516810758884,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004367836984117562,
      "loss": 4.7312,
      "mean_token_accuracy": 0.24110322594642639,
      "num_tokens": 60122233.0,
      "step": 26215
    },
    {
      "entropy": 5.052870035171509,
      "epoch": 2.5187319884726227,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043675987048863,
      "loss": 4.717,
      "mean_token_accuracy": 0.2405463919043541,
      "num_tokens": 60133144.0,
      "step": 26220
    },
    {
      "entropy": 5.088471841812134,
      "epoch": 2.519212295869356,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004367360388099201,
      "loss": 4.8145,
      "mean_token_accuracy": 0.23353252112865447,
      "num_tokens": 60144052.0,
      "step": 26225
    },
    {
      "entropy": 5.1301658153533936,
      "epoch": 2.5196926032660905,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004367122033761796,
      "loss": 4.7773,
      "mean_token_accuracy": 0.2302298367023468,
      "num_tokens": 60155333.0,
      "step": 26230
    },
    {
      "entropy": 5.08112325668335,
      "epoch": 2.5201729106628243,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004366883641879618,
      "loss": 4.7469,
      "mean_token_accuracy": 0.23786429166793824,
      "num_tokens": 60166473.0,
      "step": 26235
    },
    {
      "entropy": 4.918370246887207,
      "epoch": 2.520653218059558,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043666452124582034,
      "loss": 4.6048,
      "mean_token_accuracy": 0.24657426327466964,
      "num_tokens": 60177914.0,
      "step": 26240
    },
    {
      "entropy": 5.063231515884399,
      "epoch": 2.521133525456292,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004366406745503088,
      "loss": 4.8096,
      "mean_token_accuracy": 0.23611213266849518,
      "num_tokens": 60190414.0,
      "step": 26245
    },
    {
      "entropy": 5.214793682098389,
      "epoch": 2.521613832853026,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004366168241019807,
      "loss": 4.9411,
      "mean_token_accuracy": 0.2186445713043213,
      "num_tokens": 60202686.0,
      "step": 26250
    },
    {
      "entropy": 5.1330914974212645,
      "epoch": 2.52209414024976,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004365929699013899,
      "loss": 4.7502,
      "mean_token_accuracy": 0.23110028058290483,
      "num_tokens": 60214244.0,
      "step": 26255
    },
    {
      "entropy": 5.088634634017945,
      "epoch": 2.5225744476464937,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004365691119490902,
      "loss": 4.7457,
      "mean_token_accuracy": 0.23977120071649552,
      "num_tokens": 60225057.0,
      "step": 26260
    },
    {
      "entropy": 5.12257833480835,
      "epoch": 2.5230547550432276,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004365452502456354,
      "loss": 4.842,
      "mean_token_accuracy": 0.22596798688173295,
      "num_tokens": 60236060.0,
      "step": 26265
    },
    {
      "entropy": 5.124188470840454,
      "epoch": 2.5235350624399615,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004365213847915796,
      "loss": 4.7518,
      "mean_token_accuracy": 0.23500738441944122,
      "num_tokens": 60247493.0,
      "step": 26270
    },
    {
      "entropy": 5.000863409042358,
      "epoch": 2.5240153698366954,
      "grad_norm": 1.0,
      "learning_rate": 0.00043649751558747695,
      "loss": 4.616,
      "mean_token_accuracy": 0.24898105561733247,
      "num_tokens": 60258525.0,
      "step": 26275
    },
    {
      "entropy": 5.086121082305908,
      "epoch": 2.5244956772334293,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043647364263388143,
      "loss": 4.7402,
      "mean_token_accuracy": 0.235128752887249,
      "num_tokens": 60270466.0,
      "step": 26280
    },
    {
      "entropy": 5.106005477905273,
      "epoch": 2.524975984630163,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004364497659313475,
      "loss": 4.7356,
      "mean_token_accuracy": 0.23440536707639695,
      "num_tokens": 60282078.0,
      "step": 26285
    },
    {
      "entropy": 5.119178342819214,
      "epoch": 2.525456292026897,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004364258854804294,
      "loss": 4.7429,
      "mean_token_accuracy": 0.2262744352221489,
      "num_tokens": 60293569.0,
      "step": 26290
    },
    {
      "entropy": 4.998501348495483,
      "epoch": 2.5259365994236314,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004364020012816815,
      "loss": 4.6859,
      "mean_token_accuracy": 0.23972582221031188,
      "num_tokens": 60305840.0,
      "step": 26295
    },
    {
      "entropy": 5.092244434356689,
      "epoch": 2.526416906820365,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004363781133356584,
      "loss": 4.7743,
      "mean_token_accuracy": 0.2365038201212883,
      "num_tokens": 60318519.0,
      "step": 26300
    },
    {
      "entropy": 4.974600791931152,
      "epoch": 2.526897214217099,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004363542216429147,
      "loss": 4.6512,
      "mean_token_accuracy": 0.2425345078110695,
      "num_tokens": 60328984.0,
      "step": 26305
    },
    {
      "entropy": 5.187969446182251,
      "epoch": 2.527377521613833,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004363303262040051,
      "loss": 4.927,
      "mean_token_accuracy": 0.22392195016145705,
      "num_tokens": 60339846.0,
      "step": 26310
    },
    {
      "entropy": 5.075377798080444,
      "epoch": 2.527857829010567,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043630642701948446,
      "loss": 4.7868,
      "mean_token_accuracy": 0.23351622819900514,
      "num_tokens": 60351336.0,
      "step": 26315
    },
    {
      "entropy": 4.99967942237854,
      "epoch": 2.5283381364073008,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043628252408990756,
      "loss": 4.6784,
      "mean_token_accuracy": 0.24314217120409012,
      "num_tokens": 60362794.0,
      "step": 26320
    },
    {
      "entropy": 5.060900354385376,
      "epoch": 2.5288184438040346,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043625861741582926,
      "loss": 4.7356,
      "mean_token_accuracy": 0.23782730847597122,
      "num_tokens": 60374176.0,
      "step": 26325
    },
    {
      "entropy": 4.995164585113526,
      "epoch": 2.5292987512007685,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043623470699780483,
      "loss": 4.7049,
      "mean_token_accuracy": 0.24306674152612687,
      "num_tokens": 60385414.0,
      "step": 26330
    },
    {
      "entropy": 5.06729040145874,
      "epoch": 2.5297790585975024,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004362107928363892,
      "loss": 4.7239,
      "mean_token_accuracy": 0.23631048649549485,
      "num_tokens": 60397738.0,
      "step": 26335
    },
    {
      "entropy": 5.006624221801758,
      "epoch": 2.5302593659942363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004361868749321377,
      "loss": 4.6644,
      "mean_token_accuracy": 0.23926791846752166,
      "num_tokens": 60408554.0,
      "step": 26340
    },
    {
      "entropy": 5.0375893115997314,
      "epoch": 2.53073967339097,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004361629532856055,
      "loss": 4.7033,
      "mean_token_accuracy": 0.23968394845724106,
      "num_tokens": 60420000.0,
      "step": 26345
    },
    {
      "entropy": 5.079583311080933,
      "epoch": 2.531219980787704,
      "grad_norm": 1.296875,
      "learning_rate": 0.00043613902789734816,
      "loss": 4.6497,
      "mean_token_accuracy": 0.24417918920516968,
      "num_tokens": 60431639.0,
      "step": 26350
    },
    {
      "entropy": 5.077229309082031,
      "epoch": 2.531700288184438,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000436115098767921,
      "loss": 4.8228,
      "mean_token_accuracy": 0.22979058176279069,
      "num_tokens": 60442401.0,
      "step": 26355
    },
    {
      "entropy": 5.152002620697021,
      "epoch": 2.532180595581172,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043609116589787974,
      "loss": 4.8599,
      "mean_token_accuracy": 0.23126125037670137,
      "num_tokens": 60455530.0,
      "step": 26360
    },
    {
      "entropy": 5.172606039047241,
      "epoch": 2.5326609029779057,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004360672292877799,
      "loss": 4.8526,
      "mean_token_accuracy": 0.2274441674351692,
      "num_tokens": 60466380.0,
      "step": 26365
    },
    {
      "entropy": 5.059221601486206,
      "epoch": 2.53314121037464,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043604328893817726,
      "loss": 4.7556,
      "mean_token_accuracy": 0.23823178112506865,
      "num_tokens": 60477309.0,
      "step": 26370
    },
    {
      "entropy": 5.042608165740967,
      "epoch": 2.5336215177713735,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043601934484962775,
      "loss": 4.7948,
      "mean_token_accuracy": 0.2289544850587845,
      "num_tokens": 60489181.0,
      "step": 26375
    },
    {
      "entropy": 5.135474777221679,
      "epoch": 2.534101825168108,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004359953970226871,
      "loss": 4.7625,
      "mean_token_accuracy": 0.23967625796794892,
      "num_tokens": 60501994.0,
      "step": 26380
    },
    {
      "entropy": 5.077267742156982,
      "epoch": 2.5345821325648417,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043597144545791134,
      "loss": 4.6708,
      "mean_token_accuracy": 0.239340677857399,
      "num_tokens": 60512769.0,
      "step": 26385
    },
    {
      "entropy": 5.096933794021607,
      "epoch": 2.5350624399615755,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004359474901558567,
      "loss": 4.7548,
      "mean_token_accuracy": 0.23242665380239486,
      "num_tokens": 60523965.0,
      "step": 26390
    },
    {
      "entropy": 5.062889623641968,
      "epoch": 2.5355427473583094,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004359235311170792,
      "loss": 4.7344,
      "mean_token_accuracy": 0.23619223535060882,
      "num_tokens": 60534892.0,
      "step": 26395
    },
    {
      "entropy": 5.161354827880859,
      "epoch": 2.5360230547550433,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004358995683421352,
      "loss": 4.7633,
      "mean_token_accuracy": 0.24060899168252944,
      "num_tokens": 60545744.0,
      "step": 26400
    },
    {
      "entropy": 4.93488392829895,
      "epoch": 2.536503362151777,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043587560183158095,
      "loss": 4.6195,
      "mean_token_accuracy": 0.24342428296804428,
      "num_tokens": 60556830.0,
      "step": 26405
    },
    {
      "entropy": 5.035302782058716,
      "epoch": 2.536983669548511,
      "grad_norm": 1.078125,
      "learning_rate": 0.000435851631585973,
      "loss": 4.7728,
      "mean_token_accuracy": 0.22826349288225173,
      "num_tokens": 60567360.0,
      "step": 26410
    },
    {
      "entropy": 5.111351346969604,
      "epoch": 2.537463976945245,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004358276576058677,
      "loss": 4.7751,
      "mean_token_accuracy": 0.23481452018022536,
      "num_tokens": 60577915.0,
      "step": 26415
    },
    {
      "entropy": 5.082102584838867,
      "epoch": 2.537944284341979,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004358036798918218,
      "loss": 4.7602,
      "mean_token_accuracy": 0.2304319828748703,
      "num_tokens": 60589160.0,
      "step": 26420
    },
    {
      "entropy": 5.050460624694824,
      "epoch": 2.5384245917387127,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004357796984443919,
      "loss": 4.7648,
      "mean_token_accuracy": 0.2334355279803276,
      "num_tokens": 60599573.0,
      "step": 26425
    },
    {
      "entropy": 5.073406982421875,
      "epoch": 2.5389048991354466,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043575571326413484,
      "loss": 4.7516,
      "mean_token_accuracy": 0.23805388808250427,
      "num_tokens": 60610886.0,
      "step": 26430
    },
    {
      "entropy": 5.131623601913452,
      "epoch": 2.5393852065321805,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004357317243516075,
      "loss": 4.793,
      "mean_token_accuracy": 0.2362929791212082,
      "num_tokens": 60623056.0,
      "step": 26435
    },
    {
      "entropy": 5.075936698913575,
      "epoch": 2.5398655139289144,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043570773170736676,
      "loss": 4.7472,
      "mean_token_accuracy": 0.23712879419326782,
      "num_tokens": 60634358.0,
      "step": 26440
    },
    {
      "entropy": 5.060122394561768,
      "epoch": 2.5403458213256487,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043568373533196976,
      "loss": 4.794,
      "mean_token_accuracy": 0.23575481325387954,
      "num_tokens": 60644611.0,
      "step": 26445
    },
    {
      "entropy": 5.074518442153931,
      "epoch": 2.540826128722382,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043565973522597344,
      "loss": 4.7957,
      "mean_token_accuracy": 0.2318127155303955,
      "num_tokens": 60655873.0,
      "step": 26450
    },
    {
      "entropy": 5.164845609664917,
      "epoch": 2.5413064361191164,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043563573138993524,
      "loss": 4.8363,
      "mean_token_accuracy": 0.2336576223373413,
      "num_tokens": 60668634.0,
      "step": 26455
    },
    {
      "entropy": 5.190935659408569,
      "epoch": 2.54178674351585,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004356117238244123,
      "loss": 4.8601,
      "mean_token_accuracy": 0.22981422543525695,
      "num_tokens": 60679460.0,
      "step": 26460
    },
    {
      "entropy": 5.067248201370239,
      "epoch": 2.542267050912584,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043558771252996204,
      "loss": 4.7632,
      "mean_token_accuracy": 0.24054625481367112,
      "num_tokens": 60689704.0,
      "step": 26465
    },
    {
      "entropy": 5.1145727157592775,
      "epoch": 2.542747358309318,
      "grad_norm": 1.015625,
      "learning_rate": 0.000435563697507142,
      "loss": 4.7562,
      "mean_token_accuracy": 0.234867624938488,
      "num_tokens": 60700838.0,
      "step": 26470
    },
    {
      "entropy": 5.089047193527222,
      "epoch": 2.543227665706052,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004355396787565096,
      "loss": 4.7568,
      "mean_token_accuracy": 0.23120069354772568,
      "num_tokens": 60712048.0,
      "step": 26475
    },
    {
      "entropy": 5.0163373947143555,
      "epoch": 2.543707973102786,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043551565627862257,
      "loss": 4.6685,
      "mean_token_accuracy": 0.24924635887145996,
      "num_tokens": 60723451.0,
      "step": 26480
    },
    {
      "entropy": 5.006580066680908,
      "epoch": 2.5441882804995197,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004354916300740387,
      "loss": 4.7204,
      "mean_token_accuracy": 0.24999732673168182,
      "num_tokens": 60735992.0,
      "step": 26485
    },
    {
      "entropy": 5.1205216407775875,
      "epoch": 2.5446685878962536,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004354676001433157,
      "loss": 4.7393,
      "mean_token_accuracy": 0.23482909500598909,
      "num_tokens": 60746653.0,
      "step": 26490
    },
    {
      "entropy": 5.0684116840362545,
      "epoch": 2.5451488952929875,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004354435664870116,
      "loss": 4.76,
      "mean_token_accuracy": 0.23244809359312057,
      "num_tokens": 60757884.0,
      "step": 26495
    },
    {
      "entropy": 5.061397314071655,
      "epoch": 2.5456292026897214,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043541952910568417,
      "loss": 4.7021,
      "mean_token_accuracy": 0.24131180197000504,
      "num_tokens": 60769243.0,
      "step": 26500
    },
    {
      "entropy": 5.137715673446655,
      "epoch": 2.5461095100864553,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004353954879998916,
      "loss": 4.879,
      "mean_token_accuracy": 0.2243967607617378,
      "num_tokens": 60781168.0,
      "step": 26505
    },
    {
      "entropy": 5.088244390487671,
      "epoch": 2.546589817483189,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004353714431701922,
      "loss": 4.7353,
      "mean_token_accuracy": 0.24488245844841003,
      "num_tokens": 60792743.0,
      "step": 26510
    },
    {
      "entropy": 5.044963836669922,
      "epoch": 2.547070124879923,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004353473946171441,
      "loss": 4.6679,
      "mean_token_accuracy": 0.2428569421172142,
      "num_tokens": 60804251.0,
      "step": 26515
    },
    {
      "entropy": 5.104335975646973,
      "epoch": 2.547550432276657,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043532334234130547,
      "loss": 4.7778,
      "mean_token_accuracy": 0.23514495193958282,
      "num_tokens": 60815567.0,
      "step": 26520
    },
    {
      "entropy": 5.05094838142395,
      "epoch": 2.5480307396733908,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043529928634323503,
      "loss": 4.7878,
      "mean_token_accuracy": 0.23000932335853577,
      "num_tokens": 60827028.0,
      "step": 26525
    },
    {
      "entropy": 5.007453346252442,
      "epoch": 2.548511047070125,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043527522662349113,
      "loss": 4.6764,
      "mean_token_accuracy": 0.24555480033159255,
      "num_tokens": 60839158.0,
      "step": 26530
    },
    {
      "entropy": 5.152102088928222,
      "epoch": 2.5489913544668585,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004352511631826324,
      "loss": 4.8428,
      "mean_token_accuracy": 0.22248231768608093,
      "num_tokens": 60850022.0,
      "step": 26535
    },
    {
      "entropy": 5.033304357528687,
      "epoch": 2.549471661863593,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004352270960212175,
      "loss": 4.6658,
      "mean_token_accuracy": 0.2389891266822815,
      "num_tokens": 60861259.0,
      "step": 26540
    },
    {
      "entropy": 5.181062412261963,
      "epoch": 2.5499519692603267,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004352030251398052,
      "loss": 4.7369,
      "mean_token_accuracy": 0.23330518901348113,
      "num_tokens": 60871838.0,
      "step": 26545
    },
    {
      "entropy": 5.039798974990845,
      "epoch": 2.5504322766570606,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043517895053895434,
      "loss": 4.7001,
      "mean_token_accuracy": 0.24335480630397796,
      "num_tokens": 60883939.0,
      "step": 26550
    },
    {
      "entropy": 5.1065668106079105,
      "epoch": 2.5509125840537945,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004351548722192239,
      "loss": 4.8083,
      "mean_token_accuracy": 0.23062790781259537,
      "num_tokens": 60895799.0,
      "step": 26555
    },
    {
      "entropy": 5.102820777893067,
      "epoch": 2.5513928914505284,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004351307901811729,
      "loss": 4.6864,
      "mean_token_accuracy": 0.23638022392988206,
      "num_tokens": 60906330.0,
      "step": 26560
    },
    {
      "entropy": 5.034135150909424,
      "epoch": 2.5518731988472623,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004351067044253605,
      "loss": 4.7094,
      "mean_token_accuracy": 0.23628147542476655,
      "num_tokens": 60917613.0,
      "step": 26565
    },
    {
      "entropy": 5.087821912765503,
      "epoch": 2.552353506243996,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043508261495234577,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2272471770644188,
      "num_tokens": 60929116.0,
      "step": 26570
    },
    {
      "entropy": 5.037799930572509,
      "epoch": 2.55283381364073,
      "grad_norm": 0.890625,
      "learning_rate": 0.000435058521762688,
      "loss": 4.6207,
      "mean_token_accuracy": 0.24015882164239882,
      "num_tokens": 60940455.0,
      "step": 26575
    },
    {
      "entropy": 5.095242738723755,
      "epoch": 2.553314121037464,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004350344248569467,
      "loss": 4.8529,
      "mean_token_accuracy": 0.2301182210445404,
      "num_tokens": 60953606.0,
      "step": 26580
    },
    {
      "entropy": 5.078183937072754,
      "epoch": 2.553794428434198,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004350103242356813,
      "loss": 4.7426,
      "mean_token_accuracy": 0.23629889041185378,
      "num_tokens": 60964487.0,
      "step": 26585
    },
    {
      "entropy": 5.20848536491394,
      "epoch": 2.5542747358309317,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004349862198994513,
      "loss": 4.8916,
      "mean_token_accuracy": 0.22890950441360475,
      "num_tokens": 60976249.0,
      "step": 26590
    },
    {
      "entropy": 5.148565626144409,
      "epoch": 2.5547550432276656,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004349621118488163,
      "loss": 4.7529,
      "mean_token_accuracy": 0.2371564120054245,
      "num_tokens": 60987581.0,
      "step": 26595
    },
    {
      "entropy": 5.058470010757446,
      "epoch": 2.5552353506243994,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004349380000843361,
      "loss": 4.7542,
      "mean_token_accuracy": 0.2341224804520607,
      "num_tokens": 60999286.0,
      "step": 26600
    },
    {
      "entropy": 4.996645593643189,
      "epoch": 2.5557156580211338,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004349138846065704,
      "loss": 4.6631,
      "mean_token_accuracy": 0.24514297246932984,
      "num_tokens": 61010207.0,
      "step": 26605
    },
    {
      "entropy": 5.0348193645477295,
      "epoch": 2.556195965417867,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004348897654160791,
      "loss": 4.7539,
      "mean_token_accuracy": 0.24269762635231018,
      "num_tokens": 61021842.0,
      "step": 26610
    },
    {
      "entropy": 5.149652481079102,
      "epoch": 2.5566762728146015,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004348656425134223,
      "loss": 4.8099,
      "mean_token_accuracy": 0.22697034925222398,
      "num_tokens": 61033344.0,
      "step": 26615
    },
    {
      "entropy": 5.007366132736206,
      "epoch": 2.5571565802113354,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043484151589916,
      "loss": 4.7048,
      "mean_token_accuracy": 0.2413291871547699,
      "num_tokens": 61044920.0,
      "step": 26620
    },
    {
      "entropy": 5.059661960601806,
      "epoch": 2.5576368876080693,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004348173855738523,
      "loss": 4.6805,
      "mean_token_accuracy": 0.24257395565509796,
      "num_tokens": 61056132.0,
      "step": 26625
    },
    {
      "entropy": 5.02752685546875,
      "epoch": 2.558117195004803,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004347932515380595,
      "loss": 4.7259,
      "mean_token_accuracy": 0.23945114761590958,
      "num_tokens": 61068590.0,
      "step": 26630
    },
    {
      "entropy": 5.117930936813354,
      "epoch": 2.558597502401537,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004347691137923418,
      "loss": 4.7991,
      "mean_token_accuracy": 0.23748404085636138,
      "num_tokens": 61080011.0,
      "step": 26635
    },
    {
      "entropy": 5.084970760345459,
      "epoch": 2.559077809798271,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004347449723372598,
      "loss": 4.6691,
      "mean_token_accuracy": 0.23757386654615403,
      "num_tokens": 61091448.0,
      "step": 26640
    },
    {
      "entropy": 5.019001197814942,
      "epoch": 2.559558117195005,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004347208271733738,
      "loss": 4.6237,
      "mean_token_accuracy": 0.24922345131635665,
      "num_tokens": 61102507.0,
      "step": 26645
    },
    {
      "entropy": 5.02917890548706,
      "epoch": 2.5600384245917387,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004346966783012445,
      "loss": 4.7336,
      "mean_token_accuracy": 0.2349156394600868,
      "num_tokens": 61114577.0,
      "step": 26650
    },
    {
      "entropy": 5.042965936660766,
      "epoch": 2.5605187319884726,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043467252572143247,
      "loss": 4.6124,
      "mean_token_accuracy": 0.25262097716331483,
      "num_tokens": 61126044.0,
      "step": 26655
    },
    {
      "entropy": 5.1082672595977785,
      "epoch": 2.5609990393852065,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043464836943449866,
      "loss": 4.8484,
      "mean_token_accuracy": 0.2236395686864853,
      "num_tokens": 61137880.0,
      "step": 26660
    },
    {
      "entropy": 5.094914197921753,
      "epoch": 2.5614793467819403,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004346242094410036,
      "loss": 4.8045,
      "mean_token_accuracy": 0.23229206651449202,
      "num_tokens": 61149103.0,
      "step": 26665
    },
    {
      "entropy": 5.049566411972046,
      "epoch": 2.561959654178674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004346000457415085,
      "loss": 4.6914,
      "mean_token_accuracy": 0.23526936173439025,
      "num_tokens": 61161235.0,
      "step": 26670
    },
    {
      "entropy": 5.0815019607543945,
      "epoch": 2.562439961575408,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043457587833657424,
      "loss": 4.7599,
      "mean_token_accuracy": 0.23779678344726562,
      "num_tokens": 61172497.0,
      "step": 26675
    },
    {
      "entropy": 5.19757080078125,
      "epoch": 2.5629202689721424,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043455170722676194,
      "loss": 4.9117,
      "mean_token_accuracy": 0.22393706440925598,
      "num_tokens": 61183795.0,
      "step": 26680
    },
    {
      "entropy": 5.084181261062622,
      "epoch": 2.563400576368876,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004345275324126328,
      "loss": 4.7002,
      "mean_token_accuracy": 0.24545634537935257,
      "num_tokens": 61195909.0,
      "step": 26685
    },
    {
      "entropy": 5.0928281307220455,
      "epoch": 2.56388088376561,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043450335389474796,
      "loss": 4.7144,
      "mean_token_accuracy": 0.2377787634730339,
      "num_tokens": 61207017.0,
      "step": 26690
    },
    {
      "entropy": 5.073605012893677,
      "epoch": 2.5643611911623436,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004344791716736689,
      "loss": 4.7985,
      "mean_token_accuracy": 0.22642171531915664,
      "num_tokens": 61219172.0,
      "step": 26695
    },
    {
      "entropy": 5.118867015838623,
      "epoch": 2.564841498559078,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043445498574995705,
      "loss": 4.802,
      "mean_token_accuracy": 0.23948947340250015,
      "num_tokens": 61232612.0,
      "step": 26700
    },
    {
      "entropy": 5.101803922653199,
      "epoch": 2.565321805955812,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043443079612417394,
      "loss": 4.7998,
      "mean_token_accuracy": 0.22712824046611785,
      "num_tokens": 61243627.0,
      "step": 26705
    },
    {
      "entropy": 5.080418682098388,
      "epoch": 2.5658021133525457,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004344066027968812,
      "loss": 4.8152,
      "mean_token_accuracy": 0.22968849688768386,
      "num_tokens": 61254516.0,
      "step": 26710
    },
    {
      "entropy": 5.171592378616333,
      "epoch": 2.5662824207492796,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043438240576864034,
      "loss": 4.7655,
      "mean_token_accuracy": 0.2321400985121727,
      "num_tokens": 61266764.0,
      "step": 26715
    },
    {
      "entropy": 5.05295205116272,
      "epoch": 2.5667627281460135,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004343582050400134,
      "loss": 4.6703,
      "mean_token_accuracy": 0.24498352855443956,
      "num_tokens": 61278088.0,
      "step": 26720
    },
    {
      "entropy": 5.017121601104736,
      "epoch": 2.5672430355427474,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004343340006115621,
      "loss": 4.7548,
      "mean_token_accuracy": 0.23750760555267333,
      "num_tokens": 61289558.0,
      "step": 26725
    },
    {
      "entropy": 4.993280649185181,
      "epoch": 2.5677233429394812,
      "grad_norm": 1.0,
      "learning_rate": 0.0004343097924838483,
      "loss": 4.6765,
      "mean_token_accuracy": 0.245581017434597,
      "num_tokens": 61301815.0,
      "step": 26730
    },
    {
      "entropy": 5.2157275676727295,
      "epoch": 2.568203650336215,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004342855806574344,
      "loss": 4.8389,
      "mean_token_accuracy": 0.22863384932279587,
      "num_tokens": 61312216.0,
      "step": 26735
    },
    {
      "entropy": 5.108919286727906,
      "epoch": 2.568683957732949,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004342613651328822,
      "loss": 4.737,
      "mean_token_accuracy": 0.2324855864048004,
      "num_tokens": 61322485.0,
      "step": 26740
    },
    {
      "entropy": 5.118260049819947,
      "epoch": 2.569164265129683,
      "grad_norm": 1.078125,
      "learning_rate": 0.000434237145910754,
      "loss": 4.8027,
      "mean_token_accuracy": 0.22565922737121583,
      "num_tokens": 61334654.0,
      "step": 26745
    },
    {
      "entropy": 5.040866327285767,
      "epoch": 2.5696445725264168,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043421292299161213,
      "loss": 4.7026,
      "mean_token_accuracy": 0.2429600849747658,
      "num_tokens": 61346563.0,
      "step": 26750
    },
    {
      "entropy": 5.071473836898804,
      "epoch": 2.570124879923151,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00043418869637601887,
      "loss": 4.7546,
      "mean_token_accuracy": 0.23752598017454146,
      "num_tokens": 61360121.0,
      "step": 26755
    },
    {
      "entropy": 5.125632476806641,
      "epoch": 2.5706051873198845,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043416446606453686,
      "loss": 4.7096,
      "mean_token_accuracy": 0.24460220485925674,
      "num_tokens": 61371340.0,
      "step": 26760
    },
    {
      "entropy": 5.077202844619751,
      "epoch": 2.571085494716619,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004341402320577285,
      "loss": 4.703,
      "mean_token_accuracy": 0.2368769496679306,
      "num_tokens": 61382571.0,
      "step": 26765
    },
    {
      "entropy": 4.999770021438598,
      "epoch": 2.5715658021133523,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004341159943561566,
      "loss": 4.6127,
      "mean_token_accuracy": 0.24607600420713424,
      "num_tokens": 61394548.0,
      "step": 26770
    },
    {
      "entropy": 5.024379301071167,
      "epoch": 2.5720461095100866,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004340917529603837,
      "loss": 4.6935,
      "mean_token_accuracy": 0.2400648385286331,
      "num_tokens": 61405668.0,
      "step": 26775
    },
    {
      "entropy": 5.076011228561401,
      "epoch": 2.5725264169068205,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004340675078709728,
      "loss": 4.7695,
      "mean_token_accuracy": 0.23179153800010682,
      "num_tokens": 61417396.0,
      "step": 26780
    },
    {
      "entropy": 5.0324851989746096,
      "epoch": 2.5730067243035544,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004340432590884866,
      "loss": 4.6655,
      "mean_token_accuracy": 0.24478673040866852,
      "num_tokens": 61427981.0,
      "step": 26785
    },
    {
      "entropy": 5.060857248306275,
      "epoch": 2.5734870317002883,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043401900661348825,
      "loss": 4.7414,
      "mean_token_accuracy": 0.2322687953710556,
      "num_tokens": 61439011.0,
      "step": 26790
    },
    {
      "entropy": 5.109488487243652,
      "epoch": 2.573967339097022,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043399475044654073,
      "loss": 4.8109,
      "mean_token_accuracy": 0.23873028308153152,
      "num_tokens": 61450567.0,
      "step": 26795
    },
    {
      "entropy": 5.09658842086792,
      "epoch": 2.574447646493756,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043397049058820726,
      "loss": 4.7271,
      "mean_token_accuracy": 0.23578271716833116,
      "num_tokens": 61461563.0,
      "step": 26800
    },
    {
      "entropy": 5.056091785430908,
      "epoch": 2.57492795389049,
      "grad_norm": 0.953125,
      "learning_rate": 0.000433946227039051,
      "loss": 4.7548,
      "mean_token_accuracy": 0.23778523057699202,
      "num_tokens": 61472697.0,
      "step": 26805
    },
    {
      "entropy": 5.0361518383026125,
      "epoch": 2.5754082612872238,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004339219597996353,
      "loss": 4.7448,
      "mean_token_accuracy": 0.2316604658961296,
      "num_tokens": 61485877.0,
      "step": 26810
    },
    {
      "entropy": 5.035082292556763,
      "epoch": 2.5758885686839577,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004338976888705236,
      "loss": 4.7286,
      "mean_token_accuracy": 0.23069047033786774,
      "num_tokens": 61499007.0,
      "step": 26815
    },
    {
      "entropy": 5.10683650970459,
      "epoch": 2.5763688760806915,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043387341425227944,
      "loss": 4.7895,
      "mean_token_accuracy": 0.23326374888420104,
      "num_tokens": 61510901.0,
      "step": 26820
    },
    {
      "entropy": 5.061435747146606,
      "epoch": 2.5768491834774254,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004338491359454664,
      "loss": 4.692,
      "mean_token_accuracy": 0.2375418111681938,
      "num_tokens": 61523628.0,
      "step": 26825
    },
    {
      "entropy": 5.057381868362427,
      "epoch": 2.5773294908741593,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043382485395064796,
      "loss": 4.7628,
      "mean_token_accuracy": 0.23983976989984512,
      "num_tokens": 61534765.0,
      "step": 26830
    },
    {
      "entropy": 5.0775903224945065,
      "epoch": 2.577809798270893,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004338005682683881,
      "loss": 4.7733,
      "mean_token_accuracy": 0.2332910880446434,
      "num_tokens": 61547009.0,
      "step": 26835
    },
    {
      "entropy": 5.178178691864014,
      "epoch": 2.5782901056676275,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043377627889925057,
      "loss": 4.8591,
      "mean_token_accuracy": 0.22550309002399443,
      "num_tokens": 61558713.0,
      "step": 26840
    },
    {
      "entropy": 5.132937860488892,
      "epoch": 2.578770413064361,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004337519858437994,
      "loss": 4.8074,
      "mean_token_accuracy": 0.23198498040437698,
      "num_tokens": 61569841.0,
      "step": 26845
    },
    {
      "entropy": 5.06663932800293,
      "epoch": 2.5792507204610953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004337276891025984,
      "loss": 4.6872,
      "mean_token_accuracy": 0.23886698782444,
      "num_tokens": 61580878.0,
      "step": 26850
    },
    {
      "entropy": 5.12580828666687,
      "epoch": 2.579731027857829,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043370338867621184,
      "loss": 4.8635,
      "mean_token_accuracy": 0.222461299598217,
      "num_tokens": 61593328.0,
      "step": 26855
    },
    {
      "entropy": 5.115474367141724,
      "epoch": 2.580211335254563,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043367908456520387,
      "loss": 4.7936,
      "mean_token_accuracy": 0.2270260974764824,
      "num_tokens": 61605085.0,
      "step": 26860
    },
    {
      "entropy": 5.119378185272216,
      "epoch": 2.580691642651297,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004336547767701387,
      "loss": 4.7888,
      "mean_token_accuracy": 0.22784062176942826,
      "num_tokens": 61618410.0,
      "step": 26865
    },
    {
      "entropy": 5.146065950393677,
      "epoch": 2.581171950048031,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043363046529158077,
      "loss": 4.7302,
      "mean_token_accuracy": 0.23899787068367004,
      "num_tokens": 61628795.0,
      "step": 26870
    },
    {
      "entropy": 5.100282812118531,
      "epoch": 2.5816522574447647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004336061501300944,
      "loss": 4.7414,
      "mean_token_accuracy": 0.23864644318819045,
      "num_tokens": 61638769.0,
      "step": 26875
    },
    {
      "entropy": 5.026713514328003,
      "epoch": 2.5821325648414986,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004335818312862442,
      "loss": 4.7274,
      "mean_token_accuracy": 0.2441367968916893,
      "num_tokens": 61648748.0,
      "step": 26880
    },
    {
      "entropy": 4.984084033966065,
      "epoch": 2.5826128722382324,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043355750876059485,
      "loss": 4.6525,
      "mean_token_accuracy": 0.24055531173944472,
      "num_tokens": 61658733.0,
      "step": 26885
    },
    {
      "entropy": 5.038102674484253,
      "epoch": 2.5830931796349663,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043353318255371086,
      "loss": 4.7463,
      "mean_token_accuracy": 0.2341667741537094,
      "num_tokens": 61669192.0,
      "step": 26890
    },
    {
      "entropy": 5.174574327468872,
      "epoch": 2.5835734870317,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004335088526661571,
      "loss": 4.8325,
      "mean_token_accuracy": 0.22729283422231675,
      "num_tokens": 61679091.0,
      "step": 26895
    },
    {
      "entropy": 5.135823059082031,
      "epoch": 2.584053794428434,
      "grad_norm": 1.0,
      "learning_rate": 0.00043348451909849855,
      "loss": 4.7631,
      "mean_token_accuracy": 0.24136848300695418,
      "num_tokens": 61691143.0,
      "step": 26900
    },
    {
      "entropy": 5.17247428894043,
      "epoch": 2.584534101825168,
      "grad_norm": 1.125,
      "learning_rate": 0.00043346018185130006,
      "loss": 4.8476,
      "mean_token_accuracy": 0.23295264691114426,
      "num_tokens": 61703549.0,
      "step": 26905
    },
    {
      "entropy": 5.012521457672119,
      "epoch": 2.585014409221902,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043343584092512665,
      "loss": 4.6807,
      "mean_token_accuracy": 0.24200984835624695,
      "num_tokens": 61713447.0,
      "step": 26910
    },
    {
      "entropy": 5.118030595779419,
      "epoch": 2.585494716618636,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004334114963205435,
      "loss": 4.7579,
      "mean_token_accuracy": 0.2325425624847412,
      "num_tokens": 61724848.0,
      "step": 26915
    },
    {
      "entropy": 5.027098751068115,
      "epoch": 2.5859750240153696,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004333871480381158,
      "loss": 4.7503,
      "mean_token_accuracy": 0.2377648502588272,
      "num_tokens": 61734624.0,
      "step": 26920
    },
    {
      "entropy": 5.093262100219727,
      "epoch": 2.586455331412104,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004333627960784088,
      "loss": 4.7459,
      "mean_token_accuracy": 0.24287916719913483,
      "num_tokens": 61745627.0,
      "step": 26925
    },
    {
      "entropy": 5.052375078201294,
      "epoch": 2.586935638808838,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004333384404419879,
      "loss": 4.7718,
      "mean_token_accuracy": 0.23372279703617097,
      "num_tokens": 61757050.0,
      "step": 26930
    },
    {
      "entropy": 5.168138217926026,
      "epoch": 2.5874159462055717,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004333140811294186,
      "loss": 4.8681,
      "mean_token_accuracy": 0.22610276192426682,
      "num_tokens": 61768896.0,
      "step": 26935
    },
    {
      "entropy": 5.168129777908325,
      "epoch": 2.5878962536023056,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004332897181412664,
      "loss": 4.7949,
      "mean_token_accuracy": 0.2297089159488678,
      "num_tokens": 61779809.0,
      "step": 26940
    },
    {
      "entropy": 4.987856817245484,
      "epoch": 2.5883765609990395,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043326535147809696,
      "loss": 4.6752,
      "mean_token_accuracy": 0.24108424186706542,
      "num_tokens": 61791737.0,
      "step": 26945
    },
    {
      "entropy": 5.018596267700195,
      "epoch": 2.5888568683957733,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043324098114047604,
      "loss": 4.685,
      "mean_token_accuracy": 0.24312884360551834,
      "num_tokens": 61804193.0,
      "step": 26950
    },
    {
      "entropy": 5.029084300994873,
      "epoch": 2.589337175792507,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004332166071289695,
      "loss": 4.6658,
      "mean_token_accuracy": 0.2409258395433426,
      "num_tokens": 61815735.0,
      "step": 26955
    },
    {
      "entropy": 5.0597327709197994,
      "epoch": 2.589817483189241,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043319222944414304,
      "loss": 4.7283,
      "mean_token_accuracy": 0.24086202383041383,
      "num_tokens": 61827263.0,
      "step": 26960
    },
    {
      "entropy": 5.0300380229949955,
      "epoch": 2.590297790585975,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043316784808656276,
      "loss": 4.6409,
      "mean_token_accuracy": 0.24005349576473237,
      "num_tokens": 61838734.0,
      "step": 26965
    },
    {
      "entropy": 5.074756193161011,
      "epoch": 2.590778097982709,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043314346305679477,
      "loss": 4.7651,
      "mean_token_accuracy": 0.23208698630332947,
      "num_tokens": 61849935.0,
      "step": 26970
    },
    {
      "entropy": 5.003832817077637,
      "epoch": 2.5912584053794427,
      "grad_norm": 1.0,
      "learning_rate": 0.00043311907435540517,
      "loss": 4.6558,
      "mean_token_accuracy": 0.249183814227581,
      "num_tokens": 61861849.0,
      "step": 26975
    },
    {
      "entropy": 5.056755208969117,
      "epoch": 2.5917387127761766,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004330946819829601,
      "loss": 4.7105,
      "mean_token_accuracy": 0.2435007780790329,
      "num_tokens": 61873017.0,
      "step": 26980
    },
    {
      "entropy": 5.041497278213501,
      "epoch": 2.5922190201729105,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043307028594002597,
      "loss": 4.7079,
      "mean_token_accuracy": 0.23494507670402526,
      "num_tokens": 61884730.0,
      "step": 26985
    },
    {
      "entropy": 5.077803754806519,
      "epoch": 2.592699327569645,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043304588622716924,
      "loss": 4.7361,
      "mean_token_accuracy": 0.24334974139928817,
      "num_tokens": 61896726.0,
      "step": 26990
    },
    {
      "entropy": 5.01508059501648,
      "epoch": 2.5931796349663783,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004330214828449563,
      "loss": 4.7043,
      "mean_token_accuracy": 0.23559777736663817,
      "num_tokens": 61907931.0,
      "step": 26995
    },
    {
      "entropy": 5.043628120422364,
      "epoch": 2.5936599423631126,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043299707579395365,
      "loss": 4.6735,
      "mean_token_accuracy": 0.24669248312711717,
      "num_tokens": 61919001.0,
      "step": 27000
    },
    {
      "epoch": 2.5936599423631126,
      "eval_entropy": 4.9306121872204915,
      "eval_loss": 4.8611674308776855,
      "eval_mean_token_accuracy": 0.2388892080771057,
      "eval_num_tokens": 61919001.0,
      "eval_runtime": 26.4677,
      "eval_samples_per_second": 1239.812,
      "eval_steps_per_second": 154.981,
      "step": 27000
    },
    {
      "entropy": 5.186697769165039,
      "epoch": 2.594140249759846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004329726650747282,
      "loss": 4.8661,
      "mean_token_accuracy": 0.2255033016204834,
      "num_tokens": 61931948.0,
      "step": 27005
    },
    {
      "entropy": 5.1454840183258055,
      "epoch": 2.5946205571565804,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004329482506878465,
      "loss": 4.8516,
      "mean_token_accuracy": 0.22167310416698455,
      "num_tokens": 61942781.0,
      "step": 27010
    },
    {
      "entropy": 5.03824634552002,
      "epoch": 2.5951008645533142,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043292383263387536,
      "loss": 4.736,
      "mean_token_accuracy": 0.24108149856328964,
      "num_tokens": 61953981.0,
      "step": 27015
    },
    {
      "entropy": 5.074985408782959,
      "epoch": 2.595581171950048,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043289941091338187,
      "loss": 4.6974,
      "mean_token_accuracy": 0.24551328122615815,
      "num_tokens": 61965328.0,
      "step": 27020
    },
    {
      "entropy": 5.109946823120117,
      "epoch": 2.596061479346782,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004328749855269329,
      "loss": 4.7047,
      "mean_token_accuracy": 0.2375982478260994,
      "num_tokens": 61975427.0,
      "step": 27025
    },
    {
      "entropy": 5.059343767166138,
      "epoch": 2.596541786743516,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004328505564750955,
      "loss": 4.7513,
      "mean_token_accuracy": 0.2304479405283928,
      "num_tokens": 61985724.0,
      "step": 27030
    },
    {
      "entropy": 5.200712442398071,
      "epoch": 2.5970220941402498,
      "grad_norm": 0.984375,
      "learning_rate": 0.000432826123758437,
      "loss": 4.9211,
      "mean_token_accuracy": 0.223887600004673,
      "num_tokens": 61996382.0,
      "step": 27035
    },
    {
      "entropy": 5.1523332595825195,
      "epoch": 2.5975024015369836,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004328016873775244,
      "loss": 4.8002,
      "mean_token_accuracy": 0.23526711910963058,
      "num_tokens": 62008859.0,
      "step": 27040
    },
    {
      "entropy": 5.02960352897644,
      "epoch": 2.5979827089337175,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043277724733292527,
      "loss": 4.6221,
      "mean_token_accuracy": 0.24983428567647933,
      "num_tokens": 62019096.0,
      "step": 27045
    },
    {
      "entropy": 5.058312177658081,
      "epoch": 2.5984630163304514,
      "grad_norm": 1.0,
      "learning_rate": 0.0004327528036252071,
      "loss": 4.7696,
      "mean_token_accuracy": 0.23652503043413162,
      "num_tokens": 62030874.0,
      "step": 27050
    },
    {
      "entropy": 5.010758543014527,
      "epoch": 2.5989433237271853,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004327283562549371,
      "loss": 4.679,
      "mean_token_accuracy": 0.24563511162996293,
      "num_tokens": 62042620.0,
      "step": 27055
    },
    {
      "entropy": 5.104320001602173,
      "epoch": 2.599423631123919,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000432703905222683,
      "loss": 4.7622,
      "mean_token_accuracy": 0.23951544612646103,
      "num_tokens": 62054124.0,
      "step": 27060
    },
    {
      "entropy": 5.054672908782959,
      "epoch": 2.5999039385206535,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043267945052901264,
      "loss": 4.71,
      "mean_token_accuracy": 0.24130181223154068,
      "num_tokens": 62064647.0,
      "step": 27065
    },
    {
      "entropy": 5.09134612083435,
      "epoch": 2.600384245917387,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004326549921744935,
      "loss": 4.7229,
      "mean_token_accuracy": 0.24033046811819075,
      "num_tokens": 62075298.0,
      "step": 27070
    },
    {
      "entropy": 5.166427326202393,
      "epoch": 2.6008645533141213,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004326305301596937,
      "loss": 4.8722,
      "mean_token_accuracy": 0.2218060612678528,
      "num_tokens": 62087875.0,
      "step": 27075
    },
    {
      "entropy": 5.019661140441895,
      "epoch": 2.6013448607108547,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043260606448518096,
      "loss": 4.62,
      "mean_token_accuracy": 0.24793300032615662,
      "num_tokens": 62098538.0,
      "step": 27080
    },
    {
      "entropy": 4.997071743011475,
      "epoch": 2.601825168107589,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043258159515152347,
      "loss": 4.7009,
      "mean_token_accuracy": 0.23628205209970474,
      "num_tokens": 62109637.0,
      "step": 27085
    },
    {
      "entropy": 5.045969820022583,
      "epoch": 2.602305475504323,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004325571221592892,
      "loss": 4.6583,
      "mean_token_accuracy": 0.24080536663532257,
      "num_tokens": 62120471.0,
      "step": 27090
    },
    {
      "entropy": 5.060687255859375,
      "epoch": 2.6027857829010568,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043253264550904646,
      "loss": 4.7491,
      "mean_token_accuracy": 0.23810342997312545,
      "num_tokens": 62130150.0,
      "step": 27095
    },
    {
      "entropy": 5.105531120300293,
      "epoch": 2.6032660902977907,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004325081652013635,
      "loss": 4.7956,
      "mean_token_accuracy": 0.23024881333112718,
      "num_tokens": 62142382.0,
      "step": 27100
    },
    {
      "entropy": 5.076539993286133,
      "epoch": 2.6037463976945245,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043248368123680855,
      "loss": 4.7318,
      "mean_token_accuracy": 0.23700406551361083,
      "num_tokens": 62155398.0,
      "step": 27105
    },
    {
      "entropy": 5.0011015892028805,
      "epoch": 2.6042267050912584,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043245919361595026,
      "loss": 4.6637,
      "mean_token_accuracy": 0.23841592222452163,
      "num_tokens": 62167694.0,
      "step": 27110
    },
    {
      "entropy": 5.037281274795532,
      "epoch": 2.6047070124879923,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043243470233935696,
      "loss": 4.6802,
      "mean_token_accuracy": 0.24151979833841325,
      "num_tokens": 62179420.0,
      "step": 27115
    },
    {
      "entropy": 5.078219890594482,
      "epoch": 2.605187319884726,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004324102074075973,
      "loss": 4.7133,
      "mean_token_accuracy": 0.23895540982484817,
      "num_tokens": 62190881.0,
      "step": 27120
    },
    {
      "entropy": 4.98907699584961,
      "epoch": 2.60566762728146,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004323857088212402,
      "loss": 4.6337,
      "mean_token_accuracy": 0.2441602900624275,
      "num_tokens": 62201513.0,
      "step": 27125
    },
    {
      "entropy": 5.011536645889282,
      "epoch": 2.606147934678194,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004323612065808541,
      "loss": 4.6836,
      "mean_token_accuracy": 0.23701840937137603,
      "num_tokens": 62212921.0,
      "step": 27130
    },
    {
      "entropy": 5.028112173080444,
      "epoch": 2.606628242074928,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043233670068700827,
      "loss": 4.7114,
      "mean_token_accuracy": 0.24280804246664048,
      "num_tokens": 62224579.0,
      "step": 27135
    },
    {
      "entropy": 5.027622175216675,
      "epoch": 2.6071085494716617,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004323121911402713,
      "loss": 4.776,
      "mean_token_accuracy": 0.23491197675466538,
      "num_tokens": 62236181.0,
      "step": 27140
    },
    {
      "entropy": 5.112509536743164,
      "epoch": 2.6075888568683956,
      "grad_norm": 1.125,
      "learning_rate": 0.00043228767794121245,
      "loss": 4.8525,
      "mean_token_accuracy": 0.2319944530725479,
      "num_tokens": 62247607.0,
      "step": 27145
    },
    {
      "entropy": 5.158500862121582,
      "epoch": 2.60806916426513,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004322631610904006,
      "loss": 4.7518,
      "mean_token_accuracy": 0.23851727396249772,
      "num_tokens": 62259171.0,
      "step": 27150
    },
    {
      "entropy": 5.139279413223266,
      "epoch": 2.6085494716618634,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043223864058840534,
      "loss": 4.8202,
      "mean_token_accuracy": 0.23344897776842116,
      "num_tokens": 62271105.0,
      "step": 27155
    },
    {
      "entropy": 5.06716103553772,
      "epoch": 2.6090297790585977,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043221411643579557,
      "loss": 4.7387,
      "mean_token_accuracy": 0.23972565978765487,
      "num_tokens": 62282818.0,
      "step": 27160
    },
    {
      "entropy": 5.048837757110595,
      "epoch": 2.6095100864553316,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043218958863314096,
      "loss": 4.7861,
      "mean_token_accuracy": 0.2358744814991951,
      "num_tokens": 62294604.0,
      "step": 27165
    },
    {
      "entropy": 5.121849727630615,
      "epoch": 2.6099903938520654,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004321650571810109,
      "loss": 4.8116,
      "mean_token_accuracy": 0.24122051447629927,
      "num_tokens": 62305871.0,
      "step": 27170
    },
    {
      "entropy": 5.078708839416504,
      "epoch": 2.6104707012487993,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004321405220799747,
      "loss": 4.6897,
      "mean_token_accuracy": 0.24143462032079696,
      "num_tokens": 62317100.0,
      "step": 27175
    },
    {
      "entropy": 5.050977277755737,
      "epoch": 2.610951008645533,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004321159833306024,
      "loss": 4.72,
      "mean_token_accuracy": 0.2342998743057251,
      "num_tokens": 62328934.0,
      "step": 27180
    },
    {
      "entropy": 5.113052082061768,
      "epoch": 2.611431316042267,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004320914409334634,
      "loss": 4.7804,
      "mean_token_accuracy": 0.23358690738677979,
      "num_tokens": 62340812.0,
      "step": 27185
    },
    {
      "entropy": 5.06442551612854,
      "epoch": 2.611911623439001,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004320668948891276,
      "loss": 4.7271,
      "mean_token_accuracy": 0.23840461671352386,
      "num_tokens": 62352753.0,
      "step": 27190
    },
    {
      "entropy": 5.076548528671265,
      "epoch": 2.612391930835735,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043204234519816486,
      "loss": 4.7431,
      "mean_token_accuracy": 0.2379434362053871,
      "num_tokens": 62362731.0,
      "step": 27195
    },
    {
      "entropy": 5.138695287704468,
      "epoch": 2.6128722382324687,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004320177918611453,
      "loss": 4.7853,
      "mean_token_accuracy": 0.23400088101625444,
      "num_tokens": 62374299.0,
      "step": 27200
    },
    {
      "entropy": 5.05802173614502,
      "epoch": 2.6133525456292026,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00043199323487863876,
      "loss": 4.786,
      "mean_token_accuracy": 0.22948825359344482,
      "num_tokens": 62387708.0,
      "step": 27205
    },
    {
      "entropy": 5.092346429824829,
      "epoch": 2.6138328530259365,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043196867425121554,
      "loss": 4.7249,
      "mean_token_accuracy": 0.24286508858203887,
      "num_tokens": 62399139.0,
      "step": 27210
    },
    {
      "entropy": 5.129161643981933,
      "epoch": 2.6143131604226704,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043194410997944577,
      "loss": 4.73,
      "mean_token_accuracy": 0.23232386410236358,
      "num_tokens": 62411581.0,
      "step": 27215
    },
    {
      "entropy": 5.094691896438599,
      "epoch": 2.6147934678194042,
      "grad_norm": 1.0,
      "learning_rate": 0.00043191954206389985,
      "loss": 4.8442,
      "mean_token_accuracy": 0.22429100275039673,
      "num_tokens": 62423768.0,
      "step": 27220
    },
    {
      "entropy": 5.0586004734039305,
      "epoch": 2.6152737752161386,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004318949705051481,
      "loss": 4.7526,
      "mean_token_accuracy": 0.24115074723958968,
      "num_tokens": 62435716.0,
      "step": 27225
    },
    {
      "entropy": 5.0406536102294925,
      "epoch": 2.615754082612872,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004318703953037609,
      "loss": 4.6822,
      "mean_token_accuracy": 0.23782588988542558,
      "num_tokens": 62447409.0,
      "step": 27230
    },
    {
      "entropy": 5.080943822860718,
      "epoch": 2.6162343900096063,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004318458164603091,
      "loss": 4.7359,
      "mean_token_accuracy": 0.24107529670000077,
      "num_tokens": 62458900.0,
      "step": 27235
    },
    {
      "entropy": 5.074116230010986,
      "epoch": 2.61671469740634,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004318212339753631,
      "loss": 4.785,
      "mean_token_accuracy": 0.23819051533937455,
      "num_tokens": 62471433.0,
      "step": 27240
    },
    {
      "entropy": 5.075383424758911,
      "epoch": 2.617195004803074,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043179664784949375,
      "loss": 4.7842,
      "mean_token_accuracy": 0.2337319567799568,
      "num_tokens": 62482633.0,
      "step": 27245
    },
    {
      "entropy": 5.06030535697937,
      "epoch": 2.617675312199808,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004317720580832717,
      "loss": 4.6154,
      "mean_token_accuracy": 0.24855824410915375,
      "num_tokens": 62492661.0,
      "step": 27250
    },
    {
      "entropy": 5.0341184616088865,
      "epoch": 2.618155619596542,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004317474646772681,
      "loss": 4.7232,
      "mean_token_accuracy": 0.23800063729286194,
      "num_tokens": 62504470.0,
      "step": 27255
    },
    {
      "entropy": 5.1785581588745115,
      "epoch": 2.6186359269932757,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004317228676320539,
      "loss": 4.8104,
      "mean_token_accuracy": 0.22706159502267836,
      "num_tokens": 62514711.0,
      "step": 27260
    },
    {
      "entropy": 5.149213600158691,
      "epoch": 2.6191162343900096,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043169826694819987,
      "loss": 4.769,
      "mean_token_accuracy": 0.23876968622207642,
      "num_tokens": 62525939.0,
      "step": 27265
    },
    {
      "entropy": 5.042378187179565,
      "epoch": 2.6195965417867435,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004316736626262775,
      "loss": 4.7268,
      "mean_token_accuracy": 0.23843303620815276,
      "num_tokens": 62536643.0,
      "step": 27270
    },
    {
      "entropy": 5.067795991897583,
      "epoch": 2.6200768491834774,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004316490546668579,
      "loss": 4.743,
      "mean_token_accuracy": 0.24060970842838286,
      "num_tokens": 62547699.0,
      "step": 27275
    },
    {
      "entropy": 5.135523700714112,
      "epoch": 2.6205571565802113,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004316244430705124,
      "loss": 4.7632,
      "mean_token_accuracy": 0.23627875447273256,
      "num_tokens": 62557756.0,
      "step": 27280
    },
    {
      "entropy": 5.074234247207642,
      "epoch": 2.621037463976945,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004315998278378123,
      "loss": 4.7628,
      "mean_token_accuracy": 0.23726956844329833,
      "num_tokens": 62570046.0,
      "step": 27285
    },
    {
      "entropy": 5.115970325469971,
      "epoch": 2.621517771373679,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00043157520896932943,
      "loss": 4.802,
      "mean_token_accuracy": 0.2327379524707794,
      "num_tokens": 62580037.0,
      "step": 27290
    },
    {
      "entropy": 5.059601163864135,
      "epoch": 2.621998078770413,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000431550586465635,
      "loss": 4.7575,
      "mean_token_accuracy": 0.24071327596902847,
      "num_tokens": 62591442.0,
      "step": 27295
    },
    {
      "entropy": 5.148237466812134,
      "epoch": 2.6224783861671472,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043152596032730085,
      "loss": 4.787,
      "mean_token_accuracy": 0.22872916013002395,
      "num_tokens": 62603367.0,
      "step": 27300
    },
    {
      "entropy": 5.056389188766479,
      "epoch": 2.6229586935638807,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043150133055489865,
      "loss": 4.662,
      "mean_token_accuracy": 0.24290239065885544,
      "num_tokens": 62613506.0,
      "step": 27305
    },
    {
      "entropy": 5.0190692901611325,
      "epoch": 2.623439000960615,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004314766971490003,
      "loss": 4.6638,
      "mean_token_accuracy": 0.24093786776065826,
      "num_tokens": 62625389.0,
      "step": 27310
    },
    {
      "entropy": 5.069457912445069,
      "epoch": 2.6239193083573484,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004314520601101776,
      "loss": 4.7553,
      "mean_token_accuracy": 0.23744067549705505,
      "num_tokens": 62637949.0,
      "step": 27315
    },
    {
      "entropy": 5.055627727508545,
      "epoch": 2.6243996157540828,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043142741943900275,
      "loss": 4.7013,
      "mean_token_accuracy": 0.2350221246480942,
      "num_tokens": 62649952.0,
      "step": 27320
    },
    {
      "entropy": 5.1225358009338375,
      "epoch": 2.6248799231508166,
      "grad_norm": 0.9375,
      "learning_rate": 0.00043140277513604763,
      "loss": 4.832,
      "mean_token_accuracy": 0.22892586886882782,
      "num_tokens": 62661152.0,
      "step": 27325
    },
    {
      "entropy": 5.138351535797119,
      "epoch": 2.6253602305475505,
      "grad_norm": 1.125,
      "learning_rate": 0.0004313781272018845,
      "loss": 4.7944,
      "mean_token_accuracy": 0.22640667855739594,
      "num_tokens": 62671648.0,
      "step": 27330
    },
    {
      "entropy": 5.097940063476562,
      "epoch": 2.6258405379442844,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004313534756370857,
      "loss": 4.71,
      "mean_token_accuracy": 0.2341504916548729,
      "num_tokens": 62683114.0,
      "step": 27335
    },
    {
      "entropy": 5.0625701427459715,
      "epoch": 2.6263208453410183,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043132882044222336,
      "loss": 4.7148,
      "mean_token_accuracy": 0.24487661123275756,
      "num_tokens": 62693921.0,
      "step": 27340
    },
    {
      "entropy": 5.092754650115967,
      "epoch": 2.626801152737752,
      "grad_norm": 1.0,
      "learning_rate": 0.00043130416161787005,
      "loss": 4.8554,
      "mean_token_accuracy": 0.2298109158873558,
      "num_tokens": 62705342.0,
      "step": 27345
    },
    {
      "entropy": 5.136916303634644,
      "epoch": 2.627281460134486,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043127949916459823,
      "loss": 4.835,
      "mean_token_accuracy": 0.23415536284446717,
      "num_tokens": 62717359.0,
      "step": 27350
    },
    {
      "entropy": 5.132702398300171,
      "epoch": 2.62776176753122,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043125483308298053,
      "loss": 4.7867,
      "mean_token_accuracy": 0.236453677713871,
      "num_tokens": 62728824.0,
      "step": 27355
    },
    {
      "entropy": 5.007018232345581,
      "epoch": 2.628242074927954,
      "grad_norm": 1.0,
      "learning_rate": 0.0004312301633735895,
      "loss": 4.636,
      "mean_token_accuracy": 0.24752915948629378,
      "num_tokens": 62740909.0,
      "step": 27360
    },
    {
      "entropy": 5.0459123134613035,
      "epoch": 2.6287223823246877,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004312054900369981,
      "loss": 4.6741,
      "mean_token_accuracy": 0.23888866007328033,
      "num_tokens": 62751934.0,
      "step": 27365
    },
    {
      "entropy": 5.112234449386596,
      "epoch": 2.6292026897214216,
      "grad_norm": 1.046875,
      "learning_rate": 0.000431180813073779,
      "loss": 4.7497,
      "mean_token_accuracy": 0.2329994410276413,
      "num_tokens": 62762975.0,
      "step": 27370
    },
    {
      "entropy": 5.0990687847137455,
      "epoch": 2.629682997118156,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004311561324845051,
      "loss": 4.7482,
      "mean_token_accuracy": 0.23476216644048692,
      "num_tokens": 62774880.0,
      "step": 27375
    },
    {
      "entropy": 5.049647665023803,
      "epoch": 2.6301633045148893,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004311314482697496,
      "loss": 4.6857,
      "mean_token_accuracy": 0.2452871561050415,
      "num_tokens": 62787372.0,
      "step": 27380
    },
    {
      "entropy": 5.071458196640014,
      "epoch": 2.6306436119116237,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004311067604300855,
      "loss": 4.734,
      "mean_token_accuracy": 0.24340487271547318,
      "num_tokens": 62798437.0,
      "step": 27385
    },
    {
      "entropy": 5.044768762588501,
      "epoch": 2.631123919308357,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004310820689660859,
      "loss": 4.6989,
      "mean_token_accuracy": 0.24123535454273223,
      "num_tokens": 62810406.0,
      "step": 27390
    },
    {
      "entropy": 5.084887504577637,
      "epoch": 2.6316042267050914,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004310573738783242,
      "loss": 4.7637,
      "mean_token_accuracy": 0.2334263101220131,
      "num_tokens": 62822334.0,
      "step": 27395
    },
    {
      "entropy": 5.101444578170776,
      "epoch": 2.6320845341018253,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004310326751673736,
      "loss": 4.7607,
      "mean_token_accuracy": 0.23473986834287644,
      "num_tokens": 62834601.0,
      "step": 27400
    },
    {
      "entropy": 5.133258056640625,
      "epoch": 2.632564841498559,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043100797283380756,
      "loss": 4.7672,
      "mean_token_accuracy": 0.2384112611413002,
      "num_tokens": 62847086.0,
      "step": 27405
    },
    {
      "entropy": 5.091851329803466,
      "epoch": 2.633045148895293,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043098326687819973,
      "loss": 4.6861,
      "mean_token_accuracy": 0.2339031755924225,
      "num_tokens": 62858658.0,
      "step": 27410
    },
    {
      "entropy": 5.041199254989624,
      "epoch": 2.633525456292027,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004309585573011236,
      "loss": 4.7281,
      "mean_token_accuracy": 0.23748803436756133,
      "num_tokens": 62870731.0,
      "step": 27415
    },
    {
      "entropy": 5.11282172203064,
      "epoch": 2.634005763688761,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004309338441031528,
      "loss": 4.8359,
      "mean_token_accuracy": 0.23333800882101058,
      "num_tokens": 62880848.0,
      "step": 27420
    },
    {
      "entropy": 5.097059679031372,
      "epoch": 2.6344860710854947,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043090912728486135,
      "loss": 4.7453,
      "mean_token_accuracy": 0.24105052202939986,
      "num_tokens": 62892367.0,
      "step": 27425
    },
    {
      "entropy": 5.062040901184082,
      "epoch": 2.6349663784822286,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004308844068468228,
      "loss": 4.7712,
      "mean_token_accuracy": 0.23537192195653917,
      "num_tokens": 62903764.0,
      "step": 27430
    },
    {
      "entropy": 5.1292870998382565,
      "epoch": 2.6354466858789625,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043085968278961116,
      "loss": 4.8463,
      "mean_token_accuracy": 0.22185642123222352,
      "num_tokens": 62914502.0,
      "step": 27435
    },
    {
      "entropy": 5.14368314743042,
      "epoch": 2.6359269932756964,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043083495511380055,
      "loss": 4.9075,
      "mean_token_accuracy": 0.22974989265203477,
      "num_tokens": 62924793.0,
      "step": 27440
    },
    {
      "entropy": 5.025290155410767,
      "epoch": 2.6364073006724302,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043081022381996506,
      "loss": 4.7176,
      "mean_token_accuracy": 0.23724294006824492,
      "num_tokens": 62936806.0,
      "step": 27445
    },
    {
      "entropy": 5.0826396465301515,
      "epoch": 2.636887608069164,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004307854889086787,
      "loss": 4.7193,
      "mean_token_accuracy": 0.2411855787038803,
      "num_tokens": 62948957.0,
      "step": 27450
    },
    {
      "entropy": 5.067258691787719,
      "epoch": 2.637367915465898,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00043076075038051605,
      "loss": 4.7317,
      "mean_token_accuracy": 0.23826210647821428,
      "num_tokens": 62960983.0,
      "step": 27455
    },
    {
      "entropy": 4.987362670898437,
      "epoch": 2.6378482228626323,
      "grad_norm": 1.0,
      "learning_rate": 0.0004307360082360512,
      "loss": 4.6558,
      "mean_token_accuracy": 0.24699123203754425,
      "num_tokens": 62972155.0,
      "step": 27460
    },
    {
      "entropy": 5.108786487579346,
      "epoch": 2.6383285302593658,
      "grad_norm": 1.109375,
      "learning_rate": 0.00043071126247585866,
      "loss": 4.7788,
      "mean_token_accuracy": 0.23612050265073775,
      "num_tokens": 62982928.0,
      "step": 27465
    },
    {
      "entropy": 5.0416669845581055,
      "epoch": 2.6388088376561,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000430686513100513,
      "loss": 4.6815,
      "mean_token_accuracy": 0.2357673853635788,
      "num_tokens": 62993326.0,
      "step": 27470
    },
    {
      "entropy": 5.192589139938354,
      "epoch": 2.639289145052834,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043066176011058877,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22165786772966384,
      "num_tokens": 63006512.0,
      "step": 27475
    },
    {
      "entropy": 5.158627271652222,
      "epoch": 2.639769452449568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043063700350666066,
      "loss": 4.7884,
      "mean_token_accuracy": 0.240103143453598,
      "num_tokens": 63017439.0,
      "step": 27480
    },
    {
      "entropy": 5.0665487289428714,
      "epoch": 2.6402497598463017,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004306122432893036,
      "loss": 4.671,
      "mean_token_accuracy": 0.24064694792032243,
      "num_tokens": 63028186.0,
      "step": 27485
    },
    {
      "entropy": 5.023597288131714,
      "epoch": 2.6407300672430356,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043058747945909224,
      "loss": 4.73,
      "mean_token_accuracy": 0.24087975174188614,
      "num_tokens": 63039864.0,
      "step": 27490
    },
    {
      "entropy": 5.086215400695801,
      "epoch": 2.6412103746397695,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043056271201660166,
      "loss": 4.7486,
      "mean_token_accuracy": 0.23880534917116164,
      "num_tokens": 63051965.0,
      "step": 27495
    },
    {
      "entropy": 5.06019344329834,
      "epoch": 2.6416906820365034,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004305379409624068,
      "loss": 4.769,
      "mean_token_accuracy": 0.23698799014091493,
      "num_tokens": 63063116.0,
      "step": 27500
    },
    {
      "entropy": 5.092842102050781,
      "epoch": 2.6421709894332372,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004305131662970827,
      "loss": 4.7888,
      "mean_token_accuracy": 0.23076210618019105,
      "num_tokens": 63074680.0,
      "step": 27505
    },
    {
      "entropy": 5.059303903579712,
      "epoch": 2.642651296829971,
      "grad_norm": 1.0,
      "learning_rate": 0.0004304883880212048,
      "loss": 4.7257,
      "mean_token_accuracy": 0.2399595558643341,
      "num_tokens": 63086470.0,
      "step": 27510
    },
    {
      "entropy": 5.1710975646972654,
      "epoch": 2.643131604226705,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004304636061353482,
      "loss": 4.8083,
      "mean_token_accuracy": 0.22983134537935257,
      "num_tokens": 63097616.0,
      "step": 27515
    },
    {
      "entropy": 5.112563610076904,
      "epoch": 2.643611911623439,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004304388206400883,
      "loss": 4.778,
      "mean_token_accuracy": 0.235069440305233,
      "num_tokens": 63109792.0,
      "step": 27520
    },
    {
      "entropy": 5.120311832427978,
      "epoch": 2.6440922190201728,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004304140315360005,
      "loss": 4.8604,
      "mean_token_accuracy": 0.22914216965436934,
      "num_tokens": 63122491.0,
      "step": 27525
    },
    {
      "entropy": 5.174773120880127,
      "epoch": 2.6445725264169067,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004303892388236604,
      "loss": 4.7927,
      "mean_token_accuracy": 0.22593145966529846,
      "num_tokens": 63134042.0,
      "step": 27530
    },
    {
      "entropy": 5.042821598052979,
      "epoch": 2.645052833813641,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004303644425036436,
      "loss": 4.6207,
      "mean_token_accuracy": 0.24612151831388474,
      "num_tokens": 63145885.0,
      "step": 27535
    },
    {
      "entropy": 5.0365358829498295,
      "epoch": 2.6455331412103744,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00043033964257652575,
      "loss": 4.716,
      "mean_token_accuracy": 0.23677686154842376,
      "num_tokens": 63156995.0,
      "step": 27540
    },
    {
      "entropy": 5.072029638290405,
      "epoch": 2.6460134486071087,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004303148390428827,
      "loss": 4.797,
      "mean_token_accuracy": 0.22958382070064545,
      "num_tokens": 63168604.0,
      "step": 27545
    },
    {
      "entropy": 5.135651540756226,
      "epoch": 2.6464937560038426,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043029003190329023,
      "loss": 4.6977,
      "mean_token_accuracy": 0.23593846708536148,
      "num_tokens": 63181284.0,
      "step": 27550
    },
    {
      "entropy": 5.024864768981933,
      "epoch": 2.6469740634005765,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004302652211583244,
      "loss": 4.6526,
      "mean_token_accuracy": 0.24147741943597795,
      "num_tokens": 63193760.0,
      "step": 27555
    },
    {
      "entropy": 5.080756950378418,
      "epoch": 2.6474543707973104,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004302404068085612,
      "loss": 4.7483,
      "mean_token_accuracy": 0.24107961654663085,
      "num_tokens": 63205280.0,
      "step": 27560
    },
    {
      "entropy": 4.9964845180511475,
      "epoch": 2.6479346781940443,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004302155888545766,
      "loss": 4.6086,
      "mean_token_accuracy": 0.24604191929101943,
      "num_tokens": 63216253.0,
      "step": 27565
    },
    {
      "entropy": 5.0408917427062985,
      "epoch": 2.648414985590778,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000430190767296947,
      "loss": 4.7801,
      "mean_token_accuracy": 0.23574710339307786,
      "num_tokens": 63227341.0,
      "step": 27570
    },
    {
      "entropy": 5.071201419830322,
      "epoch": 2.648895292987512,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004301659421362486,
      "loss": 4.7493,
      "mean_token_accuracy": 0.23564403951168061,
      "num_tokens": 63239090.0,
      "step": 27575
    },
    {
      "entropy": 5.087450885772705,
      "epoch": 2.649375600384246,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004301411133730578,
      "loss": 4.7235,
      "mean_token_accuracy": 0.24512777626514434,
      "num_tokens": 63250077.0,
      "step": 27580
    },
    {
      "entropy": 5.027852010726929,
      "epoch": 2.64985590778098,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043011628100795093,
      "loss": 4.6816,
      "mean_token_accuracy": 0.23284629583358765,
      "num_tokens": 63261177.0,
      "step": 27585
    },
    {
      "entropy": 4.992810964584351,
      "epoch": 2.6503362151777137,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004300914450415047,
      "loss": 4.7057,
      "mean_token_accuracy": 0.2409521907567978,
      "num_tokens": 63273176.0,
      "step": 27590
    },
    {
      "entropy": 5.052920866012573,
      "epoch": 2.6508165225744476,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043006660547429565,
      "loss": 4.7795,
      "mean_token_accuracy": 0.2343478724360466,
      "num_tokens": 63284529.0,
      "step": 27595
    },
    {
      "entropy": 4.98816499710083,
      "epoch": 2.6512968299711814,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004300417623069005,
      "loss": 4.5758,
      "mean_token_accuracy": 0.25116375237703326,
      "num_tokens": 63295824.0,
      "step": 27600
    },
    {
      "entropy": 5.052648496627808,
      "epoch": 2.6517771373679153,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004300169155398959,
      "loss": 4.7304,
      "mean_token_accuracy": 0.23775628060102463,
      "num_tokens": 63307267.0,
      "step": 27605
    },
    {
      "entropy": 5.133443212509155,
      "epoch": 2.6522574447646496,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042999206517385885,
      "loss": 4.7975,
      "mean_token_accuracy": 0.23311027884483337,
      "num_tokens": 63317703.0,
      "step": 27610
    },
    {
      "entropy": 5.069368886947632,
      "epoch": 2.652737752161383,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004299672112093663,
      "loss": 4.7918,
      "mean_token_accuracy": 0.23447024524211885,
      "num_tokens": 63328677.0,
      "step": 27615
    },
    {
      "entropy": 5.174603652954102,
      "epoch": 2.6532180595581174,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042994235364699526,
      "loss": 4.873,
      "mean_token_accuracy": 0.22812706083059311,
      "num_tokens": 63339604.0,
      "step": 27620
    },
    {
      "entropy": 5.137395715713501,
      "epoch": 2.653698366954851,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004299174924873229,
      "loss": 4.8009,
      "mean_token_accuracy": 0.22735474854707718,
      "num_tokens": 63351428.0,
      "step": 27625
    },
    {
      "entropy": 5.120329999923706,
      "epoch": 2.654178674351585,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004298926277309263,
      "loss": 4.85,
      "mean_token_accuracy": 0.2302156925201416,
      "num_tokens": 63364545.0,
      "step": 27630
    },
    {
      "entropy": 5.1383545875549315,
      "epoch": 2.654658981748319,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042986775937838283,
      "loss": 4.7022,
      "mean_token_accuracy": 0.24010293632745744,
      "num_tokens": 63376063.0,
      "step": 27635
    },
    {
      "entropy": 5.0391397953033445,
      "epoch": 2.655139289145053,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004298428874302699,
      "loss": 4.6811,
      "mean_token_accuracy": 0.2477008432149887,
      "num_tokens": 63387448.0,
      "step": 27640
    },
    {
      "entropy": 5.041618967056275,
      "epoch": 2.655619596541787,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004298180118871649,
      "loss": 4.8,
      "mean_token_accuracy": 0.23564374893903733,
      "num_tokens": 63399674.0,
      "step": 27645
    },
    {
      "entropy": 5.094474744796753,
      "epoch": 2.6560999039385207,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042979313274964535,
      "loss": 4.778,
      "mean_token_accuracy": 0.23615514487028122,
      "num_tokens": 63410622.0,
      "step": 27650
    },
    {
      "entropy": 5.127264833450317,
      "epoch": 2.6565802113352546,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042976825001828897,
      "loss": 4.7698,
      "mean_token_accuracy": 0.23218918740749359,
      "num_tokens": 63422354.0,
      "step": 27655
    },
    {
      "entropy": 5.112388324737549,
      "epoch": 2.6570605187319885,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042974336369367333,
      "loss": 4.738,
      "mean_token_accuracy": 0.23901377171278,
      "num_tokens": 63433232.0,
      "step": 27660
    },
    {
      "entropy": 5.00833740234375,
      "epoch": 2.6575408261287223,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004297184737763763,
      "loss": 4.6582,
      "mean_token_accuracy": 0.243719644844532,
      "num_tokens": 63444998.0,
      "step": 27665
    },
    {
      "entropy": 5.080754804611206,
      "epoch": 2.658021133525456,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042969358026697567,
      "loss": 4.7915,
      "mean_token_accuracy": 0.23049854189157487,
      "num_tokens": 63455846.0,
      "step": 27670
    },
    {
      "entropy": 5.123204517364502,
      "epoch": 2.65850144092219,
      "grad_norm": 1.0,
      "learning_rate": 0.0004296686831660495,
      "loss": 4.7524,
      "mean_token_accuracy": 0.23260053545236586,
      "num_tokens": 63468353.0,
      "step": 27675
    },
    {
      "entropy": 5.106164360046387,
      "epoch": 2.658981748318924,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004296437824741758,
      "loss": 4.7117,
      "mean_token_accuracy": 0.24152374416589736,
      "num_tokens": 63479802.0,
      "step": 27680
    },
    {
      "entropy": 5.131553077697754,
      "epoch": 2.659462055715658,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042961887819193263,
      "loss": 4.7906,
      "mean_token_accuracy": 0.23069737255573272,
      "num_tokens": 63490856.0,
      "step": 27685
    },
    {
      "entropy": 5.1472580909729,
      "epoch": 2.6599423631123917,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004295939703198983,
      "loss": 4.8116,
      "mean_token_accuracy": 0.23062358945608138,
      "num_tokens": 63502294.0,
      "step": 27690
    },
    {
      "entropy": 5.166450500488281,
      "epoch": 2.660422670509126,
      "grad_norm": 1.0625,
      "learning_rate": 0.000429569058858651,
      "loss": 4.8154,
      "mean_token_accuracy": 0.22954794615507126,
      "num_tokens": 63513466.0,
      "step": 27695
    },
    {
      "entropy": 5.071706867218017,
      "epoch": 2.6609029779058595,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042954414380876906,
      "loss": 4.6529,
      "mean_token_accuracy": 0.2443309336900711,
      "num_tokens": 63523596.0,
      "step": 27700
    },
    {
      "entropy": 4.996892356872559,
      "epoch": 2.661383285302594,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042951922517083104,
      "loss": 4.6976,
      "mean_token_accuracy": 0.23741758167743682,
      "num_tokens": 63534468.0,
      "step": 27705
    },
    {
      "entropy": 5.055926513671875,
      "epoch": 2.6618635926993277,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004294943029454155,
      "loss": 4.714,
      "mean_token_accuracy": 0.2411523312330246,
      "num_tokens": 63546273.0,
      "step": 27710
    },
    {
      "entropy": 5.038618564605713,
      "epoch": 2.6623439000960616,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042946937713310093,
      "loss": 4.7249,
      "mean_token_accuracy": 0.23516607284545898,
      "num_tokens": 63557032.0,
      "step": 27715
    },
    {
      "entropy": 5.104950618743897,
      "epoch": 2.6628242074927955,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004294444477344661,
      "loss": 4.7583,
      "mean_token_accuracy": 0.23808481246232988,
      "num_tokens": 63568358.0,
      "step": 27720
    },
    {
      "entropy": 5.1562012195587155,
      "epoch": 2.6633045148895294,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004294195147500898,
      "loss": 4.8536,
      "mean_token_accuracy": 0.23696598410606384,
      "num_tokens": 63579716.0,
      "step": 27725
    },
    {
      "entropy": 5.022998476028443,
      "epoch": 2.6637848222862632,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042939457818055095,
      "loss": 4.6968,
      "mean_token_accuracy": 0.24504642337560653,
      "num_tokens": 63590871.0,
      "step": 27730
    },
    {
      "entropy": 5.132131195068359,
      "epoch": 2.664265129682997,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042936963802642843,
      "loss": 4.8017,
      "mean_token_accuracy": 0.23633942306041716,
      "num_tokens": 63600813.0,
      "step": 27735
    },
    {
      "entropy": 5.013393068313599,
      "epoch": 2.664745437079731,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004293446942883013,
      "loss": 4.623,
      "mean_token_accuracy": 0.24767047613859178,
      "num_tokens": 63611141.0,
      "step": 27740
    },
    {
      "entropy": 5.024427700042724,
      "epoch": 2.665225744476465,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042931974696674866,
      "loss": 4.6566,
      "mean_token_accuracy": 0.2442573681473732,
      "num_tokens": 63622172.0,
      "step": 27745
    },
    {
      "entropy": 5.029464960098267,
      "epoch": 2.6657060518731988,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004292947960623497,
      "loss": 4.766,
      "mean_token_accuracy": 0.22955911308526994,
      "num_tokens": 63634369.0,
      "step": 27750
    },
    {
      "entropy": 5.013013410568237,
      "epoch": 2.6661863592699326,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00042926984157568384,
      "loss": 4.6828,
      "mean_token_accuracy": 0.24699949026107787,
      "num_tokens": 63645142.0,
      "step": 27755
    },
    {
      "entropy": 5.040130567550659,
      "epoch": 2.6666666666666665,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042924488350733024,
      "loss": 4.6924,
      "mean_token_accuracy": 0.2465788960456848,
      "num_tokens": 63656497.0,
      "step": 27760
    },
    {
      "entropy": 5.128740072250366,
      "epoch": 2.6671469740634004,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042921992185786847,
      "loss": 4.8053,
      "mean_token_accuracy": 0.2364078238606453,
      "num_tokens": 63667795.0,
      "step": 27765
    },
    {
      "entropy": 5.033998727798462,
      "epoch": 2.6676272814601347,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042919495662787813,
      "loss": 4.7814,
      "mean_token_accuracy": 0.24103656858205796,
      "num_tokens": 63679676.0,
      "step": 27770
    },
    {
      "entropy": 5.123899698257446,
      "epoch": 2.668107588856868,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004291699878179387,
      "loss": 4.7454,
      "mean_token_accuracy": 0.2307498723268509,
      "num_tokens": 63690749.0,
      "step": 27775
    },
    {
      "entropy": 5.04762897491455,
      "epoch": 2.6685878962536025,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004291450154286299,
      "loss": 4.6518,
      "mean_token_accuracy": 0.2374918594956398,
      "num_tokens": 63702583.0,
      "step": 27780
    },
    {
      "entropy": 5.07662935256958,
      "epoch": 2.6690682036503364,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004291200394605317,
      "loss": 4.7367,
      "mean_token_accuracy": 0.23824533969163894,
      "num_tokens": 63715296.0,
      "step": 27785
    },
    {
      "entropy": 5.089427471160889,
      "epoch": 2.6695485110470702,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004290950599142237,
      "loss": 4.6695,
      "mean_token_accuracy": 0.24849221259355544,
      "num_tokens": 63726261.0,
      "step": 27790
    },
    {
      "entropy": 5.151883697509765,
      "epoch": 2.670028818443804,
      "grad_norm": 1.0,
      "learning_rate": 0.000429070076790286,
      "loss": 4.8607,
      "mean_token_accuracy": 0.22631134688854218,
      "num_tokens": 63738482.0,
      "step": 27795
    },
    {
      "entropy": 5.114259576797485,
      "epoch": 2.670509125840538,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042904509008929873,
      "loss": 4.8321,
      "mean_token_accuracy": 0.22664612531661987,
      "num_tokens": 63748899.0,
      "step": 27800
    },
    {
      "entropy": 5.157378387451172,
      "epoch": 2.670989433237272,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004290200998118417,
      "loss": 4.7715,
      "mean_token_accuracy": 0.23725060224533082,
      "num_tokens": 63760427.0,
      "step": 27805
    },
    {
      "entropy": 5.066434383392334,
      "epoch": 2.6714697406340058,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042899510595849544,
      "loss": 4.7339,
      "mean_token_accuracy": 0.24523456990718842,
      "num_tokens": 63772226.0,
      "step": 27810
    },
    {
      "entropy": 5.141171169281006,
      "epoch": 2.6719500480307397,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042897010852984004,
      "loss": 4.8289,
      "mean_token_accuracy": 0.23247911483049394,
      "num_tokens": 63782918.0,
      "step": 27815
    },
    {
      "entropy": 5.063779067993164,
      "epoch": 2.6724303554274735,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00042894510752645586,
      "loss": 4.7098,
      "mean_token_accuracy": 0.2357715204358101,
      "num_tokens": 63793461.0,
      "step": 27820
    },
    {
      "entropy": 5.025809001922608,
      "epoch": 2.6729106628242074,
      "grad_norm": 1.0,
      "learning_rate": 0.0004289201029489235,
      "loss": 4.6748,
      "mean_token_accuracy": 0.24604803919792176,
      "num_tokens": 63804648.0,
      "step": 27825
    },
    {
      "entropy": 5.095930910110473,
      "epoch": 2.6733909702209413,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004288950947978234,
      "loss": 4.7509,
      "mean_token_accuracy": 0.2331978499889374,
      "num_tokens": 63815809.0,
      "step": 27830
    },
    {
      "entropy": 5.059578943252563,
      "epoch": 2.673871277617675,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004288700830737361,
      "loss": 4.7053,
      "mean_token_accuracy": 0.24428513944149016,
      "num_tokens": 63828310.0,
      "step": 27835
    },
    {
      "entropy": 5.019851732254028,
      "epoch": 2.674351585014409,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042884506777724244,
      "loss": 4.69,
      "mean_token_accuracy": 0.23837850391864776,
      "num_tokens": 63840546.0,
      "step": 27840
    },
    {
      "entropy": 4.982884740829467,
      "epoch": 2.6748318924111434,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004288200489089231,
      "loss": 4.6206,
      "mean_token_accuracy": 0.24463415145874023,
      "num_tokens": 63852049.0,
      "step": 27845
    },
    {
      "entropy": 5.065077781677246,
      "epoch": 2.675312199807877,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000428795026469359,
      "loss": 4.716,
      "mean_token_accuracy": 0.24002256989479065,
      "num_tokens": 63863474.0,
      "step": 27850
    },
    {
      "entropy": 5.049188232421875,
      "epoch": 2.675792507204611,
      "grad_norm": 1.0625,
      "learning_rate": 0.000428770000459131,
      "loss": 4.7099,
      "mean_token_accuracy": 0.23868059664964675,
      "num_tokens": 63874906.0,
      "step": 27855
    },
    {
      "entropy": 5.045403289794922,
      "epoch": 2.6762728146013446,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004287449708788202,
      "loss": 4.7225,
      "mean_token_accuracy": 0.23574875742197038,
      "num_tokens": 63885443.0,
      "step": 27860
    },
    {
      "entropy": 5.0511750221252445,
      "epoch": 2.676753121998079,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004287199377290077,
      "loss": 4.7667,
      "mean_token_accuracy": 0.2368997722864151,
      "num_tokens": 63897772.0,
      "step": 27865
    },
    {
      "entropy": 5.187035799026489,
      "epoch": 2.677233429394813,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004286949010102748,
      "loss": 4.784,
      "mean_token_accuracy": 0.23247506469488144,
      "num_tokens": 63910409.0,
      "step": 27870
    },
    {
      "entropy": 5.0656898021698,
      "epoch": 2.6777137367915467,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004286698607232026,
      "loss": 4.7202,
      "mean_token_accuracy": 0.23347078263759613,
      "num_tokens": 63922671.0,
      "step": 27875
    },
    {
      "entropy": 5.112219524383545,
      "epoch": 2.6781940441882806,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042864481686837253,
      "loss": 4.7525,
      "mean_token_accuracy": 0.2383256658911705,
      "num_tokens": 63933964.0,
      "step": 27880
    },
    {
      "entropy": 5.114360618591308,
      "epoch": 2.6786743515850144,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042861976944636604,
      "loss": 4.7422,
      "mean_token_accuracy": 0.2366602525115013,
      "num_tokens": 63945347.0,
      "step": 27885
    },
    {
      "entropy": 5.101795673370361,
      "epoch": 2.6791546589817483,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004285947184577647,
      "loss": 4.761,
      "mean_token_accuracy": 0.23288916200399398,
      "num_tokens": 63957573.0,
      "step": 27890
    },
    {
      "entropy": 5.076748991012574,
      "epoch": 2.679634966378482,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042856966390315013,
      "loss": 4.7237,
      "mean_token_accuracy": 0.24330953359603882,
      "num_tokens": 63968666.0,
      "step": 27895
    },
    {
      "entropy": 4.9992852210998535,
      "epoch": 2.680115273775216,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004285446057831039,
      "loss": 4.7099,
      "mean_token_accuracy": 0.23870651423931122,
      "num_tokens": 63981147.0,
      "step": 27900
    },
    {
      "entropy": 5.1610933303833,
      "epoch": 2.68059558117195,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004285195440982078,
      "loss": 4.8479,
      "mean_token_accuracy": 0.22380560338497163,
      "num_tokens": 63992474.0,
      "step": 27905
    },
    {
      "entropy": 5.11380968093872,
      "epoch": 2.681075888568684,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004284944788490439,
      "loss": 4.7121,
      "mean_token_accuracy": 0.23583726435899735,
      "num_tokens": 64003804.0,
      "step": 27910
    },
    {
      "entropy": 5.022637033462525,
      "epoch": 2.6815561959654177,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004284694100361938,
      "loss": 4.7376,
      "mean_token_accuracy": 0.24203008115291597,
      "num_tokens": 64015587.0,
      "step": 27915
    },
    {
      "entropy": 5.133114290237427,
      "epoch": 2.682036503362152,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004284443376602398,
      "loss": 4.8344,
      "mean_token_accuracy": 0.23529283106327056,
      "num_tokens": 64027088.0,
      "step": 27920
    },
    {
      "entropy": 5.104663991928101,
      "epoch": 2.6825168107588855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004284192617217639,
      "loss": 4.7074,
      "mean_token_accuracy": 0.24350042045116424,
      "num_tokens": 64037050.0,
      "step": 27925
    },
    {
      "entropy": 5.0143725872039795,
      "epoch": 2.68299711815562,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004283941822213484,
      "loss": 4.7627,
      "mean_token_accuracy": 0.2344140186905861,
      "num_tokens": 64049287.0,
      "step": 27930
    },
    {
      "entropy": 5.106902456283569,
      "epoch": 2.6834774255523532,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004283690991595754,
      "loss": 4.7306,
      "mean_token_accuracy": 0.2405501648783684,
      "num_tokens": 64060578.0,
      "step": 27935
    },
    {
      "entropy": 5.135268926620483,
      "epoch": 2.6839577329490876,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00042834401253702734,
      "loss": 4.8284,
      "mean_token_accuracy": 0.22756471186876298,
      "num_tokens": 64071540.0,
      "step": 27940
    },
    {
      "entropy": 5.124144792556763,
      "epoch": 2.6844380403458215,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004283189223542866,
      "loss": 4.7488,
      "mean_token_accuracy": 0.2370336353778839,
      "num_tokens": 64084014.0,
      "step": 27945
    },
    {
      "entropy": 5.093026733398437,
      "epoch": 2.6849183477425553,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042829382861193585,
      "loss": 4.782,
      "mean_token_accuracy": 0.23569778800010682,
      "num_tokens": 64096735.0,
      "step": 27950
    },
    {
      "entropy": 5.145774412155151,
      "epoch": 2.685398655139289,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004282687313105575,
      "loss": 4.8388,
      "mean_token_accuracy": 0.2266838639974594,
      "num_tokens": 64107548.0,
      "step": 27955
    },
    {
      "entropy": 5.07238073348999,
      "epoch": 2.685878962536023,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042824363045073434,
      "loss": 4.7566,
      "mean_token_accuracy": 0.23895136564970015,
      "num_tokens": 64118433.0,
      "step": 27960
    },
    {
      "entropy": 5.112598848342896,
      "epoch": 2.686359269932757,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004282185260330491,
      "loss": 4.7367,
      "mean_token_accuracy": 0.24243899136781694,
      "num_tokens": 64129883.0,
      "step": 27965
    },
    {
      "entropy": 5.115155267715454,
      "epoch": 2.686839577329491,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042819341805808473,
      "loss": 4.804,
      "mean_token_accuracy": 0.23820121139287947,
      "num_tokens": 64141827.0,
      "step": 27970
    },
    {
      "entropy": 5.052349328994751,
      "epoch": 2.6873198847262247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042816830652642396,
      "loss": 4.7207,
      "mean_token_accuracy": 0.23847874402999877,
      "num_tokens": 64152957.0,
      "step": 27975
    },
    {
      "entropy": 5.043601608276367,
      "epoch": 2.6878001921229586,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004281431914386501,
      "loss": 4.684,
      "mean_token_accuracy": 0.23784219473600388,
      "num_tokens": 64164800.0,
      "step": 27980
    },
    {
      "entropy": 5.135885286331177,
      "epoch": 2.6882804995196925,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000428118072795346,
      "loss": 4.751,
      "mean_token_accuracy": 0.2394149973988533,
      "num_tokens": 64175941.0,
      "step": 27985
    },
    {
      "entropy": 5.076216411590576,
      "epoch": 2.6887608069164264,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042809295059709483,
      "loss": 4.7011,
      "mean_token_accuracy": 0.23967360556125641,
      "num_tokens": 64187229.0,
      "step": 27990
    },
    {
      "entropy": 5.096039819717407,
      "epoch": 2.6892411143131603,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00042806782484448,
      "loss": 4.7685,
      "mean_token_accuracy": 0.2397293418645859,
      "num_tokens": 64198328.0,
      "step": 27995
    },
    {
      "entropy": 5.058514881134033,
      "epoch": 2.689721421709894,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004280426955380848,
      "loss": 4.6657,
      "mean_token_accuracy": 0.24381324648857117,
      "num_tokens": 64207434.0,
      "step": 28000
    },
    {
      "entropy": 5.004641914367676,
      "epoch": 2.6902017291066285,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042801756267849266,
      "loss": 4.7001,
      "mean_token_accuracy": 0.2456513911485672,
      "num_tokens": 64219128.0,
      "step": 28005
    },
    {
      "entropy": 5.057278156280518,
      "epoch": 2.690682036503362,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004279924262662871,
      "loss": 4.7282,
      "mean_token_accuracy": 0.23628194779157638,
      "num_tokens": 64230664.0,
      "step": 28010
    },
    {
      "entropy": 5.094168376922608,
      "epoch": 2.6911623439000962,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004279672863020517,
      "loss": 4.7081,
      "mean_token_accuracy": 0.2431165859103203,
      "num_tokens": 64242339.0,
      "step": 28015
    },
    {
      "entropy": 5.042585849761963,
      "epoch": 2.69164265129683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042794214278637013,
      "loss": 4.7659,
      "mean_token_accuracy": 0.23016472309827804,
      "num_tokens": 64253005.0,
      "step": 28020
    },
    {
      "entropy": 5.064193058013916,
      "epoch": 2.692122958693564,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042791699571982606,
      "loss": 4.7896,
      "mean_token_accuracy": 0.23669118732213973,
      "num_tokens": 64264791.0,
      "step": 28025
    },
    {
      "entropy": 5.0561244010925295,
      "epoch": 2.692603266090298,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004278918451030035,
      "loss": 4.6788,
      "mean_token_accuracy": 0.24473610669374465,
      "num_tokens": 64274501.0,
      "step": 28030
    },
    {
      "entropy": 5.029261732101441,
      "epoch": 2.6930835734870318,
      "grad_norm": 1.0,
      "learning_rate": 0.0004278666909364863,
      "loss": 4.6539,
      "mean_token_accuracy": 0.24605002850294114,
      "num_tokens": 64284501.0,
      "step": 28035
    },
    {
      "entropy": 5.0177947044372555,
      "epoch": 2.6935638808837656,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004278415332208584,
      "loss": 4.6786,
      "mean_token_accuracy": 0.24173179864883423,
      "num_tokens": 64295627.0,
      "step": 28040
    },
    {
      "entropy": 5.155442905426026,
      "epoch": 2.6940441882804995,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042781637195670396,
      "loss": 4.7975,
      "mean_token_accuracy": 0.22887688875198364,
      "num_tokens": 64307060.0,
      "step": 28045
    },
    {
      "entropy": 5.115978527069092,
      "epoch": 2.6945244956772334,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004277912071446072,
      "loss": 4.8181,
      "mean_token_accuracy": 0.2378630295395851,
      "num_tokens": 64318329.0,
      "step": 28050
    },
    {
      "entropy": 5.1363606452941895,
      "epoch": 2.6950048030739673,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004277660387851522,
      "loss": 4.7491,
      "mean_token_accuracy": 0.2381145864725113,
      "num_tokens": 64329871.0,
      "step": 28055
    },
    {
      "entropy": 4.99898042678833,
      "epoch": 2.695485110470701,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004277408668789235,
      "loss": 4.7028,
      "mean_token_accuracy": 0.24534677863121032,
      "num_tokens": 64341950.0,
      "step": 28060
    },
    {
      "entropy": 5.06464786529541,
      "epoch": 2.695965417867435,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004277156914265054,
      "loss": 4.7755,
      "mean_token_accuracy": 0.2400926396250725,
      "num_tokens": 64353841.0,
      "step": 28065
    },
    {
      "entropy": 5.091238069534302,
      "epoch": 2.696445725264169,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004276905124284824,
      "loss": 4.7038,
      "mean_token_accuracy": 0.24164751470088958,
      "num_tokens": 64365914.0,
      "step": 28070
    },
    {
      "entropy": 5.000883626937866,
      "epoch": 2.696926032660903,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004276653298854391,
      "loss": 4.6419,
      "mean_token_accuracy": 0.24780778586864471,
      "num_tokens": 64376321.0,
      "step": 28075
    },
    {
      "entropy": 5.060827827453613,
      "epoch": 2.697406340057637,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004276401437979601,
      "loss": 4.784,
      "mean_token_accuracy": 0.2326599583029747,
      "num_tokens": 64387344.0,
      "step": 28080
    },
    {
      "entropy": 5.043079376220703,
      "epoch": 2.6978866474543706,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004276149541666303,
      "loss": 4.6918,
      "mean_token_accuracy": 0.2387310892343521,
      "num_tokens": 64398553.0,
      "step": 28085
    },
    {
      "entropy": 5.170830631256104,
      "epoch": 2.698366954851105,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042758976099203444,
      "loss": 4.8183,
      "mean_token_accuracy": 0.2365676835179329,
      "num_tokens": 64409174.0,
      "step": 28090
    },
    {
      "entropy": 5.05456576347351,
      "epoch": 2.6988472622478388,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042756456427475736,
      "loss": 4.6389,
      "mean_token_accuracy": 0.24419716149568557,
      "num_tokens": 64420840.0,
      "step": 28095
    },
    {
      "entropy": 5.059036588668823,
      "epoch": 2.6993275696445727,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004275393640153842,
      "loss": 4.6448,
      "mean_token_accuracy": 0.24552091360092163,
      "num_tokens": 64431796.0,
      "step": 28100
    },
    {
      "entropy": 5.048649549484253,
      "epoch": 2.6998078770413065,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042751416021449986,
      "loss": 4.7135,
      "mean_token_accuracy": 0.23449796289205552,
      "num_tokens": 64443675.0,
      "step": 28105
    },
    {
      "entropy": 5.05757360458374,
      "epoch": 2.7002881844380404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004274889528726896,
      "loss": 4.7034,
      "mean_token_accuracy": 0.25074206441640856,
      "num_tokens": 64455556.0,
      "step": 28110
    },
    {
      "entropy": 5.05715537071228,
      "epoch": 2.7007684918347743,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004274637419905388,
      "loss": 4.7248,
      "mean_token_accuracy": 0.2413482129573822,
      "num_tokens": 64466873.0,
      "step": 28115
    },
    {
      "entropy": 5.0684957027435305,
      "epoch": 2.701248799231508,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042743852756863253,
      "loss": 4.7472,
      "mean_token_accuracy": 0.2425170436501503,
      "num_tokens": 64478530.0,
      "step": 28120
    },
    {
      "entropy": 5.089036273956299,
      "epoch": 2.701729106628242,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004274133096075563,
      "loss": 4.7558,
      "mean_token_accuracy": 0.23780724853277208,
      "num_tokens": 64488941.0,
      "step": 28125
    },
    {
      "entropy": 5.021741580963135,
      "epoch": 2.702209414024976,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004273880881078956,
      "loss": 4.6698,
      "mean_token_accuracy": 0.24341011196374893,
      "num_tokens": 64500727.0,
      "step": 28130
    },
    {
      "entropy": 5.016228723526001,
      "epoch": 2.70268972142171,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004273628630702359,
      "loss": 4.7235,
      "mean_token_accuracy": 0.24050071984529495,
      "num_tokens": 64513381.0,
      "step": 28135
    },
    {
      "entropy": 5.121650075912475,
      "epoch": 2.7031700288184437,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042733763449516313,
      "loss": 4.7427,
      "mean_token_accuracy": 0.23917580097913743,
      "num_tokens": 64525442.0,
      "step": 28140
    },
    {
      "entropy": 5.126661777496338,
      "epoch": 2.7036503362151776,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042731240238326273,
      "loss": 4.6835,
      "mean_token_accuracy": 0.23727269917726518,
      "num_tokens": 64536008.0,
      "step": 28145
    },
    {
      "entropy": 5.054746103286743,
      "epoch": 2.7041306436119115,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042728716673512065,
      "loss": 4.8306,
      "mean_token_accuracy": 0.2331784874200821,
      "num_tokens": 64547161.0,
      "step": 28150
    },
    {
      "entropy": 5.0753453254699705,
      "epoch": 2.704610951008646,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042726192755132276,
      "loss": 4.7373,
      "mean_token_accuracy": 0.2399827718734741,
      "num_tokens": 64558152.0,
      "step": 28155
    },
    {
      "entropy": 5.187208795547486,
      "epoch": 2.7050912584053792,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042723668483245496,
      "loss": 4.8887,
      "mean_token_accuracy": 0.22610936015844346,
      "num_tokens": 64569150.0,
      "step": 28160
    },
    {
      "entropy": 5.153425216674805,
      "epoch": 2.7055715658021136,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004272114385791035,
      "loss": 4.7306,
      "mean_token_accuracy": 0.24405432045459746,
      "num_tokens": 64580417.0,
      "step": 28165
    },
    {
      "entropy": 5.0604462146759035,
      "epoch": 2.706051873198847,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042718618879185435,
      "loss": 4.6968,
      "mean_token_accuracy": 0.23447103053331375,
      "num_tokens": 64591628.0,
      "step": 28170
    },
    {
      "entropy": 5.0683653354644775,
      "epoch": 2.7065321805955813,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004271609354712938,
      "loss": 4.7175,
      "mean_token_accuracy": 0.23774780631065368,
      "num_tokens": 64602801.0,
      "step": 28175
    },
    {
      "entropy": 5.08750433921814,
      "epoch": 2.707012487992315,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004271356786180082,
      "loss": 4.7358,
      "mean_token_accuracy": 0.24326564520597457,
      "num_tokens": 64614327.0,
      "step": 28180
    },
    {
      "entropy": 5.056805038452149,
      "epoch": 2.707492795389049,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004271104182325838,
      "loss": 4.7742,
      "mean_token_accuracy": 0.23494082689285278,
      "num_tokens": 64625508.0,
      "step": 28185
    },
    {
      "entropy": 5.083606195449829,
      "epoch": 2.707973102785783,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00042708515431560723,
      "loss": 4.7965,
      "mean_token_accuracy": 0.23486795723438264,
      "num_tokens": 64637428.0,
      "step": 28190
    },
    {
      "entropy": 5.013066959381104,
      "epoch": 2.708453410182517,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000427059886867665,
      "loss": 4.7374,
      "mean_token_accuracy": 0.236607863008976,
      "num_tokens": 64648189.0,
      "step": 28195
    },
    {
      "entropy": 5.038277053833008,
      "epoch": 2.7089337175792507,
      "grad_norm": 1.0,
      "learning_rate": 0.0004270346158893436,
      "loss": 4.6811,
      "mean_token_accuracy": 0.24200159460306167,
      "num_tokens": 64659250.0,
      "step": 28200
    },
    {
      "entropy": 5.047351360321045,
      "epoch": 2.7094140249759846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042700934138123004,
      "loss": 4.6729,
      "mean_token_accuracy": 0.24861670881509781,
      "num_tokens": 64669746.0,
      "step": 28205
    },
    {
      "entropy": 5.071422147750854,
      "epoch": 2.7098943323727185,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042698406334391084,
      "loss": 4.7106,
      "mean_token_accuracy": 0.23719117641448975,
      "num_tokens": 64681758.0,
      "step": 28210
    },
    {
      "entropy": 5.142051410675049,
      "epoch": 2.7103746397694524,
      "grad_norm": 1.015625,
      "learning_rate": 0.000426958781777973,
      "loss": 4.757,
      "mean_token_accuracy": 0.23932647407054902,
      "num_tokens": 64692944.0,
      "step": 28215
    },
    {
      "entropy": 5.192587089538574,
      "epoch": 2.7108549471661862,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004269334966840035,
      "loss": 4.8913,
      "mean_token_accuracy": 0.2255040556192398,
      "num_tokens": 64705436.0,
      "step": 28220
    },
    {
      "entropy": 5.130615234375,
      "epoch": 2.71133525456292,
      "grad_norm": 1.0,
      "learning_rate": 0.00042690820806258933,
      "loss": 4.7472,
      "mean_token_accuracy": 0.23265192210674285,
      "num_tokens": 64716542.0,
      "step": 28225
    },
    {
      "entropy": 5.043447589874267,
      "epoch": 2.7118155619596545,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004268829159143176,
      "loss": 4.6783,
      "mean_token_accuracy": 0.24333194494247437,
      "num_tokens": 64727800.0,
      "step": 28230
    },
    {
      "entropy": 5.204685401916504,
      "epoch": 2.712295869356388,
      "grad_norm": 1.0,
      "learning_rate": 0.0004268576202397757,
      "loss": 4.88,
      "mean_token_accuracy": 0.22966899573802949,
      "num_tokens": 64739790.0,
      "step": 28235
    },
    {
      "entropy": 5.042264795303344,
      "epoch": 2.712776176753122,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004268323210395506,
      "loss": 4.7063,
      "mean_token_accuracy": 0.23843299299478532,
      "num_tokens": 64751031.0,
      "step": 28240
    },
    {
      "entropy": 5.033315706253052,
      "epoch": 2.7132564841498557,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042680701831423004,
      "loss": 4.7306,
      "mean_token_accuracy": 0.23770836740732193,
      "num_tokens": 64761751.0,
      "step": 28245
    },
    {
      "entropy": 5.201358032226563,
      "epoch": 2.71373679154659,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004267817120644012,
      "loss": 4.7731,
      "mean_token_accuracy": 0.23969102203845977,
      "num_tokens": 64773257.0,
      "step": 28250
    },
    {
      "entropy": 5.15490984916687,
      "epoch": 2.714217098943324,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042675640229065167,
      "loss": 4.7475,
      "mean_token_accuracy": 0.22894269078969956,
      "num_tokens": 64783409.0,
      "step": 28255
    },
    {
      "entropy": 5.074130201339722,
      "epoch": 2.7146974063400577,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042673108899356915,
      "loss": 4.7997,
      "mean_token_accuracy": 0.2282678261399269,
      "num_tokens": 64794392.0,
      "step": 28260
    },
    {
      "entropy": 5.083381843566895,
      "epoch": 2.7151777137367916,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004267057721737413,
      "loss": 4.7854,
      "mean_token_accuracy": 0.2277814745903015,
      "num_tokens": 64806289.0,
      "step": 28265
    },
    {
      "entropy": 5.073916006088257,
      "epoch": 2.7156580211335255,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004266804518317559,
      "loss": 4.7625,
      "mean_token_accuracy": 0.2352558344602585,
      "num_tokens": 64817278.0,
      "step": 28270
    },
    {
      "entropy": 5.095240831375122,
      "epoch": 2.7161383285302594,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004266551279682008,
      "loss": 4.7007,
      "mean_token_accuracy": 0.23430538922548294,
      "num_tokens": 64828889.0,
      "step": 28275
    },
    {
      "entropy": 5.010385274887085,
      "epoch": 2.7166186359269933,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004266298005836639,
      "loss": 4.7097,
      "mean_token_accuracy": 0.23801858127117156,
      "num_tokens": 64840260.0,
      "step": 28280
    },
    {
      "entropy": 5.108109331130981,
      "epoch": 2.717098943323727,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042660446967873327,
      "loss": 4.792,
      "mean_token_accuracy": 0.23156808465719222,
      "num_tokens": 64853099.0,
      "step": 28285
    },
    {
      "entropy": 5.167552900314331,
      "epoch": 2.717579250720461,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042657913525399703,
      "loss": 4.7219,
      "mean_token_accuracy": 0.24583050906658171,
      "num_tokens": 64863501.0,
      "step": 28290
    },
    {
      "entropy": 5.027602910995483,
      "epoch": 2.718059558117195,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004265537973100435,
      "loss": 4.7214,
      "mean_token_accuracy": 0.24423708766698837,
      "num_tokens": 64874071.0,
      "step": 28295
    },
    {
      "entropy": 5.14169921875,
      "epoch": 2.718539865513929,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004265284558474607,
      "loss": 4.7654,
      "mean_token_accuracy": 0.23332432061433792,
      "num_tokens": 64885238.0,
      "step": 28300
    },
    {
      "entropy": 5.144085550308228,
      "epoch": 2.7190201729106627,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042650311086683715,
      "loss": 4.8359,
      "mean_token_accuracy": 0.2360248476266861,
      "num_tokens": 64896384.0,
      "step": 28305
    },
    {
      "entropy": 5.131429958343506,
      "epoch": 2.7195004803073966,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004264777623687612,
      "loss": 4.7507,
      "mean_token_accuracy": 0.23800956457853317,
      "num_tokens": 64907219.0,
      "step": 28310
    },
    {
      "entropy": 5.084659671783447,
      "epoch": 2.719980787704131,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004264524103538214,
      "loss": 4.6934,
      "mean_token_accuracy": 0.24469608515501023,
      "num_tokens": 64918637.0,
      "step": 28315
    },
    {
      "entropy": 5.049862480163574,
      "epoch": 2.7204610951008643,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004264270548226064,
      "loss": 4.7922,
      "mean_token_accuracy": 0.24098494946956633,
      "num_tokens": 64931532.0,
      "step": 28320
    },
    {
      "entropy": 5.095480346679688,
      "epoch": 2.7209414024975986,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004264016957757048,
      "loss": 4.7435,
      "mean_token_accuracy": 0.23920888006687163,
      "num_tokens": 64943214.0,
      "step": 28325
    },
    {
      "entropy": 5.147266674041748,
      "epoch": 2.7214217098943325,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042637633321370545,
      "loss": 4.7872,
      "mean_token_accuracy": 0.23652815371751784,
      "num_tokens": 64954726.0,
      "step": 28330
    },
    {
      "entropy": 5.118739938735962,
      "epoch": 2.7219020172910664,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004263509671371971,
      "loss": 4.7897,
      "mean_token_accuracy": 0.22994960397481917,
      "num_tokens": 64967759.0,
      "step": 28335
    },
    {
      "entropy": 5.176482677459717,
      "epoch": 2.7223823246878003,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042632559754676865,
      "loss": 4.8877,
      "mean_token_accuracy": 0.22129809260368347,
      "num_tokens": 64979431.0,
      "step": 28340
    },
    {
      "entropy": 5.033841800689697,
      "epoch": 2.722862632084534,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004263002244430092,
      "loss": 4.7171,
      "mean_token_accuracy": 0.24446092844009398,
      "num_tokens": 64992148.0,
      "step": 28345
    },
    {
      "entropy": 5.060359716415405,
      "epoch": 2.723342939481268,
      "grad_norm": 1.0,
      "learning_rate": 0.0004262748478265078,
      "loss": 4.6482,
      "mean_token_accuracy": 0.2473047122359276,
      "num_tokens": 65003750.0,
      "step": 28350
    },
    {
      "entropy": 5.0065044403076175,
      "epoch": 2.723823246878002,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004262494676978537,
      "loss": 4.7,
      "mean_token_accuracy": 0.24132362455129625,
      "num_tokens": 65013921.0,
      "step": 28355
    },
    {
      "entropy": 5.036620378494263,
      "epoch": 2.724303554274736,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042622408405763607,
      "loss": 4.7058,
      "mean_token_accuracy": 0.2421443819999695,
      "num_tokens": 65025043.0,
      "step": 28360
    },
    {
      "entropy": 5.056824159622193,
      "epoch": 2.7247838616714697,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004261986969064442,
      "loss": 4.6417,
      "mean_token_accuracy": 0.24701725244522094,
      "num_tokens": 65037155.0,
      "step": 28365
    },
    {
      "entropy": 5.020095682144165,
      "epoch": 2.7252641690682036,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042617330624486753,
      "loss": 4.7307,
      "mean_token_accuracy": 0.23279804587364197,
      "num_tokens": 65048876.0,
      "step": 28370
    },
    {
      "entropy": 5.122087049484253,
      "epoch": 2.7257444764649374,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004261479120734956,
      "loss": 4.7768,
      "mean_token_accuracy": 0.23524128794670104,
      "num_tokens": 65060321.0,
      "step": 28375
    },
    {
      "entropy": 5.136285972595215,
      "epoch": 2.7262247838616713,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000426122514392918,
      "loss": 4.7833,
      "mean_token_accuracy": 0.23528686314821243,
      "num_tokens": 65073019.0,
      "step": 28380
    },
    {
      "entropy": 5.0793074607849125,
      "epoch": 2.726705091258405,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042609711320372435,
      "loss": 4.7483,
      "mean_token_accuracy": 0.23916420489549636,
      "num_tokens": 65084597.0,
      "step": 28385
    },
    {
      "entropy": 5.091636514663696,
      "epoch": 2.7271853986551395,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004260717085065045,
      "loss": 4.7168,
      "mean_token_accuracy": 0.23661622107028962,
      "num_tokens": 65097137.0,
      "step": 28390
    },
    {
      "entropy": 5.138994646072388,
      "epoch": 2.727665706051873,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042604630030184797,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22101079523563386,
      "num_tokens": 65110229.0,
      "step": 28395
    },
    {
      "entropy": 5.171103048324585,
      "epoch": 2.7281460134486073,
      "grad_norm": 0.96875,
      "learning_rate": 0.000426020888590345,
      "loss": 4.8163,
      "mean_token_accuracy": 0.2347375899553299,
      "num_tokens": 65121767.0,
      "step": 28400
    },
    {
      "entropy": 5.036467266082764,
      "epoch": 2.728626320845341,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042599547337258536,
      "loss": 4.629,
      "mean_token_accuracy": 0.25289592891931534,
      "num_tokens": 65132918.0,
      "step": 28405
    },
    {
      "entropy": 5.035719871520996,
      "epoch": 2.729106628242075,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042597005464915924,
      "loss": 4.7225,
      "mean_token_accuracy": 0.24161709994077682,
      "num_tokens": 65144384.0,
      "step": 28410
    },
    {
      "entropy": 5.2008894920349125,
      "epoch": 2.729586935638809,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042594463242065674,
      "loss": 4.8803,
      "mean_token_accuracy": 0.22375040054321288,
      "num_tokens": 65157495.0,
      "step": 28415
    },
    {
      "entropy": 5.068362283706665,
      "epoch": 2.730067243035543,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004259192066876681,
      "loss": 4.6511,
      "mean_token_accuracy": 0.24763869047164916,
      "num_tokens": 65168583.0,
      "step": 28420
    },
    {
      "entropy": 5.051852655410767,
      "epoch": 2.7305475504322767,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042589377745078354,
      "loss": 4.7557,
      "mean_token_accuracy": 0.2307921215891838,
      "num_tokens": 65179909.0,
      "step": 28425
    },
    {
      "entropy": 5.118683767318726,
      "epoch": 2.7310278578290106,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042586834471059366,
      "loss": 4.7983,
      "mean_token_accuracy": 0.2325149357318878,
      "num_tokens": 65192239.0,
      "step": 28430
    },
    {
      "entropy": 5.065749549865723,
      "epoch": 2.7315081652257445,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042584290846768867,
      "loss": 4.7007,
      "mean_token_accuracy": 0.2445114403963089,
      "num_tokens": 65203454.0,
      "step": 28435
    },
    {
      "entropy": 5.022877168655396,
      "epoch": 2.7319884726224783,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004258174687226593,
      "loss": 4.6676,
      "mean_token_accuracy": 0.24194752126932145,
      "num_tokens": 65216265.0,
      "step": 28440
    },
    {
      "entropy": 5.115630149841309,
      "epoch": 2.7324687800192122,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004257920254760962,
      "loss": 4.7014,
      "mean_token_accuracy": 0.24409917443990709,
      "num_tokens": 65227742.0,
      "step": 28445
    },
    {
      "entropy": 5.04864764213562,
      "epoch": 2.732949087415946,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004257665787285899,
      "loss": 4.7652,
      "mean_token_accuracy": 0.24315991401672363,
      "num_tokens": 65239485.0,
      "step": 28450
    },
    {
      "entropy": 5.075656223297119,
      "epoch": 2.73342939481268,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042574112848073147,
      "loss": 4.6994,
      "mean_token_accuracy": 0.23562485873699188,
      "num_tokens": 65252818.0,
      "step": 28455
    },
    {
      "entropy": 5.083769083023071,
      "epoch": 2.733909702209414,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042571567473311157,
      "loss": 4.7575,
      "mean_token_accuracy": 0.23936982005834578,
      "num_tokens": 65264636.0,
      "step": 28460
    },
    {
      "entropy": 5.054509449005127,
      "epoch": 2.734390009606148,
      "grad_norm": 1.0,
      "learning_rate": 0.0004256902174863213,
      "loss": 4.7265,
      "mean_token_accuracy": 0.238442100584507,
      "num_tokens": 65275426.0,
      "step": 28465
    },
    {
      "entropy": 5.05550971031189,
      "epoch": 2.7348703170028816,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042566475674095155,
      "loss": 4.7626,
      "mean_token_accuracy": 0.23497967422008514,
      "num_tokens": 65286241.0,
      "step": 28470
    },
    {
      "entropy": 5.0568372249603275,
      "epoch": 2.735350624399616,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004256392924975936,
      "loss": 4.7169,
      "mean_token_accuracy": 0.24288420528173446,
      "num_tokens": 65298307.0,
      "step": 28475
    },
    {
      "entropy": 5.160656309127807,
      "epoch": 2.7358309317963494,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042561382475683854,
      "loss": 4.8129,
      "mean_token_accuracy": 0.22982385009527206,
      "num_tokens": 65310529.0,
      "step": 28480
    },
    {
      "entropy": 5.061558437347412,
      "epoch": 2.7363112391930837,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004255883535192777,
      "loss": 4.7215,
      "mean_token_accuracy": 0.2362554043531418,
      "num_tokens": 65321061.0,
      "step": 28485
    },
    {
      "entropy": 5.069977140426635,
      "epoch": 2.7367915465898176,
      "grad_norm": 1.0,
      "learning_rate": 0.0004255628787855025,
      "loss": 4.7147,
      "mean_token_accuracy": 0.23849764466285706,
      "num_tokens": 65332068.0,
      "step": 28490
    },
    {
      "entropy": 5.207790470123291,
      "epoch": 2.7372718539865515,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004255374005561043,
      "loss": 4.9231,
      "mean_token_accuracy": 0.22359801679849625,
      "num_tokens": 65344484.0,
      "step": 28495
    },
    {
      "entropy": 5.068503046035767,
      "epoch": 2.7377521613832854,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042551191883167464,
      "loss": 4.7732,
      "mean_token_accuracy": 0.24218565970659256,
      "num_tokens": 65355021.0,
      "step": 28500
    },
    {
      "entropy": 5.0060282230377195,
      "epoch": 2.7382324687800192,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004254864336128052,
      "loss": 4.6425,
      "mean_token_accuracy": 0.23896047323942185,
      "num_tokens": 65367024.0,
      "step": 28505
    },
    {
      "entropy": 5.0489905834197994,
      "epoch": 2.738712776176753,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042546094490008765,
      "loss": 4.6816,
      "mean_token_accuracy": 0.24505512714385985,
      "num_tokens": 65377894.0,
      "step": 28510
    },
    {
      "entropy": 5.127585554122925,
      "epoch": 2.739193083573487,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004254354526941136,
      "loss": 4.7048,
      "mean_token_accuracy": 0.23427539616823195,
      "num_tokens": 65389627.0,
      "step": 28515
    },
    {
      "entropy": 5.037780046463013,
      "epoch": 2.739673390970221,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004254099569954751,
      "loss": 4.6358,
      "mean_token_accuracy": 0.24774316549301148,
      "num_tokens": 65399715.0,
      "step": 28520
    },
    {
      "entropy": 5.018020868301392,
      "epoch": 2.7401536983669548,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004253844578047641,
      "loss": 4.7633,
      "mean_token_accuracy": 0.2403384491801262,
      "num_tokens": 65410459.0,
      "step": 28525
    },
    {
      "entropy": 5.048092317581177,
      "epoch": 2.7406340057636887,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004253589551225725,
      "loss": 4.6846,
      "mean_token_accuracy": 0.24040603637695312,
      "num_tokens": 65422412.0,
      "step": 28530
    },
    {
      "entropy": 5.082907629013062,
      "epoch": 2.7411143131604225,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042533344894949245,
      "loss": 4.7489,
      "mean_token_accuracy": 0.23927004486322404,
      "num_tokens": 65433982.0,
      "step": 28535
    },
    {
      "entropy": 5.038143968582153,
      "epoch": 2.7415946205571564,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042530793928611605,
      "loss": 4.6884,
      "mean_token_accuracy": 0.2415604665875435,
      "num_tokens": 65445828.0,
      "step": 28540
    },
    {
      "entropy": 5.092839479446411,
      "epoch": 2.7420749279538903,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004252824261330357,
      "loss": 4.7006,
      "mean_token_accuracy": 0.24006526470184325,
      "num_tokens": 65457271.0,
      "step": 28545
    },
    {
      "entropy": 5.109500122070313,
      "epoch": 2.7425552353506246,
      "grad_norm": 1.0,
      "learning_rate": 0.00042525690949084364,
      "loss": 4.8223,
      "mean_token_accuracy": 0.23951631635427476,
      "num_tokens": 65468414.0,
      "step": 28550
    },
    {
      "entropy": 5.098453187942505,
      "epoch": 2.743035542747358,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042523138936013233,
      "loss": 4.6893,
      "mean_token_accuracy": 0.24111451506614684,
      "num_tokens": 65479454.0,
      "step": 28555
    },
    {
      "entropy": 5.141822576522827,
      "epoch": 2.7435158501440924,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042520586574149423,
      "loss": 4.8167,
      "mean_token_accuracy": 0.23694770634174347,
      "num_tokens": 65490028.0,
      "step": 28560
    },
    {
      "entropy": 5.021441650390625,
      "epoch": 2.7439961575408263,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042518033863552185,
      "loss": 4.6794,
      "mean_token_accuracy": 0.24751379787921907,
      "num_tokens": 65501929.0,
      "step": 28565
    },
    {
      "entropy": 5.08391580581665,
      "epoch": 2.74447646493756,
      "grad_norm": 0.875,
      "learning_rate": 0.0004251548080428081,
      "loss": 4.6824,
      "mean_token_accuracy": 0.2418898180127144,
      "num_tokens": 65513654.0,
      "step": 28570
    },
    {
      "entropy": 5.02668399810791,
      "epoch": 2.744956772334294,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004251292739639455,
      "loss": 4.6641,
      "mean_token_accuracy": 0.2488265410065651,
      "num_tokens": 65525219.0,
      "step": 28575
    },
    {
      "entropy": 5.036431789398193,
      "epoch": 2.745437079731028,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042510373639952694,
      "loss": 4.6644,
      "mean_token_accuracy": 0.24486118257045747,
      "num_tokens": 65536255.0,
      "step": 28580
    },
    {
      "entropy": 5.177157688140869,
      "epoch": 2.745917387127762,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042507819535014547,
      "loss": 4.8195,
      "mean_token_accuracy": 0.22799367606639862,
      "num_tokens": 65547489.0,
      "step": 28585
    },
    {
      "entropy": 5.094801950454712,
      "epoch": 2.7463976945244957,
      "grad_norm": 1.0,
      "learning_rate": 0.00042505265081639376,
      "loss": 4.7083,
      "mean_token_accuracy": 0.24274538308382035,
      "num_tokens": 65559892.0,
      "step": 28590
    },
    {
      "entropy": 5.224062538146972,
      "epoch": 2.7468780019212296,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004250271027988652,
      "loss": 4.8396,
      "mean_token_accuracy": 0.2281651183962822,
      "num_tokens": 65570742.0,
      "step": 28595
    },
    {
      "entropy": 5.146520280838013,
      "epoch": 2.7473583093179634,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042500155129815274,
      "loss": 4.8142,
      "mean_token_accuracy": 0.23403631746768952,
      "num_tokens": 65584202.0,
      "step": 28600
    },
    {
      "entropy": 5.065748262405395,
      "epoch": 2.7478386167146973,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042497599631484965,
      "loss": 4.6696,
      "mean_token_accuracy": 0.2407911717891693,
      "num_tokens": 65595744.0,
      "step": 28605
    },
    {
      "entropy": 5.030035209655762,
      "epoch": 2.748318924111431,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042495043784954926,
      "loss": 4.6796,
      "mean_token_accuracy": 0.24294204860925675,
      "num_tokens": 65607914.0,
      "step": 28610
    },
    {
      "entropy": 5.106660461425781,
      "epoch": 2.748799231508165,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000424924875902845,
      "loss": 4.7017,
      "mean_token_accuracy": 0.24037479013204574,
      "num_tokens": 65620134.0,
      "step": 28615
    },
    {
      "entropy": 5.118148565292358,
      "epoch": 2.749279538904899,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004248993104753303,
      "loss": 4.6839,
      "mean_token_accuracy": 0.23985691219568253,
      "num_tokens": 65631804.0,
      "step": 28620
    },
    {
      "entropy": 5.110472726821899,
      "epoch": 2.7497598463016333,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004248737415675987,
      "loss": 4.7751,
      "mean_token_accuracy": 0.23959431201219558,
      "num_tokens": 65644067.0,
      "step": 28625
    },
    {
      "entropy": 5.107403039932251,
      "epoch": 2.7502401536983667,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004248481691802439,
      "loss": 4.7794,
      "mean_token_accuracy": 0.23790225386619568,
      "num_tokens": 65655036.0,
      "step": 28630
    },
    {
      "entropy": 5.122264289855957,
      "epoch": 2.750720461095101,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004248225933138595,
      "loss": 4.7251,
      "mean_token_accuracy": 0.2367045447230339,
      "num_tokens": 65666827.0,
      "step": 28635
    },
    {
      "entropy": 5.137905740737915,
      "epoch": 2.751200768491835,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042479701396903945,
      "loss": 4.8076,
      "mean_token_accuracy": 0.2325873404741287,
      "num_tokens": 65677635.0,
      "step": 28640
    },
    {
      "entropy": 5.035058546066284,
      "epoch": 2.751681075888569,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004247714311463775,
      "loss": 4.7272,
      "mean_token_accuracy": 0.23685158044099808,
      "num_tokens": 65689695.0,
      "step": 28645
    },
    {
      "entropy": 5.08994722366333,
      "epoch": 2.7521613832853027,
      "grad_norm": 1.0,
      "learning_rate": 0.00042474584484646766,
      "loss": 4.661,
      "mean_token_accuracy": 0.2406650722026825,
      "num_tokens": 65700969.0,
      "step": 28650
    },
    {
      "entropy": 5.1558253288269045,
      "epoch": 2.7526416906820366,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004247202550699039,
      "loss": 4.7708,
      "mean_token_accuracy": 0.23313767462968826,
      "num_tokens": 65712743.0,
      "step": 28655
    },
    {
      "entropy": 4.929418897628784,
      "epoch": 2.7531219980787704,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004246946618172805,
      "loss": 4.5596,
      "mean_token_accuracy": 0.2533689796924591,
      "num_tokens": 65724912.0,
      "step": 28660
    },
    {
      "entropy": 5.05886549949646,
      "epoch": 2.7536023054755043,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004246690650891915,
      "loss": 4.69,
      "mean_token_accuracy": 0.24417962729930878,
      "num_tokens": 65736295.0,
      "step": 28665
    },
    {
      "entropy": 5.097932004928589,
      "epoch": 2.754082612872238,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004246434648862312,
      "loss": 4.7043,
      "mean_token_accuracy": 0.24179430305957794,
      "num_tokens": 65747032.0,
      "step": 28670
    },
    {
      "entropy": 5.094316101074218,
      "epoch": 2.754562920268972,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004246178612089941,
      "loss": 4.7784,
      "mean_token_accuracy": 0.23679492324590684,
      "num_tokens": 65759001.0,
      "step": 28675
    },
    {
      "entropy": 5.085325813293457,
      "epoch": 2.755043227665706,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004245922540580744,
      "loss": 4.7278,
      "mean_token_accuracy": 0.2365471377968788,
      "num_tokens": 65769781.0,
      "step": 28680
    },
    {
      "entropy": 5.108206176757813,
      "epoch": 2.75552353506244,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004245666434340668,
      "loss": 4.7393,
      "mean_token_accuracy": 0.23910265564918518,
      "num_tokens": 65781281.0,
      "step": 28685
    },
    {
      "entropy": 5.167341566085815,
      "epoch": 2.7560038424591737,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004245410293375659,
      "loss": 4.8588,
      "mean_token_accuracy": 0.22434034794569016,
      "num_tokens": 65792884.0,
      "step": 28690
    },
    {
      "entropy": 5.083365201950073,
      "epoch": 2.7564841498559076,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004245154117691664,
      "loss": 4.6695,
      "mean_token_accuracy": 0.24689380079507828,
      "num_tokens": 65804616.0,
      "step": 28695
    },
    {
      "entropy": 5.0317212581634525,
      "epoch": 2.756964457252642,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004244897907294628,
      "loss": 4.6755,
      "mean_token_accuracy": 0.24367987960577012,
      "num_tokens": 65814859.0,
      "step": 28700
    },
    {
      "entropy": 5.077862215042114,
      "epoch": 2.7574447646493754,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004244641662190504,
      "loss": 4.7012,
      "mean_token_accuracy": 0.24457672089338303,
      "num_tokens": 65824416.0,
      "step": 28705
    },
    {
      "entropy": 5.127820014953613,
      "epoch": 2.7579250720461097,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042443853823852376,
      "loss": 4.7831,
      "mean_token_accuracy": 0.23633504509925843,
      "num_tokens": 65835428.0,
      "step": 28710
    },
    {
      "entropy": 5.095054960250854,
      "epoch": 2.7584053794428436,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000424412906788478,
      "loss": 4.77,
      "mean_token_accuracy": 0.23419808894395827,
      "num_tokens": 65846683.0,
      "step": 28715
    },
    {
      "entropy": 5.142182493209839,
      "epoch": 2.7588856868395775,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004243872718695082,
      "loss": 4.8291,
      "mean_token_accuracy": 0.23527898639440536,
      "num_tokens": 65857850.0,
      "step": 28720
    },
    {
      "entropy": 4.989261817932129,
      "epoch": 2.7593659942363113,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042436163348220956,
      "loss": 4.6426,
      "mean_token_accuracy": 0.2543283149600029,
      "num_tokens": 65869688.0,
      "step": 28725
    },
    {
      "entropy": 5.079798984527588,
      "epoch": 2.7598463016330452,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004243359916271773,
      "loss": 4.7102,
      "mean_token_accuracy": 0.24351219087839127,
      "num_tokens": 65880705.0,
      "step": 28730
    },
    {
      "entropy": 5.092023515701294,
      "epoch": 2.760326609029779,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004243103463050067,
      "loss": 4.6846,
      "mean_token_accuracy": 0.2427138641476631,
      "num_tokens": 65894070.0,
      "step": 28735
    },
    {
      "entropy": 5.08001823425293,
      "epoch": 2.760806916426513,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004242846975162933,
      "loss": 4.807,
      "mean_token_accuracy": 0.23231401145458222,
      "num_tokens": 65905899.0,
      "step": 28740
    },
    {
      "entropy": 5.017371273040771,
      "epoch": 2.761287223823247,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042425904526163246,
      "loss": 4.7337,
      "mean_token_accuracy": 0.2338994100689888,
      "num_tokens": 65917173.0,
      "step": 28745
    },
    {
      "entropy": 5.168143606185913,
      "epoch": 2.7617675312199808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004242333895416198,
      "loss": 4.81,
      "mean_token_accuracy": 0.2328492134809494,
      "num_tokens": 65930125.0,
      "step": 28750
    },
    {
      "entropy": 5.090390586853028,
      "epoch": 2.7622478386167146,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000424207730356851,
      "loss": 4.7236,
      "mean_token_accuracy": 0.23537610173225404,
      "num_tokens": 65941569.0,
      "step": 28755
    },
    {
      "entropy": 5.056044101715088,
      "epoch": 2.7627281460134485,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004241820677079218,
      "loss": 4.688,
      "mean_token_accuracy": 0.23921059519052507,
      "num_tokens": 65953970.0,
      "step": 28760
    },
    {
      "entropy": 5.021924829483032,
      "epoch": 2.7632084534101824,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042415640159542783,
      "loss": 4.7044,
      "mean_token_accuracy": 0.24459148496389388,
      "num_tokens": 65965740.0,
      "step": 28765
    },
    {
      "entropy": 5.091464519500732,
      "epoch": 2.7636887608069163,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004241307320199652,
      "loss": 4.7232,
      "mean_token_accuracy": 0.2411563739180565,
      "num_tokens": 65975986.0,
      "step": 28770
    },
    {
      "entropy": 5.165722036361695,
      "epoch": 2.7641690682036506,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004241050589821298,
      "loss": 4.8129,
      "mean_token_accuracy": 0.23079067021608352,
      "num_tokens": 65988356.0,
      "step": 28775
    },
    {
      "entropy": 5.068029880523682,
      "epoch": 2.764649375600384,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004240793824825177,
      "loss": 4.6813,
      "mean_token_accuracy": 0.23868304193019868,
      "num_tokens": 65998526.0,
      "step": 28780
    },
    {
      "entropy": 5.022368764877319,
      "epoch": 2.7651296829971184,
      "grad_norm": 1.0,
      "learning_rate": 0.00042405370252172496,
      "loss": 4.6305,
      "mean_token_accuracy": 0.24751915037631989,
      "num_tokens": 66009605.0,
      "step": 28785
    },
    {
      "entropy": 5.041417407989502,
      "epoch": 2.765609990393852,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004240280191003479,
      "loss": 4.639,
      "mean_token_accuracy": 0.2459615722298622,
      "num_tokens": 66020794.0,
      "step": 28790
    },
    {
      "entropy": 5.0929028511047365,
      "epoch": 2.766090297790586,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004240023322189828,
      "loss": 4.7636,
      "mean_token_accuracy": 0.24240995198488235,
      "num_tokens": 66032174.0,
      "step": 28795
    },
    {
      "entropy": 5.121556138992309,
      "epoch": 2.76657060518732,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004239766418782258,
      "loss": 4.7037,
      "mean_token_accuracy": 0.24616808146238328,
      "num_tokens": 66042552.0,
      "step": 28800
    },
    {
      "entropy": 5.018472480773926,
      "epoch": 2.767050912584054,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004239509480786737,
      "loss": 4.6954,
      "mean_token_accuracy": 0.2409435585141182,
      "num_tokens": 66053935.0,
      "step": 28805
    },
    {
      "entropy": 4.985249805450439,
      "epoch": 2.7675312199807878,
      "grad_norm": 1.125,
      "learning_rate": 0.00042392525082092286,
      "loss": 4.5579,
      "mean_token_accuracy": 0.24928492307662964,
      "num_tokens": 66064957.0,
      "step": 28810
    },
    {
      "entropy": 5.086501884460449,
      "epoch": 2.7680115273775217,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004238995501055699,
      "loss": 4.767,
      "mean_token_accuracy": 0.2387574553489685,
      "num_tokens": 66076270.0,
      "step": 28815
    },
    {
      "entropy": 5.172310924530029,
      "epoch": 2.7684918347742555,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004238738459332115,
      "loss": 4.8064,
      "mean_token_accuracy": 0.23348175138235092,
      "num_tokens": 66088458.0,
      "step": 28820
    },
    {
      "entropy": 5.101270008087158,
      "epoch": 2.7689721421709894,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004238481383044445,
      "loss": 4.689,
      "mean_token_accuracy": 0.23796872943639755,
      "num_tokens": 66100114.0,
      "step": 28825
    },
    {
      "entropy": 5.074327278137207,
      "epoch": 2.7694524495677233,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042382242721986573,
      "loss": 4.7606,
      "mean_token_accuracy": 0.23964528739452362,
      "num_tokens": 66112448.0,
      "step": 28830
    },
    {
      "entropy": 5.074463510513306,
      "epoch": 2.769932756964457,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042379671268007207,
      "loss": 4.7474,
      "mean_token_accuracy": 0.24115068465471268,
      "num_tokens": 66124122.0,
      "step": 28835
    },
    {
      "entropy": 5.015971040725708,
      "epoch": 2.770413064361191,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004237709946856607,
      "loss": 4.6849,
      "mean_token_accuracy": 0.23880140036344527,
      "num_tokens": 66135375.0,
      "step": 28840
    },
    {
      "entropy": 5.106083726882934,
      "epoch": 2.770893371757925,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042374527323722836,
      "loss": 4.7822,
      "mean_token_accuracy": 0.22881529927253724,
      "num_tokens": 66148850.0,
      "step": 28845
    },
    {
      "entropy": 5.044667053222656,
      "epoch": 2.771373679154659,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042371954833537263,
      "loss": 4.6637,
      "mean_token_accuracy": 0.246357524394989,
      "num_tokens": 66160040.0,
      "step": 28850
    },
    {
      "entropy": 5.107617330551148,
      "epoch": 2.7718539865513927,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042369381998069055,
      "loss": 4.7026,
      "mean_token_accuracy": 0.24035543352365493,
      "num_tokens": 66172857.0,
      "step": 28855
    },
    {
      "entropy": 5.044072246551513,
      "epoch": 2.772334293948127,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004236680881737795,
      "loss": 4.7076,
      "mean_token_accuracy": 0.23919187039136885,
      "num_tokens": 66184546.0,
      "step": 28860
    },
    {
      "entropy": 5.090251636505127,
      "epoch": 2.7728146013448605,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004236423529152369,
      "loss": 4.6976,
      "mean_token_accuracy": 0.2405666798353195,
      "num_tokens": 66195270.0,
      "step": 28865
    },
    {
      "entropy": 5.045242881774902,
      "epoch": 2.773294908741595,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004236166142056602,
      "loss": 4.7263,
      "mean_token_accuracy": 0.23779467791318892,
      "num_tokens": 66206998.0,
      "step": 28870
    },
    {
      "entropy": 5.142754316329956,
      "epoch": 2.7737752161383287,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004235908720456471,
      "loss": 4.77,
      "mean_token_accuracy": 0.2317615658044815,
      "num_tokens": 66217645.0,
      "step": 28875
    },
    {
      "entropy": 5.03411545753479,
      "epoch": 2.7742555235350626,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004235651264357951,
      "loss": 4.6989,
      "mean_token_accuracy": 0.24760135263204575,
      "num_tokens": 66228740.0,
      "step": 28880
    },
    {
      "entropy": 4.95988450050354,
      "epoch": 2.7747358309317964,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042353937737670206,
      "loss": 4.5826,
      "mean_token_accuracy": 0.25234245508909225,
      "num_tokens": 66239459.0,
      "step": 28885
    },
    {
      "entropy": 5.067418909072876,
      "epoch": 2.7752161383285303,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004235136248689658,
      "loss": 4.7356,
      "mean_token_accuracy": 0.23987476378679276,
      "num_tokens": 66251470.0,
      "step": 28890
    },
    {
      "entropy": 5.123503065109253,
      "epoch": 2.775696445725264,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004234878689131841,
      "loss": 4.7299,
      "mean_token_accuracy": 0.2346540719270706,
      "num_tokens": 66262045.0,
      "step": 28895
    },
    {
      "entropy": 5.010038757324219,
      "epoch": 2.776176753121998,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000423462109509955,
      "loss": 4.6819,
      "mean_token_accuracy": 0.2444191887974739,
      "num_tokens": 66272077.0,
      "step": 28900
    },
    {
      "entropy": 5.039529800415039,
      "epoch": 2.776657060518732,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004234363466598765,
      "loss": 4.6875,
      "mean_token_accuracy": 0.24621677547693252,
      "num_tokens": 66283091.0,
      "step": 28905
    },
    {
      "entropy": 5.100388526916504,
      "epoch": 2.777137367915466,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00042341058036354687,
      "loss": 4.7691,
      "mean_token_accuracy": 0.23990754187107086,
      "num_tokens": 66293770.0,
      "step": 28910
    },
    {
      "entropy": 5.074999523162842,
      "epoch": 2.7776176753121997,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042338481062156424,
      "loss": 4.6901,
      "mean_token_accuracy": 0.24700502008199693,
      "num_tokens": 66305288.0,
      "step": 28915
    },
    {
      "entropy": 5.128028726577758,
      "epoch": 2.7780979827089336,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042335903743452694,
      "loss": 4.7817,
      "mean_token_accuracy": 0.23123996555805207,
      "num_tokens": 66316414.0,
      "step": 28920
    },
    {
      "entropy": 5.1606590270996096,
      "epoch": 2.7785782901056675,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004233332608030333,
      "loss": 4.8049,
      "mean_token_accuracy": 0.22987312227487564,
      "num_tokens": 66327573.0,
      "step": 28925
    },
    {
      "entropy": 5.076343441009522,
      "epoch": 2.7790585975024014,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042330748072768183,
      "loss": 4.7682,
      "mean_token_accuracy": 0.2414083570241928,
      "num_tokens": 66338950.0,
      "step": 28930
    },
    {
      "entropy": 5.02622618675232,
      "epoch": 2.7795389048991357,
      "grad_norm": 1.109375,
      "learning_rate": 0.000423281697209071,
      "loss": 4.6394,
      "mean_token_accuracy": 0.24842915534973145,
      "num_tokens": 66348814.0,
      "step": 28935
    },
    {
      "entropy": 5.043000316619873,
      "epoch": 2.780019212295869,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004232559102477995,
      "loss": 4.6418,
      "mean_token_accuracy": 0.24514192193746567,
      "num_tokens": 66360119.0,
      "step": 28940
    },
    {
      "entropy": 5.076425075531006,
      "epoch": 2.7804995196926034,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000423230119844466,
      "loss": 4.7278,
      "mean_token_accuracy": 0.237441585958004,
      "num_tokens": 66370843.0,
      "step": 28945
    },
    {
      "entropy": 5.112809896469116,
      "epoch": 2.7809798270893373,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004232043259996692,
      "loss": 4.6976,
      "mean_token_accuracy": 0.23919108808040618,
      "num_tokens": 66382420.0,
      "step": 28950
    },
    {
      "entropy": 5.054312467575073,
      "epoch": 2.781460134486071,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004231785287140081,
      "loss": 4.6966,
      "mean_token_accuracy": 0.24218073189258577,
      "num_tokens": 66394327.0,
      "step": 28955
    },
    {
      "entropy": 5.041667366027832,
      "epoch": 2.781940441882805,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004231527279880816,
      "loss": 4.7872,
      "mean_token_accuracy": 0.22885456532239914,
      "num_tokens": 66407927.0,
      "step": 28960
    },
    {
      "entropy": 5.025435495376587,
      "epoch": 2.782420749279539,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004231269238224885,
      "loss": 4.6082,
      "mean_token_accuracy": 0.2474083736538887,
      "num_tokens": 66419248.0,
      "step": 28965
    },
    {
      "entropy": 5.019194030761719,
      "epoch": 2.782901056676273,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004231011162178282,
      "loss": 4.655,
      "mean_token_accuracy": 0.24335674047470093,
      "num_tokens": 66430750.0,
      "step": 28970
    },
    {
      "entropy": 5.058510160446167,
      "epoch": 2.7833813640730067,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004230753051746998,
      "loss": 4.723,
      "mean_token_accuracy": 0.23764632046222686,
      "num_tokens": 66443792.0,
      "step": 28975
    },
    {
      "entropy": 5.089301538467407,
      "epoch": 2.7838616714697406,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042304949069370246,
      "loss": 4.6451,
      "mean_token_accuracy": 0.24611927717924117,
      "num_tokens": 66455091.0,
      "step": 28980
    },
    {
      "entropy": 5.08501238822937,
      "epoch": 2.7843419788664745,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042302367277543553,
      "loss": 4.749,
      "mean_token_accuracy": 0.2423679381608963,
      "num_tokens": 66466269.0,
      "step": 28985
    },
    {
      "entropy": 5.043744993209839,
      "epoch": 2.7848222862632084,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042299785142049855,
      "loss": 4.6954,
      "mean_token_accuracy": 0.24459513276815414,
      "num_tokens": 66476032.0,
      "step": 28990
    },
    {
      "entropy": 5.046516752243042,
      "epoch": 2.7853025936599423,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004229720266294908,
      "loss": 4.7234,
      "mean_token_accuracy": 0.2411787509918213,
      "num_tokens": 66487265.0,
      "step": 28995
    },
    {
      "entropy": 5.190783500671387,
      "epoch": 2.785782901056676,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000422946198403012,
      "loss": 4.8282,
      "mean_token_accuracy": 0.234787280857563,
      "num_tokens": 66498762.0,
      "step": 29000
    },
    {
      "entropy": 5.141618251800537,
      "epoch": 2.78626320845341,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004229203667416619,
      "loss": 4.8287,
      "mean_token_accuracy": 0.22691155821084977,
      "num_tokens": 66509455.0,
      "step": 29005
    },
    {
      "entropy": 5.163204050064087,
      "epoch": 2.7867435158501443,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042289453164604,
      "loss": 4.8838,
      "mean_token_accuracy": 0.22413814812898636,
      "num_tokens": 66520858.0,
      "step": 29010
    },
    {
      "entropy": 5.1101278305053714,
      "epoch": 2.787223823246878,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004228686931167463,
      "loss": 4.736,
      "mean_token_accuracy": 0.2328069642186165,
      "num_tokens": 66532680.0,
      "step": 29015
    },
    {
      "entropy": 5.1007989883422855,
      "epoch": 2.787704130643612,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004228428511543806,
      "loss": 4.7572,
      "mean_token_accuracy": 0.2383515641093254,
      "num_tokens": 66543554.0,
      "step": 29020
    },
    {
      "entropy": 5.062893486022949,
      "epoch": 2.7881844380403455,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042281700575954283,
      "loss": 4.6411,
      "mean_token_accuracy": 0.24427904039621354,
      "num_tokens": 66554041.0,
      "step": 29025
    },
    {
      "entropy": 5.023177766799927,
      "epoch": 2.78866474543708,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004227911569328332,
      "loss": 4.6793,
      "mean_token_accuracy": 0.24557148665189743,
      "num_tokens": 66566216.0,
      "step": 29030
    },
    {
      "entropy": 5.0690371036529545,
      "epoch": 2.7891450528338138,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004227653046748517,
      "loss": 4.7331,
      "mean_token_accuracy": 0.2369006648659706,
      "num_tokens": 66577964.0,
      "step": 29035
    },
    {
      "entropy": 5.10006365776062,
      "epoch": 2.7896253602305476,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042273944898619864,
      "loss": 4.7598,
      "mean_token_accuracy": 0.24344854801893234,
      "num_tokens": 66590172.0,
      "step": 29040
    },
    {
      "entropy": 5.009214591979981,
      "epoch": 2.7901056676272815,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042271358986747427,
      "loss": 4.67,
      "mean_token_accuracy": 0.2448774516582489,
      "num_tokens": 66601277.0,
      "step": 29045
    },
    {
      "entropy": 5.05663046836853,
      "epoch": 2.7905859750240154,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042268772731927895,
      "loss": 4.6563,
      "mean_token_accuracy": 0.24670451730489731,
      "num_tokens": 66612080.0,
      "step": 29050
    },
    {
      "entropy": 5.034569787979126,
      "epoch": 2.7910662824207493,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042266186134221317,
      "loss": 4.7237,
      "mean_token_accuracy": 0.23953957259654998,
      "num_tokens": 66623080.0,
      "step": 29055
    },
    {
      "entropy": 5.0564587116241455,
      "epoch": 2.791546589817483,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004226359919368774,
      "loss": 4.7188,
      "mean_token_accuracy": 0.23943115770816803,
      "num_tokens": 66635486.0,
      "step": 29060
    },
    {
      "entropy": 5.050566530227661,
      "epoch": 2.792026897214217,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042261011910387224,
      "loss": 4.6962,
      "mean_token_accuracy": 0.24445140659809111,
      "num_tokens": 66647137.0,
      "step": 29065
    },
    {
      "entropy": 5.1832304954528805,
      "epoch": 2.792507204610951,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004225842428437985,
      "loss": 4.8297,
      "mean_token_accuracy": 0.22726270705461502,
      "num_tokens": 66658890.0,
      "step": 29070
    },
    {
      "entropy": 5.0307353973388675,
      "epoch": 2.792987512007685,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042255836315725694,
      "loss": 4.6818,
      "mean_token_accuracy": 0.24810229539871215,
      "num_tokens": 66670706.0,
      "step": 29075
    },
    {
      "entropy": 5.088215732574463,
      "epoch": 2.7934678194044187,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004225324800448483,
      "loss": 4.734,
      "mean_token_accuracy": 0.23649260997772217,
      "num_tokens": 66681930.0,
      "step": 29080
    },
    {
      "entropy": 5.025963211059571,
      "epoch": 2.793948126801153,
      "grad_norm": 1.0,
      "learning_rate": 0.00042250659350717343,
      "loss": 4.6622,
      "mean_token_accuracy": 0.24524092227220534,
      "num_tokens": 66694224.0,
      "step": 29085
    },
    {
      "entropy": 5.135055875778198,
      "epoch": 2.7944284341978864,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042248070354483354,
      "loss": 4.7477,
      "mean_token_accuracy": 0.24674765765666962,
      "num_tokens": 66705905.0,
      "step": 29090
    },
    {
      "entropy": 5.086385774612427,
      "epoch": 2.7949087415946208,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004224548101584297,
      "loss": 4.7427,
      "mean_token_accuracy": 0.24131562560796738,
      "num_tokens": 66717189.0,
      "step": 29095
    },
    {
      "entropy": 5.0439146041870115,
      "epoch": 2.795389048991354,
      "grad_norm": 1.015625,
      "learning_rate": 0.000422428913348563,
      "loss": 4.6406,
      "mean_token_accuracy": 0.24298734217882156,
      "num_tokens": 66728066.0,
      "step": 29100
    },
    {
      "entropy": 5.06371660232544,
      "epoch": 2.7958693563880885,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004224030131158346,
      "loss": 4.736,
      "mean_token_accuracy": 0.2466472268104553,
      "num_tokens": 66738530.0,
      "step": 29105
    },
    {
      "entropy": 5.085722494125366,
      "epoch": 2.7963496637848224,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004223771094608461,
      "loss": 4.7872,
      "mean_token_accuracy": 0.23691455125808716,
      "num_tokens": 66749563.0,
      "step": 29110
    },
    {
      "entropy": 5.1065511226654055,
      "epoch": 2.7968299711815563,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004223512023841986,
      "loss": 4.7812,
      "mean_token_accuracy": 0.23812517523765564,
      "num_tokens": 66761376.0,
      "step": 29115
    },
    {
      "entropy": 5.0474021434783936,
      "epoch": 2.79731027857829,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00042232529188649374,
      "loss": 4.6478,
      "mean_token_accuracy": 0.2437370538711548,
      "num_tokens": 66774193.0,
      "step": 29120
    },
    {
      "entropy": 5.072095155715942,
      "epoch": 2.797790585975024,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004222993779683331,
      "loss": 4.7055,
      "mean_token_accuracy": 0.2380406677722931,
      "num_tokens": 66787007.0,
      "step": 29125
    },
    {
      "entropy": 5.068102312088013,
      "epoch": 2.798270893371758,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042227346063031837,
      "loss": 4.6663,
      "mean_token_accuracy": 0.2427460715174675,
      "num_tokens": 66798426.0,
      "step": 29130
    },
    {
      "entropy": 5.066515874862671,
      "epoch": 2.798751200768492,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004222475398730511,
      "loss": 4.7498,
      "mean_token_accuracy": 0.23835118860006332,
      "num_tokens": 66810843.0,
      "step": 29135
    },
    {
      "entropy": 4.954850387573242,
      "epoch": 2.7992315081652257,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004222216156971332,
      "loss": 4.6629,
      "mean_token_accuracy": 0.2464582696557045,
      "num_tokens": 66823350.0,
      "step": 29140
    },
    {
      "entropy": 5.053300380706787,
      "epoch": 2.7997118155619596,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042219568810316656,
      "loss": 4.7745,
      "mean_token_accuracy": 0.2420332983136177,
      "num_tokens": 66836830.0,
      "step": 29145
    },
    {
      "entropy": 5.140920829772949,
      "epoch": 2.8001921229586935,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004221697570917531,
      "loss": 4.8177,
      "mean_token_accuracy": 0.2228463649749756,
      "num_tokens": 66848843.0,
      "step": 29150
    },
    {
      "entropy": 5.140007781982422,
      "epoch": 2.8006724303554273,
      "grad_norm": 1.0,
      "learning_rate": 0.000422143822663495,
      "loss": 4.7566,
      "mean_token_accuracy": 0.2384007140994072,
      "num_tokens": 66858598.0,
      "step": 29155
    },
    {
      "entropy": 5.006210660934448,
      "epoch": 2.8011527377521612,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004221178848189941,
      "loss": 4.6888,
      "mean_token_accuracy": 0.23963692486286164,
      "num_tokens": 66869633.0,
      "step": 29160
    },
    {
      "entropy": 5.012665510177612,
      "epoch": 2.801633045148895,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042209194355885283,
      "loss": 4.7307,
      "mean_token_accuracy": 0.24290400594472886,
      "num_tokens": 66880501.0,
      "step": 29165
    },
    {
      "entropy": 5.155776929855347,
      "epoch": 2.8021133525456294,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004220659988836734,
      "loss": 4.7732,
      "mean_token_accuracy": 0.23145880550146103,
      "num_tokens": 66892488.0,
      "step": 29170
    },
    {
      "entropy": 5.017666864395141,
      "epoch": 2.802593659942363,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004220400507940582,
      "loss": 4.593,
      "mean_token_accuracy": 0.2593849703669548,
      "num_tokens": 66903861.0,
      "step": 29175
    },
    {
      "entropy": 5.014609384536743,
      "epoch": 2.803073967339097,
      "grad_norm": 1.0,
      "learning_rate": 0.00042201409929060955,
      "loss": 4.6879,
      "mean_token_accuracy": 0.23885919153690338,
      "num_tokens": 66915526.0,
      "step": 29180
    },
    {
      "entropy": 5.000254821777344,
      "epoch": 2.803554274735831,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004219881443739301,
      "loss": 4.6688,
      "mean_token_accuracy": 0.24003507494926452,
      "num_tokens": 66926798.0,
      "step": 29185
    },
    {
      "entropy": 5.135998201370239,
      "epoch": 2.804034582132565,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004219621860446225,
      "loss": 4.8212,
      "mean_token_accuracy": 0.2272834897041321,
      "num_tokens": 66938185.0,
      "step": 29190
    },
    {
      "entropy": 5.1163076877594,
      "epoch": 2.804514889529299,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004219362243032892,
      "loss": 4.7513,
      "mean_token_accuracy": 0.2382306858897209,
      "num_tokens": 66950423.0,
      "step": 29195
    },
    {
      "entropy": 5.090384817123413,
      "epoch": 2.8049951969260327,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042191025915053323,
      "loss": 4.6959,
      "mean_token_accuracy": 0.23326724916696548,
      "num_tokens": 66961545.0,
      "step": 29200
    },
    {
      "entropy": 5.146866655349731,
      "epoch": 2.8054755043227666,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042188429058695714,
      "loss": 4.7446,
      "mean_token_accuracy": 0.23920947313308716,
      "num_tokens": 66971932.0,
      "step": 29205
    },
    {
      "entropy": 5.02885479927063,
      "epoch": 2.8059558117195005,
      "grad_norm": 1.0,
      "learning_rate": 0.00042185831861316406,
      "loss": 4.6771,
      "mean_token_accuracy": 0.2424909308552742,
      "num_tokens": 66983830.0,
      "step": 29210
    },
    {
      "entropy": 5.0206077098846436,
      "epoch": 2.8064361191162344,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004218323432297568,
      "loss": 4.7307,
      "mean_token_accuracy": 0.23800697475671767,
      "num_tokens": 66995587.0,
      "step": 29215
    },
    {
      "entropy": 5.108063364028931,
      "epoch": 2.8069164265129682,
      "grad_norm": 1.0,
      "learning_rate": 0.00042180636443733864,
      "loss": 4.7635,
      "mean_token_accuracy": 0.23396404087543488,
      "num_tokens": 67007947.0,
      "step": 29220
    },
    {
      "entropy": 5.018198537826538,
      "epoch": 2.807396733909702,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042178038223651253,
      "loss": 4.6848,
      "mean_token_accuracy": 0.23476196974515914,
      "num_tokens": 67019708.0,
      "step": 29225
    },
    {
      "entropy": 5.052866125106812,
      "epoch": 2.807877041306436,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042175439662788195,
      "loss": 4.6697,
      "mean_token_accuracy": 0.24215113818645478,
      "num_tokens": 67030372.0,
      "step": 29230
    },
    {
      "entropy": 5.194032955169678,
      "epoch": 2.80835734870317,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042172840761204986,
      "loss": 4.8508,
      "mean_token_accuracy": 0.23494957089424134,
      "num_tokens": 67040614.0,
      "step": 29235
    },
    {
      "entropy": 5.128986024856568,
      "epoch": 2.8088376560999038,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004217024151896199,
      "loss": 4.7543,
      "mean_token_accuracy": 0.24108468741178513,
      "num_tokens": 67053409.0,
      "step": 29240
    },
    {
      "entropy": 5.1328778743743895,
      "epoch": 2.809317963496638,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00042167641936119557,
      "loss": 4.783,
      "mean_token_accuracy": 0.2343224361538887,
      "num_tokens": 67065649.0,
      "step": 29245
    },
    {
      "entropy": 5.124171352386474,
      "epoch": 2.8097982708933715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004216504201273802,
      "loss": 4.8102,
      "mean_token_accuracy": 0.23036458790302278,
      "num_tokens": 67077714.0,
      "step": 29250
    },
    {
      "entropy": 5.130268144607544,
      "epoch": 2.810278578290106,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004216244174887776,
      "loss": 4.775,
      "mean_token_accuracy": 0.23692994862794875,
      "num_tokens": 67089674.0,
      "step": 29255
    },
    {
      "entropy": 5.00474739074707,
      "epoch": 2.8107588856868397,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042159841144599145,
      "loss": 4.6766,
      "mean_token_accuracy": 0.2411862850189209,
      "num_tokens": 67101843.0,
      "step": 29260
    },
    {
      "entropy": 5.117538213729858,
      "epoch": 2.8112391930835736,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042157240199962537,
      "loss": 4.7631,
      "mean_token_accuracy": 0.23592806309461595,
      "num_tokens": 67113529.0,
      "step": 29265
    },
    {
      "entropy": 5.092918539047242,
      "epoch": 2.8117195004803075,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004215463891502834,
      "loss": 4.7008,
      "mean_token_accuracy": 0.243525955080986,
      "num_tokens": 67125067.0,
      "step": 29270
    },
    {
      "entropy": 5.018960666656494,
      "epoch": 2.8121998078770414,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00042152037289856954,
      "loss": 4.6801,
      "mean_token_accuracy": 0.24380457699298858,
      "num_tokens": 67137550.0,
      "step": 29275
    },
    {
      "entropy": 5.044099998474121,
      "epoch": 2.8126801152737753,
      "grad_norm": 1.0,
      "learning_rate": 0.00042149435324508755,
      "loss": 4.6177,
      "mean_token_accuracy": 0.24552652835845948,
      "num_tokens": 67149730.0,
      "step": 29280
    },
    {
      "entropy": 5.106616449356079,
      "epoch": 2.813160422670509,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004214683301904417,
      "loss": 4.7525,
      "mean_token_accuracy": 0.23488502204418182,
      "num_tokens": 67161050.0,
      "step": 29285
    },
    {
      "entropy": 5.004787111282349,
      "epoch": 2.813640730067243,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042144230373523624,
      "loss": 4.7243,
      "mean_token_accuracy": 0.23189249783754348,
      "num_tokens": 67173131.0,
      "step": 29290
    },
    {
      "entropy": 5.05130934715271,
      "epoch": 2.814121037463977,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004214162738800753,
      "loss": 4.739,
      "mean_token_accuracy": 0.23262506276369094,
      "num_tokens": 67183723.0,
      "step": 29295
    },
    {
      "entropy": 5.116889190673828,
      "epoch": 2.814601344860711,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004213902406255632,
      "loss": 4.7687,
      "mean_token_accuracy": 0.23901842385530472,
      "num_tokens": 67196622.0,
      "step": 29300
    },
    {
      "entropy": 5.1019322872161865,
      "epoch": 2.8150816522574447,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004213642039723044,
      "loss": 4.7054,
      "mean_token_accuracy": 0.2436388447880745,
      "num_tokens": 67208581.0,
      "step": 29305
    },
    {
      "entropy": 5.085634517669678,
      "epoch": 2.8155619596541785,
      "grad_norm": 0.875,
      "learning_rate": 0.00042133816392090343,
      "loss": 4.7527,
      "mean_token_accuracy": 0.23976353108882903,
      "num_tokens": 67220451.0,
      "step": 29310
    },
    {
      "entropy": 5.047297143936158,
      "epoch": 2.8160422670509124,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042131212047196484,
      "loss": 4.6998,
      "mean_token_accuracy": 0.23989350944757462,
      "num_tokens": 67232191.0,
      "step": 29315
    },
    {
      "entropy": 4.983439970016479,
      "epoch": 2.8165225744476468,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042128607362609317,
      "loss": 4.5883,
      "mean_token_accuracy": 0.2516318872570992,
      "num_tokens": 67242518.0,
      "step": 29320
    },
    {
      "entropy": 5.081890344619751,
      "epoch": 2.81700288184438,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042126002338389336,
      "loss": 4.8054,
      "mean_token_accuracy": 0.2331841304898262,
      "num_tokens": 67254598.0,
      "step": 29325
    },
    {
      "entropy": 5.033858299255371,
      "epoch": 2.8174831892411145,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042123396974597007,
      "loss": 4.6825,
      "mean_token_accuracy": 0.2490393653512001,
      "num_tokens": 67266054.0,
      "step": 29330
    },
    {
      "entropy": 5.142895078659057,
      "epoch": 2.817963496637848,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042120791271292823,
      "loss": 4.8281,
      "mean_token_accuracy": 0.23544473499059676,
      "num_tokens": 67277703.0,
      "step": 29335
    },
    {
      "entropy": 5.077246236801147,
      "epoch": 2.8184438040345823,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042118185228537283,
      "loss": 4.6905,
      "mean_token_accuracy": 0.24485918283462524,
      "num_tokens": 67288992.0,
      "step": 29340
    },
    {
      "entropy": 5.053110551834107,
      "epoch": 2.818924111431316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042115578846390884,
      "loss": 4.6631,
      "mean_token_accuracy": 0.24961425960063935,
      "num_tokens": 67299728.0,
      "step": 29345
    },
    {
      "entropy": 5.145263433456421,
      "epoch": 2.81940441882805,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004211297212491414,
      "loss": 4.8206,
      "mean_token_accuracy": 0.22895521223545073,
      "num_tokens": 67310836.0,
      "step": 29350
    },
    {
      "entropy": 5.101775217056274,
      "epoch": 2.819884726224784,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004211036506416759,
      "loss": 4.7518,
      "mean_token_accuracy": 0.235956834256649,
      "num_tokens": 67323426.0,
      "step": 29355
    },
    {
      "entropy": 5.082796382904053,
      "epoch": 2.820365033621518,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004210775766421173,
      "loss": 4.728,
      "mean_token_accuracy": 0.24019350707530976,
      "num_tokens": 67334909.0,
      "step": 29360
    },
    {
      "entropy": 5.034244728088379,
      "epoch": 2.8208453410182517,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004210514992510713,
      "loss": 4.6777,
      "mean_token_accuracy": 0.24068279266357423,
      "num_tokens": 67347315.0,
      "step": 29365
    },
    {
      "entropy": 5.045740842819214,
      "epoch": 2.8213256484149856,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004210254184691431,
      "loss": 4.6556,
      "mean_token_accuracy": 0.24876836389303209,
      "num_tokens": 67358092.0,
      "step": 29370
    },
    {
      "entropy": 5.057984161376953,
      "epoch": 2.8218059558117194,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042099933429693814,
      "loss": 4.813,
      "mean_token_accuracy": 0.23907537907361984,
      "num_tokens": 67369318.0,
      "step": 29375
    },
    {
      "entropy": 5.071371221542359,
      "epoch": 2.8222862632084533,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004209732467350624,
      "loss": 4.7374,
      "mean_token_accuracy": 0.23848043829202653,
      "num_tokens": 67379604.0,
      "step": 29380
    },
    {
      "entropy": 5.104182720184326,
      "epoch": 2.822766570605187,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004209471557841212,
      "loss": 4.675,
      "mean_token_accuracy": 0.24849707931280135,
      "num_tokens": 67391134.0,
      "step": 29385
    },
    {
      "entropy": 5.074014568328858,
      "epoch": 2.823246878001921,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004209210614447204,
      "loss": 4.6466,
      "mean_token_accuracy": 0.245580293238163,
      "num_tokens": 67400868.0,
      "step": 29390
    },
    {
      "entropy": 5.056183671951294,
      "epoch": 2.8237271853986554,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000420894963717466,
      "loss": 4.7007,
      "mean_token_accuracy": 0.2304675206542015,
      "num_tokens": 67412172.0,
      "step": 29395
    },
    {
      "entropy": 5.031347370147705,
      "epoch": 2.824207492795389,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004208688626029636,
      "loss": 4.6719,
      "mean_token_accuracy": 0.23947535455226898,
      "num_tokens": 67423700.0,
      "step": 29400
    },
    {
      "entropy": 5.112936782836914,
      "epoch": 2.824687800192123,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004208427581018194,
      "loss": 4.8099,
      "mean_token_accuracy": 0.23252042979002,
      "num_tokens": 67436471.0,
      "step": 29405
    },
    {
      "entropy": 5.16615629196167,
      "epoch": 2.8251681075888566,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004208166502146394,
      "loss": 4.8264,
      "mean_token_accuracy": 0.2247656390070915,
      "num_tokens": 67447832.0,
      "step": 29410
    },
    {
      "entropy": 5.0581683158874515,
      "epoch": 2.825648414985591,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042079053894202977,
      "loss": 4.7604,
      "mean_token_accuracy": 0.2396203562617302,
      "num_tokens": 67460467.0,
      "step": 29415
    },
    {
      "entropy": 5.073732709884643,
      "epoch": 2.826128722382325,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004207644242845968,
      "loss": 4.7035,
      "mean_token_accuracy": 0.23702015578746796,
      "num_tokens": 67471979.0,
      "step": 29420
    },
    {
      "entropy": 5.0750970363616945,
      "epoch": 2.8266090297790587,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004207383062429467,
      "loss": 4.7338,
      "mean_token_accuracy": 0.2350544214248657,
      "num_tokens": 67483441.0,
      "step": 29425
    },
    {
      "entropy": 5.146733188629151,
      "epoch": 2.8270893371757926,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004207121848176858,
      "loss": 4.8237,
      "mean_token_accuracy": 0.23687559366226196,
      "num_tokens": 67496691.0,
      "step": 29430
    },
    {
      "entropy": 5.123493576049805,
      "epoch": 2.8275696445725265,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042068606000942075,
      "loss": 4.7655,
      "mean_token_accuracy": 0.23255289644002913,
      "num_tokens": 67508120.0,
      "step": 29435
    },
    {
      "entropy": 5.123560047149658,
      "epoch": 2.8280499519692603,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042065993181875794,
      "loss": 4.7945,
      "mean_token_accuracy": 0.23327875137329102,
      "num_tokens": 67519699.0,
      "step": 29440
    },
    {
      "entropy": 5.067651700973511,
      "epoch": 2.8285302593659942,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000420633800246304,
      "loss": 4.7899,
      "mean_token_accuracy": 0.2285136714577675,
      "num_tokens": 67530842.0,
      "step": 29445
    },
    {
      "entropy": 5.120378160476685,
      "epoch": 2.829010566762728,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042060766529266577,
      "loss": 4.7633,
      "mean_token_accuracy": 0.23610294610261917,
      "num_tokens": 67543447.0,
      "step": 29450
    },
    {
      "entropy": 5.131119871139527,
      "epoch": 2.829490874159462,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042058152695844986,
      "loss": 4.7358,
      "mean_token_accuracy": 0.24705970883369446,
      "num_tokens": 67555339.0,
      "step": 29455
    },
    {
      "entropy": 5.052775621414185,
      "epoch": 2.829971181556196,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042055538524426317,
      "loss": 4.7071,
      "mean_token_accuracy": 0.24026860743761064,
      "num_tokens": 67566747.0,
      "step": 29460
    },
    {
      "entropy": 5.0188713550567625,
      "epoch": 2.8304514889529298,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004205292401507127,
      "loss": 4.6584,
      "mean_token_accuracy": 0.24760099351406098,
      "num_tokens": 67577690.0,
      "step": 29465
    },
    {
      "entropy": 5.187629270553589,
      "epoch": 2.8309317963496636,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004205030916784053,
      "loss": 4.8709,
      "mean_token_accuracy": 0.2247278571128845,
      "num_tokens": 67589287.0,
      "step": 29470
    },
    {
      "entropy": 5.096717786788941,
      "epoch": 2.8314121037463975,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00042047693982794824,
      "loss": 4.7042,
      "mean_token_accuracy": 0.2389528602361679,
      "num_tokens": 67602233.0,
      "step": 29475
    },
    {
      "entropy": 5.04383487701416,
      "epoch": 2.831892411143132,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042045078459994854,
      "loss": 4.6909,
      "mean_token_accuracy": 0.23940670043230056,
      "num_tokens": 67613076.0,
      "step": 29480
    },
    {
      "entropy": 5.060606670379639,
      "epoch": 2.8323727185398653,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004204246259950136,
      "loss": 4.6976,
      "mean_token_accuracy": 0.2408332720398903,
      "num_tokens": 67623517.0,
      "step": 29485
    },
    {
      "entropy": 5.124744701385498,
      "epoch": 2.8328530259365996,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042039846401375065,
      "loss": 4.7348,
      "mean_token_accuracy": 0.23576367348432542,
      "num_tokens": 67635982.0,
      "step": 29490
    },
    {
      "entropy": 5.0746763229370115,
      "epoch": 2.8333333333333335,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042037229865676714,
      "loss": 4.6307,
      "mean_token_accuracy": 0.24159572124481202,
      "num_tokens": 67646436.0,
      "step": 29495
    },
    {
      "entropy": 5.070221471786499,
      "epoch": 2.8338136407300674,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042034612992467046,
      "loss": 4.6993,
      "mean_token_accuracy": 0.24840300679206848,
      "num_tokens": 67657436.0,
      "step": 29500
    },
    {
      "entropy": 5.047015571594239,
      "epoch": 2.8342939481268012,
      "grad_norm": 1.0,
      "learning_rate": 0.0004203199578180683,
      "loss": 4.734,
      "mean_token_accuracy": 0.2353681966662407,
      "num_tokens": 67669542.0,
      "step": 29505
    },
    {
      "entropy": 5.035528230667114,
      "epoch": 2.834774255523535,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004202937823375682,
      "loss": 4.608,
      "mean_token_accuracy": 0.25431336611509325,
      "num_tokens": 67680512.0,
      "step": 29510
    },
    {
      "entropy": 5.077745246887207,
      "epoch": 2.835254562920269,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004202676034837779,
      "loss": 4.6996,
      "mean_token_accuracy": 0.23985062092542647,
      "num_tokens": 67691392.0,
      "step": 29515
    },
    {
      "entropy": 5.047430562973022,
      "epoch": 2.835734870317003,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004202414212573052,
      "loss": 4.7083,
      "mean_token_accuracy": 0.2436753734946251,
      "num_tokens": 67701674.0,
      "step": 29520
    },
    {
      "entropy": 5.05016303062439,
      "epoch": 2.8362151777137368,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042021523565875796,
      "loss": 4.7776,
      "mean_token_accuracy": 0.2354227840900421,
      "num_tokens": 67713960.0,
      "step": 29525
    },
    {
      "entropy": 5.0786608219146725,
      "epoch": 2.8366954851104706,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004201890466887442,
      "loss": 4.6832,
      "mean_token_accuracy": 0.24363634139299392,
      "num_tokens": 67725644.0,
      "step": 29530
    },
    {
      "entropy": 5.066598129272461,
      "epoch": 2.8371757925072045,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004201628543478718,
      "loss": 4.6859,
      "mean_token_accuracy": 0.2466487318277359,
      "num_tokens": 67737997.0,
      "step": 29535
    },
    {
      "entropy": 5.0946272850036625,
      "epoch": 2.8376560999039384,
      "grad_norm": 0.90625,
      "learning_rate": 0.000420136658636749,
      "loss": 4.7536,
      "mean_token_accuracy": 0.23685127645730972,
      "num_tokens": 67750379.0,
      "step": 29540
    },
    {
      "entropy": 5.088048267364502,
      "epoch": 2.8381364073006723,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004201104595559841,
      "loss": 4.7989,
      "mean_token_accuracy": 0.23146936148405076,
      "num_tokens": 67762567.0,
      "step": 29545
    },
    {
      "entropy": 5.081172323226928,
      "epoch": 2.838616714697406,
      "grad_norm": 0.921875,
      "learning_rate": 0.00042008425710618507,
      "loss": 4.7314,
      "mean_token_accuracy": 0.24004851132631302,
      "num_tokens": 67774660.0,
      "step": 29550
    },
    {
      "entropy": 5.14473032951355,
      "epoch": 2.8390970220941405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042005805128796043,
      "loss": 4.7008,
      "mean_token_accuracy": 0.243607397377491,
      "num_tokens": 67784757.0,
      "step": 29555
    },
    {
      "entropy": 5.051116466522217,
      "epoch": 2.839577329490874,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004200318421019186,
      "loss": 4.6857,
      "mean_token_accuracy": 0.24129762202501298,
      "num_tokens": 67795723.0,
      "step": 29560
    },
    {
      "entropy": 5.199991607666016,
      "epoch": 2.8400576368876083,
      "grad_norm": 1.0,
      "learning_rate": 0.000420005629548668,
      "loss": 4.8896,
      "mean_token_accuracy": 0.22865316569805144,
      "num_tokens": 67807685.0,
      "step": 29565
    },
    {
      "entropy": 5.063219451904297,
      "epoch": 2.840537944284342,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00041997941362881735,
      "loss": 4.6431,
      "mean_token_accuracy": 0.24870440661907195,
      "num_tokens": 67818887.0,
      "step": 29570
    },
    {
      "entropy": 5.066625595092773,
      "epoch": 2.841018251681076,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004199531943429752,
      "loss": 4.741,
      "mean_token_accuracy": 0.24339551627635955,
      "num_tokens": 67830210.0,
      "step": 29575
    },
    {
      "entropy": 5.04776177406311,
      "epoch": 2.84149855907781,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004199269716917502,
      "loss": 4.7302,
      "mean_token_accuracy": 0.23641683608293534,
      "num_tokens": 67841948.0,
      "step": 29580
    },
    {
      "entropy": 5.10105333328247,
      "epoch": 2.841978866474544,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004199007456757513,
      "loss": 4.6985,
      "mean_token_accuracy": 0.24282085299491882,
      "num_tokens": 67854117.0,
      "step": 29585
    },
    {
      "entropy": 5.101727437973023,
      "epoch": 2.8424591738712777,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00041987451629558743,
      "loss": 4.7659,
      "mean_token_accuracy": 0.23708426356315612,
      "num_tokens": 67866004.0,
      "step": 29590
    },
    {
      "entropy": 5.054056072235108,
      "epoch": 2.8429394812680115,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004198482835518674,
      "loss": 4.687,
      "mean_token_accuracy": 0.24716710597276687,
      "num_tokens": 67877178.0,
      "step": 29595
    },
    {
      "entropy": 5.016266679763794,
      "epoch": 2.8434197886647454,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004198220474452004,
      "loss": 4.6562,
      "mean_token_accuracy": 0.24792239516973497,
      "num_tokens": 67888183.0,
      "step": 29600
    },
    {
      "entropy": 5.162446355819702,
      "epoch": 2.8439000960614793,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004197958079761954,
      "loss": 4.7785,
      "mean_token_accuracy": 0.2301326423883438,
      "num_tokens": 67899300.0,
      "step": 29605
    },
    {
      "entropy": 5.047458839416504,
      "epoch": 2.844380403458213,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041976956514546185,
      "loss": 4.6912,
      "mean_token_accuracy": 0.24533195793628693,
      "num_tokens": 67909494.0,
      "step": 29610
    },
    {
      "entropy": 5.102247190475464,
      "epoch": 2.844860710854947,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041974331895360873,
      "loss": 4.7837,
      "mean_token_accuracy": 0.2366640269756317,
      "num_tokens": 67921241.0,
      "step": 29615
    },
    {
      "entropy": 5.041629409790039,
      "epoch": 2.845341018251681,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004197170694012456,
      "loss": 4.6511,
      "mean_token_accuracy": 0.24182143211364746,
      "num_tokens": 67933571.0,
      "step": 29620
    },
    {
      "entropy": 5.098745727539063,
      "epoch": 2.845821325648415,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004196908164889818,
      "loss": 4.77,
      "mean_token_accuracy": 0.23720728754997253,
      "num_tokens": 67944100.0,
      "step": 29625
    },
    {
      "entropy": 5.073638772964477,
      "epoch": 2.846301633045149,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004196645602174269,
      "loss": 4.6778,
      "mean_token_accuracy": 0.2411378175020218,
      "num_tokens": 67955908.0,
      "step": 29630
    },
    {
      "entropy": 5.137625074386596,
      "epoch": 2.8467819404418826,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041963830058719046,
      "loss": 4.7344,
      "mean_token_accuracy": 0.23095276206731796,
      "num_tokens": 67966763.0,
      "step": 29635
    },
    {
      "entropy": 5.025503158569336,
      "epoch": 2.847262247838617,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004196120375988822,
      "loss": 4.6692,
      "mean_token_accuracy": 0.23520204722881316,
      "num_tokens": 67977668.0,
      "step": 29640
    },
    {
      "entropy": 5.0600944519042965,
      "epoch": 2.8477425552353504,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004195857712531119,
      "loss": 4.7171,
      "mean_token_accuracy": 0.24572323262691498,
      "num_tokens": 67989429.0,
      "step": 29645
    },
    {
      "entropy": 5.154634189605713,
      "epoch": 2.8482228626320847,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004195595015504892,
      "loss": 4.8109,
      "mean_token_accuracy": 0.23560588657855988,
      "num_tokens": 68000911.0,
      "step": 29650
    },
    {
      "entropy": 5.0730242252349855,
      "epoch": 2.8487031700288186,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041953322849162415,
      "loss": 4.6727,
      "mean_token_accuracy": 0.24206048101186753,
      "num_tokens": 68012804.0,
      "step": 29655
    },
    {
      "entropy": 5.1487884521484375,
      "epoch": 2.8491834774255524,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004195069520771268,
      "loss": 4.7618,
      "mean_token_accuracy": 0.23433832824230194,
      "num_tokens": 68023508.0,
      "step": 29660
    },
    {
      "entropy": 5.1022828102111815,
      "epoch": 2.8496637848222863,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041948067230760706,
      "loss": 4.7479,
      "mean_token_accuracy": 0.23650110214948655,
      "num_tokens": 68035383.0,
      "step": 29665
    },
    {
      "entropy": 5.154181623458863,
      "epoch": 2.85014409221902,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00041945438918367513,
      "loss": 4.7537,
      "mean_token_accuracy": 0.22897733002901077,
      "num_tokens": 68045922.0,
      "step": 29670
    },
    {
      "entropy": 4.958856773376465,
      "epoch": 2.850624399615754,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041942810270594115,
      "loss": 4.6579,
      "mean_token_accuracy": 0.2417889729142189,
      "num_tokens": 68057291.0,
      "step": 29675
    },
    {
      "entropy": 5.078063440322876,
      "epoch": 2.851104707012488,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004194018128750157,
      "loss": 4.8039,
      "mean_token_accuracy": 0.2369493991136551,
      "num_tokens": 68069693.0,
      "step": 29680
    },
    {
      "entropy": 5.104892873764038,
      "epoch": 2.851585014409222,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041937551969150873,
      "loss": 4.7503,
      "mean_token_accuracy": 0.23564290553331374,
      "num_tokens": 68079696.0,
      "step": 29685
    },
    {
      "entropy": 5.0425646781921385,
      "epoch": 2.8520653218059557,
      "grad_norm": 0.890625,
      "learning_rate": 0.000419349223156031,
      "loss": 4.7068,
      "mean_token_accuracy": 0.23784703016281128,
      "num_tokens": 68091330.0,
      "step": 29690
    },
    {
      "entropy": 5.017551708221435,
      "epoch": 2.8525456292026896,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004193229232691929,
      "loss": 4.6529,
      "mean_token_accuracy": 0.24479199349880218,
      "num_tokens": 68102499.0,
      "step": 29695
    },
    {
      "entropy": 5.092421913146973,
      "epoch": 2.8530259365994235,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041929662003160504,
      "loss": 4.7143,
      "mean_token_accuracy": 0.23494544625282288,
      "num_tokens": 68114563.0,
      "step": 29700
    },
    {
      "entropy": 5.065534019470215,
      "epoch": 2.8535062439961574,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041927031344387824,
      "loss": 4.7443,
      "mean_token_accuracy": 0.24157904833555222,
      "num_tokens": 68124798.0,
      "step": 29705
    },
    {
      "entropy": 5.151731395721436,
      "epoch": 2.8539865513928913,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041924400350662304,
      "loss": 4.8107,
      "mean_token_accuracy": 0.23164553791284562,
      "num_tokens": 68136190.0,
      "step": 29710
    },
    {
      "entropy": 5.118416118621826,
      "epoch": 2.8544668587896256,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041921769022045045,
      "loss": 4.7053,
      "mean_token_accuracy": 0.24962661862373353,
      "num_tokens": 68146368.0,
      "step": 29715
    },
    {
      "entropy": 5.119978380203247,
      "epoch": 2.854947166186359,
      "grad_norm": 1.0,
      "learning_rate": 0.00041919137358597137,
      "loss": 4.8022,
      "mean_token_accuracy": 0.23396357893943787,
      "num_tokens": 68159338.0,
      "step": 29720
    },
    {
      "entropy": 5.139267206192017,
      "epoch": 2.8554274735830933,
      "grad_norm": 1.0,
      "learning_rate": 0.0004191650536037967,
      "loss": 4.7471,
      "mean_token_accuracy": 0.23061692714691162,
      "num_tokens": 68170906.0,
      "step": 29725
    },
    {
      "entropy": 5.128029489517212,
      "epoch": 2.8559077809798272,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041913873027453756,
      "loss": 4.7271,
      "mean_token_accuracy": 0.23977172672748565,
      "num_tokens": 68181963.0,
      "step": 29730
    },
    {
      "entropy": 4.984001159667969,
      "epoch": 2.856388088376561,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041911240359880517,
      "loss": 4.6335,
      "mean_token_accuracy": 0.24788206070661545,
      "num_tokens": 68193184.0,
      "step": 29735
    },
    {
      "entropy": 5.078259563446045,
      "epoch": 2.856868395773295,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041908607357721067,
      "loss": 4.7601,
      "mean_token_accuracy": 0.24390406161546707,
      "num_tokens": 68203984.0,
      "step": 29740
    },
    {
      "entropy": 5.076044082641602,
      "epoch": 2.857348703170029,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00041905974021036533,
      "loss": 4.665,
      "mean_token_accuracy": 0.24739695340394974,
      "num_tokens": 68215294.0,
      "step": 29745
    },
    {
      "entropy": 5.03832893371582,
      "epoch": 2.8578290105667628,
      "grad_norm": 1.0,
      "learning_rate": 0.00041903340349888065,
      "loss": 4.7639,
      "mean_token_accuracy": 0.2402009531855583,
      "num_tokens": 68226800.0,
      "step": 29750
    },
    {
      "entropy": 5.054838991165161,
      "epoch": 2.8583093179634966,
      "grad_norm": 0.984375,
      "learning_rate": 0.000419007063443368,
      "loss": 4.7054,
      "mean_token_accuracy": 0.24414485991001128,
      "num_tokens": 68238868.0,
      "step": 29755
    },
    {
      "entropy": 5.082677268981934,
      "epoch": 2.8587896253602305,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041898072004443906,
      "loss": 4.7377,
      "mean_token_accuracy": 0.24650082886219024,
      "num_tokens": 68250133.0,
      "step": 29760
    },
    {
      "entropy": 5.107921600341797,
      "epoch": 2.8592699327569644,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004189543733027052,
      "loss": 4.7562,
      "mean_token_accuracy": 0.23677445352077484,
      "num_tokens": 68261338.0,
      "step": 29765
    },
    {
      "entropy": 5.12339153289795,
      "epoch": 2.8597502401536983,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004189280232187783,
      "loss": 4.7213,
      "mean_token_accuracy": 0.23907239586114884,
      "num_tokens": 68272369.0,
      "step": 29770
    },
    {
      "entropy": 5.073468255996704,
      "epoch": 2.860230547550432,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004189016697932701,
      "loss": 4.744,
      "mean_token_accuracy": 0.24653567373752594,
      "num_tokens": 68283997.0,
      "step": 29775
    },
    {
      "entropy": 5.038801383972168,
      "epoch": 2.860710854947166,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004188753130267924,
      "loss": 4.6754,
      "mean_token_accuracy": 0.23701829612255096,
      "num_tokens": 68295719.0,
      "step": 29780
    },
    {
      "entropy": 5.075892734527588,
      "epoch": 2.8611911623439,
      "grad_norm": 1.125,
      "learning_rate": 0.0004188489529199572,
      "loss": 4.7506,
      "mean_token_accuracy": 0.24455582350492477,
      "num_tokens": 68307352.0,
      "step": 29785
    },
    {
      "entropy": 5.0274780750274655,
      "epoch": 2.8616714697406342,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041882258947337637,
      "loss": 4.6217,
      "mean_token_accuracy": 0.24203120172023773,
      "num_tokens": 68318889.0,
      "step": 29790
    },
    {
      "entropy": 4.977775001525879,
      "epoch": 2.8621517771373677,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041879622268766207,
      "loss": 4.605,
      "mean_token_accuracy": 0.24687892347574233,
      "num_tokens": 68329681.0,
      "step": 29795
    },
    {
      "entropy": 5.073207855224609,
      "epoch": 2.862632084534102,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004187698525634266,
      "loss": 4.7548,
      "mean_token_accuracy": 0.2422663927078247,
      "num_tokens": 68340323.0,
      "step": 29800
    },
    {
      "entropy": 4.986257457733155,
      "epoch": 2.863112391930836,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041874347910128193,
      "loss": 4.5778,
      "mean_token_accuracy": 0.24629957228899002,
      "num_tokens": 68351732.0,
      "step": 29805
    },
    {
      "entropy": 5.078960418701172,
      "epoch": 2.8635926993275698,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004187171023018406,
      "loss": 4.7431,
      "mean_token_accuracy": 0.23820737302303313,
      "num_tokens": 68362847.0,
      "step": 29810
    },
    {
      "entropy": 5.072182559967041,
      "epoch": 2.8640730067243036,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041869072216571486,
      "loss": 4.682,
      "mean_token_accuracy": 0.24200617223978044,
      "num_tokens": 68374368.0,
      "step": 29815
    },
    {
      "entropy": 5.002501916885376,
      "epoch": 2.8645533141210375,
      "grad_norm": 1.0,
      "learning_rate": 0.00041866433869351715,
      "loss": 4.6889,
      "mean_token_accuracy": 0.23933206647634506,
      "num_tokens": 68386463.0,
      "step": 29820
    },
    {
      "entropy": 5.0286425113677975,
      "epoch": 2.8650336215177714,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004186379518858602,
      "loss": 4.6649,
      "mean_token_accuracy": 0.24824004322290422,
      "num_tokens": 68397873.0,
      "step": 29825
    },
    {
      "entropy": 5.1290449619293215,
      "epoch": 2.8655139289145053,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004186115617433565,
      "loss": 4.7512,
      "mean_token_accuracy": 0.23202446848154068,
      "num_tokens": 68408561.0,
      "step": 29830
    },
    {
      "entropy": 5.070127534866333,
      "epoch": 2.865994236311239,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041858516826661876,
      "loss": 4.648,
      "mean_token_accuracy": 0.24680766016244887,
      "num_tokens": 68419878.0,
      "step": 29835
    },
    {
      "entropy": 5.078594207763672,
      "epoch": 2.866474543707973,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00041855877145625974,
      "loss": 4.7561,
      "mean_token_accuracy": 0.24475734680891037,
      "num_tokens": 68429913.0,
      "step": 29840
    },
    {
      "entropy": 5.105283260345459,
      "epoch": 2.866954851104707,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004185323713128924,
      "loss": 4.7619,
      "mean_token_accuracy": 0.24301362335681914,
      "num_tokens": 68441881.0,
      "step": 29845
    },
    {
      "entropy": 5.129477596282959,
      "epoch": 2.867435158501441,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041850596783712956,
      "loss": 4.6845,
      "mean_token_accuracy": 0.24058149755001068,
      "num_tokens": 68453191.0,
      "step": 29850
    },
    {
      "entropy": 5.087641191482544,
      "epoch": 2.8679154658981747,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004184795610295843,
      "loss": 4.71,
      "mean_token_accuracy": 0.24371355921030044,
      "num_tokens": 68464373.0,
      "step": 29855
    },
    {
      "entropy": 5.085254526138305,
      "epoch": 2.8683957732949086,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004184531508908697,
      "loss": 4.7495,
      "mean_token_accuracy": 0.2432363897562027,
      "num_tokens": 68474970.0,
      "step": 29860
    },
    {
      "entropy": 5.027404546737671,
      "epoch": 2.868876080691643,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004184267374215989,
      "loss": 4.7201,
      "mean_token_accuracy": 0.24200042188167573,
      "num_tokens": 68486587.0,
      "step": 29865
    },
    {
      "entropy": 5.077379655838013,
      "epoch": 2.8693563880883763,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000418400320622385,
      "loss": 4.7544,
      "mean_token_accuracy": 0.23684759587049484,
      "num_tokens": 68498091.0,
      "step": 29870
    },
    {
      "entropy": 5.105792236328125,
      "epoch": 2.8698366954851107,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004183739004938416,
      "loss": 4.7117,
      "mean_token_accuracy": 0.2385173872113228,
      "num_tokens": 68510431.0,
      "step": 29875
    },
    {
      "entropy": 5.015264892578125,
      "epoch": 2.870317002881844,
      "grad_norm": 1.015625,
      "learning_rate": 0.000418347477036582,
      "loss": 4.6155,
      "mean_token_accuracy": 0.2483307108283043,
      "num_tokens": 68522085.0,
      "step": 29880
    },
    {
      "entropy": 5.060935401916504,
      "epoch": 2.8707973102785784,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041832105025121956,
      "loss": 4.7195,
      "mean_token_accuracy": 0.2393105834722519,
      "num_tokens": 68533251.0,
      "step": 29885
    },
    {
      "entropy": 5.047052383422852,
      "epoch": 2.8712776176753123,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004182946201383679,
      "loss": 4.6481,
      "mean_token_accuracy": 0.24737206995487213,
      "num_tokens": 68545131.0,
      "step": 29890
    },
    {
      "entropy": 5.036336231231689,
      "epoch": 2.871757925072046,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004182681866986407,
      "loss": 4.6856,
      "mean_token_accuracy": 0.24738088697195054,
      "num_tokens": 68556838.0,
      "step": 29895
    },
    {
      "entropy": 5.063743686676025,
      "epoch": 2.87223823246878,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041824174993265165,
      "loss": 4.7015,
      "mean_token_accuracy": 0.24316679537296296,
      "num_tokens": 68570506.0,
      "step": 29900
    },
    {
      "entropy": 5.046118927001953,
      "epoch": 2.872718539865514,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041821530984101444,
      "loss": 4.6715,
      "mean_token_accuracy": 0.2431313395500183,
      "num_tokens": 68581364.0,
      "step": 29905
    },
    {
      "entropy": 5.048299026489258,
      "epoch": 2.873198847262248,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000418188866424343,
      "loss": 4.6808,
      "mean_token_accuracy": 0.23940720409154892,
      "num_tokens": 68593047.0,
      "step": 29910
    },
    {
      "entropy": 5.080627727508545,
      "epoch": 2.8736791546589817,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004181624196832513,
      "loss": 4.7235,
      "mean_token_accuracy": 0.2357817992568016,
      "num_tokens": 68604592.0,
      "step": 29915
    },
    {
      "entropy": 5.080392694473266,
      "epoch": 2.8741594620557156,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00041813596961835336,
      "loss": 4.7206,
      "mean_token_accuracy": 0.24580902606248856,
      "num_tokens": 68616566.0,
      "step": 29920
    },
    {
      "entropy": 4.978018188476563,
      "epoch": 2.8746397694524495,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041810951623026313,
      "loss": 4.5876,
      "mean_token_accuracy": 0.24634382575750352,
      "num_tokens": 68626507.0,
      "step": 29925
    },
    {
      "entropy": 4.974175977706909,
      "epoch": 2.8751200768491834,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041808305951959496,
      "loss": 4.5713,
      "mean_token_accuracy": 0.25241384357213975,
      "num_tokens": 68637748.0,
      "step": 29930
    },
    {
      "entropy": 5.037887668609619,
      "epoch": 2.8756003842459172,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000418056599486963,
      "loss": 4.7183,
      "mean_token_accuracy": 0.24089150428771972,
      "num_tokens": 68649379.0,
      "step": 29935
    },
    {
      "entropy": 5.094043636322022,
      "epoch": 2.8760806916426516,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004180301361329816,
      "loss": 4.7086,
      "mean_token_accuracy": 0.2336302936077118,
      "num_tokens": 68660834.0,
      "step": 29940
    },
    {
      "entropy": 5.057880783081055,
      "epoch": 2.876560999039385,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004180036694582651,
      "loss": 4.7295,
      "mean_token_accuracy": 0.23456861972808837,
      "num_tokens": 68673748.0,
      "step": 29945
    },
    {
      "entropy": 5.05615553855896,
      "epoch": 2.8770413064361193,
      "grad_norm": 1.0,
      "learning_rate": 0.00041797719946342813,
      "loss": 4.7085,
      "mean_token_accuracy": 0.23973922729492186,
      "num_tokens": 68684594.0,
      "step": 29950
    },
    {
      "entropy": 5.171247959136963,
      "epoch": 2.8775216138328528,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041795072614908503,
      "loss": 4.8375,
      "mean_token_accuracy": 0.22968949973583222,
      "num_tokens": 68697954.0,
      "step": 29955
    },
    {
      "entropy": 5.178190469741821,
      "epoch": 2.878001921229587,
      "grad_norm": 1.0,
      "learning_rate": 0.00041792424951585055,
      "loss": 4.852,
      "mean_token_accuracy": 0.23801036030054093,
      "num_tokens": 68708616.0,
      "step": 29960
    },
    {
      "entropy": 4.950202560424804,
      "epoch": 2.878482228626321,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041789776956433947,
      "loss": 4.5685,
      "mean_token_accuracy": 0.2543065890669823,
      "num_tokens": 68720226.0,
      "step": 29965
    },
    {
      "entropy": 5.077175140380859,
      "epoch": 2.878962536023055,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041787128629516645,
      "loss": 4.7155,
      "mean_token_accuracy": 0.2402342900633812,
      "num_tokens": 68731134.0,
      "step": 29970
    },
    {
      "entropy": 5.097321319580078,
      "epoch": 2.8794428434197887,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004178447997089464,
      "loss": 4.7627,
      "mean_token_accuracy": 0.24062544703483582,
      "num_tokens": 68743158.0,
      "step": 29975
    },
    {
      "entropy": 5.026002693176269,
      "epoch": 2.8799231508165226,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004178183098062943,
      "loss": 4.7265,
      "mean_token_accuracy": 0.23758306950330735,
      "num_tokens": 68755411.0,
      "step": 29980
    },
    {
      "entropy": 5.109521865844727,
      "epoch": 2.8804034582132565,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004177918165878251,
      "loss": 4.7681,
      "mean_token_accuracy": 0.24067500233650208,
      "num_tokens": 68767621.0,
      "step": 29985
    },
    {
      "entropy": 5.0098731994628904,
      "epoch": 2.8808837656099904,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004177653200541539,
      "loss": 4.6631,
      "mean_token_accuracy": 0.24670595079660415,
      "num_tokens": 68778861.0,
      "step": 29990
    },
    {
      "entropy": 5.095999813079834,
      "epoch": 2.8813640730067243,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000417738820205896,
      "loss": 4.7143,
      "mean_token_accuracy": 0.24267756193876266,
      "num_tokens": 68790101.0,
      "step": 29995
    },
    {
      "entropy": 4.932431030273437,
      "epoch": 2.881844380403458,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004177123170436665,
      "loss": 4.5894,
      "mean_token_accuracy": 0.24978718906641006,
      "num_tokens": 68801612.0,
      "step": 30000
    },
    {
      "epoch": 2.881844380403458,
      "eval_entropy": 4.857383004343841,
      "eval_loss": 4.815639972686768,
      "eval_mean_token_accuracy": 0.24364217947796168,
      "eval_num_tokens": 68801612.0,
      "eval_runtime": 26.584,
      "eval_samples_per_second": 1234.391,
      "eval_steps_per_second": 154.304,
      "step": 30000
    }
  ],
  "logging_steps": 5,
  "max_steps": 104090,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.05280319855232e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}