{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.3843164959378313,
  "eval_steps": 3000,
  "global_step": 27000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 7.74092288017273,
      "epoch": 0.0004415400918403391,
      "grad_norm": 0.9375,
      "learning_rate": 2e-06,
      "loss": 7.5069,
      "mean_token_accuracy": 0.0897733099758625,
      "num_tokens": 9390.0,
      "step": 5
    },
    {
      "entropy": 7.730613660812378,
      "epoch": 0.0008830801836806782,
      "grad_norm": 0.953125,
      "learning_rate": 4.5e-06,
      "loss": 7.6006,
      "mean_token_accuracy": 0.09082280173897743,
      "num_tokens": 18671.0,
      "step": 10
    },
    {
      "entropy": 7.711504316329956,
      "epoch": 0.0013246202755210173,
      "grad_norm": 0.9375,
      "learning_rate": 7e-06,
      "loss": 7.5458,
      "mean_token_accuracy": 0.09541130512952804,
      "num_tokens": 27614.0,
      "step": 15
    },
    {
      "entropy": 7.760569763183594,
      "epoch": 0.0017661603673613563,
      "grad_norm": 0.94140625,
      "learning_rate": 9.5e-06,
      "loss": 7.6561,
      "mean_token_accuracy": 0.08549572601914406,
      "num_tokens": 37850.0,
      "step": 20
    },
    {
      "entropy": 7.768425226211548,
      "epoch": 0.0022077004592016957,
      "grad_norm": 0.9375,
      "learning_rate": 1.2e-05,
      "loss": 7.5482,
      "mean_token_accuracy": 0.10141252279281616,
      "num_tokens": 47166.0,
      "step": 25
    },
    {
      "entropy": 7.844929361343384,
      "epoch": 0.0026492405510420347,
      "grad_norm": 0.88671875,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 7.5478,
      "mean_token_accuracy": 0.0926934964954853,
      "num_tokens": 55500.0,
      "step": 30
    },
    {
      "entropy": 7.811486482620239,
      "epoch": 0.0030907806428823736,
      "grad_norm": 0.88671875,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 7.5563,
      "mean_token_accuracy": 0.08848847448825836,
      "num_tokens": 63851.0,
      "step": 35
    },
    {
      "entropy": 7.876116132736206,
      "epoch": 0.0035323207347227126,
      "grad_norm": 0.9453125,
      "learning_rate": 1.95e-05,
      "loss": 7.5805,
      "mean_token_accuracy": 0.09331603050231933,
      "num_tokens": 73697.0,
      "step": 40
    },
    {
      "entropy": 7.89044713973999,
      "epoch": 0.003973860826563052,
      "grad_norm": 1.0390625,
      "learning_rate": 2.2e-05,
      "loss": 7.5408,
      "mean_token_accuracy": 0.09619837477803231,
      "num_tokens": 83000.0,
      "step": 45
    },
    {
      "entropy": 7.889915847778321,
      "epoch": 0.004415400918403391,
      "grad_norm": 0.80859375,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 7.5514,
      "mean_token_accuracy": 0.09451765418052674,
      "num_tokens": 92982.0,
      "step": 50
    },
    {
      "entropy": 7.9315505027771,
      "epoch": 0.00485694101024373,
      "grad_norm": 1.0078125,
      "learning_rate": 2.7e-05,
      "loss": 7.5183,
      "mean_token_accuracy": 0.09080160632729531,
      "num_tokens": 101455.0,
      "step": 55
    },
    {
      "entropy": 7.839016962051391,
      "epoch": 0.005298481102084069,
      "grad_norm": 0.8359375,
      "learning_rate": 2.95e-05,
      "loss": 7.5322,
      "mean_token_accuracy": 0.09154317900538445,
      "num_tokens": 110782.0,
      "step": 60
    },
    {
      "entropy": 7.82890944480896,
      "epoch": 0.005740021193924408,
      "grad_norm": 0.89453125,
      "learning_rate": 3.2e-05,
      "loss": 7.4277,
      "mean_token_accuracy": 0.09484038800001145,
      "num_tokens": 119241.0,
      "step": 65
    },
    {
      "entropy": 7.828151416778565,
      "epoch": 0.006181561285764747,
      "grad_norm": 0.94140625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 7.5235,
      "mean_token_accuracy": 0.08991453871130943,
      "num_tokens": 127903.0,
      "step": 70
    },
    {
      "entropy": 7.8103076934814455,
      "epoch": 0.006623101377605086,
      "grad_norm": 0.90234375,
      "learning_rate": 3.7e-05,
      "loss": 7.5,
      "mean_token_accuracy": 0.09334810897707939,
      "num_tokens": 137370.0,
      "step": 75
    },
    {
      "entropy": 7.8186195373535154,
      "epoch": 0.007064641469445425,
      "grad_norm": 0.84375,
      "learning_rate": 3.95e-05,
      "loss": 7.4449,
      "mean_token_accuracy": 0.09487180560827255,
      "num_tokens": 146582.0,
      "step": 80
    },
    {
      "entropy": 7.844836282730102,
      "epoch": 0.007506181561285765,
      "grad_norm": 0.9296875,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 7.4263,
      "mean_token_accuracy": 0.10111527815461159,
      "num_tokens": 154933.0,
      "step": 85
    },
    {
      "entropy": 7.830266809463501,
      "epoch": 0.007947721653126103,
      "grad_norm": 0.875,
      "learning_rate": 4.45e-05,
      "loss": 7.5282,
      "mean_token_accuracy": 0.09698798581957817,
      "num_tokens": 165157.0,
      "step": 90
    },
    {
      "entropy": 7.830750465393066,
      "epoch": 0.008389261744966443,
      "grad_norm": 0.84375,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 7.4588,
      "mean_token_accuracy": 0.0967378407716751,
      "num_tokens": 174958.0,
      "step": 95
    },
    {
      "entropy": 7.780857229232788,
      "epoch": 0.008830801836806783,
      "grad_norm": 0.93359375,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 7.5012,
      "mean_token_accuracy": 0.09829336479306221,
      "num_tokens": 184256.0,
      "step": 100
    },
    {
      "entropy": 7.829332780838013,
      "epoch": 0.009272341928647121,
      "grad_norm": 0.8984375,
      "learning_rate": 5.2e-05,
      "loss": 7.4522,
      "mean_token_accuracy": 0.09847906306385994,
      "num_tokens": 192894.0,
      "step": 105
    },
    {
      "entropy": 7.851593732833862,
      "epoch": 0.00971388202048746,
      "grad_norm": 0.96875,
      "learning_rate": 5.45e-05,
      "loss": 7.4989,
      "mean_token_accuracy": 0.097539983689785,
      "num_tokens": 202675.0,
      "step": 110
    },
    {
      "entropy": 7.8246654033660885,
      "epoch": 0.010155422112327799,
      "grad_norm": 0.90625,
      "learning_rate": 5.7e-05,
      "loss": 7.5051,
      "mean_token_accuracy": 0.09736146405339241,
      "num_tokens": 212261.0,
      "step": 115
    },
    {
      "entropy": 7.855020570755005,
      "epoch": 0.010596962204168139,
      "grad_norm": 0.90234375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 7.582,
      "mean_token_accuracy": 0.09346140772104264,
      "num_tokens": 222329.0,
      "step": 120
    },
    {
      "entropy": 7.888062286376953,
      "epoch": 0.011038502296008477,
      "grad_norm": 0.984375,
      "learning_rate": 6.2e-05,
      "loss": 7.5076,
      "mean_token_accuracy": 0.08999184891581535,
      "num_tokens": 231247.0,
      "step": 125
    },
    {
      "entropy": 7.855972003936768,
      "epoch": 0.011480042387848817,
      "grad_norm": 1.0859375,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.4591,
      "mean_token_accuracy": 0.10294558182358741,
      "num_tokens": 239978.0,
      "step": 130
    },
    {
      "entropy": 7.773235321044922,
      "epoch": 0.011921582479689156,
      "grad_norm": 0.9375,
      "learning_rate": 6.7e-05,
      "loss": 7.4769,
      "mean_token_accuracy": 0.09530732408165932,
      "num_tokens": 249735.0,
      "step": 135
    },
    {
      "entropy": 7.8279619216918945,
      "epoch": 0.012363122571529495,
      "grad_norm": 1.15625,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.5184,
      "mean_token_accuracy": 0.0965347282588482,
      "num_tokens": 259369.0,
      "step": 140
    },
    {
      "entropy": 7.942261266708374,
      "epoch": 0.012804662663369834,
      "grad_norm": 0.91015625,
      "learning_rate": 7.2e-05,
      "loss": 7.5286,
      "mean_token_accuracy": 0.09435783103108406,
      "num_tokens": 268645.0,
      "step": 145
    },
    {
      "entropy": 7.800592231750488,
      "epoch": 0.013246202755210172,
      "grad_norm": 0.96875,
      "learning_rate": 7.45e-05,
      "loss": 7.4194,
      "mean_token_accuracy": 0.10034767463803292,
      "num_tokens": 276667.0,
      "step": 150
    },
    {
      "entropy": 7.797780895233155,
      "epoch": 0.013687742847050512,
      "grad_norm": 0.87890625,
      "learning_rate": 7.7e-05,
      "loss": 7.4485,
      "mean_token_accuracy": 0.1016088955104351,
      "num_tokens": 286017.0,
      "step": 155
    },
    {
      "entropy": 7.838827800750733,
      "epoch": 0.01412928293889085,
      "grad_norm": 0.8984375,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.4773,
      "mean_token_accuracy": 0.0957870528101921,
      "num_tokens": 295631.0,
      "step": 160
    },
    {
      "entropy": 7.811876916885376,
      "epoch": 0.01457082303073119,
      "grad_norm": 0.91015625,
      "learning_rate": 8.2e-05,
      "loss": 7.4769,
      "mean_token_accuracy": 0.10056489408016205,
      "num_tokens": 304704.0,
      "step": 165
    },
    {
      "entropy": 7.836099576950073,
      "epoch": 0.01501236312257153,
      "grad_norm": 0.921875,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.506,
      "mean_token_accuracy": 0.09576743617653846,
      "num_tokens": 314195.0,
      "step": 170
    },
    {
      "entropy": 7.863115644454956,
      "epoch": 0.015453903214411868,
      "grad_norm": 1.078125,
      "learning_rate": 8.7e-05,
      "loss": 7.5172,
      "mean_token_accuracy": 0.09137798845767975,
      "num_tokens": 323379.0,
      "step": 175
    },
    {
      "entropy": 7.841438961029053,
      "epoch": 0.015895443306252206,
      "grad_norm": 0.9140625,
      "learning_rate": 8.95e-05,
      "loss": 7.4337,
      "mean_token_accuracy": 0.10475980639457702,
      "num_tokens": 332322.0,
      "step": 180
    },
    {
      "entropy": 7.761782026290893,
      "epoch": 0.016336983398092548,
      "grad_norm": 1.0,
      "learning_rate": 9.2e-05,
      "loss": 7.4066,
      "mean_token_accuracy": 0.10371800139546394,
      "num_tokens": 341735.0,
      "step": 185
    },
    {
      "entropy": 7.892005348205567,
      "epoch": 0.016778523489932886,
      "grad_norm": 0.96875,
      "learning_rate": 9.45e-05,
      "loss": 7.5071,
      "mean_token_accuracy": 0.09807236269116401,
      "num_tokens": 351209.0,
      "step": 190
    },
    {
      "entropy": 7.800058889389038,
      "epoch": 0.017220063581773224,
      "grad_norm": 0.91796875,
      "learning_rate": 9.7e-05,
      "loss": 7.4442,
      "mean_token_accuracy": 0.10099716112017632,
      "num_tokens": 360467.0,
      "step": 195
    },
    {
      "entropy": 7.8116179466247555,
      "epoch": 0.017661603673613566,
      "grad_norm": 0.8828125,
      "learning_rate": 9.95e-05,
      "loss": 7.4501,
      "mean_token_accuracy": 0.10495151057839394,
      "num_tokens": 370361.0,
      "step": 200
    },
    {
      "entropy": 7.931649780273437,
      "epoch": 0.018103143765453904,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000102,
      "loss": 7.496,
      "mean_token_accuracy": 0.0929229848086834,
      "num_tokens": 380366.0,
      "step": 205
    },
    {
      "entropy": 7.765449905395508,
      "epoch": 0.018544683857294242,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.4157,
      "mean_token_accuracy": 0.1013551875948906,
      "num_tokens": 390690.0,
      "step": 210
    },
    {
      "entropy": 7.802529335021973,
      "epoch": 0.01898622394913458,
      "grad_norm": 0.953125,
      "learning_rate": 0.000107,
      "loss": 7.4996,
      "mean_token_accuracy": 0.09649925082921981,
      "num_tokens": 400722.0,
      "step": 215
    },
    {
      "entropy": 7.8250895023345945,
      "epoch": 0.01942776404097492,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0001095,
      "loss": 7.3837,
      "mean_token_accuracy": 0.1022213265299797,
      "num_tokens": 410223.0,
      "step": 220
    },
    {
      "entropy": 7.719676780700683,
      "epoch": 0.01986930413281526,
      "grad_norm": 0.953125,
      "learning_rate": 0.000112,
      "loss": 7.4551,
      "mean_token_accuracy": 0.10060659795999527,
      "num_tokens": 420214.0,
      "step": 225
    },
    {
      "entropy": 7.863075590133667,
      "epoch": 0.020310844224655598,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0001145,
      "loss": 7.4203,
      "mean_token_accuracy": 0.10097097754478454,
      "num_tokens": 429407.0,
      "step": 230
    },
    {
      "entropy": 7.792139530181885,
      "epoch": 0.02075238431649594,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.3518,
      "mean_token_accuracy": 0.0997156098484993,
      "num_tokens": 438403.0,
      "step": 235
    },
    {
      "entropy": 7.722495126724243,
      "epoch": 0.021193924408336277,
      "grad_norm": 1.0625,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.404,
      "mean_token_accuracy": 0.10603894293308258,
      "num_tokens": 447466.0,
      "step": 240
    },
    {
      "entropy": 7.889057874679565,
      "epoch": 0.021635464500176615,
      "grad_norm": 1.09375,
      "learning_rate": 0.000122,
      "loss": 7.5207,
      "mean_token_accuracy": 0.09140023961663246,
      "num_tokens": 457141.0,
      "step": 245
    },
    {
      "entropy": 7.798206424713134,
      "epoch": 0.022077004592016954,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0001245,
      "loss": 7.3654,
      "mean_token_accuracy": 0.10516906753182412,
      "num_tokens": 465708.0,
      "step": 250
    },
    {
      "entropy": 7.7851112365722654,
      "epoch": 0.022518544683857295,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000127,
      "loss": 7.4083,
      "mean_token_accuracy": 0.10015119016170501,
      "num_tokens": 475369.0,
      "step": 255
    },
    {
      "entropy": 7.747976684570313,
      "epoch": 0.022960084775697633,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0001295,
      "loss": 7.3678,
      "mean_token_accuracy": 0.10004653111100197,
      "num_tokens": 484249.0,
      "step": 260
    },
    {
      "entropy": 7.840033864974975,
      "epoch": 0.02340162486753797,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000132,
      "loss": 7.4932,
      "mean_token_accuracy": 0.09563242942094803,
      "num_tokens": 493303.0,
      "step": 265
    },
    {
      "entropy": 7.812006616592408,
      "epoch": 0.023843164959378313,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.449,
      "mean_token_accuracy": 0.10385405197739601,
      "num_tokens": 501503.0,
      "step": 270
    },
    {
      "entropy": 7.883935546875,
      "epoch": 0.02428470505121865,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.3669,
      "mean_token_accuracy": 0.10403206944465637,
      "num_tokens": 509661.0,
      "step": 275
    },
    {
      "entropy": 7.698855304718018,
      "epoch": 0.02472624514305899,
      "grad_norm": 1.0,
      "learning_rate": 0.0001395,
      "loss": 7.3695,
      "mean_token_accuracy": 0.10018852427601814,
      "num_tokens": 519464.0,
      "step": 280
    },
    {
      "entropy": 7.810435962677002,
      "epoch": 0.025167785234899327,
      "grad_norm": 1.21875,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.3442,
      "mean_token_accuracy": 0.10638531595468521,
      "num_tokens": 527968.0,
      "step": 285
    },
    {
      "entropy": 7.726656579971314,
      "epoch": 0.02560932532673967,
      "grad_norm": 1.015625,
      "learning_rate": 0.0001445,
      "loss": 7.4085,
      "mean_token_accuracy": 0.10700022876262665,
      "num_tokens": 537234.0,
      "step": 290
    },
    {
      "entropy": 7.8001549243927,
      "epoch": 0.026050865418580007,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000147,
      "loss": 7.3728,
      "mean_token_accuracy": 0.10392982810735703,
      "num_tokens": 546398.0,
      "step": 295
    },
    {
      "entropy": 7.735007047653198,
      "epoch": 0.026492405510420345,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0001495,
      "loss": 7.3857,
      "mean_token_accuracy": 0.10330862030386925,
      "num_tokens": 555362.0,
      "step": 300
    },
    {
      "entropy": 7.846773290634156,
      "epoch": 0.026933945602260687,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000152,
      "loss": 7.4235,
      "mean_token_accuracy": 0.09694105237722397,
      "num_tokens": 564575.0,
      "step": 305
    },
    {
      "entropy": 7.824136352539062,
      "epoch": 0.027375485694101025,
      "grad_norm": 1.09375,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.4528,
      "mean_token_accuracy": 0.10274629592895508,
      "num_tokens": 573915.0,
      "step": 310
    },
    {
      "entropy": 7.747010231018066,
      "epoch": 0.027817025785941363,
      "grad_norm": 1.109375,
      "learning_rate": 0.000157,
      "loss": 7.3458,
      "mean_token_accuracy": 0.10164012908935546,
      "num_tokens": 583216.0,
      "step": 315
    },
    {
      "entropy": 7.775792646408081,
      "epoch": 0.0282585658777817,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0001595,
      "loss": 7.3946,
      "mean_token_accuracy": 0.10511558204889297,
      "num_tokens": 591955.0,
      "step": 320
    },
    {
      "entropy": 7.6797911643981935,
      "epoch": 0.028700105969622042,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000162,
      "loss": 7.3672,
      "mean_token_accuracy": 0.103659038990736,
      "num_tokens": 600999.0,
      "step": 325
    },
    {
      "entropy": 7.804133558273316,
      "epoch": 0.02914164606146238,
      "grad_norm": 1.0,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.2868,
      "mean_token_accuracy": 0.11126004680991172,
      "num_tokens": 609478.0,
      "step": 330
    },
    {
      "entropy": 7.633207130432129,
      "epoch": 0.02958318615330272,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.3672,
      "mean_token_accuracy": 0.10404063463211059,
      "num_tokens": 618348.0,
      "step": 335
    },
    {
      "entropy": 7.855530452728272,
      "epoch": 0.03002472624514306,
      "grad_norm": 1.25,
      "learning_rate": 0.00016950000000000003,
      "loss": 7.4232,
      "mean_token_accuracy": 0.10137763321399688,
      "num_tokens": 628548.0,
      "step": 340
    },
    {
      "entropy": 7.745834493637085,
      "epoch": 0.0304662663369834,
      "grad_norm": 1.09375,
      "learning_rate": 0.00017199999999999998,
      "loss": 7.3651,
      "mean_token_accuracy": 0.10630879923701286,
      "num_tokens": 637489.0,
      "step": 345
    },
    {
      "entropy": 7.684433507919311,
      "epoch": 0.030907806428823736,
      "grad_norm": 1.125,
      "learning_rate": 0.00017449999999999999,
      "loss": 7.3503,
      "mean_token_accuracy": 0.10281694158911706,
      "num_tokens": 646715.0,
      "step": 350
    },
    {
      "entropy": 7.803950309753418,
      "epoch": 0.031349346520664075,
      "grad_norm": 1.015625,
      "learning_rate": 0.000177,
      "loss": 7.3499,
      "mean_token_accuracy": 0.10527043938636779,
      "num_tokens": 656858.0,
      "step": 355
    },
    {
      "entropy": 7.707482957839966,
      "epoch": 0.03179088661250441,
      "grad_norm": 1.140625,
      "learning_rate": 0.0001795,
      "loss": 7.3208,
      "mean_token_accuracy": 0.10721890702843666,
      "num_tokens": 665968.0,
      "step": 360
    },
    {
      "entropy": 7.764640760421753,
      "epoch": 0.03223242670434476,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000182,
      "loss": 7.315,
      "mean_token_accuracy": 0.10370069816708564,
      "num_tokens": 674295.0,
      "step": 365
    },
    {
      "entropy": 7.695739364624023,
      "epoch": 0.032673966796185096,
      "grad_norm": 1.0,
      "learning_rate": 0.0001845,
      "loss": 7.2613,
      "mean_token_accuracy": 0.10737213939428329,
      "num_tokens": 683559.0,
      "step": 370
    },
    {
      "entropy": 7.558203125,
      "epoch": 0.033115506888025434,
      "grad_norm": 1.328125,
      "learning_rate": 0.000187,
      "loss": 7.2997,
      "mean_token_accuracy": 0.10474916398525239,
      "num_tokens": 692402.0,
      "step": 375
    },
    {
      "entropy": 7.85990538597107,
      "epoch": 0.03355704697986577,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0001895,
      "loss": 7.4051,
      "mean_token_accuracy": 0.09768549725413322,
      "num_tokens": 702052.0,
      "step": 380
    },
    {
      "entropy": 7.703890705108643,
      "epoch": 0.03399858707170611,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000192,
      "loss": 7.3332,
      "mean_token_accuracy": 0.10592006072402001,
      "num_tokens": 711926.0,
      "step": 385
    },
    {
      "entropy": 7.764595699310303,
      "epoch": 0.03444012716354645,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0001945,
      "loss": 7.2823,
      "mean_token_accuracy": 0.09970764666795731,
      "num_tokens": 720948.0,
      "step": 390
    },
    {
      "entropy": 7.651279592514038,
      "epoch": 0.034881667255386786,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00019700000000000002,
      "loss": 7.3228,
      "mean_token_accuracy": 0.11015688627958298,
      "num_tokens": 729611.0,
      "step": 395
    },
    {
      "entropy": 7.79816370010376,
      "epoch": 0.03532320734722713,
      "grad_norm": 1.171875,
      "learning_rate": 0.00019950000000000002,
      "loss": 7.3073,
      "mean_token_accuracy": 0.09714326411485671,
      "num_tokens": 738433.0,
      "step": 400
    },
    {
      "entropy": 7.687422800064087,
      "epoch": 0.03576474743906747,
      "grad_norm": 1.078125,
      "learning_rate": 0.000202,
      "loss": 7.3504,
      "mean_token_accuracy": 0.103574138879776,
      "num_tokens": 747310.0,
      "step": 405
    },
    {
      "entropy": 7.690566205978394,
      "epoch": 0.03620628753090781,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00020449999999999998,
      "loss": 7.3092,
      "mean_token_accuracy": 0.10628200694918633,
      "num_tokens": 756362.0,
      "step": 410
    },
    {
      "entropy": 7.803203916549682,
      "epoch": 0.036647827622748146,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000207,
      "loss": 7.3251,
      "mean_token_accuracy": 0.09963449612259864,
      "num_tokens": 764978.0,
      "step": 415
    },
    {
      "entropy": 7.557148265838623,
      "epoch": 0.037089367714588484,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002095,
      "loss": 7.299,
      "mean_token_accuracy": 0.10668339729309081,
      "num_tokens": 774058.0,
      "step": 420
    },
    {
      "entropy": 7.685351037979126,
      "epoch": 0.03753090780642882,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000212,
      "loss": 7.2881,
      "mean_token_accuracy": 0.10115964189171792,
      "num_tokens": 783332.0,
      "step": 425
    },
    {
      "entropy": 7.726617670059204,
      "epoch": 0.03797244789826916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002145,
      "loss": 7.3501,
      "mean_token_accuracy": 0.10238557234406472,
      "num_tokens": 792481.0,
      "step": 430
    },
    {
      "entropy": 7.598264503479004,
      "epoch": 0.038413987990109505,
      "grad_norm": 1.203125,
      "learning_rate": 0.00021700000000000002,
      "loss": 7.2437,
      "mean_token_accuracy": 0.11149554774165153,
      "num_tokens": 801396.0,
      "step": 435
    },
    {
      "entropy": 7.771391916275024,
      "epoch": 0.03885552808194984,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002195,
      "loss": 7.3027,
      "mean_token_accuracy": 0.10071089640259742,
      "num_tokens": 809903.0,
      "step": 440
    },
    {
      "entropy": 7.631546688079834,
      "epoch": 0.03929706817379018,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000222,
      "loss": 7.2639,
      "mean_token_accuracy": 0.10943293869495392,
      "num_tokens": 819144.0,
      "step": 445
    },
    {
      "entropy": 7.66679310798645,
      "epoch": 0.03973860826563052,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002245,
      "loss": 7.2976,
      "mean_token_accuracy": 0.10613659396767616,
      "num_tokens": 828881.0,
      "step": 450
    },
    {
      "entropy": 7.639922952651977,
      "epoch": 0.04018014835747086,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00022700000000000002,
      "loss": 7.224,
      "mean_token_accuracy": 0.09946361035108567,
      "num_tokens": 837588.0,
      "step": 455
    },
    {
      "entropy": 7.582935905456543,
      "epoch": 0.040621688449311195,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00022950000000000002,
      "loss": 7.2197,
      "mean_token_accuracy": 0.1077489398419857,
      "num_tokens": 847002.0,
      "step": 460
    },
    {
      "entropy": 7.601295232772827,
      "epoch": 0.041063228541151534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023200000000000003,
      "loss": 7.2374,
      "mean_token_accuracy": 0.10669372826814652,
      "num_tokens": 855791.0,
      "step": 465
    },
    {
      "entropy": 7.683427000045777,
      "epoch": 0.04150476863299188,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00023449999999999998,
      "loss": 7.2737,
      "mean_token_accuracy": 0.10692826434969901,
      "num_tokens": 865392.0,
      "step": 470
    },
    {
      "entropy": 7.632460880279541,
      "epoch": 0.04194630872483222,
      "grad_norm": 1.140625,
      "learning_rate": 0.000237,
      "loss": 7.3431,
      "mean_token_accuracy": 0.10185316428542138,
      "num_tokens": 874807.0,
      "step": 475
    },
    {
      "entropy": 7.7053466796875,
      "epoch": 0.042387848816672555,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002395,
      "loss": 7.3151,
      "mean_token_accuracy": 0.107633588463068,
      "num_tokens": 883928.0,
      "step": 480
    },
    {
      "entropy": 7.573175430297852,
      "epoch": 0.04282938890851289,
      "grad_norm": 1.046875,
      "learning_rate": 0.000242,
      "loss": 7.2153,
      "mean_token_accuracy": 0.10783804580569267,
      "num_tokens": 893047.0,
      "step": 485
    },
    {
      "entropy": 7.575657176971435,
      "epoch": 0.04327092900035323,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002445,
      "loss": 7.2168,
      "mean_token_accuracy": 0.1058282122015953,
      "num_tokens": 901645.0,
      "step": 490
    },
    {
      "entropy": 7.620817947387695,
      "epoch": 0.04371246909219357,
      "grad_norm": 1.234375,
      "learning_rate": 0.000247,
      "loss": 7.2325,
      "mean_token_accuracy": 0.10362469628453255,
      "num_tokens": 911169.0,
      "step": 495
    },
    {
      "entropy": 7.675788545608521,
      "epoch": 0.04415400918403391,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002495,
      "loss": 7.239,
      "mean_token_accuracy": 0.09903389513492585,
      "num_tokens": 921382.0,
      "step": 500
    },
    {
      "entropy": 7.509087562561035,
      "epoch": 0.04459554927587425,
      "grad_norm": 1.15625,
      "learning_rate": 0.000252,
      "loss": 7.1784,
      "mean_token_accuracy": 0.10705869123339654,
      "num_tokens": 930409.0,
      "step": 505
    },
    {
      "entropy": 7.563630390167236,
      "epoch": 0.04503708936771459,
      "grad_norm": 1.125,
      "learning_rate": 0.0002545,
      "loss": 7.1437,
      "mean_token_accuracy": 0.11103534623980522,
      "num_tokens": 939049.0,
      "step": 510
    },
    {
      "entropy": 7.656319618225098,
      "epoch": 0.04547862945955493,
      "grad_norm": 1.140625,
      "learning_rate": 0.000257,
      "loss": 7.1669,
      "mean_token_accuracy": 0.10674394220113755,
      "num_tokens": 947575.0,
      "step": 515
    },
    {
      "entropy": 7.617980527877807,
      "epoch": 0.045920169551395267,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0002595,
      "loss": 7.3263,
      "mean_token_accuracy": 0.10124524608254433,
      "num_tokens": 957848.0,
      "step": 520
    },
    {
      "entropy": 7.594845724105835,
      "epoch": 0.046361709643235605,
      "grad_norm": 1.21875,
      "learning_rate": 0.000262,
      "loss": 7.1638,
      "mean_token_accuracy": 0.1080893911421299,
      "num_tokens": 966521.0,
      "step": 525
    },
    {
      "entropy": 7.610377073287964,
      "epoch": 0.04680324973507594,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026450000000000003,
      "loss": 7.2181,
      "mean_token_accuracy": 0.10909054726362229,
      "num_tokens": 975827.0,
      "step": 530
    },
    {
      "entropy": 7.499869298934937,
      "epoch": 0.04724478982691628,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00026700000000000004,
      "loss": 7.1793,
      "mean_token_accuracy": 0.11137327626347542,
      "num_tokens": 985292.0,
      "step": 535
    },
    {
      "entropy": 7.5960760593414305,
      "epoch": 0.047686329918756626,
      "grad_norm": 1.15625,
      "learning_rate": 0.00026950000000000005,
      "loss": 7.154,
      "mean_token_accuracy": 0.10950936675071717,
      "num_tokens": 994791.0,
      "step": 540
    },
    {
      "entropy": 7.469157934188843,
      "epoch": 0.048127870010596964,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00027200000000000005,
      "loss": 7.1317,
      "mean_token_accuracy": 0.11670516580343246,
      "num_tokens": 1003700.0,
      "step": 545
    },
    {
      "entropy": 7.523545932769776,
      "epoch": 0.0485694101024373,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002745,
      "loss": 7.1673,
      "mean_token_accuracy": 0.11030299812555314,
      "num_tokens": 1012682.0,
      "step": 550
    },
    {
      "entropy": 7.583120203018188,
      "epoch": 0.04901095019427764,
      "grad_norm": 1.453125,
      "learning_rate": 0.000277,
      "loss": 7.1875,
      "mean_token_accuracy": 0.1056957058608532,
      "num_tokens": 1021018.0,
      "step": 555
    },
    {
      "entropy": 7.558827114105225,
      "epoch": 0.04945249028611798,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002795,
      "loss": 7.0584,
      "mean_token_accuracy": 0.1091762512922287,
      "num_tokens": 1029744.0,
      "step": 560
    },
    {
      "entropy": 7.384686899185181,
      "epoch": 0.049894030377958316,
      "grad_norm": 1.21875,
      "learning_rate": 0.00028199999999999997,
      "loss": 7.1435,
      "mean_token_accuracy": 0.11267905756831169,
      "num_tokens": 1038805.0,
      "step": 565
    },
    {
      "entropy": 7.607424783706665,
      "epoch": 0.050335570469798654,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002845,
      "loss": 7.1996,
      "mean_token_accuracy": 0.10786581486463546,
      "num_tokens": 1047656.0,
      "step": 570
    },
    {
      "entropy": 7.606092405319214,
      "epoch": 0.050777110561639,
      "grad_norm": 1.15625,
      "learning_rate": 0.000287,
      "loss": 7.1792,
      "mean_token_accuracy": 0.11288331896066665,
      "num_tokens": 1056598.0,
      "step": 575
    },
    {
      "entropy": 7.5449169158935545,
      "epoch": 0.05121865065347934,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002895,
      "loss": 7.1883,
      "mean_token_accuracy": 0.10918098837137222,
      "num_tokens": 1065226.0,
      "step": 580
    },
    {
      "entropy": 7.479248285293579,
      "epoch": 0.051660190745319676,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000292,
      "loss": 7.1217,
      "mean_token_accuracy": 0.11170855611562729,
      "num_tokens": 1074661.0,
      "step": 585
    },
    {
      "entropy": 7.5761857509613035,
      "epoch": 0.052101730837160014,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002945,
      "loss": 7.2093,
      "mean_token_accuracy": 0.1055817686021328,
      "num_tokens": 1083921.0,
      "step": 590
    },
    {
      "entropy": 7.475902557373047,
      "epoch": 0.05254327092900035,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000297,
      "loss": 7.1024,
      "mean_token_accuracy": 0.12069160491228104,
      "num_tokens": 1093399.0,
      "step": 595
    },
    {
      "entropy": 7.477249240875244,
      "epoch": 0.05298481102084069,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0002995,
      "loss": 7.1806,
      "mean_token_accuracy": 0.1091334879398346,
      "num_tokens": 1104065.0,
      "step": 600
    },
    {
      "entropy": 7.454396152496338,
      "epoch": 0.05342635111268103,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000302,
      "loss": 7.0759,
      "mean_token_accuracy": 0.11466379538178444,
      "num_tokens": 1112995.0,
      "step": 605
    },
    {
      "entropy": 7.4687096118927006,
      "epoch": 0.05386789120452137,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003045,
      "loss": 7.1202,
      "mean_token_accuracy": 0.10947456136345864,
      "num_tokens": 1121637.0,
      "step": 610
    },
    {
      "entropy": 7.428070116043091,
      "epoch": 0.05430943129636171,
      "grad_norm": 1.140625,
      "learning_rate": 0.000307,
      "loss": 7.061,
      "mean_token_accuracy": 0.1107931189239025,
      "num_tokens": 1131166.0,
      "step": 615
    },
    {
      "entropy": 7.479091644287109,
      "epoch": 0.05475097138820205,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003095,
      "loss": 7.0598,
      "mean_token_accuracy": 0.11267657652497291,
      "num_tokens": 1140888.0,
      "step": 620
    },
    {
      "entropy": 7.516183996200562,
      "epoch": 0.05519251148004239,
      "grad_norm": 1.140625,
      "learning_rate": 0.000312,
      "loss": 7.2209,
      "mean_token_accuracy": 0.10797442942857742,
      "num_tokens": 1150278.0,
      "step": 625
    },
    {
      "entropy": 7.450537157058716,
      "epoch": 0.055634051571882726,
      "grad_norm": 1.359375,
      "learning_rate": 0.0003145,
      "loss": 7.1601,
      "mean_token_accuracy": 0.10592598244547843,
      "num_tokens": 1159348.0,
      "step": 630
    },
    {
      "entropy": 7.588924980163574,
      "epoch": 0.056075591663723064,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000317,
      "loss": 7.1788,
      "mean_token_accuracy": 0.10933139771223069,
      "num_tokens": 1168883.0,
      "step": 635
    },
    {
      "entropy": 7.51312780380249,
      "epoch": 0.0565171317555634,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003195,
      "loss": 7.0915,
      "mean_token_accuracy": 0.1159443736076355,
      "num_tokens": 1178572.0,
      "step": 640
    },
    {
      "entropy": 7.437539958953858,
      "epoch": 0.05695867184740375,
      "grad_norm": 1.0625,
      "learning_rate": 0.000322,
      "loss": 7.17,
      "mean_token_accuracy": 0.10953533723950386,
      "num_tokens": 1188909.0,
      "step": 645
    },
    {
      "entropy": 7.465111255645752,
      "epoch": 0.057400211939244085,
      "grad_norm": 1.234375,
      "learning_rate": 0.00032450000000000003,
      "loss": 7.0672,
      "mean_token_accuracy": 0.11020829901099205,
      "num_tokens": 1197705.0,
      "step": 650
    },
    {
      "entropy": 7.397136974334717,
      "epoch": 0.05784175203108442,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.9218,
      "mean_token_accuracy": 0.11375543102622032,
      "num_tokens": 1206351.0,
      "step": 655
    },
    {
      "entropy": 7.38561692237854,
      "epoch": 0.05828329212292476,
      "grad_norm": 1.203125,
      "learning_rate": 0.00032950000000000004,
      "loss": 7.0438,
      "mean_token_accuracy": 0.11089925691485406,
      "num_tokens": 1214984.0,
      "step": 660
    },
    {
      "entropy": 7.4726481437683105,
      "epoch": 0.0587248322147651,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033200000000000005,
      "loss": 7.1247,
      "mean_token_accuracy": 0.11059069335460663,
      "num_tokens": 1224485.0,
      "step": 665
    },
    {
      "entropy": 7.461330795288086,
      "epoch": 0.05916637230660544,
      "grad_norm": 1.140625,
      "learning_rate": 0.00033450000000000005,
      "loss": 7.1386,
      "mean_token_accuracy": 0.10253455564379692,
      "num_tokens": 1233560.0,
      "step": 670
    },
    {
      "entropy": 7.473884534835816,
      "epoch": 0.05960791239844578,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000337,
      "loss": 7.0729,
      "mean_token_accuracy": 0.10994852259755135,
      "num_tokens": 1242628.0,
      "step": 675
    },
    {
      "entropy": 7.392604112625122,
      "epoch": 0.06004945249028612,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0003395,
      "loss": 6.9809,
      "mean_token_accuracy": 0.1146723486483097,
      "num_tokens": 1251004.0,
      "step": 680
    },
    {
      "entropy": 7.428996276855469,
      "epoch": 0.06049099258212646,
      "grad_norm": 1.03125,
      "learning_rate": 0.000342,
      "loss": 7.1634,
      "mean_token_accuracy": 0.11115597635507583,
      "num_tokens": 1260344.0,
      "step": 685
    },
    {
      "entropy": 7.401163053512573,
      "epoch": 0.0609325326739668,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034449999999999997,
      "loss": 7.0696,
      "mean_token_accuracy": 0.11442951783537865,
      "num_tokens": 1269988.0,
      "step": 690
    },
    {
      "entropy": 7.402554750442505,
      "epoch": 0.061374072765807135,
      "grad_norm": 1.21875,
      "learning_rate": 0.000347,
      "loss": 7.0428,
      "mean_token_accuracy": 0.11391758620738983,
      "num_tokens": 1280912.0,
      "step": 695
    },
    {
      "entropy": 7.406301021575928,
      "epoch": 0.06181561285764747,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003495,
      "loss": 7.0467,
      "mean_token_accuracy": 0.11682257801294327,
      "num_tokens": 1289684.0,
      "step": 700
    },
    {
      "entropy": 7.3641218662261965,
      "epoch": 0.06225715294948781,
      "grad_norm": 1.265625,
      "learning_rate": 0.000352,
      "loss": 7.0803,
      "mean_token_accuracy": 0.10842933505773544,
      "num_tokens": 1298853.0,
      "step": 705
    },
    {
      "entropy": 7.3719359874725345,
      "epoch": 0.06269869304132815,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003545,
      "loss": 7.0386,
      "mean_token_accuracy": 0.11791565641760826,
      "num_tokens": 1309112.0,
      "step": 710
    },
    {
      "entropy": 7.369606351852417,
      "epoch": 0.0631402331331685,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000357,
      "loss": 7.0135,
      "mean_token_accuracy": 0.11014628857374191,
      "num_tokens": 1319064.0,
      "step": 715
    },
    {
      "entropy": 7.296480321884156,
      "epoch": 0.06358177322500883,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003595,
      "loss": 6.9938,
      "mean_token_accuracy": 0.11697245836257934,
      "num_tokens": 1327889.0,
      "step": 720
    },
    {
      "entropy": 7.403489398956299,
      "epoch": 0.06402331331684917,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000362,
      "loss": 7.0083,
      "mean_token_accuracy": 0.1104284293949604,
      "num_tokens": 1337241.0,
      "step": 725
    },
    {
      "entropy": 7.347339487075805,
      "epoch": 0.06446485340868952,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003645,
      "loss": 7.0718,
      "mean_token_accuracy": 0.11292454376816749,
      "num_tokens": 1346527.0,
      "step": 730
    },
    {
      "entropy": 7.347748756408691,
      "epoch": 0.06490639350052985,
      "grad_norm": 1.09375,
      "learning_rate": 0.000367,
      "loss": 7.0639,
      "mean_token_accuracy": 0.11358218491077424,
      "num_tokens": 1355677.0,
      "step": 735
    },
    {
      "entropy": 7.3271379470825195,
      "epoch": 0.06534793359237019,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003695,
      "loss": 7.0468,
      "mean_token_accuracy": 0.1157820999622345,
      "num_tokens": 1364874.0,
      "step": 740
    },
    {
      "entropy": 7.400119829177856,
      "epoch": 0.06578947368421052,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000372,
      "loss": 6.9839,
      "mean_token_accuracy": 0.12082682698965072,
      "num_tokens": 1373717.0,
      "step": 745
    },
    {
      "entropy": 7.417098903656006,
      "epoch": 0.06623101377605087,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003745,
      "loss": 6.9526,
      "mean_token_accuracy": 0.11417729035019875,
      "num_tokens": 1382767.0,
      "step": 750
    },
    {
      "entropy": 7.194312381744385,
      "epoch": 0.0666725538678912,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000377,
      "loss": 6.9799,
      "mean_token_accuracy": 0.11586436182260514,
      "num_tokens": 1391190.0,
      "step": 755
    },
    {
      "entropy": 7.339599084854126,
      "epoch": 0.06711409395973154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003795,
      "loss": 6.9817,
      "mean_token_accuracy": 0.1170587882399559,
      "num_tokens": 1400722.0,
      "step": 760
    },
    {
      "entropy": 7.400132036209106,
      "epoch": 0.06755563405157189,
      "grad_norm": 1.140625,
      "learning_rate": 0.000382,
      "loss": 7.0091,
      "mean_token_accuracy": 0.10940207540988922,
      "num_tokens": 1409328.0,
      "step": 765
    },
    {
      "entropy": 7.115442705154419,
      "epoch": 0.06799717414341222,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003845,
      "loss": 6.8837,
      "mean_token_accuracy": 0.12403405010700226,
      "num_tokens": 1418504.0,
      "step": 770
    },
    {
      "entropy": 7.437279510498047,
      "epoch": 0.06843871423525257,
      "grad_norm": 1.15625,
      "learning_rate": 0.00038700000000000003,
      "loss": 7.0359,
      "mean_token_accuracy": 0.10912163555622101,
      "num_tokens": 1427690.0,
      "step": 775
    },
    {
      "entropy": 7.26606240272522,
      "epoch": 0.0688802543270929,
      "grad_norm": 1.140625,
      "learning_rate": 0.00038950000000000003,
      "loss": 7.0132,
      "mean_token_accuracy": 0.11617867797613143,
      "num_tokens": 1436798.0,
      "step": 780
    },
    {
      "entropy": 7.26966495513916,
      "epoch": 0.06932179441893324,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039200000000000004,
      "loss": 7.01,
      "mean_token_accuracy": 0.11528639197349548,
      "num_tokens": 1446357.0,
      "step": 785
    },
    {
      "entropy": 7.302881193161011,
      "epoch": 0.06976333451077357,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.9315,
      "mean_token_accuracy": 0.11601672917604447,
      "num_tokens": 1455998.0,
      "step": 790
    },
    {
      "entropy": 7.307020282745361,
      "epoch": 0.07020487460261392,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.9813,
      "mean_token_accuracy": 0.11031874865293503,
      "num_tokens": 1465237.0,
      "step": 795
    },
    {
      "entropy": 7.243889856338501,
      "epoch": 0.07064641469445426,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003995,
      "loss": 6.9248,
      "mean_token_accuracy": 0.12101015448570251,
      "num_tokens": 1474363.0,
      "step": 800
    },
    {
      "entropy": 7.248915958404541,
      "epoch": 0.0710879547862946,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000402,
      "loss": 6.9906,
      "mean_token_accuracy": 0.12257542759180069,
      "num_tokens": 1483379.0,
      "step": 805
    },
    {
      "entropy": 7.34154782295227,
      "epoch": 0.07152949487813494,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004045,
      "loss": 6.9828,
      "mean_token_accuracy": 0.11638684496283531,
      "num_tokens": 1492507.0,
      "step": 810
    },
    {
      "entropy": 7.212753200531006,
      "epoch": 0.07197103496997527,
      "grad_norm": 1.25,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.8604,
      "mean_token_accuracy": 0.12233690097928047,
      "num_tokens": 1500888.0,
      "step": 815
    },
    {
      "entropy": 7.181662464141846,
      "epoch": 0.07241257506181561,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004095,
      "loss": 6.9234,
      "mean_token_accuracy": 0.11673298925161361,
      "num_tokens": 1510310.0,
      "step": 820
    },
    {
      "entropy": 7.263974905014038,
      "epoch": 0.07285411515365595,
      "grad_norm": 1.171875,
      "learning_rate": 0.000412,
      "loss": 6.8977,
      "mean_token_accuracy": 0.12379544451832772,
      "num_tokens": 1519427.0,
      "step": 825
    },
    {
      "entropy": 7.25563235282898,
      "epoch": 0.07329565524549629,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004145,
      "loss": 6.9332,
      "mean_token_accuracy": 0.11141304299235344,
      "num_tokens": 1529456.0,
      "step": 830
    },
    {
      "entropy": 7.141549444198608,
      "epoch": 0.07373719533733664,
      "grad_norm": 1.15625,
      "learning_rate": 0.000417,
      "loss": 6.8617,
      "mean_token_accuracy": 0.11971206367015838,
      "num_tokens": 1537695.0,
      "step": 835
    },
    {
      "entropy": 7.242198467254639,
      "epoch": 0.07417873542917697,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004195,
      "loss": 7.0075,
      "mean_token_accuracy": 0.11687812730669975,
      "num_tokens": 1547511.0,
      "step": 840
    },
    {
      "entropy": 7.309858417510986,
      "epoch": 0.07462027552101731,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000422,
      "loss": 6.9839,
      "mean_token_accuracy": 0.11879770755767823,
      "num_tokens": 1557035.0,
      "step": 845
    },
    {
      "entropy": 7.211091947555542,
      "epoch": 0.07506181561285764,
      "grad_norm": 1.0,
      "learning_rate": 0.0004245,
      "loss": 6.9346,
      "mean_token_accuracy": 0.12483425140380859,
      "num_tokens": 1566773.0,
      "step": 850
    },
    {
      "entropy": 7.208451414108277,
      "epoch": 0.07550335570469799,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000427,
      "loss": 6.927,
      "mean_token_accuracy": 0.11267467886209488,
      "num_tokens": 1576873.0,
      "step": 855
    },
    {
      "entropy": 7.186734676361084,
      "epoch": 0.07594489579653832,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004295,
      "loss": 6.9684,
      "mean_token_accuracy": 0.11325102299451828,
      "num_tokens": 1586170.0,
      "step": 860
    },
    {
      "entropy": 7.32284631729126,
      "epoch": 0.07638643588837866,
      "grad_norm": 1.328125,
      "learning_rate": 0.000432,
      "loss": 7.0153,
      "mean_token_accuracy": 0.11636607497930526,
      "num_tokens": 1596054.0,
      "step": 865
    },
    {
      "entropy": 7.200122261047364,
      "epoch": 0.07682797598021901,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004345,
      "loss": 6.8751,
      "mean_token_accuracy": 0.1230311393737793,
      "num_tokens": 1604544.0,
      "step": 870
    },
    {
      "entropy": 7.244127464294434,
      "epoch": 0.07726951607205934,
      "grad_norm": 1.171875,
      "learning_rate": 0.000437,
      "loss": 6.9023,
      "mean_token_accuracy": 0.12475202232599258,
      "num_tokens": 1614580.0,
      "step": 875
    },
    {
      "entropy": 7.1907594203948975,
      "epoch": 0.07771105616389969,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004395,
      "loss": 6.9948,
      "mean_token_accuracy": 0.115571129322052,
      "num_tokens": 1624701.0,
      "step": 880
    },
    {
      "entropy": 7.136907148361206,
      "epoch": 0.07815259625574002,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000442,
      "loss": 6.9193,
      "mean_token_accuracy": 0.11373112797737121,
      "num_tokens": 1634085.0,
      "step": 885
    },
    {
      "entropy": 7.195565986633301,
      "epoch": 0.07859413634758036,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004445,
      "loss": 6.8968,
      "mean_token_accuracy": 0.11859833672642708,
      "num_tokens": 1643190.0,
      "step": 890
    },
    {
      "entropy": 7.176092052459717,
      "epoch": 0.0790356764394207,
      "grad_norm": 1.03125,
      "learning_rate": 0.000447,
      "loss": 6.9171,
      "mean_token_accuracy": 0.11816032081842423,
      "num_tokens": 1652705.0,
      "step": 895
    },
    {
      "entropy": 7.225932455062866,
      "epoch": 0.07947721653126104,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.9033,
      "mean_token_accuracy": 0.11247398778796196,
      "num_tokens": 1662210.0,
      "step": 900
    },
    {
      "entropy": 7.222397232055664,
      "epoch": 0.07991875662310138,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.931,
      "mean_token_accuracy": 0.11510396897792816,
      "num_tokens": 1671893.0,
      "step": 905
    },
    {
      "entropy": 7.168801355361938,
      "epoch": 0.08036029671494171,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.8842,
      "mean_token_accuracy": 0.12047431021928787,
      "num_tokens": 1681217.0,
      "step": 910
    },
    {
      "entropy": 7.189701175689697,
      "epoch": 0.08080183680678206,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.9145,
      "mean_token_accuracy": 0.11493082717061043,
      "num_tokens": 1690447.0,
      "step": 915
    },
    {
      "entropy": 7.148307180404663,
      "epoch": 0.08124337689862239,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045950000000000006,
      "loss": 7.001,
      "mean_token_accuracy": 0.10863983184099198,
      "num_tokens": 1700355.0,
      "step": 920
    },
    {
      "entropy": 7.2164154052734375,
      "epoch": 0.08168491699046274,
      "grad_norm": 1.015625,
      "learning_rate": 0.000462,
      "loss": 6.906,
      "mean_token_accuracy": 0.11343243196606637,
      "num_tokens": 1709449.0,
      "step": 925
    },
    {
      "entropy": 7.191054010391236,
      "epoch": 0.08212645708230307,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004645,
      "loss": 6.9311,
      "mean_token_accuracy": 0.11408023163676262,
      "num_tokens": 1718838.0,
      "step": 930
    },
    {
      "entropy": 7.111620044708252,
      "epoch": 0.08256799717414341,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000467,
      "loss": 6.8844,
      "mean_token_accuracy": 0.12342581376433373,
      "num_tokens": 1728594.0,
      "step": 935
    },
    {
      "entropy": 7.233243608474732,
      "epoch": 0.08300953726598376,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004695,
      "loss": 6.9305,
      "mean_token_accuracy": 0.12092314288020134,
      "num_tokens": 1738814.0,
      "step": 940
    },
    {
      "entropy": 7.168387937545776,
      "epoch": 0.08345107735782409,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000472,
      "loss": 6.8546,
      "mean_token_accuracy": 0.12510684356093407,
      "num_tokens": 1747644.0,
      "step": 945
    },
    {
      "entropy": 7.201950550079346,
      "epoch": 0.08389261744966443,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004745,
      "loss": 6.9079,
      "mean_token_accuracy": 0.11906211525201797,
      "num_tokens": 1757489.0,
      "step": 950
    },
    {
      "entropy": 7.130580997467041,
      "epoch": 0.08433415754150476,
      "grad_norm": 1.03125,
      "learning_rate": 0.000477,
      "loss": 6.8296,
      "mean_token_accuracy": 0.12410564720630646,
      "num_tokens": 1767546.0,
      "step": 955
    },
    {
      "entropy": 7.093601322174072,
      "epoch": 0.08477569763334511,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004795,
      "loss": 6.8928,
      "mean_token_accuracy": 0.10743151605129242,
      "num_tokens": 1776035.0,
      "step": 960
    },
    {
      "entropy": 7.145628499984741,
      "epoch": 0.08521723772518544,
      "grad_norm": 0.953125,
      "learning_rate": 0.000482,
      "loss": 6.953,
      "mean_token_accuracy": 0.11847842559218406,
      "num_tokens": 1786161.0,
      "step": 965
    },
    {
      "entropy": 7.274262189865112,
      "epoch": 0.08565877781702579,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004845,
      "loss": 7.007,
      "mean_token_accuracy": 0.11388442665338516,
      "num_tokens": 1796093.0,
      "step": 970
    },
    {
      "entropy": 7.080674505233764,
      "epoch": 0.08610031790886613,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000487,
      "loss": 6.8546,
      "mean_token_accuracy": 0.11734381765127182,
      "num_tokens": 1805574.0,
      "step": 975
    },
    {
      "entropy": 7.141940927505493,
      "epoch": 0.08654185800070646,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004895,
      "loss": 6.8924,
      "mean_token_accuracy": 0.11525135561823845,
      "num_tokens": 1815175.0,
      "step": 980
    },
    {
      "entropy": 7.089342260360718,
      "epoch": 0.08698339809254681,
      "grad_norm": 1.03125,
      "learning_rate": 0.000492,
      "loss": 6.8082,
      "mean_token_accuracy": 0.12424175813794136,
      "num_tokens": 1824683.0,
      "step": 985
    },
    {
      "entropy": 7.1276007175445555,
      "epoch": 0.08742493818438714,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004945,
      "loss": 6.7918,
      "mean_token_accuracy": 0.1199628159403801,
      "num_tokens": 1833174.0,
      "step": 990
    },
    {
      "entropy": 7.066866636276245,
      "epoch": 0.08786647827622748,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000497,
      "loss": 6.8796,
      "mean_token_accuracy": 0.11747556179761887,
      "num_tokens": 1842409.0,
      "step": 995
    },
    {
      "entropy": 7.007741022109985,
      "epoch": 0.08830801836806781,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995,
      "loss": 6.7861,
      "mean_token_accuracy": 0.12650212571024894,
      "num_tokens": 1851441.0,
      "step": 1000
    },
    {
      "entropy": 7.240553998947144,
      "epoch": 0.08874955845990816,
      "grad_norm": 1.140625,
      "learning_rate": 0.000499999998589561,
      "loss": 6.9076,
      "mean_token_accuracy": 0.11784921512007714,
      "num_tokens": 1861188.0,
      "step": 1005
    },
    {
      "entropy": 7.046397066116333,
      "epoch": 0.0891910985517485,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999999928596523,
      "loss": 6.8583,
      "mean_token_accuracy": 0.12223245799541474,
      "num_tokens": 1870284.0,
      "step": 1010
    },
    {
      "entropy": 7.0789531707763675,
      "epoch": 0.08963263864358884,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999999827221219,
      "loss": 6.8283,
      "mean_token_accuracy": 0.12814989313483238,
      "num_tokens": 1879744.0,
      "step": 1015
    },
    {
      "entropy": 7.0941243171691895,
      "epoch": 0.09007417873542918,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999999681769696,
      "loss": 6.8228,
      "mean_token_accuracy": 0.12476407438516617,
      "num_tokens": 1889241.0,
      "step": 1020
    },
    {
      "entropy": 7.043097877502442,
      "epoch": 0.09051571882726951,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000499999949224196,
      "loss": 6.8413,
      "mean_token_accuracy": 0.12211177572607994,
      "num_tokens": 1898247.0,
      "step": 1025
    },
    {
      "entropy": 7.068183851242066,
      "epoch": 0.09095725891910986,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999999258638013,
      "loss": 6.7999,
      "mean_token_accuracy": 0.122597586363554,
      "num_tokens": 1907559.0,
      "step": 1030
    },
    {
      "entropy": 7.042861270904541,
      "epoch": 0.09139879901095019,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999998980957861,
      "loss": 6.8116,
      "mean_token_accuracy": 0.12296328842639923,
      "num_tokens": 1917072.0,
      "step": 1035
    },
    {
      "entropy": 7.02711238861084,
      "epoch": 0.09184033910279053,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999998659201508,
      "loss": 6.7577,
      "mean_token_accuracy": 0.1265648201107979,
      "num_tokens": 1926597.0,
      "step": 1040
    },
    {
      "entropy": 7.008166885375976,
      "epoch": 0.09228187919463088,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999998293368961,
      "loss": 6.8148,
      "mean_token_accuracy": 0.11956329345703125,
      "num_tokens": 1935978.0,
      "step": 1045
    },
    {
      "entropy": 7.03216814994812,
      "epoch": 0.09272341928647121,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999997883460227,
      "loss": 6.7551,
      "mean_token_accuracy": 0.12221006751060486,
      "num_tokens": 1944424.0,
      "step": 1050
    },
    {
      "entropy": 7.016500616073609,
      "epoch": 0.09316495937831155,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999997429475314,
      "loss": 6.7199,
      "mean_token_accuracy": 0.1299702636897564,
      "num_tokens": 1953844.0,
      "step": 1055
    },
    {
      "entropy": 7.052721452713013,
      "epoch": 0.09360649947015189,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999996931414232,
      "loss": 6.8195,
      "mean_token_accuracy": 0.12436865568161011,
      "num_tokens": 1963974.0,
      "step": 1060
    },
    {
      "entropy": 7.030517196655273,
      "epoch": 0.09404803956199223,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999996389276988,
      "loss": 6.7732,
      "mean_token_accuracy": 0.13136502131819724,
      "num_tokens": 1973466.0,
      "step": 1065
    },
    {
      "entropy": 6.936943769454956,
      "epoch": 0.09448957965383256,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999995803063596,
      "loss": 6.7879,
      "mean_token_accuracy": 0.1239612028002739,
      "num_tokens": 1983478.0,
      "step": 1070
    },
    {
      "entropy": 7.092060518264771,
      "epoch": 0.0949311197456729,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999995172774065,
      "loss": 6.8703,
      "mean_token_accuracy": 0.12069909051060676,
      "num_tokens": 1992775.0,
      "step": 1075
    },
    {
      "entropy": 7.053188705444336,
      "epoch": 0.09537265983751325,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999994498408408,
      "loss": 6.8269,
      "mean_token_accuracy": 0.11928088068962098,
      "num_tokens": 2002526.0,
      "step": 1080
    },
    {
      "entropy": 7.058347511291504,
      "epoch": 0.09581419992935358,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999993779966639,
      "loss": 6.8484,
      "mean_token_accuracy": 0.11959519013762474,
      "num_tokens": 2012476.0,
      "step": 1085
    },
    {
      "entropy": 6.85009355545044,
      "epoch": 0.09625574002119393,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004999993017448771,
      "loss": 6.6854,
      "mean_token_accuracy": 0.1360420987010002,
      "num_tokens": 2021252.0,
      "step": 1090
    },
    {
      "entropy": 7.145733261108399,
      "epoch": 0.09669728011303426,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999992210854821,
      "loss": 6.8438,
      "mean_token_accuracy": 0.1258312664926052,
      "num_tokens": 2031438.0,
      "step": 1095
    },
    {
      "entropy": 7.019267272949219,
      "epoch": 0.0971388202048746,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999991360184801,
      "loss": 6.883,
      "mean_token_accuracy": 0.12097008153796196,
      "num_tokens": 2041319.0,
      "step": 1100
    },
    {
      "entropy": 7.073525238037109,
      "epoch": 0.09758036029671494,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999990465438731,
      "loss": 6.862,
      "mean_token_accuracy": 0.1219029612839222,
      "num_tokens": 2052060.0,
      "step": 1105
    },
    {
      "entropy": 6.95486044883728,
      "epoch": 0.09802190038855528,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999989526616628,
      "loss": 6.7495,
      "mean_token_accuracy": 0.12949764505028724,
      "num_tokens": 2061331.0,
      "step": 1110
    },
    {
      "entropy": 6.9770708084106445,
      "epoch": 0.09846344048039563,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999988543718509,
      "loss": 6.7566,
      "mean_token_accuracy": 0.12212359756231309,
      "num_tokens": 2070006.0,
      "step": 1115
    },
    {
      "entropy": 7.0353333950042725,
      "epoch": 0.09890498057223596,
      "grad_norm": 1.125,
      "learning_rate": 0.0004999987516744394,
      "loss": 6.7158,
      "mean_token_accuracy": 0.13030195459723473,
      "num_tokens": 2079089.0,
      "step": 1120
    },
    {
      "entropy": 6.971569681167603,
      "epoch": 0.0993465206640763,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999986445694303,
      "loss": 6.7236,
      "mean_token_accuracy": 0.12542433887720109,
      "num_tokens": 2087237.0,
      "step": 1125
    },
    {
      "entropy": 6.899944829940796,
      "epoch": 0.09978806075591663,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999985330568258,
      "loss": 6.6822,
      "mean_token_accuracy": 0.13330597206950187,
      "num_tokens": 2096598.0,
      "step": 1130
    },
    {
      "entropy": 7.01187915802002,
      "epoch": 0.10022960084775698,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999984171366278,
      "loss": 6.7788,
      "mean_token_accuracy": 0.11807065233588218,
      "num_tokens": 2106106.0,
      "step": 1135
    },
    {
      "entropy": 6.9801586151123045,
      "epoch": 0.10067114093959731,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499998296808839,
      "loss": 6.7582,
      "mean_token_accuracy": 0.11982692703604698,
      "num_tokens": 2115478.0,
      "step": 1140
    },
    {
      "entropy": 7.019436502456665,
      "epoch": 0.10111268103143765,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004999981720734615,
      "loss": 6.7836,
      "mean_token_accuracy": 0.12526557743549346,
      "num_tokens": 2124439.0,
      "step": 1145
    },
    {
      "entropy": 7.01093807220459,
      "epoch": 0.101554221123278,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004999980429304977,
      "loss": 6.796,
      "mean_token_accuracy": 0.11945406049489975,
      "num_tokens": 2133718.0,
      "step": 1150
    },
    {
      "entropy": 6.833093214035034,
      "epoch": 0.10199576121511833,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999979093799502,
      "loss": 6.6939,
      "mean_token_accuracy": 0.12498606964945794,
      "num_tokens": 2142978.0,
      "step": 1155
    },
    {
      "entropy": 6.946839380264282,
      "epoch": 0.10243730130695868,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999977714218217,
      "loss": 6.7688,
      "mean_token_accuracy": 0.12323360666632652,
      "num_tokens": 2152250.0,
      "step": 1160
    },
    {
      "entropy": 6.99408917427063,
      "epoch": 0.102878841398799,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999976290561147,
      "loss": 6.7613,
      "mean_token_accuracy": 0.12313725724816323,
      "num_tokens": 2161620.0,
      "step": 1165
    },
    {
      "entropy": 6.988214206695557,
      "epoch": 0.10332038149063935,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999974822828322,
      "loss": 6.7814,
      "mean_token_accuracy": 0.12046999335289002,
      "num_tokens": 2170856.0,
      "step": 1170
    },
    {
      "entropy": 6.9620466232299805,
      "epoch": 0.10376192158247968,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499997331101977,
      "loss": 6.7547,
      "mean_token_accuracy": 0.12075523287057877,
      "num_tokens": 2180926.0,
      "step": 1175
    },
    {
      "entropy": 6.9654961109161375,
      "epoch": 0.10420346167432003,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499997175513552,
      "loss": 6.7462,
      "mean_token_accuracy": 0.12285592630505562,
      "num_tokens": 2190248.0,
      "step": 1180
    },
    {
      "entropy": 6.9309080123901365,
      "epoch": 0.10464500176616037,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999970155175603,
      "loss": 6.7514,
      "mean_token_accuracy": 0.12348200678825379,
      "num_tokens": 2199833.0,
      "step": 1185
    },
    {
      "entropy": 6.959126377105713,
      "epoch": 0.1050865418580007,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499996851114005,
      "loss": 6.7184,
      "mean_token_accuracy": 0.12861018776893615,
      "num_tokens": 2208240.0,
      "step": 1190
    },
    {
      "entropy": 6.932129383087158,
      "epoch": 0.10552808194984105,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999966823028894,
      "loss": 6.7618,
      "mean_token_accuracy": 0.11978181153535843,
      "num_tokens": 2218758.0,
      "step": 1195
    },
    {
      "entropy": 6.9820332527160645,
      "epoch": 0.10596962204168138,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999965090842168,
      "loss": 6.7626,
      "mean_token_accuracy": 0.1300593949854374,
      "num_tokens": 2228443.0,
      "step": 1200
    },
    {
      "entropy": 6.858966636657715,
      "epoch": 0.10641116213352173,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999963314579905,
      "loss": 6.6107,
      "mean_token_accuracy": 0.1355985902249813,
      "num_tokens": 2236787.0,
      "step": 1205
    },
    {
      "entropy": 6.947892761230468,
      "epoch": 0.10685270222536206,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999961494242139,
      "loss": 6.7746,
      "mean_token_accuracy": 0.12228678241372108,
      "num_tokens": 2247089.0,
      "step": 1210
    },
    {
      "entropy": 6.849541854858399,
      "epoch": 0.1072942423172024,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999959629828908,
      "loss": 6.7035,
      "mean_token_accuracy": 0.12353255078196526,
      "num_tokens": 2256045.0,
      "step": 1215
    },
    {
      "entropy": 7.002861881256104,
      "epoch": 0.10773578240904275,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999957721340248,
      "loss": 6.7841,
      "mean_token_accuracy": 0.12372039780020713,
      "num_tokens": 2265835.0,
      "step": 1220
    },
    {
      "entropy": 6.845533561706543,
      "epoch": 0.10817732250088308,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999955768776195,
      "loss": 6.7134,
      "mean_token_accuracy": 0.12301751002669334,
      "num_tokens": 2275318.0,
      "step": 1225
    },
    {
      "entropy": 6.958423137664795,
      "epoch": 0.10861886259272342,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999953772136788,
      "loss": 6.7049,
      "mean_token_accuracy": 0.12736163064837455,
      "num_tokens": 2284821.0,
      "step": 1230
    },
    {
      "entropy": 6.942495965957642,
      "epoch": 0.10906040268456375,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999951731422068,
      "loss": 6.6712,
      "mean_token_accuracy": 0.12646759897470475,
      "num_tokens": 2294013.0,
      "step": 1235
    },
    {
      "entropy": 6.917871713638306,
      "epoch": 0.1095019427764041,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999949646632072,
      "loss": 6.6694,
      "mean_token_accuracy": 0.12900351211428643,
      "num_tokens": 2302727.0,
      "step": 1240
    },
    {
      "entropy": 6.849976634979248,
      "epoch": 0.10994348286824443,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999947517766842,
      "loss": 6.6906,
      "mean_token_accuracy": 0.12817103341221808,
      "num_tokens": 2312032.0,
      "step": 1245
    },
    {
      "entropy": 7.057008552551269,
      "epoch": 0.11038502296008477,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000499994534482642,
      "loss": 6.7819,
      "mean_token_accuracy": 0.12531876042485238,
      "num_tokens": 2321839.0,
      "step": 1250
    },
    {
      "entropy": 6.899663734436035,
      "epoch": 0.11082656305192512,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999943127810847,
      "loss": 6.7555,
      "mean_token_accuracy": 0.1212481401860714,
      "num_tokens": 2331255.0,
      "step": 1255
    },
    {
      "entropy": 6.816301393508911,
      "epoch": 0.11126810314376545,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999940866720169,
      "loss": 6.5529,
      "mean_token_accuracy": 0.13424451053142547,
      "num_tokens": 2340038.0,
      "step": 1260
    },
    {
      "entropy": 6.8964927196502686,
      "epoch": 0.1117096432356058,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999938561554429,
      "loss": 6.6959,
      "mean_token_accuracy": 0.12449696063995361,
      "num_tokens": 2348901.0,
      "step": 1265
    },
    {
      "entropy": 6.875371694564819,
      "epoch": 0.11215118332744613,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999936212313672,
      "loss": 6.7697,
      "mean_token_accuracy": 0.119696693867445,
      "num_tokens": 2358842.0,
      "step": 1270
    },
    {
      "entropy": 6.907081985473633,
      "epoch": 0.11259272341928647,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999933818997943,
      "loss": 6.6654,
      "mean_token_accuracy": 0.12900050207972527,
      "num_tokens": 2368650.0,
      "step": 1275
    },
    {
      "entropy": 6.866778802871704,
      "epoch": 0.1130342635111268,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999931381607292,
      "loss": 6.7041,
      "mean_token_accuracy": 0.1312815934419632,
      "num_tokens": 2377916.0,
      "step": 1280
    },
    {
      "entropy": 6.936831331253051,
      "epoch": 0.11347580360296715,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999928900141764,
      "loss": 6.6683,
      "mean_token_accuracy": 0.12833984717726707,
      "num_tokens": 2387507.0,
      "step": 1285
    },
    {
      "entropy": 6.833027648925781,
      "epoch": 0.1139173436948075,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499992637460141,
      "loss": 6.7166,
      "mean_token_accuracy": 0.12722958698868753,
      "num_tokens": 2396148.0,
      "step": 1290
    },
    {
      "entropy": 6.854029512405395,
      "epoch": 0.11435888378664782,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999923804986275,
      "loss": 6.5844,
      "mean_token_accuracy": 0.12824990823864937,
      "num_tokens": 2404891.0,
      "step": 1295
    },
    {
      "entropy": 6.922542858123779,
      "epoch": 0.11480042387848817,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999921191296415,
      "loss": 6.6288,
      "mean_token_accuracy": 0.12811262235045434,
      "num_tokens": 2414406.0,
      "step": 1300
    },
    {
      "entropy": 6.860514211654663,
      "epoch": 0.1152419639703285,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999918533531877,
      "loss": 6.7126,
      "mean_token_accuracy": 0.12714724242687225,
      "num_tokens": 2424363.0,
      "step": 1305
    },
    {
      "entropy": 6.917534399032593,
      "epoch": 0.11568350406216885,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999915831692714,
      "loss": 6.6534,
      "mean_token_accuracy": 0.12873802334070206,
      "num_tokens": 2433753.0,
      "step": 1310
    },
    {
      "entropy": 6.790404272079468,
      "epoch": 0.11612504415400918,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999913085778981,
      "loss": 6.6769,
      "mean_token_accuracy": 0.12445072457194328,
      "num_tokens": 2443275.0,
      "step": 1315
    },
    {
      "entropy": 6.909478187561035,
      "epoch": 0.11656658424584952,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999910295790729,
      "loss": 6.6931,
      "mean_token_accuracy": 0.12268302440643311,
      "num_tokens": 2452510.0,
      "step": 1320
    },
    {
      "entropy": 6.915535879135132,
      "epoch": 0.11700812433768987,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999907461728014,
      "loss": 6.7807,
      "mean_token_accuracy": 0.1234104610979557,
      "num_tokens": 2462742.0,
      "step": 1325
    },
    {
      "entropy": 6.835855340957641,
      "epoch": 0.1174496644295302,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999904583590893,
      "loss": 6.6549,
      "mean_token_accuracy": 0.12718312442302704,
      "num_tokens": 2471409.0,
      "step": 1330
    },
    {
      "entropy": 6.876492547988891,
      "epoch": 0.11789120452137054,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999901661379418,
      "loss": 6.6053,
      "mean_token_accuracy": 0.1328142687678337,
      "num_tokens": 2481011.0,
      "step": 1335
    },
    {
      "entropy": 6.79636025428772,
      "epoch": 0.11833274461321087,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999898695093652,
      "loss": 6.6995,
      "mean_token_accuracy": 0.12458944991230965,
      "num_tokens": 2490664.0,
      "step": 1340
    },
    {
      "entropy": 6.888507127761841,
      "epoch": 0.11877428470505122,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999895684733648,
      "loss": 6.6235,
      "mean_token_accuracy": 0.13585169985890388,
      "num_tokens": 2499799.0,
      "step": 1345
    },
    {
      "entropy": 6.845640707015991,
      "epoch": 0.11921582479689156,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004999892630299467,
      "loss": 6.6025,
      "mean_token_accuracy": 0.12698834463953973,
      "num_tokens": 2508780.0,
      "step": 1350
    },
    {
      "entropy": 6.898168849945068,
      "epoch": 0.1196573648887319,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999889531791171,
      "loss": 6.6341,
      "mean_token_accuracy": 0.12324970662593841,
      "num_tokens": 2517741.0,
      "step": 1355
    },
    {
      "entropy": 6.750218534469605,
      "epoch": 0.12009890498057224,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999886389208817,
      "loss": 6.7105,
      "mean_token_accuracy": 0.12327531054615974,
      "num_tokens": 2528742.0,
      "step": 1360
    },
    {
      "entropy": 6.880941247940063,
      "epoch": 0.12054044507241257,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999883202552468,
      "loss": 6.6444,
      "mean_token_accuracy": 0.1296772688627243,
      "num_tokens": 2538609.0,
      "step": 1365
    },
    {
      "entropy": 6.886623859405518,
      "epoch": 0.12098198516425292,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999879971822189,
      "loss": 6.6274,
      "mean_token_accuracy": 0.12891870513558387,
      "num_tokens": 2547772.0,
      "step": 1370
    },
    {
      "entropy": 6.6913806915283205,
      "epoch": 0.12142352525609325,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999876697018038,
      "loss": 6.6008,
      "mean_token_accuracy": 0.13324561789631845,
      "num_tokens": 2556114.0,
      "step": 1375
    },
    {
      "entropy": 6.90076265335083,
      "epoch": 0.1218650653479336,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999873378140085,
      "loss": 6.6339,
      "mean_token_accuracy": 0.12988200038671494,
      "num_tokens": 2566814.0,
      "step": 1380
    },
    {
      "entropy": 6.784169101715088,
      "epoch": 0.12230660543977394,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004999870015188389,
      "loss": 6.6031,
      "mean_token_accuracy": 0.13186766728758811,
      "num_tokens": 2576030.0,
      "step": 1385
    },
    {
      "entropy": 6.762492084503174,
      "epoch": 0.12274814553161427,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999866608163021,
      "loss": 6.6231,
      "mean_token_accuracy": 0.13228361681103706,
      "num_tokens": 2585756.0,
      "step": 1390
    },
    {
      "entropy": 6.82820086479187,
      "epoch": 0.12318968562345461,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999863157064045,
      "loss": 6.6779,
      "mean_token_accuracy": 0.1298144727945328,
      "num_tokens": 2595676.0,
      "step": 1395
    },
    {
      "entropy": 6.848227310180664,
      "epoch": 0.12363122571529495,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999859661891529,
      "loss": 6.6748,
      "mean_token_accuracy": 0.1304032675921917,
      "num_tokens": 2606197.0,
      "step": 1400
    },
    {
      "entropy": 6.879591560363769,
      "epoch": 0.12407276580713529,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999856122645543,
      "loss": 6.5967,
      "mean_token_accuracy": 0.12973913177847862,
      "num_tokens": 2615311.0,
      "step": 1405
    },
    {
      "entropy": 6.734095478057862,
      "epoch": 0.12451430589897562,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004999852539326154,
      "loss": 6.5243,
      "mean_token_accuracy": 0.1325017362833023,
      "num_tokens": 2624074.0,
      "step": 1410
    },
    {
      "entropy": 6.758558845520019,
      "epoch": 0.12495584599081597,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999848911933434,
      "loss": 6.6522,
      "mean_token_accuracy": 0.13497036173939705,
      "num_tokens": 2633877.0,
      "step": 1415
    },
    {
      "entropy": 6.745906019210816,
      "epoch": 0.1253973860826563,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999845240467453,
      "loss": 6.575,
      "mean_token_accuracy": 0.12966776564717292,
      "num_tokens": 2643330.0,
      "step": 1420
    },
    {
      "entropy": 6.863449001312256,
      "epoch": 0.12583892617449666,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999841524928282,
      "loss": 6.5664,
      "mean_token_accuracy": 0.13387957438826562,
      "num_tokens": 2652070.0,
      "step": 1425
    },
    {
      "entropy": 6.756294584274292,
      "epoch": 0.126280466266337,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004999837765315997,
      "loss": 6.6242,
      "mean_token_accuracy": 0.12831888422369958,
      "num_tokens": 2660546.0,
      "step": 1430
    },
    {
      "entropy": 6.676768016815186,
      "epoch": 0.12672200635817732,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999833961630669,
      "loss": 6.507,
      "mean_token_accuracy": 0.13364252746105193,
      "num_tokens": 2669938.0,
      "step": 1435
    },
    {
      "entropy": 6.824239492416382,
      "epoch": 0.12716354645001765,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999830113872374,
      "loss": 6.637,
      "mean_token_accuracy": 0.13026153072714805,
      "num_tokens": 2679814.0,
      "step": 1440
    },
    {
      "entropy": 6.675518751144409,
      "epoch": 0.127605086541858,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999826222041186,
      "loss": 6.5447,
      "mean_token_accuracy": 0.13804391250014306,
      "num_tokens": 2688733.0,
      "step": 1445
    },
    {
      "entropy": 6.745492744445801,
      "epoch": 0.12804662663369834,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999822286137182,
      "loss": 6.5346,
      "mean_token_accuracy": 0.13509366363286973,
      "num_tokens": 2697744.0,
      "step": 1450
    },
    {
      "entropy": 6.658346796035767,
      "epoch": 0.12848816672553867,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999818306160439,
      "loss": 6.4971,
      "mean_token_accuracy": 0.13757269233465194,
      "num_tokens": 2707037.0,
      "step": 1455
    },
    {
      "entropy": 6.859276628494262,
      "epoch": 0.12892970681737903,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999814282111034,
      "loss": 6.6552,
      "mean_token_accuracy": 0.13125866502523423,
      "num_tokens": 2717345.0,
      "step": 1460
    },
    {
      "entropy": 6.818531513214111,
      "epoch": 0.12937124690921936,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999810213989047,
      "loss": 6.5966,
      "mean_token_accuracy": 0.13112419992685317,
      "num_tokens": 2726892.0,
      "step": 1465
    },
    {
      "entropy": 6.735688495635986,
      "epoch": 0.1298127870010597,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999806101794558,
      "loss": 6.572,
      "mean_token_accuracy": 0.13323748409748076,
      "num_tokens": 2736479.0,
      "step": 1470
    },
    {
      "entropy": 6.835882139205933,
      "epoch": 0.13025432709290002,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999801945527648,
      "loss": 6.6304,
      "mean_token_accuracy": 0.12691461741924287,
      "num_tokens": 2745998.0,
      "step": 1475
    },
    {
      "entropy": 6.753359127044678,
      "epoch": 0.13069586718474038,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999797745188395,
      "loss": 6.6159,
      "mean_token_accuracy": 0.1317651830613613,
      "num_tokens": 2754346.0,
      "step": 1480
    },
    {
      "entropy": 6.801856279373169,
      "epoch": 0.13113740727658071,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999793500776886,
      "loss": 6.5416,
      "mean_token_accuracy": 0.13924193307757377,
      "num_tokens": 2763391.0,
      "step": 1485
    },
    {
      "entropy": 6.706637144088745,
      "epoch": 0.13157894736842105,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999789212293201,
      "loss": 6.5867,
      "mean_token_accuracy": 0.13430250883102418,
      "num_tokens": 2772764.0,
      "step": 1490
    },
    {
      "entropy": 6.875431776046753,
      "epoch": 0.1320204874602614,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999784879737423,
      "loss": 6.6781,
      "mean_token_accuracy": 0.1260899782180786,
      "num_tokens": 2782312.0,
      "step": 1495
    },
    {
      "entropy": 6.746471691131592,
      "epoch": 0.13246202755210174,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999780503109642,
      "loss": 6.5979,
      "mean_token_accuracy": 0.1290306605398655,
      "num_tokens": 2791159.0,
      "step": 1500
    },
    {
      "entropy": 6.704893779754639,
      "epoch": 0.13290356764394207,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999776082409939,
      "loss": 6.408,
      "mean_token_accuracy": 0.13926204219460486,
      "num_tokens": 2799319.0,
      "step": 1505
    },
    {
      "entropy": 6.6507806301116945,
      "epoch": 0.1333451077357824,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999771617638401,
      "loss": 6.5335,
      "mean_token_accuracy": 0.13334701284766198,
      "num_tokens": 2807401.0,
      "step": 1510
    },
    {
      "entropy": 6.7505804061889645,
      "epoch": 0.13378664782762276,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999767108795118,
      "loss": 6.6075,
      "mean_token_accuracy": 0.1309545248746872,
      "num_tokens": 2817734.0,
      "step": 1515
    },
    {
      "entropy": 6.790192747116089,
      "epoch": 0.1342281879194631,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999762555880176,
      "loss": 6.5906,
      "mean_token_accuracy": 0.12997578456997871,
      "num_tokens": 2828235.0,
      "step": 1520
    },
    {
      "entropy": 6.771645879745483,
      "epoch": 0.13466972801130342,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999757958893666,
      "loss": 6.6309,
      "mean_token_accuracy": 0.1289198786020279,
      "num_tokens": 2837453.0,
      "step": 1525
    },
    {
      "entropy": 6.712240839004517,
      "epoch": 0.13511126810314378,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999753317835677,
      "loss": 6.587,
      "mean_token_accuracy": 0.12751804441213607,
      "num_tokens": 2847055.0,
      "step": 1530
    },
    {
      "entropy": 6.758566999435425,
      "epoch": 0.1355528081949841,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999748632706299,
      "loss": 6.5597,
      "mean_token_accuracy": 0.13620577305555343,
      "num_tokens": 2857101.0,
      "step": 1535
    },
    {
      "entropy": 6.699764680862427,
      "epoch": 0.13599434828682444,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999743903505626,
      "loss": 6.4845,
      "mean_token_accuracy": 0.13134869784116746,
      "num_tokens": 2866685.0,
      "step": 1540
    },
    {
      "entropy": 6.813828182220459,
      "epoch": 0.13643588837866477,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999739130233749,
      "loss": 6.5821,
      "mean_token_accuracy": 0.1299674078822136,
      "num_tokens": 2876022.0,
      "step": 1545
    },
    {
      "entropy": 6.708414220809937,
      "epoch": 0.13687742847050513,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999734312890761,
      "loss": 6.5192,
      "mean_token_accuracy": 0.12855476662516593,
      "num_tokens": 2885560.0,
      "step": 1550
    },
    {
      "entropy": 6.690548992156982,
      "epoch": 0.13731896856234546,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999729451476757,
      "loss": 6.5568,
      "mean_token_accuracy": 0.13441056087613107,
      "num_tokens": 2894686.0,
      "step": 1555
    },
    {
      "entropy": 6.749054193496704,
      "epoch": 0.1377605086541858,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999724545991835,
      "loss": 6.5599,
      "mean_token_accuracy": 0.1379253476858139,
      "num_tokens": 2904390.0,
      "step": 1560
    },
    {
      "entropy": 6.715701723098755,
      "epoch": 0.13820204874602615,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999719596436086,
      "loss": 6.6185,
      "mean_token_accuracy": 0.13158311247825621,
      "num_tokens": 2913311.0,
      "step": 1565
    },
    {
      "entropy": 6.777320003509521,
      "epoch": 0.13864358883786648,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999714602809611,
      "loss": 6.4916,
      "mean_token_accuracy": 0.13447055071592331,
      "num_tokens": 2923196.0,
      "step": 1570
    },
    {
      "entropy": 6.681281852722168,
      "epoch": 0.1390851289297068,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999709565112506,
      "loss": 6.5513,
      "mean_token_accuracy": 0.13382177129387857,
      "num_tokens": 2932813.0,
      "step": 1575
    },
    {
      "entropy": 6.7223258972167965,
      "epoch": 0.13952666902154715,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499970448334487,
      "loss": 6.4762,
      "mean_token_accuracy": 0.12845111936330794,
      "num_tokens": 2942694.0,
      "step": 1580
    },
    {
      "entropy": 6.633160924911499,
      "epoch": 0.1399682091133875,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999699357506803,
      "loss": 6.4203,
      "mean_token_accuracy": 0.14310574680566787,
      "num_tokens": 2950932.0,
      "step": 1585
    },
    {
      "entropy": 6.733539772033692,
      "epoch": 0.14040974920522784,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999694187598406,
      "loss": 6.5208,
      "mean_token_accuracy": 0.1363385871052742,
      "num_tokens": 2960102.0,
      "step": 1590
    },
    {
      "entropy": 6.716689872741699,
      "epoch": 0.14085128929706817,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004999688973619777,
      "loss": 6.5161,
      "mean_token_accuracy": 0.1320077233016491,
      "num_tokens": 2969968.0,
      "step": 1595
    },
    {
      "entropy": 6.624239683151245,
      "epoch": 0.14129282938890853,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999683715571022,
      "loss": 6.4732,
      "mean_token_accuracy": 0.14136387258768082,
      "num_tokens": 2978880.0,
      "step": 1600
    },
    {
      "entropy": 6.666266059875488,
      "epoch": 0.14173436948074886,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999678413452242,
      "loss": 6.5087,
      "mean_token_accuracy": 0.1332849942147732,
      "num_tokens": 2988369.0,
      "step": 1605
    },
    {
      "entropy": 6.669563817977905,
      "epoch": 0.1421759095725892,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999673067263542,
      "loss": 6.5549,
      "mean_token_accuracy": 0.1297246351838112,
      "num_tokens": 2997070.0,
      "step": 1610
    },
    {
      "entropy": 6.7209991931915285,
      "epoch": 0.14261744966442952,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999667677005026,
      "loss": 6.4881,
      "mean_token_accuracy": 0.1361941583454609,
      "num_tokens": 3006547.0,
      "step": 1615
    },
    {
      "entropy": 6.670996856689453,
      "epoch": 0.14305898975626988,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999662242676799,
      "loss": 6.5007,
      "mean_token_accuracy": 0.13505242466926576,
      "num_tokens": 3015821.0,
      "step": 1620
    },
    {
      "entropy": 6.667819404602051,
      "epoch": 0.1435005298481102,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999656764278968,
      "loss": 6.4863,
      "mean_token_accuracy": 0.13409159779548646,
      "num_tokens": 3024750.0,
      "step": 1625
    },
    {
      "entropy": 6.699454641342163,
      "epoch": 0.14394206993995054,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999651241811642,
      "loss": 6.4543,
      "mean_token_accuracy": 0.13378062024712561,
      "num_tokens": 3033345.0,
      "step": 1630
    },
    {
      "entropy": 6.63138108253479,
      "epoch": 0.1443836100317909,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999645675274925,
      "loss": 6.4314,
      "mean_token_accuracy": 0.1369638502597809,
      "num_tokens": 3042060.0,
      "step": 1635
    },
    {
      "entropy": 6.6189075946807865,
      "epoch": 0.14482515012363123,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999640064668931,
      "loss": 6.584,
      "mean_token_accuracy": 0.1317050538957119,
      "num_tokens": 3052490.0,
      "step": 1640
    },
    {
      "entropy": 6.820568704605103,
      "epoch": 0.14526669021547156,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999634409993766,
      "loss": 6.5606,
      "mean_token_accuracy": 0.1320354349911213,
      "num_tokens": 3061934.0,
      "step": 1645
    },
    {
      "entropy": 6.609713792800903,
      "epoch": 0.1457082303073119,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999628711249544,
      "loss": 6.4512,
      "mean_token_accuracy": 0.14322613999247552,
      "num_tokens": 3070890.0,
      "step": 1650
    },
    {
      "entropy": 6.7396402835845945,
      "epoch": 0.14614977039915225,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999622968436373,
      "loss": 6.4715,
      "mean_token_accuracy": 0.13257212415337563,
      "num_tokens": 3079933.0,
      "step": 1655
    },
    {
      "entropy": 6.579079866409302,
      "epoch": 0.14659131049099258,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999617181554369,
      "loss": 6.5248,
      "mean_token_accuracy": 0.13561063706874849,
      "num_tokens": 3089910.0,
      "step": 1660
    },
    {
      "entropy": 6.757257795333862,
      "epoch": 0.1470328505828329,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004999611350603643,
      "loss": 6.5275,
      "mean_token_accuracy": 0.13599081262946128,
      "num_tokens": 3098676.0,
      "step": 1665
    },
    {
      "entropy": 6.642300748825074,
      "epoch": 0.14747439067467327,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499960547558431,
      "loss": 6.6165,
      "mean_token_accuracy": 0.13092149198055267,
      "num_tokens": 3109055.0,
      "step": 1670
    },
    {
      "entropy": 6.597188997268677,
      "epoch": 0.1479159307665136,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999599556496486,
      "loss": 6.4706,
      "mean_token_accuracy": 0.13985515162348747,
      "num_tokens": 3117517.0,
      "step": 1675
    },
    {
      "entropy": 6.640227365493774,
      "epoch": 0.14835747085835393,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999593593340286,
      "loss": 6.463,
      "mean_token_accuracy": 0.14289223924279212,
      "num_tokens": 3126606.0,
      "step": 1680
    },
    {
      "entropy": 6.657995748519897,
      "epoch": 0.14879901095019427,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999587586115826,
      "loss": 6.4902,
      "mean_token_accuracy": 0.1365136206150055,
      "num_tokens": 3135444.0,
      "step": 1685
    },
    {
      "entropy": 6.682824659347534,
      "epoch": 0.14924055104203462,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999581534823226,
      "loss": 6.4214,
      "mean_token_accuracy": 0.14298791959881782,
      "num_tokens": 3144967.0,
      "step": 1690
    },
    {
      "entropy": 6.649335050582886,
      "epoch": 0.14968209113387496,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999575439462601,
      "loss": 6.4563,
      "mean_token_accuracy": 0.13346426337957382,
      "num_tokens": 3153898.0,
      "step": 1695
    },
    {
      "entropy": 6.639866065979004,
      "epoch": 0.1501236312257153,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999569300034075,
      "loss": 6.5099,
      "mean_token_accuracy": 0.13122038245201112,
      "num_tokens": 3162341.0,
      "step": 1700
    },
    {
      "entropy": 6.621887493133545,
      "epoch": 0.15056517131755565,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999563116537764,
      "loss": 6.3038,
      "mean_token_accuracy": 0.1461081884801388,
      "num_tokens": 3171420.0,
      "step": 1705
    },
    {
      "entropy": 6.58403959274292,
      "epoch": 0.15100671140939598,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999556888973792,
      "loss": 6.3453,
      "mean_token_accuracy": 0.14070172980427742,
      "num_tokens": 3180221.0,
      "step": 1710
    },
    {
      "entropy": 6.7422089099884035,
      "epoch": 0.1514482515012363,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999550617342279,
      "loss": 6.5005,
      "mean_token_accuracy": 0.1325171984732151,
      "num_tokens": 3189156.0,
      "step": 1715
    },
    {
      "entropy": 6.5561034202575685,
      "epoch": 0.15188979159307664,
      "grad_norm": 1.0,
      "learning_rate": 0.000499954430164335,
      "loss": 6.4253,
      "mean_token_accuracy": 0.13670367896556854,
      "num_tokens": 3199870.0,
      "step": 1720
    },
    {
      "entropy": 6.655413389205933,
      "epoch": 0.152331331684917,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999537941877127,
      "loss": 6.4231,
      "mean_token_accuracy": 0.14037118926644326,
      "num_tokens": 3208815.0,
      "step": 1725
    },
    {
      "entropy": 6.575525569915771,
      "epoch": 0.15277287177675733,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999531538043735,
      "loss": 6.4845,
      "mean_token_accuracy": 0.137246023863554,
      "num_tokens": 3218692.0,
      "step": 1730
    },
    {
      "entropy": 6.627863502502441,
      "epoch": 0.15321441186859766,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999525090143298,
      "loss": 6.4482,
      "mean_token_accuracy": 0.14504025802016257,
      "num_tokens": 3227604.0,
      "step": 1735
    },
    {
      "entropy": 6.625754451751709,
      "epoch": 0.15365595196043802,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999518598175946,
      "loss": 6.4675,
      "mean_token_accuracy": 0.14471371322870255,
      "num_tokens": 3237912.0,
      "step": 1740
    },
    {
      "entropy": 6.527244615554809,
      "epoch": 0.15409749205227835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999512062141805,
      "loss": 6.3302,
      "mean_token_accuracy": 0.14219417944550514,
      "num_tokens": 3246003.0,
      "step": 1745
    },
    {
      "entropy": 6.632448387145996,
      "epoch": 0.15453903214411868,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999505482040999,
      "loss": 6.4719,
      "mean_token_accuracy": 0.13460535928606987,
      "num_tokens": 3256363.0,
      "step": 1750
    },
    {
      "entropy": 6.610141611099243,
      "epoch": 0.154980572235959,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999498857873662,
      "loss": 6.4661,
      "mean_token_accuracy": 0.13534337058663368,
      "num_tokens": 3265822.0,
      "step": 1755
    },
    {
      "entropy": 6.538852977752685,
      "epoch": 0.15542211232779937,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999492189639921,
      "loss": 6.3683,
      "mean_token_accuracy": 0.14053185135126114,
      "num_tokens": 3274614.0,
      "step": 1760
    },
    {
      "entropy": 6.577081727981567,
      "epoch": 0.1558636524196397,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999485477339907,
      "loss": 6.4384,
      "mean_token_accuracy": 0.13609417378902436,
      "num_tokens": 3283800.0,
      "step": 1765
    },
    {
      "entropy": 6.623507022857666,
      "epoch": 0.15630519251148003,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999478720973753,
      "loss": 6.3582,
      "mean_token_accuracy": 0.14612039923667908,
      "num_tokens": 3293221.0,
      "step": 1770
    },
    {
      "entropy": 6.631183099746704,
      "epoch": 0.1567467326033204,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499947192054159,
      "loss": 6.5087,
      "mean_token_accuracy": 0.13212059661746026,
      "num_tokens": 3302852.0,
      "step": 1775
    },
    {
      "entropy": 6.668994379043579,
      "epoch": 0.15718827269516072,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499946507604355,
      "loss": 6.3761,
      "mean_token_accuracy": 0.14372102692723274,
      "num_tokens": 3311752.0,
      "step": 1780
    },
    {
      "entropy": 6.522418546676636,
      "epoch": 0.15762981278700106,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499945818747977,
      "loss": 6.4622,
      "mean_token_accuracy": 0.14355819597840308,
      "num_tokens": 3321339.0,
      "step": 1785
    },
    {
      "entropy": 6.667537021636963,
      "epoch": 0.1580713528788414,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999451254850383,
      "loss": 6.4756,
      "mean_token_accuracy": 0.13703957572579384,
      "num_tokens": 3330269.0,
      "step": 1790
    },
    {
      "entropy": 6.634947395324707,
      "epoch": 0.15851289297068175,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999444278155525,
      "loss": 6.4032,
      "mean_token_accuracy": 0.14509668350219726,
      "num_tokens": 3340770.0,
      "step": 1795
    },
    {
      "entropy": 6.548423147201538,
      "epoch": 0.15895443306252208,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999437257395333,
      "loss": 6.4725,
      "mean_token_accuracy": 0.1425343744456768,
      "num_tokens": 3349976.0,
      "step": 1800
    },
    {
      "entropy": 6.667039012908935,
      "epoch": 0.1593959731543624,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999430192569944,
      "loss": 6.5242,
      "mean_token_accuracy": 0.13105336651206018,
      "num_tokens": 3359764.0,
      "step": 1805
    },
    {
      "entropy": 6.581896781921387,
      "epoch": 0.15983751324620277,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999423083679498,
      "loss": 6.405,
      "mean_token_accuracy": 0.13858939930796624,
      "num_tokens": 3369939.0,
      "step": 1810
    },
    {
      "entropy": 6.633963489532471,
      "epoch": 0.1602790533380431,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999415930724133,
      "loss": 6.5557,
      "mean_token_accuracy": 0.1334507644176483,
      "num_tokens": 3381326.0,
      "step": 1815
    },
    {
      "entropy": 6.706169319152832,
      "epoch": 0.16072059342988343,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999408733703988,
      "loss": 6.5135,
      "mean_token_accuracy": 0.13054400235414504,
      "num_tokens": 3391016.0,
      "step": 1820
    },
    {
      "entropy": 6.57712345123291,
      "epoch": 0.16116213352172376,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999401492619207,
      "loss": 6.2754,
      "mean_token_accuracy": 0.14139460623264313,
      "num_tokens": 3400294.0,
      "step": 1825
    },
    {
      "entropy": 6.533513784408569,
      "epoch": 0.16160367361356412,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999394207469928,
      "loss": 6.4737,
      "mean_token_accuracy": 0.13292715698480606,
      "num_tokens": 3409685.0,
      "step": 1830
    },
    {
      "entropy": 6.695759248733521,
      "epoch": 0.16204521370540445,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999386878256297,
      "loss": 6.3887,
      "mean_token_accuracy": 0.13601208254694938,
      "num_tokens": 3418946.0,
      "step": 1835
    },
    {
      "entropy": 6.4282214641571045,
      "epoch": 0.16248675379724478,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999379504978457,
      "loss": 6.3763,
      "mean_token_accuracy": 0.13835799768567086,
      "num_tokens": 3428245.0,
      "step": 1840
    },
    {
      "entropy": 6.614362144470215,
      "epoch": 0.16292829388908514,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499937208763655,
      "loss": 6.3958,
      "mean_token_accuracy": 0.14421852231025695,
      "num_tokens": 3437580.0,
      "step": 1845
    },
    {
      "entropy": 6.64262170791626,
      "epoch": 0.16336983398092547,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999364626230724,
      "loss": 6.3701,
      "mean_token_accuracy": 0.14203060865402223,
      "num_tokens": 3446393.0,
      "step": 1850
    },
    {
      "entropy": 6.477195358276367,
      "epoch": 0.1638113740727658,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999357120761124,
      "loss": 6.4748,
      "mean_token_accuracy": 0.1370338521897793,
      "num_tokens": 3455956.0,
      "step": 1855
    },
    {
      "entropy": 6.65583610534668,
      "epoch": 0.16425291416460613,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999349571227898,
      "loss": 6.4634,
      "mean_token_accuracy": 0.1417573131620884,
      "num_tokens": 3465722.0,
      "step": 1860
    },
    {
      "entropy": 6.579311370849609,
      "epoch": 0.1646944542564465,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999341977631193,
      "loss": 6.4233,
      "mean_token_accuracy": 0.13521963283419608,
      "num_tokens": 3475467.0,
      "step": 1865
    },
    {
      "entropy": 6.563374280929565,
      "epoch": 0.16513599434828682,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999334339971157,
      "loss": 6.3294,
      "mean_token_accuracy": 0.140533210337162,
      "num_tokens": 3484931.0,
      "step": 1870
    },
    {
      "entropy": 6.542880487442017,
      "epoch": 0.16557753444012716,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999326658247942,
      "loss": 6.4374,
      "mean_token_accuracy": 0.14069321602582932,
      "num_tokens": 3494001.0,
      "step": 1875
    },
    {
      "entropy": 6.596772766113281,
      "epoch": 0.16601907453196751,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999318932461696,
      "loss": 6.3622,
      "mean_token_accuracy": 0.14485045671463012,
      "num_tokens": 3503021.0,
      "step": 1880
    },
    {
      "entropy": 6.477844142913819,
      "epoch": 0.16646061462380785,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999311162612571,
      "loss": 6.4034,
      "mean_token_accuracy": 0.1405440613627434,
      "num_tokens": 3513005.0,
      "step": 1885
    },
    {
      "entropy": 6.664361810684204,
      "epoch": 0.16690215471564818,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499930334870072,
      "loss": 6.468,
      "mean_token_accuracy": 0.14140931516885757,
      "num_tokens": 3523219.0,
      "step": 1890
    },
    {
      "entropy": 6.607090282440185,
      "epoch": 0.1673436948074885,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999295490726296,
      "loss": 6.4475,
      "mean_token_accuracy": 0.13736767545342446,
      "num_tokens": 3532917.0,
      "step": 1895
    },
    {
      "entropy": 6.548594951629639,
      "epoch": 0.16778523489932887,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999287588689453,
      "loss": 6.4314,
      "mean_token_accuracy": 0.14416549280285834,
      "num_tokens": 3542845.0,
      "step": 1900
    },
    {
      "entropy": 6.615040254592896,
      "epoch": 0.1682267749911692,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999279642590344,
      "loss": 6.4644,
      "mean_token_accuracy": 0.14099383279681205,
      "num_tokens": 3552126.0,
      "step": 1905
    },
    {
      "entropy": 6.59620246887207,
      "epoch": 0.16866831508300953,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999271652429127,
      "loss": 6.4683,
      "mean_token_accuracy": 0.13365241289138793,
      "num_tokens": 3561254.0,
      "step": 1910
    },
    {
      "entropy": 6.603018665313721,
      "epoch": 0.1691098551748499,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999263618205958,
      "loss": 6.3064,
      "mean_token_accuracy": 0.14801083952188493,
      "num_tokens": 3569781.0,
      "step": 1915
    },
    {
      "entropy": 6.507657670974732,
      "epoch": 0.16955139526669022,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999255539920993,
      "loss": 6.4154,
      "mean_token_accuracy": 0.13945700079202653,
      "num_tokens": 3579664.0,
      "step": 1920
    },
    {
      "entropy": 6.607545375823975,
      "epoch": 0.16999293535853055,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999247417574391,
      "loss": 6.4491,
      "mean_token_accuracy": 0.14012959450483323,
      "num_tokens": 3588671.0,
      "step": 1925
    },
    {
      "entropy": 6.5801286697387695,
      "epoch": 0.17043447545037088,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999239251166312,
      "loss": 6.3639,
      "mean_token_accuracy": 0.14608858227729798,
      "num_tokens": 3597656.0,
      "step": 1930
    },
    {
      "entropy": 6.465410947799683,
      "epoch": 0.17087601554221124,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999231040696914,
      "loss": 6.373,
      "mean_token_accuracy": 0.14378879442811013,
      "num_tokens": 3608017.0,
      "step": 1935
    },
    {
      "entropy": 6.607225227355957,
      "epoch": 0.17131755563405157,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999222786166361,
      "loss": 6.4412,
      "mean_token_accuracy": 0.14090094193816186,
      "num_tokens": 3618189.0,
      "step": 1940
    },
    {
      "entropy": 6.596541976928711,
      "epoch": 0.1717590957258919,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999214487574812,
      "loss": 6.3977,
      "mean_token_accuracy": 0.1351962171494961,
      "num_tokens": 3627211.0,
      "step": 1945
    },
    {
      "entropy": 6.564425277709961,
      "epoch": 0.17220063581773226,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999206144922431,
      "loss": 6.331,
      "mean_token_accuracy": 0.1392312839627266,
      "num_tokens": 3636781.0,
      "step": 1950
    },
    {
      "entropy": 6.521996641159058,
      "epoch": 0.1726421759095726,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499919775820938,
      "loss": 6.4204,
      "mean_token_accuracy": 0.1457922250032425,
      "num_tokens": 3644891.0,
      "step": 1955
    },
    {
      "entropy": 6.611178350448609,
      "epoch": 0.17308371600141292,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999189327435825,
      "loss": 6.4558,
      "mean_token_accuracy": 0.13934372514486312,
      "num_tokens": 3655477.0,
      "step": 1960
    },
    {
      "entropy": 6.615030193328858,
      "epoch": 0.17352525609325326,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999180852601929,
      "loss": 6.4555,
      "mean_token_accuracy": 0.1424385577440262,
      "num_tokens": 3664542.0,
      "step": 1965
    },
    {
      "entropy": 6.556003856658935,
      "epoch": 0.17396679618509361,
      "grad_norm": 1.140625,
      "learning_rate": 0.000499917233370786,
      "loss": 6.3637,
      "mean_token_accuracy": 0.13885778263211251,
      "num_tokens": 3673806.0,
      "step": 1970
    },
    {
      "entropy": 6.537166738510132,
      "epoch": 0.17440833627693395,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999163770753784,
      "loss": 6.3706,
      "mean_token_accuracy": 0.14099680185317992,
      "num_tokens": 3683238.0,
      "step": 1975
    },
    {
      "entropy": 6.504315948486328,
      "epoch": 0.17484987636877428,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004999155163739869,
      "loss": 6.3407,
      "mean_token_accuracy": 0.14536889716982843,
      "num_tokens": 3692161.0,
      "step": 1980
    },
    {
      "entropy": 6.581552124023437,
      "epoch": 0.17529141646061464,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999146512666284,
      "loss": 6.367,
      "mean_token_accuracy": 0.1379326641559601,
      "num_tokens": 3701431.0,
      "step": 1985
    },
    {
      "entropy": 6.487573671340942,
      "epoch": 0.17573295655245497,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004999137817533197,
      "loss": 6.3461,
      "mean_token_accuracy": 0.14381761997938156,
      "num_tokens": 3710963.0,
      "step": 1990
    },
    {
      "entropy": 6.506614542007446,
      "epoch": 0.1761744966442953,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999129078340779,
      "loss": 6.3353,
      "mean_token_accuracy": 0.14846789166331292,
      "num_tokens": 3720177.0,
      "step": 1995
    },
    {
      "entropy": 6.530578517913819,
      "epoch": 0.17661603673613563,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999120295089202,
      "loss": 6.2971,
      "mean_token_accuracy": 0.14959900006651877,
      "num_tokens": 3728123.0,
      "step": 2000
    },
    {
      "entropy": 6.542397546768188,
      "epoch": 0.177057576827976,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999111467778639,
      "loss": 6.4341,
      "mean_token_accuracy": 0.1391107402741909,
      "num_tokens": 3736869.0,
      "step": 2005
    },
    {
      "entropy": 6.53966269493103,
      "epoch": 0.17749911691981632,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000499910259640926,
      "loss": 6.3237,
      "mean_token_accuracy": 0.14294100552797318,
      "num_tokens": 3745830.0,
      "step": 2010
    },
    {
      "entropy": 6.51638708114624,
      "epoch": 0.17794065701165665,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499909368098124,
      "loss": 6.3564,
      "mean_token_accuracy": 0.13725989684462547,
      "num_tokens": 3755019.0,
      "step": 2015
    },
    {
      "entropy": 6.511772108078003,
      "epoch": 0.178382197103497,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999084721494754,
      "loss": 6.331,
      "mean_token_accuracy": 0.14094301015138627,
      "num_tokens": 3764814.0,
      "step": 2020
    },
    {
      "entropy": 6.452651882171631,
      "epoch": 0.17882373719533734,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999075717949978,
      "loss": 6.2835,
      "mean_token_accuracy": 0.14623287171125413,
      "num_tokens": 3774258.0,
      "step": 2025
    },
    {
      "entropy": 6.427514886856079,
      "epoch": 0.17926527728717767,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004999066670347089,
      "loss": 6.3158,
      "mean_token_accuracy": 0.1432214617729187,
      "num_tokens": 3783961.0,
      "step": 2030
    },
    {
      "entropy": 6.497758913040161,
      "epoch": 0.179706817379018,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999057578686261,
      "loss": 6.2899,
      "mean_token_accuracy": 0.14475978389382363,
      "num_tokens": 3792417.0,
      "step": 2035
    },
    {
      "entropy": 6.472495174407959,
      "epoch": 0.18014835747085836,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999048442967675,
      "loss": 6.2812,
      "mean_token_accuracy": 0.14342265501618384,
      "num_tokens": 3801479.0,
      "step": 2040
    },
    {
      "entropy": 6.572735977172852,
      "epoch": 0.1805898975626987,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999039263191508,
      "loss": 6.3581,
      "mean_token_accuracy": 0.13960720375180244,
      "num_tokens": 3810799.0,
      "step": 2045
    },
    {
      "entropy": 6.510642623901367,
      "epoch": 0.18103143765453902,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999030039357943,
      "loss": 6.3896,
      "mean_token_accuracy": 0.13761023208498954,
      "num_tokens": 3820966.0,
      "step": 2050
    },
    {
      "entropy": 6.4944600582122805,
      "epoch": 0.18147297774637938,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999020771467158,
      "loss": 6.3647,
      "mean_token_accuracy": 0.1407288655638695,
      "num_tokens": 3829247.0,
      "step": 2055
    },
    {
      "entropy": 6.609596920013428,
      "epoch": 0.1819145178382197,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999011459519335,
      "loss": 6.3036,
      "mean_token_accuracy": 0.14792885035276412,
      "num_tokens": 3838114.0,
      "step": 2060
    },
    {
      "entropy": 6.411144733428955,
      "epoch": 0.18235605793006004,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999002103514655,
      "loss": 6.3792,
      "mean_token_accuracy": 0.142097420245409,
      "num_tokens": 3848075.0,
      "step": 2065
    },
    {
      "entropy": 6.5771557807922365,
      "epoch": 0.18279759802190038,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998992703453304,
      "loss": 6.4163,
      "mean_token_accuracy": 0.14099472612142563,
      "num_tokens": 3857934.0,
      "step": 2070
    },
    {
      "entropy": 6.490412330627441,
      "epoch": 0.18323913811374073,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998983259335466,
      "loss": 6.2769,
      "mean_token_accuracy": 0.15167167782783508,
      "num_tokens": 3866707.0,
      "step": 2075
    },
    {
      "entropy": 6.511580276489258,
      "epoch": 0.18368067820558107,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998973771161324,
      "loss": 6.3342,
      "mean_token_accuracy": 0.14083316251635553,
      "num_tokens": 3875233.0,
      "step": 2080
    },
    {
      "entropy": 6.495502281188965,
      "epoch": 0.1841222182974214,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998964238931065,
      "loss": 6.3311,
      "mean_token_accuracy": 0.1447795733809471,
      "num_tokens": 3885173.0,
      "step": 2085
    },
    {
      "entropy": 6.511646032333374,
      "epoch": 0.18456375838926176,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998954662644876,
      "loss": 6.3103,
      "mean_token_accuracy": 0.14369324892759322,
      "num_tokens": 3894198.0,
      "step": 2090
    },
    {
      "entropy": 6.465791845321656,
      "epoch": 0.1850052984811021,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998945042302943,
      "loss": 6.3043,
      "mean_token_accuracy": 0.14370952174067497,
      "num_tokens": 3904076.0,
      "step": 2095
    },
    {
      "entropy": 6.529692268371582,
      "epoch": 0.18544683857294242,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998935377905457,
      "loss": 6.414,
      "mean_token_accuracy": 0.13739102557301522,
      "num_tokens": 3913204.0,
      "step": 2100
    },
    {
      "entropy": 6.59678544998169,
      "epoch": 0.18588837866478275,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998925669452605,
      "loss": 6.3665,
      "mean_token_accuracy": 0.14387816935777664,
      "num_tokens": 3922148.0,
      "step": 2105
    },
    {
      "entropy": 6.451368236541748,
      "epoch": 0.1863299187566231,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998915916944579,
      "loss": 6.336,
      "mean_token_accuracy": 0.14496995508670807,
      "num_tokens": 3931333.0,
      "step": 2110
    },
    {
      "entropy": 6.508979606628418,
      "epoch": 0.18677145884846344,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998906120381568,
      "loss": 6.2869,
      "mean_token_accuracy": 0.14773214906454085,
      "num_tokens": 3941061.0,
      "step": 2115
    },
    {
      "entropy": 6.536288022994995,
      "epoch": 0.18721299894030377,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998896279763766,
      "loss": 6.3841,
      "mean_token_accuracy": 0.13904453590512275,
      "num_tokens": 3950075.0,
      "step": 2120
    },
    {
      "entropy": 6.4915399074554445,
      "epoch": 0.18765453903214413,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998886395091365,
      "loss": 6.2552,
      "mean_token_accuracy": 0.1421836346387863,
      "num_tokens": 3958885.0,
      "step": 2125
    },
    {
      "entropy": 6.379503393173218,
      "epoch": 0.18809607912398446,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998876466364559,
      "loss": 6.3307,
      "mean_token_accuracy": 0.1476896308362484,
      "num_tokens": 3968218.0,
      "step": 2130
    },
    {
      "entropy": 6.450522613525391,
      "epoch": 0.1885376192158248,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998866493583541,
      "loss": 6.2918,
      "mean_token_accuracy": 0.14630952328443528,
      "num_tokens": 3977435.0,
      "step": 2135
    },
    {
      "entropy": 6.4365739822387695,
      "epoch": 0.18897915930766512,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998856476748509,
      "loss": 6.2738,
      "mean_token_accuracy": 0.14805593118071556,
      "num_tokens": 3986608.0,
      "step": 2140
    },
    {
      "entropy": 6.498980379104614,
      "epoch": 0.18942069939950548,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998846415859656,
      "loss": 6.2719,
      "mean_token_accuracy": 0.14355492740869522,
      "num_tokens": 3996087.0,
      "step": 2145
    },
    {
      "entropy": 6.4271481990814205,
      "epoch": 0.1898622394913458,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004998836310917182,
      "loss": 6.3259,
      "mean_token_accuracy": 0.13990697860717774,
      "num_tokens": 4006257.0,
      "step": 2150
    },
    {
      "entropy": 6.432379961013794,
      "epoch": 0.19030377958318614,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998826161921282,
      "loss": 6.3248,
      "mean_token_accuracy": 0.1525232471525669,
      "num_tokens": 4015904.0,
      "step": 2155
    },
    {
      "entropy": 6.395722532272339,
      "epoch": 0.1907453196750265,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998815968872157,
      "loss": 6.3357,
      "mean_token_accuracy": 0.14217145144939422,
      "num_tokens": 4025417.0,
      "step": 2160
    },
    {
      "entropy": 6.434484338760376,
      "epoch": 0.19118685976686683,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998805731770007,
      "loss": 6.194,
      "mean_token_accuracy": 0.16098422557115555,
      "num_tokens": 4035181.0,
      "step": 2165
    },
    {
      "entropy": 6.466515350341797,
      "epoch": 0.19162839985870717,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499879545061503,
      "loss": 6.3656,
      "mean_token_accuracy": 0.14232496544718742,
      "num_tokens": 4045112.0,
      "step": 2170
    },
    {
      "entropy": 6.544062900543213,
      "epoch": 0.1920699399505475,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998785125407432,
      "loss": 6.468,
      "mean_token_accuracy": 0.13045382052659987,
      "num_tokens": 4054566.0,
      "step": 2175
    },
    {
      "entropy": 6.538053798675537,
      "epoch": 0.19251148004238786,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499877475614741,
      "loss": 6.2829,
      "mean_token_accuracy": 0.14255435317754744,
      "num_tokens": 4063960.0,
      "step": 2180
    },
    {
      "entropy": 6.318663692474365,
      "epoch": 0.1929530201342282,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998764342835169,
      "loss": 6.2248,
      "mean_token_accuracy": 0.14946237206459045,
      "num_tokens": 4072620.0,
      "step": 2185
    },
    {
      "entropy": 6.361247396469116,
      "epoch": 0.19339456022606852,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998753885470915,
      "loss": 6.1939,
      "mean_token_accuracy": 0.15080117732286452,
      "num_tokens": 4081590.0,
      "step": 2190
    },
    {
      "entropy": 6.448186588287354,
      "epoch": 0.19383610031790888,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998743384054851,
      "loss": 6.3243,
      "mean_token_accuracy": 0.15017441809177398,
      "num_tokens": 4090758.0,
      "step": 2195
    },
    {
      "entropy": 6.359209203720093,
      "epoch": 0.1942776404097492,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998732838587183,
      "loss": 6.1452,
      "mean_token_accuracy": 0.1547905460000038,
      "num_tokens": 4099281.0,
      "step": 2200
    },
    {
      "entropy": 6.484091281890869,
      "epoch": 0.19471918050158954,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998722249068118,
      "loss": 6.2579,
      "mean_token_accuracy": 0.14498598128557205,
      "num_tokens": 4108953.0,
      "step": 2205
    },
    {
      "entropy": 6.412591505050659,
      "epoch": 0.19516072059342987,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998711615497863,
      "loss": 6.365,
      "mean_token_accuracy": 0.1421401210129261,
      "num_tokens": 4118799.0,
      "step": 2210
    },
    {
      "entropy": 6.515138959884643,
      "epoch": 0.19560226068527023,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998700937876626,
      "loss": 6.3215,
      "mean_token_accuracy": 0.14281034767627715,
      "num_tokens": 4127862.0,
      "step": 2215
    },
    {
      "entropy": 6.504071760177612,
      "epoch": 0.19604380077711056,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998690216204615,
      "loss": 6.4025,
      "mean_token_accuracy": 0.13254041969776154,
      "num_tokens": 4139029.0,
      "step": 2220
    },
    {
      "entropy": 6.468188190460205,
      "epoch": 0.1964853408689509,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998679450482043,
      "loss": 6.2211,
      "mean_token_accuracy": 0.15794688761234282,
      "num_tokens": 4148257.0,
      "step": 2225
    },
    {
      "entropy": 6.399339389801026,
      "epoch": 0.19692688096079125,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499866864070912,
      "loss": 6.1777,
      "mean_token_accuracy": 0.14184310287237167,
      "num_tokens": 4157626.0,
      "step": 2230
    },
    {
      "entropy": 6.394889450073242,
      "epoch": 0.19736842105263158,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998657786886056,
      "loss": 6.4048,
      "mean_token_accuracy": 0.1460250087082386,
      "num_tokens": 4166804.0,
      "step": 2235
    },
    {
      "entropy": 6.561651229858398,
      "epoch": 0.1978099611444719,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998646889013066,
      "loss": 6.2711,
      "mean_token_accuracy": 0.14743182435631752,
      "num_tokens": 4175701.0,
      "step": 2240
    },
    {
      "entropy": 6.471790933609009,
      "epoch": 0.19825150123631224,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998635947090362,
      "loss": 6.3589,
      "mean_token_accuracy": 0.1456327997148037,
      "num_tokens": 4184711.0,
      "step": 2245
    },
    {
      "entropy": 6.440358591079712,
      "epoch": 0.1986930413281526,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004998624961118158,
      "loss": 6.3148,
      "mean_token_accuracy": 0.1504688784480095,
      "num_tokens": 4193931.0,
      "step": 2250
    },
    {
      "entropy": 6.443251514434815,
      "epoch": 0.19913458141999293,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499861393109667,
      "loss": 6.1933,
      "mean_token_accuracy": 0.15764214992523193,
      "num_tokens": 4203000.0,
      "step": 2255
    },
    {
      "entropy": 6.3125227928161625,
      "epoch": 0.19957612151183327,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998602857026114,
      "loss": 6.214,
      "mean_token_accuracy": 0.1494602829217911,
      "num_tokens": 4211977.0,
      "step": 2260
    },
    {
      "entropy": 6.4371459007263185,
      "epoch": 0.20001766160367362,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004998591738906708,
      "loss": 6.2855,
      "mean_token_accuracy": 0.15524563640356065,
      "num_tokens": 4220375.0,
      "step": 2265
    },
    {
      "entropy": 6.506232404708863,
      "epoch": 0.20045920169551396,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998580576738668,
      "loss": 6.3022,
      "mean_token_accuracy": 0.14764061123132705,
      "num_tokens": 4230506.0,
      "step": 2270
    },
    {
      "entropy": 6.4268731594085695,
      "epoch": 0.2009007417873543,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998569370522213,
      "loss": 6.2711,
      "mean_token_accuracy": 0.1487172946333885,
      "num_tokens": 4240235.0,
      "step": 2275
    },
    {
      "entropy": 6.3987833023071286,
      "epoch": 0.20134228187919462,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004998558120257563,
      "loss": 6.3105,
      "mean_token_accuracy": 0.1501203641295433,
      "num_tokens": 4249775.0,
      "step": 2280
    },
    {
      "entropy": 6.428017330169678,
      "epoch": 0.20178382197103498,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998546825944938,
      "loss": 6.2596,
      "mean_token_accuracy": 0.14586075246334076,
      "num_tokens": 4258523.0,
      "step": 2285
    },
    {
      "entropy": 6.456875276565552,
      "epoch": 0.2022253620628753,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000499853548758456,
      "loss": 6.2549,
      "mean_token_accuracy": 0.14920931309461594,
      "num_tokens": 4267683.0,
      "step": 2290
    },
    {
      "entropy": 6.502974796295166,
      "epoch": 0.20266690215471564,
      "grad_norm": 1.5,
      "learning_rate": 0.000499852410517665,
      "loss": 6.3396,
      "mean_token_accuracy": 0.1415552243590355,
      "num_tokens": 4276903.0,
      "step": 2295
    },
    {
      "entropy": 6.390913295745849,
      "epoch": 0.203108442246556,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004998512678721431,
      "loss": 6.2765,
      "mean_token_accuracy": 0.14179632887244226,
      "num_tokens": 4287257.0,
      "step": 2300
    },
    {
      "entropy": 6.467247819900512,
      "epoch": 0.20354998233839633,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499850120821913,
      "loss": 6.2994,
      "mean_token_accuracy": 0.1429171495139599,
      "num_tokens": 4297345.0,
      "step": 2305
    },
    {
      "entropy": 6.504390478134155,
      "epoch": 0.20399152243023666,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998489693669967,
      "loss": 6.2774,
      "mean_token_accuracy": 0.14350235760211943,
      "num_tokens": 4306533.0,
      "step": 2310
    },
    {
      "entropy": 6.430126476287842,
      "epoch": 0.204433062522077,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499847813507417,
      "loss": 6.3142,
      "mean_token_accuracy": 0.14808136820793152,
      "num_tokens": 4316338.0,
      "step": 2315
    },
    {
      "entropy": 6.418506526947022,
      "epoch": 0.20487460261391735,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998466532431966,
      "loss": 6.2461,
      "mean_token_accuracy": 0.14408156126737595,
      "num_tokens": 4326585.0,
      "step": 2320
    },
    {
      "entropy": 6.502306318283081,
      "epoch": 0.20531614270575768,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998454885743581,
      "loss": 6.4021,
      "mean_token_accuracy": 0.1425124689936638,
      "num_tokens": 4336490.0,
      "step": 2325
    },
    {
      "entropy": 6.4134602546691895,
      "epoch": 0.205757682797598,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998443195009242,
      "loss": 6.266,
      "mean_token_accuracy": 0.14605539664626122,
      "num_tokens": 4346264.0,
      "step": 2330
    },
    {
      "entropy": 6.444290590286255,
      "epoch": 0.20619922288943837,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998431460229182,
      "loss": 6.261,
      "mean_token_accuracy": 0.14759050458669662,
      "num_tokens": 4355102.0,
      "step": 2335
    },
    {
      "entropy": 6.518517303466797,
      "epoch": 0.2066407629812787,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998419681403627,
      "loss": 6.4309,
      "mean_token_accuracy": 0.139105836302042,
      "num_tokens": 4365569.0,
      "step": 2340
    },
    {
      "entropy": 6.428114986419677,
      "epoch": 0.20708230307311903,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998407858532809,
      "loss": 6.2633,
      "mean_token_accuracy": 0.1455647498369217,
      "num_tokens": 4375437.0,
      "step": 2345
    },
    {
      "entropy": 6.348156213760376,
      "epoch": 0.20752384316495937,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499839599161696,
      "loss": 6.2444,
      "mean_token_accuracy": 0.15116954892873763,
      "num_tokens": 4384420.0,
      "step": 2350
    },
    {
      "entropy": 6.430051231384278,
      "epoch": 0.20796538325679972,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004998384080656314,
      "loss": 6.1993,
      "mean_token_accuracy": 0.15055150538682938,
      "num_tokens": 4393730.0,
      "step": 2355
    },
    {
      "entropy": 6.349585485458374,
      "epoch": 0.20840692334864006,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00049983721256511,
      "loss": 6.1842,
      "mean_token_accuracy": 0.14877034425735475,
      "num_tokens": 4402731.0,
      "step": 2360
    },
    {
      "entropy": 6.3974512100219725,
      "epoch": 0.2088484634404804,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998360126601556,
      "loss": 6.2405,
      "mean_token_accuracy": 0.1482672870159149,
      "num_tokens": 4411606.0,
      "step": 2365
    },
    {
      "entropy": 6.3687732219696045,
      "epoch": 0.20929000353232075,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998348083507916,
      "loss": 6.3177,
      "mean_token_accuracy": 0.14799434095621108,
      "num_tokens": 4421685.0,
      "step": 2370
    },
    {
      "entropy": 6.44785623550415,
      "epoch": 0.20973154362416108,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998335996370416,
      "loss": 6.2883,
      "mean_token_accuracy": 0.14761801213026046,
      "num_tokens": 4431765.0,
      "step": 2375
    },
    {
      "entropy": 6.392046165466309,
      "epoch": 0.2101730837160014,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998323865189291,
      "loss": 6.2439,
      "mean_token_accuracy": 0.14551597088575363,
      "num_tokens": 4441191.0,
      "step": 2380
    },
    {
      "entropy": 6.444688892364502,
      "epoch": 0.21061462380784174,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998311689964781,
      "loss": 6.3302,
      "mean_token_accuracy": 0.138855642080307,
      "num_tokens": 4450156.0,
      "step": 2385
    },
    {
      "entropy": 6.510595750808716,
      "epoch": 0.2110561638996821,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998299470697125,
      "loss": 6.3199,
      "mean_token_accuracy": 0.15451483130455018,
      "num_tokens": 4459466.0,
      "step": 2390
    },
    {
      "entropy": 6.423422145843506,
      "epoch": 0.21149770399152243,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998287207386559,
      "loss": 6.3232,
      "mean_token_accuracy": 0.14224404096603394,
      "num_tokens": 4468539.0,
      "step": 2395
    },
    {
      "entropy": 6.416548156738282,
      "epoch": 0.21193924408336276,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998274900033326,
      "loss": 6.2395,
      "mean_token_accuracy": 0.15238402634859086,
      "num_tokens": 4477579.0,
      "step": 2400
    },
    {
      "entropy": 6.291043853759765,
      "epoch": 0.21238078417520312,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998262548637667,
      "loss": 6.1989,
      "mean_token_accuracy": 0.15245457291603087,
      "num_tokens": 4486800.0,
      "step": 2405
    },
    {
      "entropy": 6.308730173110962,
      "epoch": 0.21282232426704345,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998250153199822,
      "loss": 6.1484,
      "mean_token_accuracy": 0.15403092205524443,
      "num_tokens": 4495985.0,
      "step": 2410
    },
    {
      "entropy": 6.466254568099975,
      "epoch": 0.21326386435888378,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998237713720036,
      "loss": 6.3239,
      "mean_token_accuracy": 0.14338542893528938,
      "num_tokens": 4504944.0,
      "step": 2415
    },
    {
      "entropy": 6.381775999069214,
      "epoch": 0.2137054044507241,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998225230198552,
      "loss": 6.1987,
      "mean_token_accuracy": 0.1538494348526001,
      "num_tokens": 4515402.0,
      "step": 2420
    },
    {
      "entropy": 6.330594635009765,
      "epoch": 0.21414694454256447,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998212702635614,
      "loss": 6.2685,
      "mean_token_accuracy": 0.14354601204395295,
      "num_tokens": 4525009.0,
      "step": 2425
    },
    {
      "entropy": 6.5298998832702635,
      "epoch": 0.2145884846344048,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004998200131031469,
      "loss": 6.3256,
      "mean_token_accuracy": 0.14529723376035691,
      "num_tokens": 4534460.0,
      "step": 2430
    },
    {
      "entropy": 6.389695167541504,
      "epoch": 0.21503002472624513,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998187515386361,
      "loss": 6.1518,
      "mean_token_accuracy": 0.1601933628320694,
      "num_tokens": 4543748.0,
      "step": 2435
    },
    {
      "entropy": 6.282786703109741,
      "epoch": 0.2154715648180855,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004998174855700538,
      "loss": 6.2753,
      "mean_token_accuracy": 0.15319369435310365,
      "num_tokens": 4552722.0,
      "step": 2440
    },
    {
      "entropy": 6.487105512619019,
      "epoch": 0.21591310490992582,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004998162151974248,
      "loss": 6.1183,
      "mean_token_accuracy": 0.15194425731897354,
      "num_tokens": 4561607.0,
      "step": 2445
    },
    {
      "entropy": 6.372227382659912,
      "epoch": 0.21635464500176615,
      "grad_norm": 0.859375,
      "learning_rate": 0.000499814940420774,
      "loss": 6.4008,
      "mean_token_accuracy": 0.13625949174165725,
      "num_tokens": 4572524.0,
      "step": 2450
    },
    {
      "entropy": 6.387830066680908,
      "epoch": 0.21679618509360649,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998136612401266,
      "loss": 6.2112,
      "mean_token_accuracy": 0.1517648920416832,
      "num_tokens": 4581601.0,
      "step": 2455
    },
    {
      "entropy": 6.404532241821289,
      "epoch": 0.21723772518544684,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998123776555071,
      "loss": 6.2609,
      "mean_token_accuracy": 0.1447027012705803,
      "num_tokens": 4591795.0,
      "step": 2460
    },
    {
      "entropy": 6.37783408164978,
      "epoch": 0.21767926527728718,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998110896669412,
      "loss": 6.2334,
      "mean_token_accuracy": 0.14586359560489653,
      "num_tokens": 4600745.0,
      "step": 2465
    },
    {
      "entropy": 6.4204151153564455,
      "epoch": 0.2181208053691275,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998097972744539,
      "loss": 6.2744,
      "mean_token_accuracy": 0.1438829004764557,
      "num_tokens": 4610490.0,
      "step": 2470
    },
    {
      "entropy": 6.344004535675049,
      "epoch": 0.21856234546096787,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998085004780705,
      "loss": 6.1917,
      "mean_token_accuracy": 0.1555175483226776,
      "num_tokens": 4619511.0,
      "step": 2475
    },
    {
      "entropy": 6.415195465087891,
      "epoch": 0.2190038855528082,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998071992778164,
      "loss": 6.2515,
      "mean_token_accuracy": 0.14207181110978126,
      "num_tokens": 4628186.0,
      "step": 2480
    },
    {
      "entropy": 6.438273286819458,
      "epoch": 0.21944542564464853,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499805893673717,
      "loss": 6.2691,
      "mean_token_accuracy": 0.14731226563453675,
      "num_tokens": 4637431.0,
      "step": 2485
    },
    {
      "entropy": 6.356544446945191,
      "epoch": 0.21988696573648886,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998045836657982,
      "loss": 6.1291,
      "mean_token_accuracy": 0.15215693414211273,
      "num_tokens": 4646627.0,
      "step": 2490
    },
    {
      "entropy": 6.363220739364624,
      "epoch": 0.22032850582832922,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998032692540853,
      "loss": 6.2228,
      "mean_token_accuracy": 0.14926744401454925,
      "num_tokens": 4656095.0,
      "step": 2495
    },
    {
      "entropy": 6.444636392593384,
      "epoch": 0.22077004592016955,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998019504386044,
      "loss": 6.2983,
      "mean_token_accuracy": 0.1443503238260746,
      "num_tokens": 4665807.0,
      "step": 2500
    },
    {
      "entropy": 6.39345760345459,
      "epoch": 0.22121158601200988,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998006272193809,
      "loss": 6.2515,
      "mean_token_accuracy": 0.14276793897151946,
      "num_tokens": 4674459.0,
      "step": 2505
    },
    {
      "entropy": 6.420822715759277,
      "epoch": 0.22165312610385024,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004997992995964412,
      "loss": 6.3335,
      "mean_token_accuracy": 0.14485810697078705,
      "num_tokens": 4684063.0,
      "step": 2510
    },
    {
      "entropy": 6.430441427230835,
      "epoch": 0.22209466619569057,
      "grad_norm": 1.125,
      "learning_rate": 0.0004997979675698109,
      "loss": 6.2237,
      "mean_token_accuracy": 0.1505151480436325,
      "num_tokens": 4692807.0,
      "step": 2515
    },
    {
      "entropy": 6.493818712234497,
      "epoch": 0.2225362062875309,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997966311395164,
      "loss": 6.1509,
      "mean_token_accuracy": 0.15266127437353133,
      "num_tokens": 4701100.0,
      "step": 2520
    },
    {
      "entropy": 6.261106777191162,
      "epoch": 0.22297774637937123,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997952903055836,
      "loss": 6.2014,
      "mean_token_accuracy": 0.1452627256512642,
      "num_tokens": 4710697.0,
      "step": 2525
    },
    {
      "entropy": 6.3538493633270265,
      "epoch": 0.2234192864712116,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499793945068039,
      "loss": 6.1951,
      "mean_token_accuracy": 0.14937613308429717,
      "num_tokens": 4718745.0,
      "step": 2530
    },
    {
      "entropy": 6.333649778366089,
      "epoch": 0.22386082656305192,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997925954269088,
      "loss": 6.173,
      "mean_token_accuracy": 0.15181030780076982,
      "num_tokens": 4728056.0,
      "step": 2535
    },
    {
      "entropy": 6.434943294525146,
      "epoch": 0.22430236665489225,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004997912413822196,
      "loss": 6.2935,
      "mean_token_accuracy": 0.15063572451472282,
      "num_tokens": 4737605.0,
      "step": 2540
    },
    {
      "entropy": 6.293275022506714,
      "epoch": 0.2247439067467326,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997898829339979,
      "loss": 6.114,
      "mean_token_accuracy": 0.16079342663288115,
      "num_tokens": 4746168.0,
      "step": 2545
    },
    {
      "entropy": 6.297460889816284,
      "epoch": 0.22518544683857294,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049978852008227,
      "loss": 6.1821,
      "mean_token_accuracy": 0.1518242985010147,
      "num_tokens": 4755072.0,
      "step": 2550
    },
    {
      "entropy": 6.327752351760864,
      "epoch": 0.22562698693041328,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000499787152827063,
      "loss": 6.192,
      "mean_token_accuracy": 0.15824129432439804,
      "num_tokens": 4764580.0,
      "step": 2555
    },
    {
      "entropy": 6.290549898147583,
      "epoch": 0.2260685270222536,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997857811684035,
      "loss": 6.2145,
      "mean_token_accuracy": 0.15303274542093276,
      "num_tokens": 4774135.0,
      "step": 2560
    },
    {
      "entropy": 6.410637426376343,
      "epoch": 0.22651006711409397,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004997844051063183,
      "loss": 6.3043,
      "mean_token_accuracy": 0.14438919797539712,
      "num_tokens": 4784733.0,
      "step": 2565
    },
    {
      "entropy": 6.41192512512207,
      "epoch": 0.2269516072059343,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997830246408346,
      "loss": 6.2387,
      "mean_token_accuracy": 0.15428498163819312,
      "num_tokens": 4795327.0,
      "step": 2570
    },
    {
      "entropy": 6.278723859786988,
      "epoch": 0.22739314729777463,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997816397719791,
      "loss": 6.2157,
      "mean_token_accuracy": 0.14635045304894448,
      "num_tokens": 4804314.0,
      "step": 2575
    },
    {
      "entropy": 6.441279935836792,
      "epoch": 0.227834687389615,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997802504997792,
      "loss": 6.303,
      "mean_token_accuracy": 0.14009749442338942,
      "num_tokens": 4813637.0,
      "step": 2580
    },
    {
      "entropy": 6.366417980194091,
      "epoch": 0.22827622748145532,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004997788568242621,
      "loss": 6.1695,
      "mean_token_accuracy": 0.15038952380418777,
      "num_tokens": 4823094.0,
      "step": 2585
    },
    {
      "entropy": 6.292829847335815,
      "epoch": 0.22871776757329565,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000499777458745455,
      "loss": 6.1123,
      "mean_token_accuracy": 0.1579282909631729,
      "num_tokens": 4833199.0,
      "step": 2590
    },
    {
      "entropy": 6.3866489887237545,
      "epoch": 0.22915930766513598,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004997760562633853,
      "loss": 6.2102,
      "mean_token_accuracy": 0.15025776177644729,
      "num_tokens": 4842970.0,
      "step": 2595
    },
    {
      "entropy": 6.37343578338623,
      "epoch": 0.22960084775697634,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997746493780804,
      "loss": 6.2792,
      "mean_token_accuracy": 0.14459849298000335,
      "num_tokens": 4852043.0,
      "step": 2600
    },
    {
      "entropy": 6.356409358978271,
      "epoch": 0.23004238784881667,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499773238089568,
      "loss": 6.1844,
      "mean_token_accuracy": 0.14483380317687988,
      "num_tokens": 4862232.0,
      "step": 2605
    },
    {
      "entropy": 6.343675518035889,
      "epoch": 0.230483927940657,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997718223978758,
      "loss": 6.1204,
      "mean_token_accuracy": 0.14981301128864288,
      "num_tokens": 4871186.0,
      "step": 2610
    },
    {
      "entropy": 6.302071237564087,
      "epoch": 0.23092546803249736,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997704023030315,
      "loss": 6.2134,
      "mean_token_accuracy": 0.1568485602736473,
      "num_tokens": 4879974.0,
      "step": 2615
    },
    {
      "entropy": 6.37543830871582,
      "epoch": 0.2313670081243377,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997689778050627,
      "loss": 6.2599,
      "mean_token_accuracy": 0.15443349331617356,
      "num_tokens": 4890300.0,
      "step": 2620
    },
    {
      "entropy": 6.287172794342041,
      "epoch": 0.23180854821617802,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997675489039975,
      "loss": 6.2494,
      "mean_token_accuracy": 0.1468710109591484,
      "num_tokens": 4900428.0,
      "step": 2625
    },
    {
      "entropy": 6.366749954223633,
      "epoch": 0.23225008830801835,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997661155998638,
      "loss": 6.2476,
      "mean_token_accuracy": 0.15147435963153838,
      "num_tokens": 4910092.0,
      "step": 2630
    },
    {
      "entropy": 6.374930906295776,
      "epoch": 0.2326916283998587,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997646778926898,
      "loss": 6.2313,
      "mean_token_accuracy": 0.14748895019292832,
      "num_tokens": 4919593.0,
      "step": 2635
    },
    {
      "entropy": 6.320575475692749,
      "epoch": 0.23313316849169904,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004997632357825035,
      "loss": 6.1889,
      "mean_token_accuracy": 0.15245203971862792,
      "num_tokens": 4929098.0,
      "step": 2640
    },
    {
      "entropy": 6.37444224357605,
      "epoch": 0.23357470858353938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997617892693333,
      "loss": 6.2685,
      "mean_token_accuracy": 0.14796690195798873,
      "num_tokens": 4938265.0,
      "step": 2645
    },
    {
      "entropy": 6.381702041625976,
      "epoch": 0.23401624867537973,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997603383532075,
      "loss": 6.2157,
      "mean_token_accuracy": 0.15212348997592925,
      "num_tokens": 4946694.0,
      "step": 2650
    },
    {
      "entropy": 6.345135974884033,
      "epoch": 0.23445778876722007,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997588830341545,
      "loss": 6.2007,
      "mean_token_accuracy": 0.1495561882853508,
      "num_tokens": 4955296.0,
      "step": 2655
    },
    {
      "entropy": 6.262336206436157,
      "epoch": 0.2348993288590604,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997574233122028,
      "loss": 6.1783,
      "mean_token_accuracy": 0.15376680493354797,
      "num_tokens": 4964409.0,
      "step": 2660
    },
    {
      "entropy": 6.344006538391113,
      "epoch": 0.23534086895090076,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997559591873809,
      "loss": 6.1848,
      "mean_token_accuracy": 0.15347870737314223,
      "num_tokens": 4973449.0,
      "step": 2665
    },
    {
      "entropy": 6.325773239135742,
      "epoch": 0.2357824090427411,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997544906597178,
      "loss": 6.1774,
      "mean_token_accuracy": 0.1539867863059044,
      "num_tokens": 4983057.0,
      "step": 2670
    },
    {
      "entropy": 6.2636034965515135,
      "epoch": 0.23622394913458142,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004997530177292418,
      "loss": 6.2376,
      "mean_token_accuracy": 0.14284619241952895,
      "num_tokens": 4991950.0,
      "step": 2675
    },
    {
      "entropy": 6.425727081298828,
      "epoch": 0.23666548922642175,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997515403959823,
      "loss": 6.1908,
      "mean_token_accuracy": 0.15085240155458451,
      "num_tokens": 5001042.0,
      "step": 2680
    },
    {
      "entropy": 6.364514875411987,
      "epoch": 0.2371070293182621,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997500586599677,
      "loss": 6.1239,
      "mean_token_accuracy": 0.15567631870508195,
      "num_tokens": 5009827.0,
      "step": 2685
    },
    {
      "entropy": 6.238472986221313,
      "epoch": 0.23754856941010244,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004997485725212274,
      "loss": 6.1721,
      "mean_token_accuracy": 0.15332650095224382,
      "num_tokens": 5018708.0,
      "step": 2690
    },
    {
      "entropy": 6.277514266967773,
      "epoch": 0.23799010950194277,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997470819797903,
      "loss": 6.0688,
      "mean_token_accuracy": 0.1563563570380211,
      "num_tokens": 5027522.0,
      "step": 2695
    },
    {
      "entropy": 6.287735462188721,
      "epoch": 0.23843164959378313,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997455870356857,
      "loss": 6.1948,
      "mean_token_accuracy": 0.15198080092668534,
      "num_tokens": 5035755.0,
      "step": 2700
    },
    {
      "entropy": 6.307868242263794,
      "epoch": 0.23887318968562346,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997440876889429,
      "loss": 6.1377,
      "mean_token_accuracy": 0.15334794968366622,
      "num_tokens": 5045289.0,
      "step": 2705
    },
    {
      "entropy": 6.278635931015015,
      "epoch": 0.2393147297774638,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004997425839395913,
      "loss": 6.1641,
      "mean_token_accuracy": 0.1568767547607422,
      "num_tokens": 5053774.0,
      "step": 2710
    },
    {
      "entropy": 6.4377528667449955,
      "epoch": 0.23975626986930412,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004997410757876602,
      "loss": 6.2444,
      "mean_token_accuracy": 0.14437243491411209,
      "num_tokens": 5062911.0,
      "step": 2715
    },
    {
      "entropy": 6.3234247207641605,
      "epoch": 0.24019780996114448,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997395632331793,
      "loss": 6.1091,
      "mean_token_accuracy": 0.15637961626052857,
      "num_tokens": 5072107.0,
      "step": 2720
    },
    {
      "entropy": 6.226778221130371,
      "epoch": 0.2406393500529848,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997380462761781,
      "loss": 6.0586,
      "mean_token_accuracy": 0.15822814106941224,
      "num_tokens": 5080588.0,
      "step": 2725
    },
    {
      "entropy": 6.362985134124756,
      "epoch": 0.24108089014482514,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004997365249166864,
      "loss": 6.2812,
      "mean_token_accuracy": 0.148160120844841,
      "num_tokens": 5090262.0,
      "step": 2730
    },
    {
      "entropy": 6.364387369155883,
      "epoch": 0.2415224302366655,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997349991547342,
      "loss": 6.1672,
      "mean_token_accuracy": 0.15980131253600122,
      "num_tokens": 5099285.0,
      "step": 2735
    },
    {
      "entropy": 6.311175918579101,
      "epoch": 0.24196397032850583,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997334689903509,
      "loss": 6.2297,
      "mean_token_accuracy": 0.1506420761346817,
      "num_tokens": 5109115.0,
      "step": 2740
    },
    {
      "entropy": 6.381357097625733,
      "epoch": 0.24240551042034617,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004997319344235668,
      "loss": 6.2471,
      "mean_token_accuracy": 0.15130761936306952,
      "num_tokens": 5117977.0,
      "step": 2745
    },
    {
      "entropy": 6.371263027191162,
      "epoch": 0.2428470505121865,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499730395454412,
      "loss": 6.1953,
      "mean_token_accuracy": 0.15422819852828978,
      "num_tokens": 5127457.0,
      "step": 2750
    },
    {
      "entropy": 6.326935291290283,
      "epoch": 0.24328859060402686,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997288520829166,
      "loss": 6.2614,
      "mean_token_accuracy": 0.14421235620975495,
      "num_tokens": 5137310.0,
      "step": 2755
    },
    {
      "entropy": 6.359060192108155,
      "epoch": 0.2437301306958672,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997273043091107,
      "loss": 6.1923,
      "mean_token_accuracy": 0.1464222900569439,
      "num_tokens": 5146963.0,
      "step": 2760
    },
    {
      "entropy": 6.268268156051636,
      "epoch": 0.24417167078770752,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997257521330248,
      "loss": 6.1424,
      "mean_token_accuracy": 0.14788112491369249,
      "num_tokens": 5155521.0,
      "step": 2765
    },
    {
      "entropy": 6.351244401931763,
      "epoch": 0.24461321087954788,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997241955546892,
      "loss": 6.1424,
      "mean_token_accuracy": 0.1483145385980606,
      "num_tokens": 5165182.0,
      "step": 2770
    },
    {
      "entropy": 6.31810302734375,
      "epoch": 0.2450547509713882,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997226345741343,
      "loss": 6.1684,
      "mean_token_accuracy": 0.15033391118049622,
      "num_tokens": 5175511.0,
      "step": 2775
    },
    {
      "entropy": 6.3378277778625485,
      "epoch": 0.24549629106322854,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499721069191391,
      "loss": 6.1013,
      "mean_token_accuracy": 0.1578337162733078,
      "num_tokens": 5184772.0,
      "step": 2780
    },
    {
      "entropy": 6.310634183883667,
      "epoch": 0.24593783115506887,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004997194994064896,
      "loss": 6.1866,
      "mean_token_accuracy": 0.1496903583407402,
      "num_tokens": 5195136.0,
      "step": 2785
    },
    {
      "entropy": 6.332416534423828,
      "epoch": 0.24637937124690923,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499717925219461,
      "loss": 6.1167,
      "mean_token_accuracy": 0.15987685918807984,
      "num_tokens": 5203163.0,
      "step": 2790
    },
    {
      "entropy": 6.345263051986694,
      "epoch": 0.24682091133874956,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997163466303362,
      "loss": 6.2061,
      "mean_token_accuracy": 0.1503751039505005,
      "num_tokens": 5213233.0,
      "step": 2795
    },
    {
      "entropy": 6.292828702926636,
      "epoch": 0.2472624514305899,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499714763639146,
      "loss": 6.1372,
      "mean_token_accuracy": 0.15086597949266434,
      "num_tokens": 5222940.0,
      "step": 2800
    },
    {
      "entropy": 6.270041036605835,
      "epoch": 0.24770399152243025,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997131762459211,
      "loss": 6.1676,
      "mean_token_accuracy": 0.15103059709072114,
      "num_tokens": 5232263.0,
      "step": 2805
    },
    {
      "entropy": 6.331594228744507,
      "epoch": 0.24814553161427058,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997115844506932,
      "loss": 6.1405,
      "mean_token_accuracy": 0.15087736994028092,
      "num_tokens": 5241536.0,
      "step": 2810
    },
    {
      "entropy": 6.34823522567749,
      "epoch": 0.2485870717061109,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997099882534929,
      "loss": 6.1704,
      "mean_token_accuracy": 0.14839458093047142,
      "num_tokens": 5250702.0,
      "step": 2815
    },
    {
      "entropy": 6.377634048461914,
      "epoch": 0.24902861179795124,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004997083876543519,
      "loss": 6.1759,
      "mean_token_accuracy": 0.1505747467279434,
      "num_tokens": 5259811.0,
      "step": 2820
    },
    {
      "entropy": 6.377023410797119,
      "epoch": 0.2494701518897916,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004997067826533014,
      "loss": 6.2543,
      "mean_token_accuracy": 0.1441645845770836,
      "num_tokens": 5270518.0,
      "step": 2825
    },
    {
      "entropy": 6.403309440612793,
      "epoch": 0.24991169198163193,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004997051732503726,
      "loss": 6.1311,
      "mean_token_accuracy": 0.15377675592899323,
      "num_tokens": 5279538.0,
      "step": 2830
    },
    {
      "entropy": 6.22372031211853,
      "epoch": 0.25035323207347226,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004997035594455975,
      "loss": 6.1674,
      "mean_token_accuracy": 0.1481279417872429,
      "num_tokens": 5289633.0,
      "step": 2835
    },
    {
      "entropy": 6.387079381942749,
      "epoch": 0.2507947721653126,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004997019412390074,
      "loss": 6.2575,
      "mean_token_accuracy": 0.15361315459012986,
      "num_tokens": 5299148.0,
      "step": 2840
    },
    {
      "entropy": 6.307563066482544,
      "epoch": 0.2512363122571529,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499700318630634,
      "loss": 6.1357,
      "mean_token_accuracy": 0.15210598707199097,
      "num_tokens": 5309090.0,
      "step": 2845
    },
    {
      "entropy": 6.273115110397339,
      "epoch": 0.2516778523489933,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996986916205092,
      "loss": 6.2278,
      "mean_token_accuracy": 0.15018689632415771,
      "num_tokens": 5318798.0,
      "step": 2850
    },
    {
      "entropy": 6.345541620254517,
      "epoch": 0.25211939244083365,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996970602086648,
      "loss": 6.0964,
      "mean_token_accuracy": 0.15610066950321197,
      "num_tokens": 5327915.0,
      "step": 2855
    },
    {
      "entropy": 6.211011552810669,
      "epoch": 0.252560932532674,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996954243951327,
      "loss": 6.1209,
      "mean_token_accuracy": 0.15709759593009948,
      "num_tokens": 5336970.0,
      "step": 2860
    },
    {
      "entropy": 6.323676109313965,
      "epoch": 0.2530024726245143,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004996937841799451,
      "loss": 6.0769,
      "mean_token_accuracy": 0.15997228324413298,
      "num_tokens": 5345167.0,
      "step": 2865
    },
    {
      "entropy": 6.199875545501709,
      "epoch": 0.25344401271635464,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004996921395631342,
      "loss": 6.071,
      "mean_token_accuracy": 0.1565924033522606,
      "num_tokens": 5353399.0,
      "step": 2870
    },
    {
      "entropy": 6.27602071762085,
      "epoch": 0.25388555280819497,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499690490544732,
      "loss": 6.1828,
      "mean_token_accuracy": 0.1513873428106308,
      "num_tokens": 5363224.0,
      "step": 2875
    },
    {
      "entropy": 6.339487314224243,
      "epoch": 0.2543270929000353,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996888371247707,
      "loss": 6.1923,
      "mean_token_accuracy": 0.14602998048067092,
      "num_tokens": 5372274.0,
      "step": 2880
    },
    {
      "entropy": 6.343375635147095,
      "epoch": 0.2547686329918757,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499687179303283,
      "loss": 6.1815,
      "mean_token_accuracy": 0.15472851544618607,
      "num_tokens": 5380240.0,
      "step": 2885
    },
    {
      "entropy": 6.228068780899048,
      "epoch": 0.255210173083716,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996855170803012,
      "loss": 6.0344,
      "mean_token_accuracy": 0.15991738438606262,
      "num_tokens": 5389390.0,
      "step": 2890
    },
    {
      "entropy": 6.28302755355835,
      "epoch": 0.25565171317555635,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996838504558581,
      "loss": 6.1902,
      "mean_token_accuracy": 0.1521752417087555,
      "num_tokens": 5399425.0,
      "step": 2895
    },
    {
      "entropy": 6.355008220672607,
      "epoch": 0.2560932532673967,
      "grad_norm": 0.96875,
      "learning_rate": 0.000499682179429986,
      "loss": 6.2011,
      "mean_token_accuracy": 0.1508815959095955,
      "num_tokens": 5408717.0,
      "step": 2900
    },
    {
      "entropy": 6.280868196487427,
      "epoch": 0.256534793359237,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996805040027178,
      "loss": 6.1368,
      "mean_token_accuracy": 0.14726324751973152,
      "num_tokens": 5418475.0,
      "step": 2905
    },
    {
      "entropy": 6.2741169929504395,
      "epoch": 0.25697633345107734,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004996788241740863,
      "loss": 6.1707,
      "mean_token_accuracy": 0.1536627657711506,
      "num_tokens": 5428403.0,
      "step": 2910
    },
    {
      "entropy": 6.269294166564942,
      "epoch": 0.2574178735429177,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996771399441243,
      "loss": 6.2226,
      "mean_token_accuracy": 0.14502449780702592,
      "num_tokens": 5437347.0,
      "step": 2915
    },
    {
      "entropy": 6.407036876678466,
      "epoch": 0.25785941363475806,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004996754513128652,
      "loss": 6.1309,
      "mean_token_accuracy": 0.15732353776693345,
      "num_tokens": 5446804.0,
      "step": 2920
    },
    {
      "entropy": 6.222485494613648,
      "epoch": 0.2583009537265984,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996737582803416,
      "loss": 6.0819,
      "mean_token_accuracy": 0.15406210720539093,
      "num_tokens": 5455888.0,
      "step": 2925
    },
    {
      "entropy": 6.3086316108703615,
      "epoch": 0.2587424938184387,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996720608465868,
      "loss": 6.0772,
      "mean_token_accuracy": 0.15540871918201446,
      "num_tokens": 5463977.0,
      "step": 2930
    },
    {
      "entropy": 6.277329921722412,
      "epoch": 0.25918403391027905,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004996703590116342,
      "loss": 6.1841,
      "mean_token_accuracy": 0.14695459753274917,
      "num_tokens": 5473780.0,
      "step": 2935
    },
    {
      "entropy": 6.239596223831176,
      "epoch": 0.2596255740021194,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996686527755171,
      "loss": 6.0776,
      "mean_token_accuracy": 0.15587782710790635,
      "num_tokens": 5482151.0,
      "step": 2940
    },
    {
      "entropy": 6.316288614273072,
      "epoch": 0.2600671140939597,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996669421382687,
      "loss": 6.0935,
      "mean_token_accuracy": 0.16187871843576432,
      "num_tokens": 5491103.0,
      "step": 2945
    },
    {
      "entropy": 6.257398414611816,
      "epoch": 0.26050865418580005,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996652270999228,
      "loss": 6.1026,
      "mean_token_accuracy": 0.1524055764079094,
      "num_tokens": 5500367.0,
      "step": 2950
    },
    {
      "entropy": 6.286735200881958,
      "epoch": 0.26095019427764043,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996635076605128,
      "loss": 6.1432,
      "mean_token_accuracy": 0.15162961781024933,
      "num_tokens": 5509631.0,
      "step": 2955
    },
    {
      "entropy": 6.3608544826507565,
      "epoch": 0.26139173436948077,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996617838200725,
      "loss": 6.1488,
      "mean_token_accuracy": 0.1510395437479019,
      "num_tokens": 5518635.0,
      "step": 2960
    },
    {
      "entropy": 6.2053210735321045,
      "epoch": 0.2618332744613211,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004996600555786357,
      "loss": 6.1166,
      "mean_token_accuracy": 0.1505479171872139,
      "num_tokens": 5527696.0,
      "step": 2965
    },
    {
      "entropy": 6.343273687362671,
      "epoch": 0.26227481455316143,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996583229362362,
      "loss": 6.077,
      "mean_token_accuracy": 0.15462881475687026,
      "num_tokens": 5536632.0,
      "step": 2970
    },
    {
      "entropy": 6.356432819366455,
      "epoch": 0.26271635464500176,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996565858929078,
      "loss": 6.1634,
      "mean_token_accuracy": 0.15574948489665985,
      "num_tokens": 5545825.0,
      "step": 2975
    },
    {
      "entropy": 6.2175110340118405,
      "epoch": 0.2631578947368421,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004996548444486847,
      "loss": 6.0344,
      "mean_token_accuracy": 0.1576728031039238,
      "num_tokens": 5555158.0,
      "step": 2980
    },
    {
      "entropy": 6.174654150009156,
      "epoch": 0.2635994348286824,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004996530986036008,
      "loss": 5.9889,
      "mean_token_accuracy": 0.16068323999643325,
      "num_tokens": 5564218.0,
      "step": 2985
    },
    {
      "entropy": 6.207144069671631,
      "epoch": 0.2640409749205228,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004996513483576907,
      "loss": 6.1208,
      "mean_token_accuracy": 0.15933286994695664,
      "num_tokens": 5572760.0,
      "step": 2990
    },
    {
      "entropy": 6.368608522415161,
      "epoch": 0.26448251501236314,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996495937109884,
      "loss": 6.1754,
      "mean_token_accuracy": 0.14925839453935624,
      "num_tokens": 5581660.0,
      "step": 2995
    },
    {
      "entropy": 6.248563766479492,
      "epoch": 0.26492405510420347,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996478346635283,
      "loss": 6.0906,
      "mean_token_accuracy": 0.15518097579479218,
      "num_tokens": 5590664.0,
      "step": 3000
    },
    {
      "epoch": 0.26492405510420347,
      "eval_entropy": 5.978648100514914,
      "eval_loss": 6.153229713439941,
      "eval_mean_token_accuracy": 0.1595988018887533,
      "eval_num_tokens": 5590664.0,
      "eval_runtime": 20.3151,
      "eval_samples_per_second": 1738.363,
      "eval_steps_per_second": 217.326,
      "step": 3000
    },
    {
      "entropy": 6.155827331542969,
      "epoch": 0.2653655951960438,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996460712153448,
      "loss": 5.9704,
      "mean_token_accuracy": 0.16879431158304214,
      "num_tokens": 5598727.0,
      "step": 3005
    },
    {
      "entropy": 6.174825668334961,
      "epoch": 0.26580713528788413,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004996443033664726,
      "loss": 6.1576,
      "mean_token_accuracy": 0.15344538390636445,
      "num_tokens": 5608549.0,
      "step": 3010
    },
    {
      "entropy": 6.312784004211426,
      "epoch": 0.26624867537972446,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996425311169463,
      "loss": 6.0951,
      "mean_token_accuracy": 0.16055944710969924,
      "num_tokens": 5617628.0,
      "step": 3015
    },
    {
      "entropy": 6.304688596725464,
      "epoch": 0.2666902154715648,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996407544668005,
      "loss": 6.1541,
      "mean_token_accuracy": 0.1486222580075264,
      "num_tokens": 5627182.0,
      "step": 3020
    },
    {
      "entropy": 6.258115196228028,
      "epoch": 0.2671317555634052,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996389734160701,
      "loss": 6.1827,
      "mean_token_accuracy": 0.15313406884670258,
      "num_tokens": 5636413.0,
      "step": 3025
    },
    {
      "entropy": 6.3286293029785154,
      "epoch": 0.2675732956552455,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00049963718796479,
      "loss": 6.1165,
      "mean_token_accuracy": 0.15353906974196435,
      "num_tokens": 5646025.0,
      "step": 3030
    },
    {
      "entropy": 6.2249290466308596,
      "epoch": 0.26801483574708584,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996353981129952,
      "loss": 6.1401,
      "mean_token_accuracy": 0.15831294059753417,
      "num_tokens": 5656577.0,
      "step": 3035
    },
    {
      "entropy": 6.266787099838257,
      "epoch": 0.2684563758389262,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004996336038607206,
      "loss": 6.1096,
      "mean_token_accuracy": 0.15671551674604417,
      "num_tokens": 5665163.0,
      "step": 3040
    },
    {
      "entropy": 6.256934928894043,
      "epoch": 0.2688979159307665,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996318052080015,
      "loss": 6.1384,
      "mean_token_accuracy": 0.1572331577539444,
      "num_tokens": 5675196.0,
      "step": 3045
    },
    {
      "entropy": 6.291875505447388,
      "epoch": 0.26933945602260684,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996300021548731,
      "loss": 6.0604,
      "mean_token_accuracy": 0.1568915903568268,
      "num_tokens": 5684361.0,
      "step": 3050
    },
    {
      "entropy": 6.193565845489502,
      "epoch": 0.26978099611444717,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996281947013707,
      "loss": 6.063,
      "mean_token_accuracy": 0.15895331352949144,
      "num_tokens": 5694187.0,
      "step": 3055
    },
    {
      "entropy": 6.237375640869141,
      "epoch": 0.27022253620628756,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004996263828475298,
      "loss": 6.1115,
      "mean_token_accuracy": 0.14960144013166427,
      "num_tokens": 5702858.0,
      "step": 3060
    },
    {
      "entropy": 6.268642711639404,
      "epoch": 0.2706640762981279,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004996245665933857,
      "loss": 6.0305,
      "mean_token_accuracy": 0.16427914202213287,
      "num_tokens": 5712727.0,
      "step": 3065
    },
    {
      "entropy": 6.29313907623291,
      "epoch": 0.2711056163899682,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996227459389741,
      "loss": 6.1448,
      "mean_token_accuracy": 0.15978289842605592,
      "num_tokens": 5722904.0,
      "step": 3070
    },
    {
      "entropy": 6.289960145950317,
      "epoch": 0.27154715648180855,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996209208843307,
      "loss": 6.2333,
      "mean_token_accuracy": 0.15185119658708573,
      "num_tokens": 5732228.0,
      "step": 3075
    },
    {
      "entropy": 6.30490255355835,
      "epoch": 0.2719886965736489,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004996190914294912,
      "loss": 6.1829,
      "mean_token_accuracy": 0.1575074926018715,
      "num_tokens": 5743548.0,
      "step": 3080
    },
    {
      "entropy": 6.295343112945557,
      "epoch": 0.2724302366654892,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996172575744914,
      "loss": 6.1796,
      "mean_token_accuracy": 0.146508327126503,
      "num_tokens": 5754112.0,
      "step": 3085
    },
    {
      "entropy": 6.300486421585083,
      "epoch": 0.27287177675732954,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004996154193193673,
      "loss": 6.0018,
      "mean_token_accuracy": 0.16367141306400299,
      "num_tokens": 5763377.0,
      "step": 3090
    },
    {
      "entropy": 6.170904731750488,
      "epoch": 0.27331331684916993,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996135766641549,
      "loss": 5.9686,
      "mean_token_accuracy": 0.16288323402404786,
      "num_tokens": 5772000.0,
      "step": 3095
    },
    {
      "entropy": 6.214351940155029,
      "epoch": 0.27375485694101026,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996117296088903,
      "loss": 6.0917,
      "mean_token_accuracy": 0.1531004711985588,
      "num_tokens": 5782192.0,
      "step": 3100
    },
    {
      "entropy": 6.343245601654052,
      "epoch": 0.2741963970328506,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004996098781536095,
      "loss": 6.1403,
      "mean_token_accuracy": 0.15874934047460557,
      "num_tokens": 5791163.0,
      "step": 3105
    },
    {
      "entropy": 6.304137182235718,
      "epoch": 0.2746379371246909,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996080222983492,
      "loss": 6.1539,
      "mean_token_accuracy": 0.15088921338319777,
      "num_tokens": 5801815.0,
      "step": 3110
    },
    {
      "entropy": 6.357846736907959,
      "epoch": 0.27507947721653125,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996061620431453,
      "loss": 6.1611,
      "mean_token_accuracy": 0.152420574426651,
      "num_tokens": 5811545.0,
      "step": 3115
    },
    {
      "entropy": 6.260142993927002,
      "epoch": 0.2755210173083716,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004996042973880344,
      "loss": 6.1806,
      "mean_token_accuracy": 0.15066104382276535,
      "num_tokens": 5820285.0,
      "step": 3120
    },
    {
      "entropy": 6.208466100692749,
      "epoch": 0.2759625574002119,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004996024283330532,
      "loss": 6.0189,
      "mean_token_accuracy": 0.15600398033857346,
      "num_tokens": 5829284.0,
      "step": 3125
    },
    {
      "entropy": 6.209367752075195,
      "epoch": 0.2764040974920523,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499600554878238,
      "loss": 6.1824,
      "mean_token_accuracy": 0.1472199648618698,
      "num_tokens": 5838973.0,
      "step": 3130
    },
    {
      "entropy": 6.292955923080444,
      "epoch": 0.27684563758389263,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004995986770236258,
      "loss": 6.0744,
      "mean_token_accuracy": 0.1584090195596218,
      "num_tokens": 5847434.0,
      "step": 3135
    },
    {
      "entropy": 6.26265139579773,
      "epoch": 0.27728717767573297,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995967947692533,
      "loss": 6.0732,
      "mean_token_accuracy": 0.1618253469467163,
      "num_tokens": 5856464.0,
      "step": 3140
    },
    {
      "entropy": 6.268027257919312,
      "epoch": 0.2777287177675733,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004995949081151571,
      "loss": 6.1406,
      "mean_token_accuracy": 0.15451844334602355,
      "num_tokens": 5865892.0,
      "step": 3145
    },
    {
      "entropy": 6.259841060638427,
      "epoch": 0.2781702578594136,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995930170613746,
      "loss": 6.107,
      "mean_token_accuracy": 0.16154323220252992,
      "num_tokens": 5874606.0,
      "step": 3150
    },
    {
      "entropy": 6.247693395614624,
      "epoch": 0.27861179795125396,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995911216079425,
      "loss": 6.0967,
      "mean_token_accuracy": 0.15935523062944412,
      "num_tokens": 5883853.0,
      "step": 3155
    },
    {
      "entropy": 6.314785385131836,
      "epoch": 0.2790533380430943,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004995892217548981,
      "loss": 6.1328,
      "mean_token_accuracy": 0.15520500987768174,
      "num_tokens": 5892841.0,
      "step": 3160
    },
    {
      "entropy": 6.289192247390747,
      "epoch": 0.2794948781349347,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995873175022786,
      "loss": 6.0993,
      "mean_token_accuracy": 0.15804312825202943,
      "num_tokens": 5901966.0,
      "step": 3165
    },
    {
      "entropy": 6.273761415481568,
      "epoch": 0.279936418226775,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004995854088501213,
      "loss": 6.0034,
      "mean_token_accuracy": 0.1539376214146614,
      "num_tokens": 5910003.0,
      "step": 3170
    },
    {
      "entropy": 6.177750110626221,
      "epoch": 0.28037795831861534,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995834957984634,
      "loss": 6.1634,
      "mean_token_accuracy": 0.15334712862968444,
      "num_tokens": 5920307.0,
      "step": 3175
    },
    {
      "entropy": 6.249305629730225,
      "epoch": 0.28081949841045567,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995815783473428,
      "loss": 6.0055,
      "mean_token_accuracy": 0.16042215675115584,
      "num_tokens": 5929875.0,
      "step": 3180
    },
    {
      "entropy": 6.254670238494873,
      "epoch": 0.281261038502296,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995796564967967,
      "loss": 5.9757,
      "mean_token_accuracy": 0.15762123465538025,
      "num_tokens": 5938570.0,
      "step": 3185
    },
    {
      "entropy": 6.167326974868774,
      "epoch": 0.28170257859413633,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995777302468628,
      "loss": 6.127,
      "mean_token_accuracy": 0.1609175220131874,
      "num_tokens": 5947693.0,
      "step": 3190
    },
    {
      "entropy": 6.24321608543396,
      "epoch": 0.28214411868597666,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995757995975789,
      "loss": 6.2097,
      "mean_token_accuracy": 0.15263804495334626,
      "num_tokens": 5957377.0,
      "step": 3195
    },
    {
      "entropy": 6.332187747955322,
      "epoch": 0.28258565877781705,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995738645489828,
      "loss": 6.1274,
      "mean_token_accuracy": 0.1612716645002365,
      "num_tokens": 5966443.0,
      "step": 3200
    },
    {
      "entropy": 6.200443315505981,
      "epoch": 0.2830271988696574,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004995719251011124,
      "loss": 6.0412,
      "mean_token_accuracy": 0.15479797273874282,
      "num_tokens": 5975027.0,
      "step": 3205
    },
    {
      "entropy": 6.266233110427857,
      "epoch": 0.2834687389614977,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995699812540058,
      "loss": 6.1083,
      "mean_token_accuracy": 0.1580372929573059,
      "num_tokens": 5983722.0,
      "step": 3210
    },
    {
      "entropy": 6.205762672424316,
      "epoch": 0.28391027905333804,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499568033007701,
      "loss": 6.1233,
      "mean_token_accuracy": 0.15123755782842635,
      "num_tokens": 5993358.0,
      "step": 3215
    },
    {
      "entropy": 6.283669900894165,
      "epoch": 0.2843518191451784,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004995660803622361,
      "loss": 6.0542,
      "mean_token_accuracy": 0.16230252236127854,
      "num_tokens": 6002743.0,
      "step": 3220
    },
    {
      "entropy": 6.320738506317139,
      "epoch": 0.2847933592370187,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004995641233176494,
      "loss": 6.2182,
      "mean_token_accuracy": 0.1505182772874832,
      "num_tokens": 6013727.0,
      "step": 3225
    },
    {
      "entropy": 6.313286638259887,
      "epoch": 0.28523489932885904,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995621618739792,
      "loss": 6.162,
      "mean_token_accuracy": 0.15084221735596656,
      "num_tokens": 6023397.0,
      "step": 3230
    },
    {
      "entropy": 6.2680644512176515,
      "epoch": 0.2856764394206994,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499560196031264,
      "loss": 6.0331,
      "mean_token_accuracy": 0.16098945140838622,
      "num_tokens": 6032453.0,
      "step": 3235
    },
    {
      "entropy": 6.175032663345337,
      "epoch": 0.28611797951253976,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004995582257895423,
      "loss": 6.02,
      "mean_token_accuracy": 0.15336361676454544,
      "num_tokens": 6042441.0,
      "step": 3240
    },
    {
      "entropy": 6.2402129650115965,
      "epoch": 0.2865595196043801,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995562511488528,
      "loss": 6.0108,
      "mean_token_accuracy": 0.1634404867887497,
      "num_tokens": 6050881.0,
      "step": 3245
    },
    {
      "entropy": 6.239614248275757,
      "epoch": 0.2870010596962204,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995542721092337,
      "loss": 6.0702,
      "mean_token_accuracy": 0.159418722987175,
      "num_tokens": 6060804.0,
      "step": 3250
    },
    {
      "entropy": 6.330715084075928,
      "epoch": 0.28744259978806075,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004995522886707244,
      "loss": 6.1609,
      "mean_token_accuracy": 0.15356300622224808,
      "num_tokens": 6070774.0,
      "step": 3255
    },
    {
      "entropy": 6.253082895278931,
      "epoch": 0.2878841398799011,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995503008333634,
      "loss": 6.1559,
      "mean_token_accuracy": 0.15287251621484757,
      "num_tokens": 6080158.0,
      "step": 3260
    },
    {
      "entropy": 6.263114023208618,
      "epoch": 0.2883256799717414,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995483085971897,
      "loss": 6.05,
      "mean_token_accuracy": 0.15794130861759187,
      "num_tokens": 6089183.0,
      "step": 3265
    },
    {
      "entropy": 6.209271335601807,
      "epoch": 0.2887672200635818,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995463119622424,
      "loss": 6.1544,
      "mean_token_accuracy": 0.15566999167203904,
      "num_tokens": 6098536.0,
      "step": 3270
    },
    {
      "entropy": 6.124404239654541,
      "epoch": 0.28920876015542213,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995443109285604,
      "loss": 5.9011,
      "mean_token_accuracy": 0.1713046222925186,
      "num_tokens": 6107745.0,
      "step": 3275
    },
    {
      "entropy": 6.262866878509522,
      "epoch": 0.28965030024726246,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004995423054961832,
      "loss": 6.0899,
      "mean_token_accuracy": 0.15983563363552095,
      "num_tokens": 6117512.0,
      "step": 3280
    },
    {
      "entropy": 6.150252914428711,
      "epoch": 0.2900918403391028,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049954029566515,
      "loss": 6.1128,
      "mean_token_accuracy": 0.16109101474285126,
      "num_tokens": 6126030.0,
      "step": 3285
    },
    {
      "entropy": 6.30514931678772,
      "epoch": 0.2905333804309431,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004995382814355,
      "loss": 6.1308,
      "mean_token_accuracy": 0.1520018458366394,
      "num_tokens": 6134888.0,
      "step": 3290
    },
    {
      "entropy": 6.29309344291687,
      "epoch": 0.29097492052278345,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004995362628072728,
      "loss": 6.0945,
      "mean_token_accuracy": 0.16048415303230285,
      "num_tokens": 6144274.0,
      "step": 3295
    },
    {
      "entropy": 6.152520227432251,
      "epoch": 0.2914164606146238,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995342397805078,
      "loss": 6.109,
      "mean_token_accuracy": 0.1570252701640129,
      "num_tokens": 6153406.0,
      "step": 3300
    },
    {
      "entropy": 6.257608938217163,
      "epoch": 0.29185800070646417,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004995322123552448,
      "loss": 6.0337,
      "mean_token_accuracy": 0.15978619009256362,
      "num_tokens": 6162743.0,
      "step": 3305
    },
    {
      "entropy": 6.219382572174072,
      "epoch": 0.2922995407983045,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995301805315235,
      "loss": 6.0253,
      "mean_token_accuracy": 0.16270643919706346,
      "num_tokens": 6171997.0,
      "step": 3310
    },
    {
      "entropy": 6.169751071929932,
      "epoch": 0.29274108089014483,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004995281443093837,
      "loss": 6.0379,
      "mean_token_accuracy": 0.16006233543157578,
      "num_tokens": 6181275.0,
      "step": 3315
    },
    {
      "entropy": 6.232631635665894,
      "epoch": 0.29318262098198516,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995261036888652,
      "loss": 6.1001,
      "mean_token_accuracy": 0.1559043176472187,
      "num_tokens": 6191640.0,
      "step": 3320
    },
    {
      "entropy": 6.329467058181763,
      "epoch": 0.2936241610738255,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995240586700081,
      "loss": 6.1076,
      "mean_token_accuracy": 0.15200517624616622,
      "num_tokens": 6201508.0,
      "step": 3325
    },
    {
      "entropy": 6.1505632400512695,
      "epoch": 0.2940657011656658,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004995220092528522,
      "loss": 5.9859,
      "mean_token_accuracy": 0.1673257976770401,
      "num_tokens": 6209902.0,
      "step": 3330
    },
    {
      "entropy": 6.255099248886109,
      "epoch": 0.29450724125750616,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499519955437438,
      "loss": 6.1516,
      "mean_token_accuracy": 0.14913926422595977,
      "num_tokens": 6219760.0,
      "step": 3335
    },
    {
      "entropy": 6.2619555473327635,
      "epoch": 0.29494878134934654,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995178972238054,
      "loss": 6.1247,
      "mean_token_accuracy": 0.15912023782730103,
      "num_tokens": 6228721.0,
      "step": 3340
    },
    {
      "entropy": 6.183184289932251,
      "epoch": 0.2953903214411869,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499515834611995,
      "loss": 5.9324,
      "mean_token_accuracy": 0.1727195903658867,
      "num_tokens": 6237070.0,
      "step": 3345
    },
    {
      "entropy": 6.158340120315552,
      "epoch": 0.2958318615330272,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995137676020472,
      "loss": 6.0929,
      "mean_token_accuracy": 0.15666336119174956,
      "num_tokens": 6245659.0,
      "step": 3350
    },
    {
      "entropy": 6.225800704956055,
      "epoch": 0.29627340162486754,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995116961940023,
      "loss": 6.0472,
      "mean_token_accuracy": 0.16301591396331788,
      "num_tokens": 6255175.0,
      "step": 3355
    },
    {
      "entropy": 6.226327657699585,
      "epoch": 0.29671494171670787,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004995096203879009,
      "loss": 6.0769,
      "mean_token_accuracy": 0.15317506715655327,
      "num_tokens": 6264962.0,
      "step": 3360
    },
    {
      "entropy": 6.216853618621826,
      "epoch": 0.2971564818085482,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995075401837837,
      "loss": 5.9968,
      "mean_token_accuracy": 0.16241736561059952,
      "num_tokens": 6273411.0,
      "step": 3365
    },
    {
      "entropy": 6.083216285705566,
      "epoch": 0.29759802190038853,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004995054555816915,
      "loss": 5.9669,
      "mean_token_accuracy": 0.16453457474708558,
      "num_tokens": 6282618.0,
      "step": 3370
    },
    {
      "entropy": 6.205840110778809,
      "epoch": 0.2980395619922289,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004995033665816651,
      "loss": 5.9706,
      "mean_token_accuracy": 0.1685549646615982,
      "num_tokens": 6292008.0,
      "step": 3375
    },
    {
      "entropy": 6.129777622222901,
      "epoch": 0.29848110208406925,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004995012731837454,
      "loss": 5.9386,
      "mean_token_accuracy": 0.1638485983014107,
      "num_tokens": 6301079.0,
      "step": 3380
    },
    {
      "entropy": 6.172431135177613,
      "epoch": 0.2989226421759096,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004994991753879736,
      "loss": 6.1135,
      "mean_token_accuracy": 0.15206168442964554,
      "num_tokens": 6310543.0,
      "step": 3385
    },
    {
      "entropy": 6.267715644836426,
      "epoch": 0.2993641822677499,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004994970731943904,
      "loss": 6.0568,
      "mean_token_accuracy": 0.15798714011907578,
      "num_tokens": 6320037.0,
      "step": 3390
    },
    {
      "entropy": 6.189213752746582,
      "epoch": 0.29980572235959024,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004994949666030374,
      "loss": 6.0594,
      "mean_token_accuracy": 0.1617031455039978,
      "num_tokens": 6328948.0,
      "step": 3395
    },
    {
      "entropy": 6.132642126083374,
      "epoch": 0.3002472624514306,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994928556139557,
      "loss": 5.9311,
      "mean_token_accuracy": 0.16827442049980162,
      "num_tokens": 6337449.0,
      "step": 3400
    },
    {
      "entropy": 6.117817974090576,
      "epoch": 0.3006888025432709,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994907402271865,
      "loss": 5.9745,
      "mean_token_accuracy": 0.1659398391842842,
      "num_tokens": 6346130.0,
      "step": 3405
    },
    {
      "entropy": 6.122855234146118,
      "epoch": 0.3011303426351113,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994886204427715,
      "loss": 6.0239,
      "mean_token_accuracy": 0.16507700830698013,
      "num_tokens": 6354973.0,
      "step": 3410
    },
    {
      "entropy": 6.271668004989624,
      "epoch": 0.3015718827269516,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994864962607519,
      "loss": 6.1041,
      "mean_token_accuracy": 0.15557809472084044,
      "num_tokens": 6364107.0,
      "step": 3415
    },
    {
      "entropy": 6.2369715690612795,
      "epoch": 0.30201342281879195,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994843676811697,
      "loss": 6.0198,
      "mean_token_accuracy": 0.1597614735364914,
      "num_tokens": 6372859.0,
      "step": 3420
    },
    {
      "entropy": 6.088519191741943,
      "epoch": 0.3024549629106323,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994822347040664,
      "loss": 5.8991,
      "mean_token_accuracy": 0.17260994017124176,
      "num_tokens": 6381818.0,
      "step": 3425
    },
    {
      "entropy": 6.1546448230743405,
      "epoch": 0.3028965030024726,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994800973294837,
      "loss": 6.0699,
      "mean_token_accuracy": 0.15940487533807754,
      "num_tokens": 6391460.0,
      "step": 3430
    },
    {
      "entropy": 6.290472555160522,
      "epoch": 0.30333804309431295,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994779555574636,
      "loss": 6.1024,
      "mean_token_accuracy": 0.15706333965063096,
      "num_tokens": 6401461.0,
      "step": 3435
    },
    {
      "entropy": 6.211634826660156,
      "epoch": 0.3037795831861533,
      "grad_norm": 1.0625,
      "learning_rate": 0.000499475809388048,
      "loss": 5.9738,
      "mean_token_accuracy": 0.15956642478704453,
      "num_tokens": 6410435.0,
      "step": 3440
    },
    {
      "entropy": 6.20361361503601,
      "epoch": 0.30422112327799367,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499473658821279,
      "loss": 6.1684,
      "mean_token_accuracy": 0.15328484475612641,
      "num_tokens": 6420710.0,
      "step": 3445
    },
    {
      "entropy": 6.279194784164429,
      "epoch": 0.304662663369834,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004994715038571986,
      "loss": 6.0188,
      "mean_token_accuracy": 0.1564917415380478,
      "num_tokens": 6429882.0,
      "step": 3450
    },
    {
      "entropy": 6.139882755279541,
      "epoch": 0.30510420346167433,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004994693444958493,
      "loss": 5.9924,
      "mean_token_accuracy": 0.16666608154773713,
      "num_tokens": 6439183.0,
      "step": 3455
    },
    {
      "entropy": 6.22991909980774,
      "epoch": 0.30554574355351466,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499467180737273,
      "loss": 6.1876,
      "mean_token_accuracy": 0.15393830388784407,
      "num_tokens": 6448460.0,
      "step": 3460
    },
    {
      "entropy": 6.241871309280396,
      "epoch": 0.305987283645355,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994650125815124,
      "loss": 6.0346,
      "mean_token_accuracy": 0.16356185972690582,
      "num_tokens": 6457687.0,
      "step": 3465
    },
    {
      "entropy": 6.1781104564666744,
      "epoch": 0.3064288237371953,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994628400286097,
      "loss": 5.9815,
      "mean_token_accuracy": 0.16612761467695236,
      "num_tokens": 6467057.0,
      "step": 3470
    },
    {
      "entropy": 6.136842584609985,
      "epoch": 0.30687036382903565,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994606630786078,
      "loss": 6.0576,
      "mean_token_accuracy": 0.16337907910346985,
      "num_tokens": 6477484.0,
      "step": 3475
    },
    {
      "entropy": 6.187231159210205,
      "epoch": 0.30731190392087604,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004994584817315492,
      "loss": 6.1329,
      "mean_token_accuracy": 0.15671780258417128,
      "num_tokens": 6488381.0,
      "step": 3480
    },
    {
      "entropy": 6.257808637619019,
      "epoch": 0.30775344401271637,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004994562959874765,
      "loss": 6.0139,
      "mean_token_accuracy": 0.16207417994737625,
      "num_tokens": 6498279.0,
      "step": 3485
    },
    {
      "entropy": 6.21480975151062,
      "epoch": 0.3081949841045567,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004994541058464326,
      "loss": 6.0169,
      "mean_token_accuracy": 0.16068682819604874,
      "num_tokens": 6508008.0,
      "step": 3490
    },
    {
      "entropy": 6.236310052871704,
      "epoch": 0.30863652419639703,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004994519113084605,
      "loss": 6.0915,
      "mean_token_accuracy": 0.15919461101293564,
      "num_tokens": 6517263.0,
      "step": 3495
    },
    {
      "entropy": 6.244526529312134,
      "epoch": 0.30907806428823736,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994497123736029,
      "loss": 6.1388,
      "mean_token_accuracy": 0.15009350925683976,
      "num_tokens": 6527682.0,
      "step": 3500
    },
    {
      "entropy": 6.20444884300232,
      "epoch": 0.3095196043800777,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004994475090419034,
      "loss": 6.0139,
      "mean_token_accuracy": 0.16034948080778122,
      "num_tokens": 6537143.0,
      "step": 3505
    },
    {
      "entropy": 6.1589967727661135,
      "epoch": 0.309961144471918,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004994453013134047,
      "loss": 6.004,
      "mean_token_accuracy": 0.16135213673114776,
      "num_tokens": 6546561.0,
      "step": 3510
    },
    {
      "entropy": 6.153254127502441,
      "epoch": 0.3104026845637584,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004994430891881502,
      "loss": 6.0215,
      "mean_token_accuracy": 0.1592785008251667,
      "num_tokens": 6555806.0,
      "step": 3515
    },
    {
      "entropy": 6.202929401397705,
      "epoch": 0.31084422465559874,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994408726661832,
      "loss": 6.05,
      "mean_token_accuracy": 0.1624141365289688,
      "num_tokens": 6564836.0,
      "step": 3520
    },
    {
      "entropy": 6.133888912200928,
      "epoch": 0.3112857647474391,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004994386517475472,
      "loss": 6.0371,
      "mean_token_accuracy": 0.155925951898098,
      "num_tokens": 6574997.0,
      "step": 3525
    },
    {
      "entropy": 6.208050298690796,
      "epoch": 0.3117273048392794,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994364264322856,
      "loss": 6.0597,
      "mean_token_accuracy": 0.15896133482456207,
      "num_tokens": 6584589.0,
      "step": 3530
    },
    {
      "entropy": 6.167923545837402,
      "epoch": 0.31216884493111974,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004994341967204421,
      "loss": 6.0051,
      "mean_token_accuracy": 0.16373383700847627,
      "num_tokens": 6592898.0,
      "step": 3535
    },
    {
      "entropy": 6.255459213256836,
      "epoch": 0.31261038502296007,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004994319626120603,
      "loss": 6.0143,
      "mean_token_accuracy": 0.1714682549238205,
      "num_tokens": 6602454.0,
      "step": 3540
    },
    {
      "entropy": 6.140329837799072,
      "epoch": 0.3130519251148004,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994297241071841,
      "loss": 6.0189,
      "mean_token_accuracy": 0.16352686434984207,
      "num_tokens": 6612454.0,
      "step": 3545
    },
    {
      "entropy": 6.176025867462158,
      "epoch": 0.3134934652066408,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499427481205857,
      "loss": 6.0204,
      "mean_token_accuracy": 0.16263296753168105,
      "num_tokens": 6621286.0,
      "step": 3550
    },
    {
      "entropy": 6.158733367919922,
      "epoch": 0.3139350052984811,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004994252339081234,
      "loss": 5.9611,
      "mean_token_accuracy": 0.16285568177700044,
      "num_tokens": 6629684.0,
      "step": 3555
    },
    {
      "entropy": 6.135843944549561,
      "epoch": 0.31437654539032145,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000499422982214027,
      "loss": 6.0543,
      "mean_token_accuracy": 0.1660478115081787,
      "num_tokens": 6638836.0,
      "step": 3560
    },
    {
      "entropy": 6.2141636371612545,
      "epoch": 0.3148180854821618,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994207261236121,
      "loss": 5.9942,
      "mean_token_accuracy": 0.16404083073139192,
      "num_tokens": 6647337.0,
      "step": 3565
    },
    {
      "entropy": 6.096640634536743,
      "epoch": 0.3152596255740021,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994184656369227,
      "loss": 5.9733,
      "mean_token_accuracy": 0.15804241746664047,
      "num_tokens": 6657246.0,
      "step": 3570
    },
    {
      "entropy": 6.225153303146362,
      "epoch": 0.31570116566584244,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004994162007540033,
      "loss": 6.1202,
      "mean_token_accuracy": 0.15196825563907623,
      "num_tokens": 6667981.0,
      "step": 3575
    },
    {
      "entropy": 6.204136848449707,
      "epoch": 0.3161427057576828,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994139314748981,
      "loss": 5.9703,
      "mean_token_accuracy": 0.16491541713476182,
      "num_tokens": 6677462.0,
      "step": 3580
    },
    {
      "entropy": 6.090105485916138,
      "epoch": 0.31658424584952316,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004994116577996517,
      "loss": 6.007,
      "mean_token_accuracy": 0.17120114117860794,
      "num_tokens": 6687225.0,
      "step": 3585
    },
    {
      "entropy": 6.172569704055786,
      "epoch": 0.3170257859413635,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004994093797283084,
      "loss": 6.086,
      "mean_token_accuracy": 0.15607115626335144,
      "num_tokens": 6696345.0,
      "step": 3590
    },
    {
      "entropy": 6.24049186706543,
      "epoch": 0.3174673260332038,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994070972609132,
      "loss": 6.0527,
      "mean_token_accuracy": 0.16193179190158843,
      "num_tokens": 6706207.0,
      "step": 3595
    },
    {
      "entropy": 6.173488807678223,
      "epoch": 0.31790886612504415,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994048103975103,
      "loss": 5.9788,
      "mean_token_accuracy": 0.15975290089845656,
      "num_tokens": 6714430.0,
      "step": 3600
    },
    {
      "entropy": 6.130007123947143,
      "epoch": 0.3183504062168845,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499402519138145,
      "loss": 5.9923,
      "mean_token_accuracy": 0.1620108649134636,
      "num_tokens": 6724139.0,
      "step": 3605
    },
    {
      "entropy": 6.143885087966919,
      "epoch": 0.3187919463087248,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004994002234828619,
      "loss": 6.0353,
      "mean_token_accuracy": 0.16159311830997466,
      "num_tokens": 6733525.0,
      "step": 3610
    },
    {
      "entropy": 6.262012004852295,
      "epoch": 0.31923348640056515,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499397923431706,
      "loss": 6.1456,
      "mean_token_accuracy": 0.15688765943050384,
      "num_tokens": 6743738.0,
      "step": 3615
    },
    {
      "entropy": 6.226923990249634,
      "epoch": 0.31967502649240553,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004993956189847226,
      "loss": 5.9809,
      "mean_token_accuracy": 0.1651754342019558,
      "num_tokens": 6752900.0,
      "step": 3620
    },
    {
      "entropy": 6.146596193313599,
      "epoch": 0.32011656658424587,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993933101419565,
      "loss": 5.9722,
      "mean_token_accuracy": 0.1666112557053566,
      "num_tokens": 6761776.0,
      "step": 3625
    },
    {
      "entropy": 6.1981529712677,
      "epoch": 0.3205581066760862,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993909969034531,
      "loss": 6.12,
      "mean_token_accuracy": 0.15697512328624724,
      "num_tokens": 6771543.0,
      "step": 3630
    },
    {
      "entropy": 6.192965984344482,
      "epoch": 0.3209996467679265,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993886792692576,
      "loss": 6.005,
      "mean_token_accuracy": 0.16011940240859984,
      "num_tokens": 6780010.0,
      "step": 3635
    },
    {
      "entropy": 6.1778465747833256,
      "epoch": 0.32144118685976686,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993863572394156,
      "loss": 6.0593,
      "mean_token_accuracy": 0.1610572248697281,
      "num_tokens": 6788830.0,
      "step": 3640
    },
    {
      "entropy": 6.20711760520935,
      "epoch": 0.3218827269516072,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993840308139724,
      "loss": 6.0473,
      "mean_token_accuracy": 0.16559676080942154,
      "num_tokens": 6799121.0,
      "step": 3645
    },
    {
      "entropy": 6.192859411239624,
      "epoch": 0.3223242670434475,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993816999929738,
      "loss": 5.9948,
      "mean_token_accuracy": 0.15838830024003983,
      "num_tokens": 6808519.0,
      "step": 3650
    },
    {
      "entropy": 6.124808979034424,
      "epoch": 0.3227658071352879,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993793647764651,
      "loss": 6.0069,
      "mean_token_accuracy": 0.17118934392929078,
      "num_tokens": 6817583.0,
      "step": 3655
    },
    {
      "entropy": 6.23282527923584,
      "epoch": 0.32320734722712824,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993770251644923,
      "loss": 6.0196,
      "mean_token_accuracy": 0.162643201649189,
      "num_tokens": 6827732.0,
      "step": 3660
    },
    {
      "entropy": 6.109570169448853,
      "epoch": 0.32364888731896857,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993746811571013,
      "loss": 5.9775,
      "mean_token_accuracy": 0.16294157058000563,
      "num_tokens": 6836903.0,
      "step": 3665
    },
    {
      "entropy": 6.069127988815308,
      "epoch": 0.3240904274108089,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993723327543379,
      "loss": 5.986,
      "mean_token_accuracy": 0.17082992941141129,
      "num_tokens": 6845684.0,
      "step": 3670
    },
    {
      "entropy": 6.141482591629028,
      "epoch": 0.32453196750264923,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499369979956248,
      "loss": 5.9745,
      "mean_token_accuracy": 0.16510514169931412,
      "num_tokens": 6854650.0,
      "step": 3675
    },
    {
      "entropy": 6.195809412002563,
      "epoch": 0.32497350759448956,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993676227628779,
      "loss": 5.9775,
      "mean_token_accuracy": 0.15800541490316392,
      "num_tokens": 6864340.0,
      "step": 3680
    },
    {
      "entropy": 6.201231145858765,
      "epoch": 0.3254150476863299,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993652611742736,
      "loss": 6.022,
      "mean_token_accuracy": 0.16170618385076524,
      "num_tokens": 6873244.0,
      "step": 3685
    },
    {
      "entropy": 6.065858983993531,
      "epoch": 0.3258565877781703,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004993628951904815,
      "loss": 5.8965,
      "mean_token_accuracy": 0.16670404374599457,
      "num_tokens": 6882392.0,
      "step": 3690
    },
    {
      "entropy": 6.172443294525147,
      "epoch": 0.3262981278700106,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993605248115479,
      "loss": 6.1129,
      "mean_token_accuracy": 0.14900100082159043,
      "num_tokens": 6891515.0,
      "step": 3695
    },
    {
      "entropy": 6.265958166122436,
      "epoch": 0.32673966796185094,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993581500375191,
      "loss": 6.0042,
      "mean_token_accuracy": 0.16194652020931244,
      "num_tokens": 6900104.0,
      "step": 3700
    },
    {
      "entropy": 6.21707010269165,
      "epoch": 0.3271812080536913,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004993557708684417,
      "loss": 6.1181,
      "mean_token_accuracy": 0.15311858505010606,
      "num_tokens": 6910443.0,
      "step": 3705
    },
    {
      "entropy": 6.130673027038574,
      "epoch": 0.3276227481455316,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993533873043625,
      "loss": 5.983,
      "mean_token_accuracy": 0.16641464680433274,
      "num_tokens": 6919882.0,
      "step": 3710
    },
    {
      "entropy": 6.2687431335449215,
      "epoch": 0.32806428823737194,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499350999345328,
      "loss": 6.0932,
      "mean_token_accuracy": 0.15396894812583922,
      "num_tokens": 6929757.0,
      "step": 3715
    },
    {
      "entropy": 6.1528526782989506,
      "epoch": 0.32850582832921227,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499348606991385,
      "loss": 6.0822,
      "mean_token_accuracy": 0.15250903815031053,
      "num_tokens": 6938578.0,
      "step": 3720
    },
    {
      "entropy": 6.175461196899414,
      "epoch": 0.32894736842105265,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993462102425805,
      "loss": 5.9653,
      "mean_token_accuracy": 0.16177183985710145,
      "num_tokens": 6947920.0,
      "step": 3725
    },
    {
      "entropy": 6.154925155639648,
      "epoch": 0.329388908512893,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993438090989612,
      "loss": 5.9195,
      "mean_token_accuracy": 0.16460449397563934,
      "num_tokens": 6957291.0,
      "step": 3730
    },
    {
      "entropy": 6.145351982116699,
      "epoch": 0.3298304486047333,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004993414035605743,
      "loss": 5.9816,
      "mean_token_accuracy": 0.171938194334507,
      "num_tokens": 6966504.0,
      "step": 3735
    },
    {
      "entropy": 6.169538497924805,
      "epoch": 0.33027198869657365,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004993389936274669,
      "loss": 6.0167,
      "mean_token_accuracy": 0.16437650918960572,
      "num_tokens": 6976234.0,
      "step": 3740
    },
    {
      "entropy": 6.15001654624939,
      "epoch": 0.330713528788414,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004993365792996862,
      "loss": 6.0641,
      "mean_token_accuracy": 0.15483692288398743,
      "num_tokens": 6986372.0,
      "step": 3745
    },
    {
      "entropy": 6.160980892181397,
      "epoch": 0.3311550688802543,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993341605772795,
      "loss": 5.9526,
      "mean_token_accuracy": 0.167180797457695,
      "num_tokens": 6995553.0,
      "step": 3750
    },
    {
      "entropy": 6.0821184635162355,
      "epoch": 0.33159660897209464,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004993317374602941,
      "loss": 5.8782,
      "mean_token_accuracy": 0.1738182157278061,
      "num_tokens": 7005675.0,
      "step": 3755
    },
    {
      "entropy": 6.085787343978882,
      "epoch": 0.33203814906393503,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993293099487777,
      "loss": 5.9976,
      "mean_token_accuracy": 0.16012882590293884,
      "num_tokens": 7014953.0,
      "step": 3760
    },
    {
      "entropy": 6.188743925094604,
      "epoch": 0.33247968915577536,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004993268780427776,
      "loss": 6.0033,
      "mean_token_accuracy": 0.17121922522783278,
      "num_tokens": 7025075.0,
      "step": 3765
    },
    {
      "entropy": 6.118736314773559,
      "epoch": 0.3329212292476157,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004993244417423416,
      "loss": 6.0144,
      "mean_token_accuracy": 0.16219213157892226,
      "num_tokens": 7034286.0,
      "step": 3770
    },
    {
      "entropy": 6.19470705986023,
      "epoch": 0.333362769339456,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993220010475174,
      "loss": 6.0445,
      "mean_token_accuracy": 0.15993663668632507,
      "num_tokens": 7043957.0,
      "step": 3775
    },
    {
      "entropy": 6.16349458694458,
      "epoch": 0.33380430943129635,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004993195559583526,
      "loss": 5.966,
      "mean_token_accuracy": 0.1597854785621166,
      "num_tokens": 7053170.0,
      "step": 3780
    },
    {
      "entropy": 6.016193008422851,
      "epoch": 0.3342458495231367,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993171064748954,
      "loss": 5.955,
      "mean_token_accuracy": 0.16348500102758406,
      "num_tokens": 7062469.0,
      "step": 3785
    },
    {
      "entropy": 6.167666959762573,
      "epoch": 0.334687389614977,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004993146525971937,
      "loss": 5.9591,
      "mean_token_accuracy": 0.16554407477378846,
      "num_tokens": 7071463.0,
      "step": 3790
    },
    {
      "entropy": 6.229743194580078,
      "epoch": 0.3351289297068174,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004993121943252955,
      "loss": 6.0586,
      "mean_token_accuracy": 0.1564462423324585,
      "num_tokens": 7081574.0,
      "step": 3795
    },
    {
      "entropy": 6.085719156265259,
      "epoch": 0.33557046979865773,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993097316592489,
      "loss": 5.9676,
      "mean_token_accuracy": 0.16341703832149507,
      "num_tokens": 7090835.0,
      "step": 3800
    },
    {
      "entropy": 6.035996341705323,
      "epoch": 0.33601200989049806,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004993072645991023,
      "loss": 5.906,
      "mean_token_accuracy": 0.1561555340886116,
      "num_tokens": 7100521.0,
      "step": 3805
    },
    {
      "entropy": 6.160156917572022,
      "epoch": 0.3364535499823384,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499304793144904,
      "loss": 5.9702,
      "mean_token_accuracy": 0.16260155886411667,
      "num_tokens": 7109011.0,
      "step": 3810
    },
    {
      "entropy": 6.203392267227173,
      "epoch": 0.3368950900741787,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993023172967022,
      "loss": 5.9987,
      "mean_token_accuracy": 0.15889298021793366,
      "num_tokens": 7119128.0,
      "step": 3815
    },
    {
      "entropy": 6.194024133682251,
      "epoch": 0.33733663016601906,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992998370545458,
      "loss": 6.081,
      "mean_token_accuracy": 0.15163458064198493,
      "num_tokens": 7128313.0,
      "step": 3820
    },
    {
      "entropy": 6.201267099380493,
      "epoch": 0.3377781702578594,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992973524184831,
      "loss": 6.0549,
      "mean_token_accuracy": 0.15506500452756883,
      "num_tokens": 7137567.0,
      "step": 3825
    },
    {
      "entropy": 6.092210531234741,
      "epoch": 0.3382197103496998,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992948633885627,
      "loss": 6.0193,
      "mean_token_accuracy": 0.16242285072803497,
      "num_tokens": 7147254.0,
      "step": 3830
    },
    {
      "entropy": 6.069639110565186,
      "epoch": 0.3386612504415401,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992923699648335,
      "loss": 5.9133,
      "mean_token_accuracy": 0.17447974383831025,
      "num_tokens": 7156215.0,
      "step": 3835
    },
    {
      "entropy": 6.222735929489136,
      "epoch": 0.33910279053338044,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992898721473445,
      "loss": 6.0054,
      "mean_token_accuracy": 0.15784852206707,
      "num_tokens": 7165653.0,
      "step": 3840
    },
    {
      "entropy": 6.0844096660614015,
      "epoch": 0.33954433062522077,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992873699361444,
      "loss": 5.9012,
      "mean_token_accuracy": 0.1631088748574257,
      "num_tokens": 7174797.0,
      "step": 3845
    },
    {
      "entropy": 6.192289066314697,
      "epoch": 0.3399858707170611,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004992848633312822,
      "loss": 5.9327,
      "mean_token_accuracy": 0.1653936520218849,
      "num_tokens": 7184040.0,
      "step": 3850
    },
    {
      "entropy": 6.110203790664673,
      "epoch": 0.34042741080890143,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004992823523328071,
      "loss": 5.8661,
      "mean_token_accuracy": 0.17056092023849487,
      "num_tokens": 7193533.0,
      "step": 3855
    },
    {
      "entropy": 6.074834299087525,
      "epoch": 0.34086895090074176,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992798369407684,
      "loss": 5.9999,
      "mean_token_accuracy": 0.16645675301551818,
      "num_tokens": 7203355.0,
      "step": 3860
    },
    {
      "entropy": 6.240177392959595,
      "epoch": 0.34131049099258215,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992773171552152,
      "loss": 5.9403,
      "mean_token_accuracy": 0.16228992938995362,
      "num_tokens": 7212569.0,
      "step": 3865
    },
    {
      "entropy": 6.043364143371582,
      "epoch": 0.3417520310844225,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004992747929761968,
      "loss": 5.968,
      "mean_token_accuracy": 0.16697075814008713,
      "num_tokens": 7221886.0,
      "step": 3870
    },
    {
      "entropy": 6.094742727279663,
      "epoch": 0.3421935711762628,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992722644037628,
      "loss": 6.0245,
      "mean_token_accuracy": 0.16425753831863404,
      "num_tokens": 7232065.0,
      "step": 3875
    },
    {
      "entropy": 6.176693868637085,
      "epoch": 0.34263511126810314,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004992697314379628,
      "loss": 5.9362,
      "mean_token_accuracy": 0.16599779576063156,
      "num_tokens": 7242287.0,
      "step": 3880
    },
    {
      "entropy": 6.141386938095093,
      "epoch": 0.3430766513599435,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992671940788462,
      "loss": 5.9315,
      "mean_token_accuracy": 0.16158146411180496,
      "num_tokens": 7250638.0,
      "step": 3885
    },
    {
      "entropy": 6.137431240081787,
      "epoch": 0.3435181914517838,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992646523264628,
      "loss": 5.9567,
      "mean_token_accuracy": 0.1678780198097229,
      "num_tokens": 7259526.0,
      "step": 3890
    },
    {
      "entropy": 6.091816043853759,
      "epoch": 0.34395973154362414,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992621061808625,
      "loss": 5.9422,
      "mean_token_accuracy": 0.1681588336825371,
      "num_tokens": 7269260.0,
      "step": 3895
    },
    {
      "entropy": 6.137960958480835,
      "epoch": 0.3444012716354645,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499259555642095,
      "loss": 5.9608,
      "mean_token_accuracy": 0.1592722475528717,
      "num_tokens": 7279061.0,
      "step": 3900
    },
    {
      "entropy": 6.152037239074707,
      "epoch": 0.34484281172730485,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992570007102104,
      "loss": 5.9434,
      "mean_token_accuracy": 0.16730886548757554,
      "num_tokens": 7289721.0,
      "step": 3905
    },
    {
      "entropy": 6.210768556594848,
      "epoch": 0.3452843518191452,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004992544413852587,
      "loss": 6.0913,
      "mean_token_accuracy": 0.15217040181159974,
      "num_tokens": 7299836.0,
      "step": 3910
    },
    {
      "entropy": 6.154385423660278,
      "epoch": 0.3457258919109855,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00049925187766729,
      "loss": 6.1203,
      "mean_token_accuracy": 0.1559758186340332,
      "num_tokens": 7309306.0,
      "step": 3915
    },
    {
      "entropy": 6.210061311721802,
      "epoch": 0.34616743200282585,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004992493095563545,
      "loss": 5.9828,
      "mean_token_accuracy": 0.16974906474351883,
      "num_tokens": 7318486.0,
      "step": 3920
    },
    {
      "entropy": 6.071922349929809,
      "epoch": 0.3466089720946662,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992467370525026,
      "loss": 5.94,
      "mean_token_accuracy": 0.16524540185928344,
      "num_tokens": 7327764.0,
      "step": 3925
    },
    {
      "entropy": 6.158257675170899,
      "epoch": 0.3470505121865065,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004992441601557848,
      "loss": 6.0295,
      "mean_token_accuracy": 0.16136583387851716,
      "num_tokens": 7337118.0,
      "step": 3930
    },
    {
      "entropy": 6.115141105651856,
      "epoch": 0.3474920522783469,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992415788662514,
      "loss": 5.97,
      "mean_token_accuracy": 0.1630101129412651,
      "num_tokens": 7345973.0,
      "step": 3935
    },
    {
      "entropy": 6.12622766494751,
      "epoch": 0.34793359237018723,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004992389931839529,
      "loss": 5.9695,
      "mean_token_accuracy": 0.16391944289207458,
      "num_tokens": 7355776.0,
      "step": 3940
    },
    {
      "entropy": 6.120577955245972,
      "epoch": 0.34837513246202756,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004992364031089401,
      "loss": 5.9826,
      "mean_token_accuracy": 0.16477784961462022,
      "num_tokens": 7364820.0,
      "step": 3945
    },
    {
      "entropy": 6.1659424781799315,
      "epoch": 0.3488166725538679,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992338086412636,
      "loss": 5.9967,
      "mean_token_accuracy": 0.16261015236377716,
      "num_tokens": 7373881.0,
      "step": 3950
    },
    {
      "entropy": 6.100357103347778,
      "epoch": 0.3492582126457082,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004992312097809744,
      "loss": 5.9439,
      "mean_token_accuracy": 0.1656390517950058,
      "num_tokens": 7383095.0,
      "step": 3955
    },
    {
      "entropy": 6.111528301239014,
      "epoch": 0.34969975273754855,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992286065281234,
      "loss": 5.9943,
      "mean_token_accuracy": 0.16113595068454742,
      "num_tokens": 7392702.0,
      "step": 3960
    },
    {
      "entropy": 6.163757038116455,
      "epoch": 0.3501412928293889,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992259988827614,
      "loss": 6.1647,
      "mean_token_accuracy": 0.15409360080957413,
      "num_tokens": 7402560.0,
      "step": 3965
    },
    {
      "entropy": 6.163148641586304,
      "epoch": 0.35058283292122927,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992233868449397,
      "loss": 5.9105,
      "mean_token_accuracy": 0.17125627696514129,
      "num_tokens": 7412375.0,
      "step": 3970
    },
    {
      "entropy": 6.125844812393188,
      "epoch": 0.3510243730130696,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992207704147093,
      "loss": 5.9829,
      "mean_token_accuracy": 0.1594072923064232,
      "num_tokens": 7420734.0,
      "step": 3975
    },
    {
      "entropy": 6.08319845199585,
      "epoch": 0.35146591310490993,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004992181495921216,
      "loss": 5.9549,
      "mean_token_accuracy": 0.1678524360060692,
      "num_tokens": 7430156.0,
      "step": 3980
    },
    {
      "entropy": 6.203280305862426,
      "epoch": 0.35190745319675026,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004992155243772277,
      "loss": 5.9672,
      "mean_token_accuracy": 0.17124156951904296,
      "num_tokens": 7439189.0,
      "step": 3985
    },
    {
      "entropy": 6.162478065490722,
      "epoch": 0.3523489932885906,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004992128947700795,
      "loss": 5.9733,
      "mean_token_accuracy": 0.16030417680740355,
      "num_tokens": 7448768.0,
      "step": 3990
    },
    {
      "entropy": 6.085399436950683,
      "epoch": 0.3527905333804309,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499210260770728,
      "loss": 5.9384,
      "mean_token_accuracy": 0.16325445622205734,
      "num_tokens": 7458468.0,
      "step": 3995
    },
    {
      "entropy": 6.143225765228271,
      "epoch": 0.35323207347227126,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499207622379225,
      "loss": 5.9964,
      "mean_token_accuracy": 0.1624176785349846,
      "num_tokens": 7467351.0,
      "step": 4000
    },
    {
      "entropy": 6.191954469680786,
      "epoch": 0.35367361356411164,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004992049795956222,
      "loss": 5.9732,
      "mean_token_accuracy": 0.15985945910215377,
      "num_tokens": 7477116.0,
      "step": 4005
    },
    {
      "entropy": 6.096588611602783,
      "epoch": 0.354115153655952,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004992023324199715,
      "loss": 5.9035,
      "mean_token_accuracy": 0.1694460391998291,
      "num_tokens": 7485216.0,
      "step": 4010
    },
    {
      "entropy": 6.05253791809082,
      "epoch": 0.3545566937477923,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991996808523245,
      "loss": 6.0587,
      "mean_token_accuracy": 0.15506483614444733,
      "num_tokens": 7495084.0,
      "step": 4015
    },
    {
      "entropy": 6.274289941787719,
      "epoch": 0.35499823383963264,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991970248927332,
      "loss": 6.0609,
      "mean_token_accuracy": 0.16138618141412736,
      "num_tokens": 7503219.0,
      "step": 4020
    },
    {
      "entropy": 6.107459831237793,
      "epoch": 0.35543977393147297,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991943645412498,
      "loss": 5.8498,
      "mean_token_accuracy": 0.17398339360952378,
      "num_tokens": 7511961.0,
      "step": 4025
    },
    {
      "entropy": 6.018071126937866,
      "epoch": 0.3558813140233133,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991916997979263,
      "loss": 5.9675,
      "mean_token_accuracy": 0.16527593061327933,
      "num_tokens": 7521927.0,
      "step": 4030
    },
    {
      "entropy": 6.0901021480560305,
      "epoch": 0.35632285411515363,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991890306628149,
      "loss": 5.9106,
      "mean_token_accuracy": 0.16498993933200837,
      "num_tokens": 7531093.0,
      "step": 4035
    },
    {
      "entropy": 6.112301921844482,
      "epoch": 0.356764394206994,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991863571359678,
      "loss": 5.8843,
      "mean_token_accuracy": 0.1641499012708664,
      "num_tokens": 7539813.0,
      "step": 4040
    },
    {
      "entropy": 6.191416263580322,
      "epoch": 0.35720593429883435,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004991836792174376,
      "loss": 6.0347,
      "mean_token_accuracy": 0.15737137496471404,
      "num_tokens": 7548954.0,
      "step": 4045
    },
    {
      "entropy": 6.059986782073975,
      "epoch": 0.3576474743906747,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991809969072765,
      "loss": 5.8989,
      "mean_token_accuracy": 0.1670374259352684,
      "num_tokens": 7558275.0,
      "step": 4050
    },
    {
      "entropy": 6.059090328216553,
      "epoch": 0.358089014482515,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991783102055371,
      "loss": 5.9351,
      "mean_token_accuracy": 0.17020276039838791,
      "num_tokens": 7567870.0,
      "step": 4055
    },
    {
      "entropy": 6.14141616821289,
      "epoch": 0.35853055457435534,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004991756191122723,
      "loss": 5.8914,
      "mean_token_accuracy": 0.17240921556949615,
      "num_tokens": 7577485.0,
      "step": 4060
    },
    {
      "entropy": 6.102657413482666,
      "epoch": 0.3589720946661957,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004991729236275346,
      "loss": 5.9435,
      "mean_token_accuracy": 0.16436953246593475,
      "num_tokens": 7587398.0,
      "step": 4065
    },
    {
      "entropy": 6.044604969024658,
      "epoch": 0.359413634758036,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991702237513768,
      "loss": 5.9639,
      "mean_token_accuracy": 0.16127175837755203,
      "num_tokens": 7596310.0,
      "step": 4070
    },
    {
      "entropy": 6.149261903762818,
      "epoch": 0.3598551748498764,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004991675194838517,
      "loss": 5.9934,
      "mean_token_accuracy": 0.15606556236743926,
      "num_tokens": 7606971.0,
      "step": 4075
    },
    {
      "entropy": 6.130791330337525,
      "epoch": 0.3602967149417167,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004991648108250125,
      "loss": 5.9699,
      "mean_token_accuracy": 0.162691630423069,
      "num_tokens": 7616636.0,
      "step": 4080
    },
    {
      "entropy": 6.138277244567871,
      "epoch": 0.36073825503355705,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499162097774912,
      "loss": 5.9991,
      "mean_token_accuracy": 0.16216389685869217,
      "num_tokens": 7625128.0,
      "step": 4085
    },
    {
      "entropy": 6.10425295829773,
      "epoch": 0.3611797951253974,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991593803336037,
      "loss": 5.9044,
      "mean_token_accuracy": 0.1682908684015274,
      "num_tokens": 7634224.0,
      "step": 4090
    },
    {
      "entropy": 6.075386428833008,
      "epoch": 0.3616213352172377,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004991566585011405,
      "loss": 5.8935,
      "mean_token_accuracy": 0.1718817874789238,
      "num_tokens": 7643392.0,
      "step": 4095
    },
    {
      "entropy": 6.03798017501831,
      "epoch": 0.36206287530907805,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991539322775758,
      "loss": 5.9139,
      "mean_token_accuracy": 0.16705035418272018,
      "num_tokens": 7652491.0,
      "step": 4100
    },
    {
      "entropy": 6.292382764816284,
      "epoch": 0.3625044154009184,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004991512016629632,
      "loss": 6.1198,
      "mean_token_accuracy": 0.14770500659942626,
      "num_tokens": 7663192.0,
      "step": 4105
    },
    {
      "entropy": 6.111117219924926,
      "epoch": 0.36294595549275877,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004991484666573558,
      "loss": 5.9498,
      "mean_token_accuracy": 0.16446800976991655,
      "num_tokens": 7672801.0,
      "step": 4110
    },
    {
      "entropy": 6.096119546890259,
      "epoch": 0.3633874955845991,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004991457272608077,
      "loss": 5.9759,
      "mean_token_accuracy": 0.16090452373027803,
      "num_tokens": 7683013.0,
      "step": 4115
    },
    {
      "entropy": 6.168970012664795,
      "epoch": 0.3638290356764394,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991429834733721,
      "loss": 6.0052,
      "mean_token_accuracy": 0.1668308272957802,
      "num_tokens": 7692760.0,
      "step": 4120
    },
    {
      "entropy": 6.156150960922242,
      "epoch": 0.36427057576827976,
      "grad_norm": 1.1875,
      "learning_rate": 0.000499140235295103,
      "loss": 5.9444,
      "mean_token_accuracy": 0.16200983226299287,
      "num_tokens": 7702005.0,
      "step": 4125
    },
    {
      "entropy": 6.12065167427063,
      "epoch": 0.3647121158601201,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991374827260542,
      "loss": 5.9973,
      "mean_token_accuracy": 0.16509999632835387,
      "num_tokens": 7711604.0,
      "step": 4130
    },
    {
      "entropy": 6.022671031951904,
      "epoch": 0.3651536559519604,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991347257662795,
      "loss": 5.8832,
      "mean_token_accuracy": 0.16578579545021058,
      "num_tokens": 7720434.0,
      "step": 4135
    },
    {
      "entropy": 6.131404399871826,
      "epoch": 0.36559519604380075,
      "grad_norm": 0.84765625,
      "learning_rate": 0.000499131964415833,
      "loss": 5.9468,
      "mean_token_accuracy": 0.167755126953125,
      "num_tokens": 7730733.0,
      "step": 4140
    },
    {
      "entropy": 6.139104986190796,
      "epoch": 0.36603673613564114,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004991291986747689,
      "loss": 5.9964,
      "mean_token_accuracy": 0.16309818848967553,
      "num_tokens": 7740394.0,
      "step": 4145
    },
    {
      "entropy": 6.125427103042602,
      "epoch": 0.36647827622748147,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991264285431412,
      "loss": 5.962,
      "mean_token_accuracy": 0.1680387258529663,
      "num_tokens": 7748919.0,
      "step": 4150
    },
    {
      "entropy": 6.123518419265747,
      "epoch": 0.3669198163193218,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004991236540210041,
      "loss": 5.974,
      "mean_token_accuracy": 0.16233558654785157,
      "num_tokens": 7758591.0,
      "step": 4155
    },
    {
      "entropy": 6.1502306938171385,
      "epoch": 0.36736135641116213,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991208751084122,
      "loss": 6.0083,
      "mean_token_accuracy": 0.15538841038942336,
      "num_tokens": 7768269.0,
      "step": 4160
    },
    {
      "entropy": 6.204276371002197,
      "epoch": 0.36780289650300246,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004991180918054199,
      "loss": 6.0402,
      "mean_token_accuracy": 0.15984537899494172,
      "num_tokens": 7776980.0,
      "step": 4165
    },
    {
      "entropy": 6.127771854400635,
      "epoch": 0.3682444365948428,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991153041120815,
      "loss": 5.941,
      "mean_token_accuracy": 0.1670564740896225,
      "num_tokens": 7786664.0,
      "step": 4170
    },
    {
      "entropy": 6.0249744892120365,
      "epoch": 0.3686859766866831,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004991125120284519,
      "loss": 5.8206,
      "mean_token_accuracy": 0.17915644198656083,
      "num_tokens": 7794981.0,
      "step": 4175
    },
    {
      "entropy": 6.056008577346802,
      "epoch": 0.3691275167785235,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991097155545856,
      "loss": 5.9086,
      "mean_token_accuracy": 0.16700488328933716,
      "num_tokens": 7804850.0,
      "step": 4180
    },
    {
      "entropy": 6.102103614807129,
      "epoch": 0.36956905687036384,
      "grad_norm": 1.0,
      "learning_rate": 0.0004991069146905374,
      "loss": 5.9142,
      "mean_token_accuracy": 0.1711506649851799,
      "num_tokens": 7814117.0,
      "step": 4185
    },
    {
      "entropy": 6.098058605194092,
      "epoch": 0.3700105969622042,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991041094363621,
      "loss": 5.9473,
      "mean_token_accuracy": 0.16947417855262756,
      "num_tokens": 7823384.0,
      "step": 4190
    },
    {
      "entropy": 6.091346216201782,
      "epoch": 0.3704521370540445,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004991012997921149,
      "loss": 5.9529,
      "mean_token_accuracy": 0.16536522805690765,
      "num_tokens": 7834072.0,
      "step": 4195
    },
    {
      "entropy": 6.12990288734436,
      "epoch": 0.37089367714588484,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990984857578506,
      "loss": 5.9442,
      "mean_token_accuracy": 0.16497403979301453,
      "num_tokens": 7843334.0,
      "step": 4200
    },
    {
      "entropy": 5.994000387191773,
      "epoch": 0.37133521723772517,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990956673336245,
      "loss": 5.865,
      "mean_token_accuracy": 0.17073259353637696,
      "num_tokens": 7853187.0,
      "step": 4205
    },
    {
      "entropy": 6.0792876243591305,
      "epoch": 0.3717767573295655,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990928445194917,
      "loss": 5.9617,
      "mean_token_accuracy": 0.1619952842593193,
      "num_tokens": 7862262.0,
      "step": 4210
    },
    {
      "entropy": 6.071585988998413,
      "epoch": 0.3722182974214059,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990900173155074,
      "loss": 5.8975,
      "mean_token_accuracy": 0.16614274084568023,
      "num_tokens": 7870725.0,
      "step": 4215
    },
    {
      "entropy": 6.031388998031616,
      "epoch": 0.3726598375132462,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990871857217273,
      "loss": 5.8055,
      "mean_token_accuracy": 0.1678528070449829,
      "num_tokens": 7879492.0,
      "step": 4220
    },
    {
      "entropy": 6.110537576675415,
      "epoch": 0.37310137760508655,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004990843497382066,
      "loss": 5.9115,
      "mean_token_accuracy": 0.16305805295705794,
      "num_tokens": 7888710.0,
      "step": 4225
    },
    {
      "entropy": 6.0610274314880375,
      "epoch": 0.3735429176969269,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004990815093650009,
      "loss": 5.9409,
      "mean_token_accuracy": 0.16872991621494293,
      "num_tokens": 7898313.0,
      "step": 4230
    },
    {
      "entropy": 6.102960062026978,
      "epoch": 0.3739844577887672,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004990786646021659,
      "loss": 5.9634,
      "mean_token_accuracy": 0.16352523416280745,
      "num_tokens": 7906726.0,
      "step": 4235
    },
    {
      "entropy": 6.1046594142913815,
      "epoch": 0.37442599788060754,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990758154497573,
      "loss": 5.8937,
      "mean_token_accuracy": 0.17032238245010375,
      "num_tokens": 7915554.0,
      "step": 4240
    },
    {
      "entropy": 6.101940011978149,
      "epoch": 0.3748675379724479,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004990729619078309,
      "loss": 6.0112,
      "mean_token_accuracy": 0.16436887234449388,
      "num_tokens": 7924732.0,
      "step": 4245
    },
    {
      "entropy": 6.106853342056274,
      "epoch": 0.37530907806428826,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990701039764427,
      "loss": 5.9303,
      "mean_token_accuracy": 0.1686663657426834,
      "num_tokens": 7933808.0,
      "step": 4250
    },
    {
      "entropy": 6.086512899398803,
      "epoch": 0.3757506181561286,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990672416556485,
      "loss": 5.8807,
      "mean_token_accuracy": 0.1658458963036537,
      "num_tokens": 7942558.0,
      "step": 4255
    },
    {
      "entropy": 6.064576911926269,
      "epoch": 0.3761921582479689,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004990643749455045,
      "loss": 5.9793,
      "mean_token_accuracy": 0.16173869967460633,
      "num_tokens": 7951951.0,
      "step": 4260
    },
    {
      "entropy": 6.199822521209716,
      "epoch": 0.37663369833980925,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990615038460667,
      "loss": 5.8825,
      "mean_token_accuracy": 0.16723221093416213,
      "num_tokens": 7960543.0,
      "step": 4265
    },
    {
      "entropy": 6.080804681777954,
      "epoch": 0.3770752384316496,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004990586283573916,
      "loss": 5.8682,
      "mean_token_accuracy": 0.1681565523147583,
      "num_tokens": 7970273.0,
      "step": 4270
    },
    {
      "entropy": 6.062951707839966,
      "epoch": 0.3775167785234899,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990557484795355,
      "loss": 5.9628,
      "mean_token_accuracy": 0.167630435526371,
      "num_tokens": 7978888.0,
      "step": 4275
    },
    {
      "entropy": 6.145836114883423,
      "epoch": 0.37795831861533025,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004990528642125545,
      "loss": 5.9531,
      "mean_token_accuracy": 0.15775933712720872,
      "num_tokens": 7988843.0,
      "step": 4280
    },
    {
      "entropy": 6.107465028762817,
      "epoch": 0.37839985870717063,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004990499755565055,
      "loss": 5.9345,
      "mean_token_accuracy": 0.16518889516592025,
      "num_tokens": 7998403.0,
      "step": 4285
    },
    {
      "entropy": 6.123253870010376,
      "epoch": 0.37884139879901096,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004990470825114448,
      "loss": 5.9763,
      "mean_token_accuracy": 0.1672004759311676,
      "num_tokens": 8007493.0,
      "step": 4290
    },
    {
      "entropy": 6.10452790260315,
      "epoch": 0.3792829388908513,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004990441850774292,
      "loss": 5.9169,
      "mean_token_accuracy": 0.1715584471821785,
      "num_tokens": 8016399.0,
      "step": 4295
    },
    {
      "entropy": 6.0872495651245115,
      "epoch": 0.3797244789826916,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004990412832545155,
      "loss": 5.9401,
      "mean_token_accuracy": 0.1636321425437927,
      "num_tokens": 8025891.0,
      "step": 4300
    },
    {
      "entropy": 6.022824382781982,
      "epoch": 0.38016601907453196,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990383770427603,
      "loss": 5.8622,
      "mean_token_accuracy": 0.1687710165977478,
      "num_tokens": 8034733.0,
      "step": 4305
    },
    {
      "entropy": 6.052893352508545,
      "epoch": 0.3806075591663723,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004990354664422209,
      "loss": 5.9131,
      "mean_token_accuracy": 0.17057955414056777,
      "num_tokens": 8044443.0,
      "step": 4310
    },
    {
      "entropy": 6.0984782695770265,
      "epoch": 0.3810490992582126,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990325514529541,
      "loss": 5.9031,
      "mean_token_accuracy": 0.17352878600358962,
      "num_tokens": 8053920.0,
      "step": 4315
    },
    {
      "entropy": 6.143695259094239,
      "epoch": 0.381490639350053,
      "grad_norm": 0.875,
      "learning_rate": 0.0004990296320750169,
      "loss": 5.8922,
      "mean_token_accuracy": 0.16723661124706268,
      "num_tokens": 8063455.0,
      "step": 4320
    },
    {
      "entropy": 6.028673410415649,
      "epoch": 0.38193217944189334,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990267083084667,
      "loss": 5.9811,
      "mean_token_accuracy": 0.16257134675979615,
      "num_tokens": 8073353.0,
      "step": 4325
    },
    {
      "entropy": 6.11106915473938,
      "epoch": 0.38237371953373367,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004990237801533607,
      "loss": 5.9733,
      "mean_token_accuracy": 0.17016141563653947,
      "num_tokens": 8082164.0,
      "step": 4330
    },
    {
      "entropy": 6.187211561203003,
      "epoch": 0.382815259625574,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004990208476097562,
      "loss": 5.993,
      "mean_token_accuracy": 0.16434455662965775,
      "num_tokens": 8090790.0,
      "step": 4335
    },
    {
      "entropy": 6.228871631622314,
      "epoch": 0.38325679971741433,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990179106777109,
      "loss": 6.0236,
      "mean_token_accuracy": 0.1600963294506073,
      "num_tokens": 8100412.0,
      "step": 4340
    },
    {
      "entropy": 6.002315425872803,
      "epoch": 0.38369833980925466,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990149693572819,
      "loss": 5.982,
      "mean_token_accuracy": 0.16522495746612548,
      "num_tokens": 8110002.0,
      "step": 4345
    },
    {
      "entropy": 6.066628885269165,
      "epoch": 0.384139879901095,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004990120236485271,
      "loss": 5.8771,
      "mean_token_accuracy": 0.16987511217594148,
      "num_tokens": 8118898.0,
      "step": 4350
    },
    {
      "entropy": 6.056738805770874,
      "epoch": 0.3845814199929354,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990090735515043,
      "loss": 5.9422,
      "mean_token_accuracy": 0.15970152467489243,
      "num_tokens": 8128207.0,
      "step": 4355
    },
    {
      "entropy": 6.094995737075806,
      "epoch": 0.3850229600847757,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000499006119066271,
      "loss": 5.9157,
      "mean_token_accuracy": 0.16543019860982894,
      "num_tokens": 8137317.0,
      "step": 4360
    },
    {
      "entropy": 6.137070083618164,
      "epoch": 0.38546450017661604,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004990031601928854,
      "loss": 5.9609,
      "mean_token_accuracy": 0.16855712682008744,
      "num_tokens": 8146519.0,
      "step": 4365
    },
    {
      "entropy": 6.0023377418518065,
      "epoch": 0.3859060402684564,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004990001969314051,
      "loss": 5.9722,
      "mean_token_accuracy": 0.16350623667240144,
      "num_tokens": 8155916.0,
      "step": 4370
    },
    {
      "entropy": 6.148402261734009,
      "epoch": 0.3863475803602967,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004989972292818884,
      "loss": 5.9791,
      "mean_token_accuracy": 0.16953045129776,
      "num_tokens": 8165251.0,
      "step": 4375
    },
    {
      "entropy": 6.093685007095337,
      "epoch": 0.38678912045213704,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004989942572443934,
      "loss": 5.973,
      "mean_token_accuracy": 0.16652768552303315,
      "num_tokens": 8174761.0,
      "step": 4380
    },
    {
      "entropy": 6.088238048553467,
      "epoch": 0.38723066054397737,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989912808189784,
      "loss": 5.9255,
      "mean_token_accuracy": 0.1713149756193161,
      "num_tokens": 8182980.0,
      "step": 4385
    },
    {
      "entropy": 6.1683008670806885,
      "epoch": 0.38767220063581775,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989883000057013,
      "loss": 5.9693,
      "mean_token_accuracy": 0.15927564948797227,
      "num_tokens": 8193306.0,
      "step": 4390
    },
    {
      "entropy": 6.058181524276733,
      "epoch": 0.3881137407276581,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000498985314804621,
      "loss": 5.8593,
      "mean_token_accuracy": 0.16985781043767928,
      "num_tokens": 8203446.0,
      "step": 4395
    },
    {
      "entropy": 6.02800440788269,
      "epoch": 0.3885552808194984,
      "grad_norm": 0.875,
      "learning_rate": 0.0004989823252157958,
      "loss": 5.9169,
      "mean_token_accuracy": 0.16310855597257615,
      "num_tokens": 8212414.0,
      "step": 4400
    },
    {
      "entropy": 6.018846893310547,
      "epoch": 0.38899682091133875,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004989793312392841,
      "loss": 5.8359,
      "mean_token_accuracy": 0.17101139277219773,
      "num_tokens": 8221509.0,
      "step": 4405
    },
    {
      "entropy": 6.062700700759888,
      "epoch": 0.3894383610031791,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004989763328751448,
      "loss": 5.9423,
      "mean_token_accuracy": 0.1717799872159958,
      "num_tokens": 8230017.0,
      "step": 4410
    },
    {
      "entropy": 6.0845201969146725,
      "epoch": 0.3898799010950194,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989733301234365,
      "loss": 5.8743,
      "mean_token_accuracy": 0.168140971660614,
      "num_tokens": 8238853.0,
      "step": 4415
    },
    {
      "entropy": 6.119135093688965,
      "epoch": 0.39032144118685974,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000498970322984218,
      "loss": 5.9573,
      "mean_token_accuracy": 0.1616509214043617,
      "num_tokens": 8248112.0,
      "step": 4420
    },
    {
      "entropy": 6.067855501174927,
      "epoch": 0.39076298127870013,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004989673114575483,
      "loss": 5.9268,
      "mean_token_accuracy": 0.16173332184553146,
      "num_tokens": 8257166.0,
      "step": 4425
    },
    {
      "entropy": 6.099668502807617,
      "epoch": 0.39120452137054046,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004989642955434863,
      "loss": 5.9336,
      "mean_token_accuracy": 0.1655343070626259,
      "num_tokens": 8266229.0,
      "step": 4430
    },
    {
      "entropy": 6.116667413711548,
      "epoch": 0.3916460614623808,
      "grad_norm": 0.875,
      "learning_rate": 0.0004989612752420912,
      "loss": 5.9014,
      "mean_token_accuracy": 0.16495247632265092,
      "num_tokens": 8275102.0,
      "step": 4435
    },
    {
      "entropy": 5.985392999649048,
      "epoch": 0.3920876015542211,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498958250553422,
      "loss": 5.8257,
      "mean_token_accuracy": 0.17586065083742142,
      "num_tokens": 8283847.0,
      "step": 4440
    },
    {
      "entropy": 6.055251932144165,
      "epoch": 0.39252914164606145,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004989552214775381,
      "loss": 5.9606,
      "mean_token_accuracy": 0.16416369527578353,
      "num_tokens": 8292622.0,
      "step": 4445
    },
    {
      "entropy": 6.125879526138306,
      "epoch": 0.3929706817379018,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989521880144988,
      "loss": 5.8634,
      "mean_token_accuracy": 0.17751139402389526,
      "num_tokens": 8301026.0,
      "step": 4450
    },
    {
      "entropy": 6.148255681991577,
      "epoch": 0.3934122218297421,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989491501643635,
      "loss": 6.1177,
      "mean_token_accuracy": 0.15533054769039153,
      "num_tokens": 8310977.0,
      "step": 4455
    },
    {
      "entropy": 6.070802021026611,
      "epoch": 0.3938537619215825,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989461079271916,
      "loss": 5.8891,
      "mean_token_accuracy": 0.17560941129922866,
      "num_tokens": 8319391.0,
      "step": 4460
    },
    {
      "entropy": 6.047619771957398,
      "epoch": 0.39429530201342283,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989430613030429,
      "loss": 5.8413,
      "mean_token_accuracy": 0.16868884712457657,
      "num_tokens": 8328639.0,
      "step": 4465
    },
    {
      "entropy": 6.128910255432129,
      "epoch": 0.39473684210526316,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000498940010291977,
      "loss": 5.9599,
      "mean_token_accuracy": 0.16220096349716187,
      "num_tokens": 8338190.0,
      "step": 4470
    },
    {
      "entropy": 6.021871757507324,
      "epoch": 0.3951783821971035,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004989369548940536,
      "loss": 5.8946,
      "mean_token_accuracy": 0.16686442196369172,
      "num_tokens": 8346874.0,
      "step": 4475
    },
    {
      "entropy": 6.107971143722534,
      "epoch": 0.3956199222889438,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989338951093327,
      "loss": 5.8898,
      "mean_token_accuracy": 0.17452280074357987,
      "num_tokens": 8356446.0,
      "step": 4480
    },
    {
      "entropy": 6.077869462966919,
      "epoch": 0.39606146238078416,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004989308309378741,
      "loss": 5.8862,
      "mean_token_accuracy": 0.17257023006677627,
      "num_tokens": 8365694.0,
      "step": 4485
    },
    {
      "entropy": 6.063984346389771,
      "epoch": 0.3965030024726245,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989277623797379,
      "loss": 5.9406,
      "mean_token_accuracy": 0.16608838438987733,
      "num_tokens": 8374282.0,
      "step": 4490
    },
    {
      "entropy": 6.068101119995117,
      "epoch": 0.3969445425644649,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989246894349841,
      "loss": 5.908,
      "mean_token_accuracy": 0.16799129843711852,
      "num_tokens": 8383315.0,
      "step": 4495
    },
    {
      "entropy": 6.085828971862793,
      "epoch": 0.3973860826563052,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989216121036732,
      "loss": 5.8716,
      "mean_token_accuracy": 0.16725074946880342,
      "num_tokens": 8392263.0,
      "step": 4500
    },
    {
      "entropy": 6.094727611541748,
      "epoch": 0.39782762274814554,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004989185303858651,
      "loss": 5.9052,
      "mean_token_accuracy": 0.17103357762098312,
      "num_tokens": 8400734.0,
      "step": 4505
    },
    {
      "entropy": 6.054353046417236,
      "epoch": 0.39826916283998587,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989154442816203,
      "loss": 5.9632,
      "mean_token_accuracy": 0.16649849265813826,
      "num_tokens": 8410635.0,
      "step": 4510
    },
    {
      "entropy": 6.089747953414917,
      "epoch": 0.3987107029318262,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004989123537909994,
      "loss": 5.9367,
      "mean_token_accuracy": 0.1698456108570099,
      "num_tokens": 8420111.0,
      "step": 4515
    },
    {
      "entropy": 6.013007402420044,
      "epoch": 0.39915224302366653,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989092589140629,
      "loss": 5.8902,
      "mean_token_accuracy": 0.1611982986330986,
      "num_tokens": 8429459.0,
      "step": 4520
    },
    {
      "entropy": 6.087127161026001,
      "epoch": 0.39959378311550686,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004989061596508712,
      "loss": 5.8813,
      "mean_token_accuracy": 0.17483796626329423,
      "num_tokens": 8438083.0,
      "step": 4525
    },
    {
      "entropy": 6.033258438110352,
      "epoch": 0.40003532320734725,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004989030560014853,
      "loss": 5.9093,
      "mean_token_accuracy": 0.17043748795986174,
      "num_tokens": 8447713.0,
      "step": 4530
    },
    {
      "entropy": 6.111389303207398,
      "epoch": 0.4004768632991876,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988999479659657,
      "loss": 5.88,
      "mean_token_accuracy": 0.16914622038602828,
      "num_tokens": 8457394.0,
      "step": 4535
    },
    {
      "entropy": 6.061598920822144,
      "epoch": 0.4009184033910279,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004988968355443737,
      "loss": 5.8502,
      "mean_token_accuracy": 0.17563896626234055,
      "num_tokens": 8467333.0,
      "step": 4540
    },
    {
      "entropy": 6.019485664367676,
      "epoch": 0.40135994348286824,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988937187367699,
      "loss": 5.9683,
      "mean_token_accuracy": 0.16143568158149718,
      "num_tokens": 8477530.0,
      "step": 4545
    },
    {
      "entropy": 6.126996564865112,
      "epoch": 0.4018014835747086,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988905975432154,
      "loss": 6.0199,
      "mean_token_accuracy": 0.16645699143409728,
      "num_tokens": 8486861.0,
      "step": 4550
    },
    {
      "entropy": 6.076916742324829,
      "epoch": 0.4022430236665489,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988874719637715,
      "loss": 5.8415,
      "mean_token_accuracy": 0.17304785847663878,
      "num_tokens": 8496541.0,
      "step": 4555
    },
    {
      "entropy": 6.047940254211426,
      "epoch": 0.40268456375838924,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988843419984994,
      "loss": 5.9254,
      "mean_token_accuracy": 0.1652942880988121,
      "num_tokens": 8505667.0,
      "step": 4560
    },
    {
      "entropy": 6.153884840011597,
      "epoch": 0.4031261038502296,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988812076474604,
      "loss": 5.8971,
      "mean_token_accuracy": 0.16712863743305206,
      "num_tokens": 8515133.0,
      "step": 4565
    },
    {
      "entropy": 6.094613790512085,
      "epoch": 0.40356764394206995,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988780689107158,
      "loss": 5.87,
      "mean_token_accuracy": 0.17001326382160187,
      "num_tokens": 8524797.0,
      "step": 4570
    },
    {
      "entropy": 5.936724328994751,
      "epoch": 0.4040091840339103,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004988749257883271,
      "loss": 5.893,
      "mean_token_accuracy": 0.1669435977935791,
      "num_tokens": 8534667.0,
      "step": 4575
    },
    {
      "entropy": 6.046252727508545,
      "epoch": 0.4044507241257506,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000498871778280356,
      "loss": 5.9061,
      "mean_token_accuracy": 0.16667770594358444,
      "num_tokens": 8543874.0,
      "step": 4580
    },
    {
      "entropy": 6.062067222595215,
      "epoch": 0.40489226421759095,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988686263868641,
      "loss": 5.8976,
      "mean_token_accuracy": 0.1650629997253418,
      "num_tokens": 8553620.0,
      "step": 4585
    },
    {
      "entropy": 6.07551589012146,
      "epoch": 0.4053338043094313,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004988654701079131,
      "loss": 5.9763,
      "mean_token_accuracy": 0.16740324944257737,
      "num_tokens": 8563857.0,
      "step": 4590
    },
    {
      "entropy": 6.113534784317016,
      "epoch": 0.4057753444012716,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004988623094435649,
      "loss": 5.9471,
      "mean_token_accuracy": 0.16143951714038848,
      "num_tokens": 8572677.0,
      "step": 4595
    },
    {
      "entropy": 6.016867923736572,
      "epoch": 0.406216884493112,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004988591443938813,
      "loss": 5.8775,
      "mean_token_accuracy": 0.1681135416030884,
      "num_tokens": 8581907.0,
      "step": 4600
    },
    {
      "entropy": 6.116613101959229,
      "epoch": 0.4066584245849523,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988559749589244,
      "loss": 5.9016,
      "mean_token_accuracy": 0.16940733939409255,
      "num_tokens": 8591701.0,
      "step": 4605
    },
    {
      "entropy": 6.058897066116333,
      "epoch": 0.40709996467679266,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988528011387563,
      "loss": 5.9476,
      "mean_token_accuracy": 0.170088429749012,
      "num_tokens": 8601054.0,
      "step": 4610
    },
    {
      "entropy": 6.01906566619873,
      "epoch": 0.407541504768633,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004988496229334392,
      "loss": 5.8599,
      "mean_token_accuracy": 0.1751582607626915,
      "num_tokens": 8610346.0,
      "step": 4615
    },
    {
      "entropy": 6.039183759689331,
      "epoch": 0.4079830448604733,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004988464403430352,
      "loss": 5.8486,
      "mean_token_accuracy": 0.1733414739370346,
      "num_tokens": 8620823.0,
      "step": 4620
    },
    {
      "entropy": 6.101271772384644,
      "epoch": 0.40842458495231365,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988432533676067,
      "loss": 6.0104,
      "mean_token_accuracy": 0.15902107506990432,
      "num_tokens": 8630184.0,
      "step": 4625
    },
    {
      "entropy": 6.092225360870361,
      "epoch": 0.408866125044154,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004988400620072163,
      "loss": 5.9966,
      "mean_token_accuracy": 0.15500882863998414,
      "num_tokens": 8640064.0,
      "step": 4630
    },
    {
      "entropy": 6.104148864746094,
      "epoch": 0.40930766513599437,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004988368662619263,
      "loss": 5.8964,
      "mean_token_accuracy": 0.16778270304203033,
      "num_tokens": 8650503.0,
      "step": 4635
    },
    {
      "entropy": 6.148624897003174,
      "epoch": 0.4097492052278347,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988336661317994,
      "loss": 5.9207,
      "mean_token_accuracy": 0.16135725378990173,
      "num_tokens": 8659125.0,
      "step": 4640
    },
    {
      "entropy": 6.012713432312012,
      "epoch": 0.41019074531967503,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004988304616168984,
      "loss": 5.946,
      "mean_token_accuracy": 0.16297379732131959,
      "num_tokens": 8668193.0,
      "step": 4645
    },
    {
      "entropy": 6.092987585067749,
      "epoch": 0.41063228541151536,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004988272527172858,
      "loss": 5.8977,
      "mean_token_accuracy": 0.16841595321893693,
      "num_tokens": 8677515.0,
      "step": 4650
    },
    {
      "entropy": 6.112038326263428,
      "epoch": 0.4110738255033557,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004988240394330246,
      "loss": 5.9334,
      "mean_token_accuracy": 0.16429426819086074,
      "num_tokens": 8687549.0,
      "step": 4655
    },
    {
      "entropy": 6.041188097000122,
      "epoch": 0.411515365595196,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004988208217641778,
      "loss": 5.9235,
      "mean_token_accuracy": 0.1723731204867363,
      "num_tokens": 8697371.0,
      "step": 4660
    },
    {
      "entropy": 6.059947109222412,
      "epoch": 0.41195690568703636,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004988175997108086,
      "loss": 5.9241,
      "mean_token_accuracy": 0.16697203665971755,
      "num_tokens": 8707193.0,
      "step": 4665
    },
    {
      "entropy": 6.085496950149536,
      "epoch": 0.41239844577887674,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004988143732729797,
      "loss": 5.9383,
      "mean_token_accuracy": 0.16928174197673798,
      "num_tokens": 8716052.0,
      "step": 4670
    },
    {
      "entropy": 6.014765739440918,
      "epoch": 0.4128399858707171,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988111424507546,
      "loss": 5.8757,
      "mean_token_accuracy": 0.17379332482814788,
      "num_tokens": 8726140.0,
      "step": 4675
    },
    {
      "entropy": 6.012660837173462,
      "epoch": 0.4132815259625574,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004988079072441964,
      "loss": 5.8323,
      "mean_token_accuracy": 0.1784330740571022,
      "num_tokens": 8735299.0,
      "step": 4680
    },
    {
      "entropy": 6.058017349243164,
      "epoch": 0.41372306605439774,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988046676533687,
      "loss": 5.9309,
      "mean_token_accuracy": 0.16514215022325515,
      "num_tokens": 8744686.0,
      "step": 4685
    },
    {
      "entropy": 6.061055088043213,
      "epoch": 0.41416460614623807,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004988014236783347,
      "loss": 5.9154,
      "mean_token_accuracy": 0.17677023261785507,
      "num_tokens": 8754942.0,
      "step": 4690
    },
    {
      "entropy": 6.153789758682251,
      "epoch": 0.4146061462380784,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004987981753191582,
      "loss": 5.9172,
      "mean_token_accuracy": 0.16312515884637832,
      "num_tokens": 8764054.0,
      "step": 4695
    },
    {
      "entropy": 5.930011558532715,
      "epoch": 0.41504768632991873,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987949225759027,
      "loss": 5.7808,
      "mean_token_accuracy": 0.17439974397420882,
      "num_tokens": 8773050.0,
      "step": 4700
    },
    {
      "entropy": 6.110511398315429,
      "epoch": 0.4154892264217591,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987916654486321,
      "loss": 5.9053,
      "mean_token_accuracy": 0.16541254967451097,
      "num_tokens": 8782476.0,
      "step": 4705
    },
    {
      "entropy": 6.106545448303223,
      "epoch": 0.41593076651359945,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004987884039374099,
      "loss": 6.0026,
      "mean_token_accuracy": 0.16032444387674333,
      "num_tokens": 8791147.0,
      "step": 4710
    },
    {
      "entropy": 6.089293193817139,
      "epoch": 0.4163723066054398,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004987851380423001,
      "loss": 5.962,
      "mean_token_accuracy": 0.1648245297372341,
      "num_tokens": 8801151.0,
      "step": 4715
    },
    {
      "entropy": 6.077416801452637,
      "epoch": 0.4168138466972801,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987818677633668,
      "loss": 5.914,
      "mean_token_accuracy": 0.1638292446732521,
      "num_tokens": 8809587.0,
      "step": 4720
    },
    {
      "entropy": 5.994811487197876,
      "epoch": 0.41725538678912044,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498778593100674,
      "loss": 5.7869,
      "mean_token_accuracy": 0.17192702740430832,
      "num_tokens": 8818168.0,
      "step": 4725
    },
    {
      "entropy": 5.973134136199951,
      "epoch": 0.4176969268809608,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004987753140542857,
      "loss": 5.8466,
      "mean_token_accuracy": 0.1726938620209694,
      "num_tokens": 8827477.0,
      "step": 4730
    },
    {
      "entropy": 6.047493362426758,
      "epoch": 0.4181384669728011,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987720306242664,
      "loss": 5.8354,
      "mean_token_accuracy": 0.17470157742500306,
      "num_tokens": 8837067.0,
      "step": 4735
    },
    {
      "entropy": 6.049209785461426,
      "epoch": 0.4185800070646415,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987687428106803,
      "loss": 5.9488,
      "mean_token_accuracy": 0.1697684958577156,
      "num_tokens": 8845790.0,
      "step": 4740
    },
    {
      "entropy": 6.073341274261475,
      "epoch": 0.4190215471564818,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004987654506135917,
      "loss": 5.882,
      "mean_token_accuracy": 0.17220733910799027,
      "num_tokens": 8855242.0,
      "step": 4745
    },
    {
      "entropy": 6.10594334602356,
      "epoch": 0.41946308724832215,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987621540330652,
      "loss": 5.906,
      "mean_token_accuracy": 0.17124312967061997,
      "num_tokens": 8864459.0,
      "step": 4750
    },
    {
      "entropy": 6.104751348495483,
      "epoch": 0.4199046273401625,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004987588530691653,
      "loss": 5.9673,
      "mean_token_accuracy": 0.16231338381767274,
      "num_tokens": 8875028.0,
      "step": 4755
    },
    {
      "entropy": 5.991171360015869,
      "epoch": 0.4203461674320028,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004987555477219569,
      "loss": 5.8217,
      "mean_token_accuracy": 0.17759732604026796,
      "num_tokens": 8883857.0,
      "step": 4760
    },
    {
      "entropy": 6.0481287956237795,
      "epoch": 0.42078770752384315,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004987522379915045,
      "loss": 5.7978,
      "mean_token_accuracy": 0.17629640698432922,
      "num_tokens": 8893499.0,
      "step": 4765
    },
    {
      "entropy": 6.0255521774292,
      "epoch": 0.4212292476156835,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000498748923877873,
      "loss": 5.8209,
      "mean_token_accuracy": 0.17817992717027664,
      "num_tokens": 8903368.0,
      "step": 4770
    },
    {
      "entropy": 6.027704524993896,
      "epoch": 0.42167078770752386,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987456053811273,
      "loss": 5.9307,
      "mean_token_accuracy": 0.16877609640359878,
      "num_tokens": 8912701.0,
      "step": 4775
    },
    {
      "entropy": 6.072631692886352,
      "epoch": 0.4221123277993642,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004987422825013325,
      "loss": 5.879,
      "mean_token_accuracy": 0.16829278767108918,
      "num_tokens": 8921962.0,
      "step": 4780
    },
    {
      "entropy": 6.109710693359375,
      "epoch": 0.4225538678912045,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987389552385536,
      "loss": 5.8662,
      "mean_token_accuracy": 0.17203690707683564,
      "num_tokens": 8931923.0,
      "step": 4785
    },
    {
      "entropy": 6.003576421737671,
      "epoch": 0.42299540798304486,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004987356235928558,
      "loss": 5.9099,
      "mean_token_accuracy": 0.16035820543766022,
      "num_tokens": 8940403.0,
      "step": 4790
    },
    {
      "entropy": 6.0650804996490475,
      "epoch": 0.4234369480748852,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987322875643044,
      "loss": 5.8359,
      "mean_token_accuracy": 0.1735415205359459,
      "num_tokens": 8949377.0,
      "step": 4795
    },
    {
      "entropy": 6.1672426700592045,
      "epoch": 0.4238784881667255,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987289471529647,
      "loss": 5.9921,
      "mean_token_accuracy": 0.15903109014034272,
      "num_tokens": 8958719.0,
      "step": 4800
    },
    {
      "entropy": 6.056859874725342,
      "epoch": 0.42432002825856585,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004987256023589022,
      "loss": 5.9481,
      "mean_token_accuracy": 0.15579934269189835,
      "num_tokens": 8968226.0,
      "step": 4805
    },
    {
      "entropy": 6.062025690078736,
      "epoch": 0.42476156835040624,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004987222531821824,
      "loss": 5.8949,
      "mean_token_accuracy": 0.16687395423650742,
      "num_tokens": 8976670.0,
      "step": 4810
    },
    {
      "entropy": 5.9757143497467045,
      "epoch": 0.42520310844224657,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987188996228709,
      "loss": 5.8281,
      "mean_token_accuracy": 0.17414531707763672,
      "num_tokens": 8986185.0,
      "step": 4815
    },
    {
      "entropy": 6.0142176151275635,
      "epoch": 0.4256446485340869,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004987155416810334,
      "loss": 5.8315,
      "mean_token_accuracy": 0.17239081412553786,
      "num_tokens": 8995124.0,
      "step": 4820
    },
    {
      "entropy": 6.050481128692627,
      "epoch": 0.42608618862592723,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004987121793567356,
      "loss": 5.9421,
      "mean_token_accuracy": 0.1673525243997574,
      "num_tokens": 9004380.0,
      "step": 4825
    },
    {
      "entropy": 6.11012635231018,
      "epoch": 0.42652772871776756,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004987088126500436,
      "loss": 5.8117,
      "mean_token_accuracy": 0.17067271023988723,
      "num_tokens": 9013791.0,
      "step": 4830
    },
    {
      "entropy": 6.019365072250366,
      "epoch": 0.4269692688096079,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498705441561023,
      "loss": 5.8511,
      "mean_token_accuracy": 0.17690863311290742,
      "num_tokens": 9023076.0,
      "step": 4835
    },
    {
      "entropy": 6.13967719078064,
      "epoch": 0.4274108089014482,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004987020660897401,
      "loss": 5.9362,
      "mean_token_accuracy": 0.16226004511117936,
      "num_tokens": 9032720.0,
      "step": 4840
    },
    {
      "entropy": 6.107657623291016,
      "epoch": 0.4278523489932886,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000498698686236261,
      "loss": 5.908,
      "mean_token_accuracy": 0.16728319376707076,
      "num_tokens": 9042652.0,
      "step": 4845
    },
    {
      "entropy": 6.1251993656158445,
      "epoch": 0.42829388908512894,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986953020006519,
      "loss": 5.9835,
      "mean_token_accuracy": 0.16414731591939927,
      "num_tokens": 9052172.0,
      "step": 4850
    },
    {
      "entropy": 6.0267579555511475,
      "epoch": 0.4287354291769693,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004986919133829788,
      "loss": 5.8619,
      "mean_token_accuracy": 0.1743416041135788,
      "num_tokens": 9061798.0,
      "step": 4855
    },
    {
      "entropy": 5.954681015014648,
      "epoch": 0.4291769692688096,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004986885203833086,
      "loss": 5.7878,
      "mean_token_accuracy": 0.17132241576910018,
      "num_tokens": 9070429.0,
      "step": 4860
    },
    {
      "entropy": 6.0399816036224365,
      "epoch": 0.42961850936064994,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986851230017075,
      "loss": 5.8992,
      "mean_token_accuracy": 0.1731168210506439,
      "num_tokens": 9080200.0,
      "step": 4865
    },
    {
      "entropy": 6.017161512374878,
      "epoch": 0.43006004945249027,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986817212382419,
      "loss": 5.8932,
      "mean_token_accuracy": 0.1726991668343544,
      "num_tokens": 9089555.0,
      "step": 4870
    },
    {
      "entropy": 6.037603855133057,
      "epoch": 0.4305015895443306,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004986783150929786,
      "loss": 5.8551,
      "mean_token_accuracy": 0.16833554804325104,
      "num_tokens": 9099091.0,
      "step": 4875
    },
    {
      "entropy": 6.059744453430175,
      "epoch": 0.430943129636171,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004986749045659845,
      "loss": 5.956,
      "mean_token_accuracy": 0.15879134684801102,
      "num_tokens": 9109008.0,
      "step": 4880
    },
    {
      "entropy": 6.035340404510498,
      "epoch": 0.4313846697280113,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986714896573261,
      "loss": 5.8079,
      "mean_token_accuracy": 0.16968925893306733,
      "num_tokens": 9117882.0,
      "step": 4885
    },
    {
      "entropy": 6.090554904937744,
      "epoch": 0.43182620981985165,
      "grad_norm": 1.0,
      "learning_rate": 0.0004986680703670704,
      "loss": 5.9384,
      "mean_token_accuracy": 0.1664467602968216,
      "num_tokens": 9126860.0,
      "step": 4890
    },
    {
      "entropy": 6.059638452529907,
      "epoch": 0.432267749911692,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986646466952845,
      "loss": 5.8479,
      "mean_token_accuracy": 0.1714713305234909,
      "num_tokens": 9135819.0,
      "step": 4895
    },
    {
      "entropy": 6.017725610733033,
      "epoch": 0.4327092900035323,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986612186420353,
      "loss": 5.8381,
      "mean_token_accuracy": 0.17259113788604735,
      "num_tokens": 9145302.0,
      "step": 4900
    },
    {
      "entropy": 6.003063488006592,
      "epoch": 0.43315083009537264,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986577862073901,
      "loss": 5.8874,
      "mean_token_accuracy": 0.16548449397087098,
      "num_tokens": 9154667.0,
      "step": 4905
    },
    {
      "entropy": 6.125681400299072,
      "epoch": 0.43359237018721297,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004986543493914159,
      "loss": 5.9333,
      "mean_token_accuracy": 0.15938420295715333,
      "num_tokens": 9164562.0,
      "step": 4910
    },
    {
      "entropy": 6.060342788696289,
      "epoch": 0.43403391027905336,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004986509081941805,
      "loss": 5.958,
      "mean_token_accuracy": 0.16365616768598557,
      "num_tokens": 9174872.0,
      "step": 4915
    },
    {
      "entropy": 5.960330772399902,
      "epoch": 0.4344754503708937,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986474626157507,
      "loss": 5.6941,
      "mean_token_accuracy": 0.1887632668018341,
      "num_tokens": 9184322.0,
      "step": 4920
    },
    {
      "entropy": 5.985001468658448,
      "epoch": 0.434916990462734,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986440126561945,
      "loss": 5.8471,
      "mean_token_accuracy": 0.17162098586559296,
      "num_tokens": 9194450.0,
      "step": 4925
    },
    {
      "entropy": 6.089518976211548,
      "epoch": 0.43535853055457435,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986405583155792,
      "loss": 5.8902,
      "mean_token_accuracy": 0.16652678400278093,
      "num_tokens": 9203658.0,
      "step": 4930
    },
    {
      "entropy": 6.0208498477935795,
      "epoch": 0.4358000706464147,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004986370995939725,
      "loss": 5.8667,
      "mean_token_accuracy": 0.1689656123518944,
      "num_tokens": 9213609.0,
      "step": 4935
    },
    {
      "entropy": 6.101765394210815,
      "epoch": 0.436241610738255,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004986336364914423,
      "loss": 5.8749,
      "mean_token_accuracy": 0.16693566143512725,
      "num_tokens": 9222704.0,
      "step": 4940
    },
    {
      "entropy": 6.069900178909302,
      "epoch": 0.43668315083009535,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986301690080564,
      "loss": 5.8994,
      "mean_token_accuracy": 0.16185810565948486,
      "num_tokens": 9231599.0,
      "step": 4945
    },
    {
      "entropy": 6.067622566223145,
      "epoch": 0.43712469092193573,
      "grad_norm": 0.875,
      "learning_rate": 0.0004986266971438826,
      "loss": 5.8988,
      "mean_token_accuracy": 0.1730104938149452,
      "num_tokens": 9241886.0,
      "step": 4950
    },
    {
      "entropy": 6.030080842971802,
      "epoch": 0.43756623101377606,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000498623220898989,
      "loss": 5.9147,
      "mean_token_accuracy": 0.1641816407442093,
      "num_tokens": 9251085.0,
      "step": 4955
    },
    {
      "entropy": 6.146408605575561,
      "epoch": 0.4380077711056164,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004986197402734436,
      "loss": 5.881,
      "mean_token_accuracy": 0.16509923040866853,
      "num_tokens": 9259601.0,
      "step": 4960
    },
    {
      "entropy": 6.047398519515991,
      "epoch": 0.4384493111974567,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986162552673148,
      "loss": 5.9035,
      "mean_token_accuracy": 0.17131353914737701,
      "num_tokens": 9268935.0,
      "step": 4965
    },
    {
      "entropy": 6.043163251876831,
      "epoch": 0.43889085128929706,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986127658806706,
      "loss": 5.9097,
      "mean_token_accuracy": 0.1679410979151726,
      "num_tokens": 9277647.0,
      "step": 4970
    },
    {
      "entropy": 6.058499574661255,
      "epoch": 0.4393323913811374,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004986092721135796,
      "loss": 5.8651,
      "mean_token_accuracy": 0.1684943124651909,
      "num_tokens": 9286610.0,
      "step": 4975
    },
    {
      "entropy": 6.079809713363647,
      "epoch": 0.4397739314729777,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986057739661101,
      "loss": 5.9556,
      "mean_token_accuracy": 0.16726439744234084,
      "num_tokens": 9295946.0,
      "step": 4980
    },
    {
      "entropy": 5.996550035476685,
      "epoch": 0.4402154715648181,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004986022714383307,
      "loss": 5.8543,
      "mean_token_accuracy": 0.16506249457597733,
      "num_tokens": 9304903.0,
      "step": 4985
    },
    {
      "entropy": 6.0311790943145756,
      "epoch": 0.44065701165665844,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985987645303099,
      "loss": 5.7545,
      "mean_token_accuracy": 0.17690206319093704,
      "num_tokens": 9313606.0,
      "step": 4990
    },
    {
      "entropy": 5.978774547576904,
      "epoch": 0.44109855174849877,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004985952532421164,
      "loss": 5.8664,
      "mean_token_accuracy": 0.16727809906005858,
      "num_tokens": 9322815.0,
      "step": 4995
    },
    {
      "entropy": 6.001399135589599,
      "epoch": 0.4415400918403391,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985917375738193,
      "loss": 5.7948,
      "mean_token_accuracy": 0.16886604577302933,
      "num_tokens": 9332630.0,
      "step": 5000
    },
    {
      "entropy": 5.953770589828491,
      "epoch": 0.44198163193217943,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985882175254871,
      "loss": 5.8331,
      "mean_token_accuracy": 0.16362490952014924,
      "num_tokens": 9342216.0,
      "step": 5005
    },
    {
      "entropy": 6.037289476394653,
      "epoch": 0.44242317202401976,
      "grad_norm": 0.875,
      "learning_rate": 0.0004985846930971887,
      "loss": 5.8741,
      "mean_token_accuracy": 0.16757041811943055,
      "num_tokens": 9352295.0,
      "step": 5010
    },
    {
      "entropy": 6.042323303222656,
      "epoch": 0.4428647121158601,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985811642889937,
      "loss": 5.8725,
      "mean_token_accuracy": 0.1704294890165329,
      "num_tokens": 9361462.0,
      "step": 5015
    },
    {
      "entropy": 6.0555445671081545,
      "epoch": 0.4433062522077005,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004985776311009705,
      "loss": 5.8104,
      "mean_token_accuracy": 0.16330993324518203,
      "num_tokens": 9370379.0,
      "step": 5020
    },
    {
      "entropy": 6.038263320922852,
      "epoch": 0.4437477922995408,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004985740935331888,
      "loss": 5.8682,
      "mean_token_accuracy": 0.17026212811470032,
      "num_tokens": 9379255.0,
      "step": 5025
    },
    {
      "entropy": 6.011457347869873,
      "epoch": 0.44418933239138114,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004985705515857177,
      "loss": 5.8265,
      "mean_token_accuracy": 0.16707434952259065,
      "num_tokens": 9389313.0,
      "step": 5030
    },
    {
      "entropy": 5.883806419372559,
      "epoch": 0.4446308724832215,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985670052586268,
      "loss": 5.6301,
      "mean_token_accuracy": 0.1897162079811096,
      "num_tokens": 9397778.0,
      "step": 5035
    },
    {
      "entropy": 6.027179956436157,
      "epoch": 0.4450724125750618,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004985634545519853,
      "loss": 5.9104,
      "mean_token_accuracy": 0.16823799014091492,
      "num_tokens": 9407831.0,
      "step": 5040
    },
    {
      "entropy": 6.010887002944946,
      "epoch": 0.44551395266690214,
      "grad_norm": 1.125,
      "learning_rate": 0.0004985598994658629,
      "loss": 5.8031,
      "mean_token_accuracy": 0.17146655917167664,
      "num_tokens": 9418458.0,
      "step": 5045
    },
    {
      "entropy": 5.885775423049926,
      "epoch": 0.44595549275874247,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004985563400003291,
      "loss": 5.7187,
      "mean_token_accuracy": 0.18735787272453308,
      "num_tokens": 9426911.0,
      "step": 5050
    },
    {
      "entropy": 5.971336841583252,
      "epoch": 0.44639703285058285,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985527761554539,
      "loss": 5.9022,
      "mean_token_accuracy": 0.17282261699438095,
      "num_tokens": 9435896.0,
      "step": 5055
    },
    {
      "entropy": 6.102570009231568,
      "epoch": 0.4468385729424232,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000498549207931307,
      "loss": 5.8913,
      "mean_token_accuracy": 0.16545256823301316,
      "num_tokens": 9445272.0,
      "step": 5060
    },
    {
      "entropy": 6.019857549667359,
      "epoch": 0.4472801130342635,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985456353279581,
      "loss": 5.872,
      "mean_token_accuracy": 0.16560457795858383,
      "num_tokens": 9455377.0,
      "step": 5065
    },
    {
      "entropy": 6.038298034667969,
      "epoch": 0.44772165312610385,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004985420583454774,
      "loss": 5.9011,
      "mean_token_accuracy": 0.16640974432229996,
      "num_tokens": 9464918.0,
      "step": 5070
    },
    {
      "entropy": 6.034631872177124,
      "epoch": 0.4481631932179442,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004985384769839349,
      "loss": 5.8991,
      "mean_token_accuracy": 0.15917696356773375,
      "num_tokens": 9473322.0,
      "step": 5075
    },
    {
      "entropy": 6.0339454174041744,
      "epoch": 0.4486047333097845,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985348912434008,
      "loss": 5.8089,
      "mean_token_accuracy": 0.17205141931772233,
      "num_tokens": 9482255.0,
      "step": 5080
    },
    {
      "entropy": 6.015864276885987,
      "epoch": 0.44904627340162484,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004985313011239452,
      "loss": 5.8479,
      "mean_token_accuracy": 0.1718486800789833,
      "num_tokens": 9491709.0,
      "step": 5085
    },
    {
      "entropy": 5.916902494430542,
      "epoch": 0.4494878134934652,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004985277066256388,
      "loss": 5.8391,
      "mean_token_accuracy": 0.17222511470317842,
      "num_tokens": 9500594.0,
      "step": 5090
    },
    {
      "entropy": 6.0166843891143795,
      "epoch": 0.44992935358530556,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985241077485515,
      "loss": 5.8302,
      "mean_token_accuracy": 0.17186546325683594,
      "num_tokens": 9509088.0,
      "step": 5095
    },
    {
      "entropy": 6.054204368591309,
      "epoch": 0.4503708936771459,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004985205044927541,
      "loss": 5.8254,
      "mean_token_accuracy": 0.17474611550569535,
      "num_tokens": 9517776.0,
      "step": 5100
    },
    {
      "entropy": 6.00213885307312,
      "epoch": 0.4508124337689862,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985168968583173,
      "loss": 5.8802,
      "mean_token_accuracy": 0.16938357502222062,
      "num_tokens": 9527080.0,
      "step": 5105
    },
    {
      "entropy": 5.956983804702759,
      "epoch": 0.45125397386082655,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985132848453114,
      "loss": 5.7284,
      "mean_token_accuracy": 0.18465249985456467,
      "num_tokens": 9536358.0,
      "step": 5110
    },
    {
      "entropy": 5.965890550613404,
      "epoch": 0.4516955139526669,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004985096684538075,
      "loss": 5.8272,
      "mean_token_accuracy": 0.16931069046258926,
      "num_tokens": 9545528.0,
      "step": 5115
    },
    {
      "entropy": 6.022493124008179,
      "epoch": 0.4521370540445072,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004985060476838763,
      "loss": 5.8626,
      "mean_token_accuracy": 0.1728214368224144,
      "num_tokens": 9554433.0,
      "step": 5120
    },
    {
      "entropy": 5.963240098953247,
      "epoch": 0.4525785941363476,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004985024225355887,
      "loss": 5.7917,
      "mean_token_accuracy": 0.16836365759372712,
      "num_tokens": 9563932.0,
      "step": 5125
    },
    {
      "entropy": 6.0266313552856445,
      "epoch": 0.45302013422818793,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004984987930090158,
      "loss": 5.8077,
      "mean_token_accuracy": 0.17788426131010054,
      "num_tokens": 9572829.0,
      "step": 5130
    },
    {
      "entropy": 5.9038893699646,
      "epoch": 0.45346167432002826,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984951591042285,
      "loss": 5.8342,
      "mean_token_accuracy": 0.17775831371545792,
      "num_tokens": 9583597.0,
      "step": 5135
    },
    {
      "entropy": 6.081721448898316,
      "epoch": 0.4539032144118686,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984915208212983,
      "loss": 5.905,
      "mean_token_accuracy": 0.1649731293320656,
      "num_tokens": 9593114.0,
      "step": 5140
    },
    {
      "entropy": 6.008826160430909,
      "epoch": 0.4543447545037089,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004984878781602964,
      "loss": 5.8081,
      "mean_token_accuracy": 0.16601281613111496,
      "num_tokens": 9601454.0,
      "step": 5145
    },
    {
      "entropy": 5.970934009552002,
      "epoch": 0.45478629459554926,
      "grad_norm": 0.78125,
      "learning_rate": 0.0004984842311212939,
      "loss": 5.845,
      "mean_token_accuracy": 0.17320733666419982,
      "num_tokens": 9611994.0,
      "step": 5150
    },
    {
      "entropy": 6.006623125076294,
      "epoch": 0.4552278346873896,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004984805797043625,
      "loss": 5.8176,
      "mean_token_accuracy": 0.17202675640583037,
      "num_tokens": 9621318.0,
      "step": 5155
    },
    {
      "entropy": 5.9953593730926515,
      "epoch": 0.45566937477923,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984769239095736,
      "loss": 5.8221,
      "mean_token_accuracy": 0.17728879749774934,
      "num_tokens": 9630270.0,
      "step": 5160
    },
    {
      "entropy": 6.006083202362061,
      "epoch": 0.4561109148710703,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984732637369989,
      "loss": 5.8543,
      "mean_token_accuracy": 0.16921540051698686,
      "num_tokens": 9640391.0,
      "step": 5165
    },
    {
      "entropy": 5.991392326354981,
      "epoch": 0.45655245496291064,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004984695991867099,
      "loss": 5.8739,
      "mean_token_accuracy": 0.16853009909391403,
      "num_tokens": 9648827.0,
      "step": 5170
    },
    {
      "entropy": 5.993249130249024,
      "epoch": 0.45699399505475097,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004984659302587788,
      "loss": 5.7851,
      "mean_token_accuracy": 0.16915750354528428,
      "num_tokens": 9657940.0,
      "step": 5175
    },
    {
      "entropy": 5.999341154098511,
      "epoch": 0.4574355351465913,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000498462256953277,
      "loss": 5.913,
      "mean_token_accuracy": 0.16240039765834807,
      "num_tokens": 9668296.0,
      "step": 5180
    },
    {
      "entropy": 6.074361944198609,
      "epoch": 0.45787707523843163,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004984585792702767,
      "loss": 5.8598,
      "mean_token_accuracy": 0.16403512805700302,
      "num_tokens": 9677914.0,
      "step": 5185
    },
    {
      "entropy": 6.065802478790284,
      "epoch": 0.45831861533027196,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004984548972098501,
      "loss": 6.0056,
      "mean_token_accuracy": 0.15658557415008545,
      "num_tokens": 9687339.0,
      "step": 5190
    },
    {
      "entropy": 5.965339040756225,
      "epoch": 0.45876015542211235,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000498451210772069,
      "loss": 5.7932,
      "mean_token_accuracy": 0.17470138892531395,
      "num_tokens": 9696889.0,
      "step": 5195
    },
    {
      "entropy": 6.006475973129272,
      "epoch": 0.4592016955139527,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004984475199570058,
      "loss": 5.7387,
      "mean_token_accuracy": 0.17829496264457703,
      "num_tokens": 9705498.0,
      "step": 5200
    },
    {
      "entropy": 6.030996513366699,
      "epoch": 0.459643235605793,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984438247647329,
      "loss": 5.8224,
      "mean_token_accuracy": 0.17188162505626678,
      "num_tokens": 9713936.0,
      "step": 5205
    },
    {
      "entropy": 6.003251218795777,
      "epoch": 0.46008477569763334,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984401251953223,
      "loss": 5.885,
      "mean_token_accuracy": 0.1700436070561409,
      "num_tokens": 9723924.0,
      "step": 5210
    },
    {
      "entropy": 6.122557973861694,
      "epoch": 0.4605263157894737,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004984364212488469,
      "loss": 5.9228,
      "mean_token_accuracy": 0.16665101200342178,
      "num_tokens": 9733427.0,
      "step": 5215
    },
    {
      "entropy": 5.952805805206299,
      "epoch": 0.460967855881314,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004984327129253789,
      "loss": 5.7258,
      "mean_token_accuracy": 0.1850974217057228,
      "num_tokens": 9742054.0,
      "step": 5220
    },
    {
      "entropy": 5.9985393524169925,
      "epoch": 0.46140939597315433,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984290002249914,
      "loss": 5.9011,
      "mean_token_accuracy": 0.1632828250527382,
      "num_tokens": 9752447.0,
      "step": 5225
    },
    {
      "entropy": 6.011101341247558,
      "epoch": 0.4618509360649947,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004984252831477567,
      "loss": 5.7107,
      "mean_token_accuracy": 0.17958685308694838,
      "num_tokens": 9760878.0,
      "step": 5230
    },
    {
      "entropy": 5.887458419799804,
      "epoch": 0.46229247615683505,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984215616937477,
      "loss": 5.7921,
      "mean_token_accuracy": 0.17403154075145721,
      "num_tokens": 9770200.0,
      "step": 5235
    },
    {
      "entropy": 6.054266548156738,
      "epoch": 0.4627340162486754,
      "grad_norm": 0.875,
      "learning_rate": 0.0004984178358630374,
      "loss": 5.8944,
      "mean_token_accuracy": 0.17439596951007844,
      "num_tokens": 9780303.0,
      "step": 5240
    },
    {
      "entropy": 5.999644231796265,
      "epoch": 0.4631755563405157,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984141056556989,
      "loss": 5.7899,
      "mean_token_accuracy": 0.17770855873823166,
      "num_tokens": 9790248.0,
      "step": 5245
    },
    {
      "entropy": 6.080256414413452,
      "epoch": 0.46361709643235605,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004984103710718051,
      "loss": 5.8931,
      "mean_token_accuracy": 0.16303637325763704,
      "num_tokens": 9799345.0,
      "step": 5250
    },
    {
      "entropy": 5.97104778289795,
      "epoch": 0.4640586365241964,
      "grad_norm": 0.953125,
      "learning_rate": 0.000498406632111429,
      "loss": 5.7595,
      "mean_token_accuracy": 0.17320360243320465,
      "num_tokens": 9808561.0,
      "step": 5255
    },
    {
      "entropy": 5.9949586391448975,
      "epoch": 0.4645001766160367,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004984028887746443,
      "loss": 5.8166,
      "mean_token_accuracy": 0.17151765376329423,
      "num_tokens": 9818324.0,
      "step": 5260
    },
    {
      "entropy": 5.9564426898956295,
      "epoch": 0.4649417167078771,
      "grad_norm": 1.0,
      "learning_rate": 0.0004983991410615239,
      "loss": 5.7875,
      "mean_token_accuracy": 0.1792932629585266,
      "num_tokens": 9827935.0,
      "step": 5265
    },
    {
      "entropy": 5.999061918258667,
      "epoch": 0.4653832567997174,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983953889721414,
      "loss": 5.8047,
      "mean_token_accuracy": 0.17208529710769654,
      "num_tokens": 9837118.0,
      "step": 5270
    },
    {
      "entropy": 6.022691106796264,
      "epoch": 0.46582479689155776,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004983916325065703,
      "loss": 5.8701,
      "mean_token_accuracy": 0.1652819335460663,
      "num_tokens": 9846197.0,
      "step": 5275
    },
    {
      "entropy": 6.097032976150513,
      "epoch": 0.4662663369833981,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983878716648842,
      "loss": 5.9403,
      "mean_token_accuracy": 0.15975897908210754,
      "num_tokens": 9856120.0,
      "step": 5280
    },
    {
      "entropy": 6.038593721389771,
      "epoch": 0.4667078770752384,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004983841064471567,
      "loss": 5.8877,
      "mean_token_accuracy": 0.16843740940093993,
      "num_tokens": 9865599.0,
      "step": 5285
    },
    {
      "entropy": 6.003615522384644,
      "epoch": 0.46714941716707875,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004983803368534617,
      "loss": 5.7839,
      "mean_token_accuracy": 0.1708475798368454,
      "num_tokens": 9876471.0,
      "step": 5290
    },
    {
      "entropy": 6.113279008865357,
      "epoch": 0.4675909572589191,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983765628838728,
      "loss": 5.8396,
      "mean_token_accuracy": 0.16985518038272857,
      "num_tokens": 9887680.0,
      "step": 5295
    },
    {
      "entropy": 6.086295223236084,
      "epoch": 0.46803249735075947,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983727845384641,
      "loss": 5.9189,
      "mean_token_accuracy": 0.1621912255883217,
      "num_tokens": 9897366.0,
      "step": 5300
    },
    {
      "entropy": 5.999127626419067,
      "epoch": 0.4684740374425998,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004983690018173096,
      "loss": 5.8913,
      "mean_token_accuracy": 0.1667093500494957,
      "num_tokens": 9907437.0,
      "step": 5305
    },
    {
      "entropy": 5.950143623352051,
      "epoch": 0.46891557753444013,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983652147204834,
      "loss": 5.6476,
      "mean_token_accuracy": 0.18681340366601945,
      "num_tokens": 9915766.0,
      "step": 5310
    },
    {
      "entropy": 5.981765604019165,
      "epoch": 0.46935711762628046,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983614232480598,
      "loss": 5.8873,
      "mean_token_accuracy": 0.16824992448091508,
      "num_tokens": 9925389.0,
      "step": 5315
    },
    {
      "entropy": 6.03397536277771,
      "epoch": 0.4697986577181208,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004983576274001127,
      "loss": 5.9331,
      "mean_token_accuracy": 0.16481596678495408,
      "num_tokens": 9935798.0,
      "step": 5320
    },
    {
      "entropy": 6.047959184646606,
      "epoch": 0.4702401978099611,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000498353827176717,
      "loss": 5.7848,
      "mean_token_accuracy": 0.17924853414297104,
      "num_tokens": 9945579.0,
      "step": 5325
    },
    {
      "entropy": 5.996797943115235,
      "epoch": 0.4706817379018015,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004983500225779466,
      "loss": 5.7843,
      "mean_token_accuracy": 0.175889290869236,
      "num_tokens": 9955248.0,
      "step": 5330
    },
    {
      "entropy": 5.982284498214722,
      "epoch": 0.47112327799364184,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983462136038764,
      "loss": 5.816,
      "mean_token_accuracy": 0.17963929623365402,
      "num_tokens": 9965078.0,
      "step": 5335
    },
    {
      "entropy": 6.037516069412232,
      "epoch": 0.4715648180854822,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983424002545809,
      "loss": 5.7763,
      "mean_token_accuracy": 0.18421670794487,
      "num_tokens": 9973932.0,
      "step": 5340
    },
    {
      "entropy": 6.022534418106079,
      "epoch": 0.4720063581773225,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004983385825301348,
      "loss": 5.8297,
      "mean_token_accuracy": 0.17369676381349564,
      "num_tokens": 9983315.0,
      "step": 5345
    },
    {
      "entropy": 5.988554048538208,
      "epoch": 0.47244789826916284,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004983347604306129,
      "loss": 5.7781,
      "mean_token_accuracy": 0.1726129099726677,
      "num_tokens": 9992987.0,
      "step": 5350
    },
    {
      "entropy": 5.954887008666992,
      "epoch": 0.47288943836100317,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983309339560899,
      "loss": 5.8602,
      "mean_token_accuracy": 0.17085562348365785,
      "num_tokens": 10002268.0,
      "step": 5355
    },
    {
      "entropy": 6.0079731941223145,
      "epoch": 0.4733309784528435,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004983271031066412,
      "loss": 5.8679,
      "mean_token_accuracy": 0.16448251605033876,
      "num_tokens": 10011772.0,
      "step": 5360
    },
    {
      "entropy": 6.009686517715454,
      "epoch": 0.4737725185446839,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004983232678823414,
      "loss": 5.868,
      "mean_token_accuracy": 0.1690191462635994,
      "num_tokens": 10021069.0,
      "step": 5365
    },
    {
      "entropy": 6.010422468185425,
      "epoch": 0.4742140586365242,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004983194282832657,
      "loss": 5.7833,
      "mean_token_accuracy": 0.17588813751935958,
      "num_tokens": 10029706.0,
      "step": 5370
    },
    {
      "entropy": 5.897851991653442,
      "epoch": 0.47465559872836455,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004983155843094895,
      "loss": 5.7517,
      "mean_token_accuracy": 0.17830378711223602,
      "num_tokens": 10039633.0,
      "step": 5375
    },
    {
      "entropy": 5.99191837310791,
      "epoch": 0.4750971388202049,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004983117359610881,
      "loss": 5.8065,
      "mean_token_accuracy": 0.1708494946360588,
      "num_tokens": 10048675.0,
      "step": 5380
    },
    {
      "entropy": 6.01197624206543,
      "epoch": 0.4755386789120452,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004983078832381367,
      "loss": 5.7896,
      "mean_token_accuracy": 0.17758023142814636,
      "num_tokens": 10057447.0,
      "step": 5385
    },
    {
      "entropy": 6.0020026683807375,
      "epoch": 0.47598021900388554,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004983040261407109,
      "loss": 5.862,
      "mean_token_accuracy": 0.16202759742736816,
      "num_tokens": 10067020.0,
      "step": 5390
    },
    {
      "entropy": 6.078158807754517,
      "epoch": 0.47642175909572587,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983001646688863,
      "loss": 5.8989,
      "mean_token_accuracy": 0.16627535820007325,
      "num_tokens": 10076466.0,
      "step": 5395
    },
    {
      "entropy": 6.0067445755004885,
      "epoch": 0.47686329918756626,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982962988227383,
      "loss": 5.7734,
      "mean_token_accuracy": 0.17013956755399703,
      "num_tokens": 10085384.0,
      "step": 5400
    },
    {
      "entropy": 5.968350505828857,
      "epoch": 0.4773048392794066,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498292428602343,
      "loss": 5.897,
      "mean_token_accuracy": 0.1669442757964134,
      "num_tokens": 10094015.0,
      "step": 5405
    },
    {
      "entropy": 5.968395566940307,
      "epoch": 0.4777463793712469,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982885540077758,
      "loss": 5.7796,
      "mean_token_accuracy": 0.17206281870603563,
      "num_tokens": 10103107.0,
      "step": 5410
    },
    {
      "entropy": 5.966505622863769,
      "epoch": 0.47818791946308725,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004982846750391129,
      "loss": 5.7145,
      "mean_token_accuracy": 0.18297846764326095,
      "num_tokens": 10112408.0,
      "step": 5415
    },
    {
      "entropy": 6.003802013397217,
      "epoch": 0.4786294595549276,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982807916964303,
      "loss": 5.7725,
      "mean_token_accuracy": 0.17201074063777924,
      "num_tokens": 10121605.0,
      "step": 5420
    },
    {
      "entropy": 5.9246360778808596,
      "epoch": 0.4790709996467679,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498276903979804,
      "loss": 5.7731,
      "mean_token_accuracy": 0.17475963085889817,
      "num_tokens": 10130484.0,
      "step": 5425
    },
    {
      "entropy": 5.951443338394165,
      "epoch": 0.47951253973860825,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049827301188931,
      "loss": 5.7344,
      "mean_token_accuracy": 0.18052613735198975,
      "num_tokens": 10140002.0,
      "step": 5430
    },
    {
      "entropy": 6.060378789901733,
      "epoch": 0.47995407983044863,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004982691154250247,
      "loss": 5.8956,
      "mean_token_accuracy": 0.17365120649337767,
      "num_tokens": 10150287.0,
      "step": 5435
    },
    {
      "entropy": 5.939712142944336,
      "epoch": 0.48039561992228896,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982652145870245,
      "loss": 5.7209,
      "mean_token_accuracy": 0.1840263232588768,
      "num_tokens": 10160615.0,
      "step": 5440
    },
    {
      "entropy": 5.987045478820801,
      "epoch": 0.4808371600141293,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004982613093753856,
      "loss": 5.8025,
      "mean_token_accuracy": 0.17736782282590866,
      "num_tokens": 10170271.0,
      "step": 5445
    },
    {
      "entropy": 5.9765979766845705,
      "epoch": 0.4812787001059696,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004982573997901847,
      "loss": 5.7628,
      "mean_token_accuracy": 0.17974358052015305,
      "num_tokens": 10179663.0,
      "step": 5450
    },
    {
      "entropy": 6.029351854324341,
      "epoch": 0.48172024019780996,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004982534858314982,
      "loss": 5.845,
      "mean_token_accuracy": 0.1695121109485626,
      "num_tokens": 10188608.0,
      "step": 5455
    },
    {
      "entropy": 6.038789987564087,
      "epoch": 0.4821617802896503,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004982495674994031,
      "loss": 5.8093,
      "mean_token_accuracy": 0.16905225962400436,
      "num_tokens": 10198462.0,
      "step": 5460
    },
    {
      "entropy": 5.905494213104248,
      "epoch": 0.4826033203814906,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982456447939758,
      "loss": 5.7539,
      "mean_token_accuracy": 0.18186217844486235,
      "num_tokens": 10208362.0,
      "step": 5465
    },
    {
      "entropy": 6.002645826339721,
      "epoch": 0.483044860473331,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004982417177152933,
      "loss": 5.7952,
      "mean_token_accuracy": 0.17442271411418914,
      "num_tokens": 10217915.0,
      "step": 5470
    },
    {
      "entropy": 6.008253574371338,
      "epoch": 0.48348640056517134,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004982377862634325,
      "loss": 5.8407,
      "mean_token_accuracy": 0.16862571984529495,
      "num_tokens": 10226913.0,
      "step": 5475
    },
    {
      "entropy": 5.935338640213013,
      "epoch": 0.48392794065701167,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004982338504384705,
      "loss": 5.8078,
      "mean_token_accuracy": 0.17525406777858735,
      "num_tokens": 10236516.0,
      "step": 5480
    },
    {
      "entropy": 5.997871446609497,
      "epoch": 0.484369480748852,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004982299102404843,
      "loss": 5.8276,
      "mean_token_accuracy": 0.18046585321426392,
      "num_tokens": 10245492.0,
      "step": 5485
    },
    {
      "entropy": 5.913363599777222,
      "epoch": 0.48481102084069233,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498225965669551,
      "loss": 5.6818,
      "mean_token_accuracy": 0.19123907536268234,
      "num_tokens": 10254094.0,
      "step": 5490
    },
    {
      "entropy": 5.917245578765869,
      "epoch": 0.48525256093253266,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982220167257482,
      "loss": 5.7687,
      "mean_token_accuracy": 0.1767420396208763,
      "num_tokens": 10263645.0,
      "step": 5495
    },
    {
      "entropy": 5.899789619445801,
      "epoch": 0.485694101024373,
      "grad_norm": 0.8203125,
      "learning_rate": 0.0004982180634091529,
      "loss": 5.7196,
      "mean_token_accuracy": 0.18229301422834396,
      "num_tokens": 10273416.0,
      "step": 5500
    },
    {
      "entropy": 5.976984024047852,
      "epoch": 0.4861356411162134,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004982141057198427,
      "loss": 5.8398,
      "mean_token_accuracy": 0.17539370357990264,
      "num_tokens": 10283194.0,
      "step": 5505
    },
    {
      "entropy": 5.983163928985595,
      "epoch": 0.4865771812080537,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004982101436578952,
      "loss": 5.8025,
      "mean_token_accuracy": 0.17517516165971755,
      "num_tokens": 10292188.0,
      "step": 5510
    },
    {
      "entropy": 5.889424133300781,
      "epoch": 0.48701872129989404,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004982061772233878,
      "loss": 5.7229,
      "mean_token_accuracy": 0.18908909112215042,
      "num_tokens": 10301539.0,
      "step": 5515
    },
    {
      "entropy": 5.965572834014893,
      "epoch": 0.4874602613917344,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982022064163984,
      "loss": 5.7355,
      "mean_token_accuracy": 0.18447270691394807,
      "num_tokens": 10310379.0,
      "step": 5520
    },
    {
      "entropy": 6.091221475601197,
      "epoch": 0.4879018014835747,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004981982312370047,
      "loss": 5.9364,
      "mean_token_accuracy": 0.16318108290433883,
      "num_tokens": 10320622.0,
      "step": 5525
    },
    {
      "entropy": 5.95211443901062,
      "epoch": 0.48834334157541504,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004981942516852847,
      "loss": 5.7453,
      "mean_token_accuracy": 0.17574492841959,
      "num_tokens": 10330525.0,
      "step": 5530
    },
    {
      "entropy": 5.932955980300903,
      "epoch": 0.48878488166725537,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004981902677613161,
      "loss": 5.7909,
      "mean_token_accuracy": 0.17560216039419174,
      "num_tokens": 10340288.0,
      "step": 5535
    },
    {
      "entropy": 6.035536384582519,
      "epoch": 0.48922642175909575,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981862794651771,
      "loss": 5.8825,
      "mean_token_accuracy": 0.1646736666560173,
      "num_tokens": 10349836.0,
      "step": 5540
    },
    {
      "entropy": 6.015440130233765,
      "epoch": 0.4896679618509361,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004981822867969459,
      "loss": 5.8088,
      "mean_token_accuracy": 0.17500588893890381,
      "num_tokens": 10359481.0,
      "step": 5545
    },
    {
      "entropy": 6.0031482696533205,
      "epoch": 0.4901095019427764,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004981782897567006,
      "loss": 5.7763,
      "mean_token_accuracy": 0.17515885084867477,
      "num_tokens": 10369837.0,
      "step": 5550
    },
    {
      "entropy": 5.974392652511597,
      "epoch": 0.49055104203461675,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004981742883445195,
      "loss": 5.7395,
      "mean_token_accuracy": 0.17979752272367477,
      "num_tokens": 10379276.0,
      "step": 5555
    },
    {
      "entropy": 6.046759605407715,
      "epoch": 0.4909925821264571,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000498170282560481,
      "loss": 5.8989,
      "mean_token_accuracy": 0.16546373218297958,
      "num_tokens": 10388878.0,
      "step": 5560
    },
    {
      "entropy": 6.05845193862915,
      "epoch": 0.4914341222182974,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004981662724046637,
      "loss": 5.8647,
      "mean_token_accuracy": 0.16656183153390886,
      "num_tokens": 10399148.0,
      "step": 5565
    },
    {
      "entropy": 5.8881865501403805,
      "epoch": 0.49187566231013774,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000498162257877146,
      "loss": 5.7904,
      "mean_token_accuracy": 0.17556624859571457,
      "num_tokens": 10409002.0,
      "step": 5570
    },
    {
      "entropy": 5.949180030822754,
      "epoch": 0.4923172024019781,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004981582389780065,
      "loss": 5.8411,
      "mean_token_accuracy": 0.17601792961359025,
      "num_tokens": 10418083.0,
      "step": 5575
    },
    {
      "entropy": 6.075670337677002,
      "epoch": 0.49275874249381846,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004981542157073241,
      "loss": 5.8447,
      "mean_token_accuracy": 0.171254663169384,
      "num_tokens": 10428134.0,
      "step": 5580
    },
    {
      "entropy": 5.985844087600708,
      "epoch": 0.4932002825856588,
      "grad_norm": 1.0,
      "learning_rate": 0.0004981501880651775,
      "loss": 5.8326,
      "mean_token_accuracy": 0.1712258756160736,
      "num_tokens": 10437436.0,
      "step": 5585
    },
    {
      "entropy": 5.9775690078735355,
      "epoch": 0.4936418226774991,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004981461560516457,
      "loss": 5.7736,
      "mean_token_accuracy": 0.17167493402957917,
      "num_tokens": 10446176.0,
      "step": 5590
    },
    {
      "entropy": 6.065248298645019,
      "epoch": 0.49408336276933945,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004981421196668075,
      "loss": 5.9154,
      "mean_token_accuracy": 0.16895957440137863,
      "num_tokens": 10455953.0,
      "step": 5595
    },
    {
      "entropy": 6.053670883178711,
      "epoch": 0.4945249028611798,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004981380789107422,
      "loss": 5.7647,
      "mean_token_accuracy": 0.17600456476211548,
      "num_tokens": 10465129.0,
      "step": 5600
    },
    {
      "entropy": 5.887149000167847,
      "epoch": 0.4949664429530201,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004981340337835287,
      "loss": 5.7692,
      "mean_token_accuracy": 0.1786765456199646,
      "num_tokens": 10474376.0,
      "step": 5605
    },
    {
      "entropy": 5.976278829574585,
      "epoch": 0.4954079830448605,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004981299842852464,
      "loss": 5.8378,
      "mean_token_accuracy": 0.17351005226373672,
      "num_tokens": 10484849.0,
      "step": 5610
    },
    {
      "entropy": 5.9031822204589846,
      "epoch": 0.49584952313670083,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004981259304159747,
      "loss": 5.7426,
      "mean_token_accuracy": 0.18532529473304749,
      "num_tokens": 10493950.0,
      "step": 5615
    },
    {
      "entropy": 5.928320455551147,
      "epoch": 0.49629106322854116,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004981218721757929,
      "loss": 5.6874,
      "mean_token_accuracy": 0.1819420799612999,
      "num_tokens": 10502890.0,
      "step": 5620
    },
    {
      "entropy": 5.977579784393311,
      "epoch": 0.4967326033203815,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004981178095647805,
      "loss": 5.6809,
      "mean_token_accuracy": 0.17615050375461577,
      "num_tokens": 10511702.0,
      "step": 5625
    },
    {
      "entropy": 5.920915508270264,
      "epoch": 0.4971741434122218,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004981137425830171,
      "loss": 5.7728,
      "mean_token_accuracy": 0.17457375377416612,
      "num_tokens": 10520212.0,
      "step": 5630
    },
    {
      "entropy": 5.900805234909058,
      "epoch": 0.49761568350406216,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981096712305825,
      "loss": 5.7985,
      "mean_token_accuracy": 0.17847424298524855,
      "num_tokens": 10529652.0,
      "step": 5635
    },
    {
      "entropy": 6.063334560394287,
      "epoch": 0.4980572235959025,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981055955075561,
      "loss": 5.856,
      "mean_token_accuracy": 0.17167602330446244,
      "num_tokens": 10537980.0,
      "step": 5640
    },
    {
      "entropy": 5.985257768630982,
      "epoch": 0.4984987636877429,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004981015154140181,
      "loss": 5.716,
      "mean_token_accuracy": 0.18206259310245515,
      "num_tokens": 10546854.0,
      "step": 5645
    },
    {
      "entropy": 5.934383678436279,
      "epoch": 0.4989403037795832,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004980974309500483,
      "loss": 5.7564,
      "mean_token_accuracy": 0.17640744894742966,
      "num_tokens": 10555691.0,
      "step": 5650
    },
    {
      "entropy": 5.953382968902588,
      "epoch": 0.49938184387142354,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004980933421157267,
      "loss": 5.8198,
      "mean_token_accuracy": 0.17145697474479676,
      "num_tokens": 10564750.0,
      "step": 5655
    },
    {
      "entropy": 5.992902565002441,
      "epoch": 0.49982338396326387,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980892489111334,
      "loss": 5.8708,
      "mean_token_accuracy": 0.17010241001844406,
      "num_tokens": 10573515.0,
      "step": 5660
    },
    {
      "entropy": 5.965075302124023,
      "epoch": 0.5002649240551043,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004980851513363486,
      "loss": 5.7427,
      "mean_token_accuracy": 0.183037468791008,
      "num_tokens": 10582730.0,
      "step": 5665
    },
    {
      "entropy": 5.837370443344116,
      "epoch": 0.5007064641469445,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980810493914526,
      "loss": 5.6517,
      "mean_token_accuracy": 0.19263290315866471,
      "num_tokens": 10591267.0,
      "step": 5670
    },
    {
      "entropy": 5.930122137069702,
      "epoch": 0.5011480042387849,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980769430765256,
      "loss": 5.7238,
      "mean_token_accuracy": 0.1784021809697151,
      "num_tokens": 10599776.0,
      "step": 5675
    },
    {
      "entropy": 5.913628816604614,
      "epoch": 0.5015895443306252,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004980728323916484,
      "loss": 5.7442,
      "mean_token_accuracy": 0.1708422601222992,
      "num_tokens": 10609670.0,
      "step": 5680
    },
    {
      "entropy": 5.911869525909424,
      "epoch": 0.5020310844224656,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980687173369009,
      "loss": 5.687,
      "mean_token_accuracy": 0.1840984344482422,
      "num_tokens": 10619032.0,
      "step": 5685
    },
    {
      "entropy": 5.96634349822998,
      "epoch": 0.5024726245143059,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004980645979123644,
      "loss": 5.7824,
      "mean_token_accuracy": 0.17758573591709137,
      "num_tokens": 10628851.0,
      "step": 5690
    },
    {
      "entropy": 5.955354261398315,
      "epoch": 0.5029141646061462,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004980604741181192,
      "loss": 5.7578,
      "mean_token_accuracy": 0.1791350230574608,
      "num_tokens": 10638335.0,
      "step": 5695
    },
    {
      "entropy": 5.9715251445770265,
      "epoch": 0.5033557046979866,
      "grad_norm": 0.875,
      "learning_rate": 0.0004980563459542461,
      "loss": 5.8441,
      "mean_token_accuracy": 0.17314625829458236,
      "num_tokens": 10648450.0,
      "step": 5700
    },
    {
      "entropy": 6.0982170581817625,
      "epoch": 0.5037972447898269,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004980522134208261,
      "loss": 5.8715,
      "mean_token_accuracy": 0.16750158816576005,
      "num_tokens": 10658700.0,
      "step": 5705
    },
    {
      "entropy": 6.037292432785034,
      "epoch": 0.5042387848816673,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980480765179401,
      "loss": 5.8785,
      "mean_token_accuracy": 0.1714808985590935,
      "num_tokens": 10667952.0,
      "step": 5710
    },
    {
      "entropy": 5.895547342300415,
      "epoch": 0.5046803249735076,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004980439352456692,
      "loss": 5.7019,
      "mean_token_accuracy": 0.17903856039047242,
      "num_tokens": 10677228.0,
      "step": 5715
    },
    {
      "entropy": 5.980213356018067,
      "epoch": 0.505121865065348,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004980397896040944,
      "loss": 5.7934,
      "mean_token_accuracy": 0.17691175937652587,
      "num_tokens": 10686183.0,
      "step": 5720
    },
    {
      "entropy": 6.027280044555664,
      "epoch": 0.5055634051571882,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980356395932969,
      "loss": 5.8348,
      "mean_token_accuracy": 0.1764046236872673,
      "num_tokens": 10695293.0,
      "step": 5725
    },
    {
      "entropy": 5.946255016326904,
      "epoch": 0.5060049452490286,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980314852133581,
      "loss": 5.7923,
      "mean_token_accuracy": 0.1797493353486061,
      "num_tokens": 10704853.0,
      "step": 5730
    },
    {
      "entropy": 5.957155179977417,
      "epoch": 0.506446485340869,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004980273264643594,
      "loss": 5.7572,
      "mean_token_accuracy": 0.1728769674897194,
      "num_tokens": 10714307.0,
      "step": 5735
    },
    {
      "entropy": 5.847817277908325,
      "epoch": 0.5068880254327093,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004980231633463822,
      "loss": 5.6558,
      "mean_token_accuracy": 0.17468281239271163,
      "num_tokens": 10723513.0,
      "step": 5740
    },
    {
      "entropy": 5.904374122619629,
      "epoch": 0.5073295655245497,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004980189958595081,
      "loss": 5.8291,
      "mean_token_accuracy": 0.16950504183769227,
      "num_tokens": 10732809.0,
      "step": 5745
    },
    {
      "entropy": 6.001342058181763,
      "epoch": 0.5077711056163899,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004980148240038186,
      "loss": 5.8531,
      "mean_token_accuracy": 0.17709660679101943,
      "num_tokens": 10742960.0,
      "step": 5750
    },
    {
      "entropy": 6.103294467926025,
      "epoch": 0.5082126457082303,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980106477793957,
      "loss": 5.9156,
      "mean_token_accuracy": 0.1706822007894516,
      "num_tokens": 10752835.0,
      "step": 5755
    },
    {
      "entropy": 5.984300756454468,
      "epoch": 0.5086541858000706,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004980064671863209,
      "loss": 5.7683,
      "mean_token_accuracy": 0.17604003846645355,
      "num_tokens": 10762139.0,
      "step": 5760
    },
    {
      "entropy": 5.982193517684936,
      "epoch": 0.509095725891911,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980022822246763,
      "loss": 5.8479,
      "mean_token_accuracy": 0.17236626744270325,
      "num_tokens": 10771961.0,
      "step": 5765
    },
    {
      "entropy": 6.084481191635132,
      "epoch": 0.5095372659837514,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004979980928945439,
      "loss": 5.8726,
      "mean_token_accuracy": 0.16543770730495452,
      "num_tokens": 10780163.0,
      "step": 5770
    },
    {
      "entropy": 6.0536665439605715,
      "epoch": 0.5099788060755917,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004979938991960056,
      "loss": 5.8429,
      "mean_token_accuracy": 0.17202573120594025,
      "num_tokens": 10788339.0,
      "step": 5775
    },
    {
      "entropy": 6.004624271392823,
      "epoch": 0.510420346167432,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004979897011291436,
      "loss": 5.8412,
      "mean_token_accuracy": 0.16892923414707184,
      "num_tokens": 10797315.0,
      "step": 5780
    },
    {
      "entropy": 6.031994342803955,
      "epoch": 0.5108618862592723,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004979854986940402,
      "loss": 5.7524,
      "mean_token_accuracy": 0.17269293516874312,
      "num_tokens": 10806864.0,
      "step": 5785
    },
    {
      "entropy": 6.027312898635865,
      "epoch": 0.5113034263511127,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979812918907777,
      "loss": 5.7983,
      "mean_token_accuracy": 0.17434913963079451,
      "num_tokens": 10816417.0,
      "step": 5790
    },
    {
      "entropy": 5.969706630706787,
      "epoch": 0.511744966442953,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004979770807194385,
      "loss": 5.8605,
      "mean_token_accuracy": 0.17248998880386351,
      "num_tokens": 10825692.0,
      "step": 5795
    },
    {
      "entropy": 5.984615087509155,
      "epoch": 0.5121865065347934,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004979728651801051,
      "loss": 5.8402,
      "mean_token_accuracy": 0.1778946951031685,
      "num_tokens": 10834652.0,
      "step": 5800
    },
    {
      "entropy": 6.018920278549194,
      "epoch": 0.5126280466266337,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004979686452728602,
      "loss": 5.828,
      "mean_token_accuracy": 0.17260900139808655,
      "num_tokens": 10844491.0,
      "step": 5805
    },
    {
      "entropy": 5.937997484207154,
      "epoch": 0.513069586718474,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004979644209977863,
      "loss": 5.7696,
      "mean_token_accuracy": 0.17836492955684663,
      "num_tokens": 10853948.0,
      "step": 5810
    },
    {
      "entropy": 5.9561933994293215,
      "epoch": 0.5135111268103144,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979601923549661,
      "loss": 5.7795,
      "mean_token_accuracy": 0.17643316686153412,
      "num_tokens": 10861745.0,
      "step": 5815
    },
    {
      "entropy": 6.041404867172242,
      "epoch": 0.5139526669021547,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979559593444826,
      "loss": 5.8554,
      "mean_token_accuracy": 0.16654422879219055,
      "num_tokens": 10871239.0,
      "step": 5820
    },
    {
      "entropy": 5.95691614151001,
      "epoch": 0.5143942069939951,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004979517219664187,
      "loss": 5.7184,
      "mean_token_accuracy": 0.1781800225377083,
      "num_tokens": 10880883.0,
      "step": 5825
    },
    {
      "entropy": 5.979454565048218,
      "epoch": 0.5148357470858353,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004979474802208572,
      "loss": 5.7715,
      "mean_token_accuracy": 0.17757208198308944,
      "num_tokens": 10890002.0,
      "step": 5830
    },
    {
      "entropy": 5.9080863952636715,
      "epoch": 0.5152772871776757,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004979432341078816,
      "loss": 5.7001,
      "mean_token_accuracy": 0.17272792011499405,
      "num_tokens": 10898299.0,
      "step": 5835
    },
    {
      "entropy": 6.025524568557739,
      "epoch": 0.5157188272695161,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004979389836275746,
      "loss": 5.91,
      "mean_token_accuracy": 0.16205677539110183,
      "num_tokens": 10907918.0,
      "step": 5840
    },
    {
      "entropy": 6.026790571212769,
      "epoch": 0.5161603673613564,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004979347287800198,
      "loss": 5.7872,
      "mean_token_accuracy": 0.1693910390138626,
      "num_tokens": 10917318.0,
      "step": 5845
    },
    {
      "entropy": 5.962571716308593,
      "epoch": 0.5166019074531968,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004979304695653005,
      "loss": 5.7259,
      "mean_token_accuracy": 0.1753496468067169,
      "num_tokens": 10926338.0,
      "step": 5850
    },
    {
      "entropy": 5.909990406036377,
      "epoch": 0.5170434475450371,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004979262059835001,
      "loss": 5.7636,
      "mean_token_accuracy": 0.17829854786396027,
      "num_tokens": 10935244.0,
      "step": 5855
    },
    {
      "entropy": 6.022508668899536,
      "epoch": 0.5174849876368774,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004979219380347021,
      "loss": 5.8021,
      "mean_token_accuracy": 0.17361603677272797,
      "num_tokens": 10943764.0,
      "step": 5860
    },
    {
      "entropy": 6.012854290008545,
      "epoch": 0.5179265277287177,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004979176657189901,
      "loss": 5.7735,
      "mean_token_accuracy": 0.17107923626899718,
      "num_tokens": 10952580.0,
      "step": 5865
    },
    {
      "entropy": 5.9548540115356445,
      "epoch": 0.5183680678205581,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004979133890364477,
      "loss": 5.804,
      "mean_token_accuracy": 0.17859138250350953,
      "num_tokens": 10961745.0,
      "step": 5870
    },
    {
      "entropy": 5.8955892562866214,
      "epoch": 0.5188096079123985,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497909107987159,
      "loss": 5.7239,
      "mean_token_accuracy": 0.17332278788089753,
      "num_tokens": 10971018.0,
      "step": 5875
    },
    {
      "entropy": 5.957081127166748,
      "epoch": 0.5192511480042388,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979048225712074,
      "loss": 5.7574,
      "mean_token_accuracy": 0.17847395539283753,
      "num_tokens": 10980070.0,
      "step": 5880
    },
    {
      "entropy": 6.00935115814209,
      "epoch": 0.5196926880960792,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004979005327886772,
      "loss": 5.7406,
      "mean_token_accuracy": 0.17666334360837938,
      "num_tokens": 10989090.0,
      "step": 5885
    },
    {
      "entropy": 5.992536735534668,
      "epoch": 0.5201342281879194,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004978962386396525,
      "loss": 5.7908,
      "mean_token_accuracy": 0.17853841334581375,
      "num_tokens": 10998652.0,
      "step": 5890
    },
    {
      "entropy": 5.848595428466797,
      "epoch": 0.5205757682797598,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004978919401242171,
      "loss": 5.6844,
      "mean_token_accuracy": 0.18879187554121019,
      "num_tokens": 11006323.0,
      "step": 5895
    },
    {
      "entropy": 5.797189474105835,
      "epoch": 0.5210173083716001,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004978876372424553,
      "loss": 5.6297,
      "mean_token_accuracy": 0.1844954565167427,
      "num_tokens": 11015054.0,
      "step": 5900
    },
    {
      "entropy": 5.956575059890747,
      "epoch": 0.5214588484634405,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978833299944515,
      "loss": 5.8139,
      "mean_token_accuracy": 0.170148167014122,
      "num_tokens": 11025787.0,
      "step": 5905
    },
    {
      "entropy": 6.084104728698731,
      "epoch": 0.5219003885552809,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978790183802901,
      "loss": 5.9223,
      "mean_token_accuracy": 0.16725478023290635,
      "num_tokens": 11036091.0,
      "step": 5910
    },
    {
      "entropy": 5.951149892807007,
      "epoch": 0.5223419286471211,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004978747024000554,
      "loss": 5.8703,
      "mean_token_accuracy": 0.16429074257612228,
      "num_tokens": 11046442.0,
      "step": 5915
    },
    {
      "entropy": 6.039041662216187,
      "epoch": 0.5227834687389615,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978703820538321,
      "loss": 5.9039,
      "mean_token_accuracy": 0.16753809303045272,
      "num_tokens": 11056907.0,
      "step": 5920
    },
    {
      "entropy": 6.0201164245605465,
      "epoch": 0.5232250088308018,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978660573417048,
      "loss": 5.7536,
      "mean_token_accuracy": 0.17897529453039168,
      "num_tokens": 11065298.0,
      "step": 5925
    },
    {
      "entropy": 5.924715948104859,
      "epoch": 0.5236665489226422,
      "grad_norm": 0.96875,
      "learning_rate": 0.000497861728263758,
      "loss": 5.7376,
      "mean_token_accuracy": 0.1859300434589386,
      "num_tokens": 11074509.0,
      "step": 5930
    },
    {
      "entropy": 5.973381519317627,
      "epoch": 0.5241080890144825,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004978573948200769,
      "loss": 5.7772,
      "mean_token_accuracy": 0.1776812806725502,
      "num_tokens": 11082433.0,
      "step": 5935
    },
    {
      "entropy": 5.972002792358398,
      "epoch": 0.5245496291063229,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004978530570107461,
      "loss": 5.7476,
      "mean_token_accuracy": 0.17605585902929305,
      "num_tokens": 11090838.0,
      "step": 5940
    },
    {
      "entropy": 5.965470218658448,
      "epoch": 0.5249911691981632,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004978487148358506,
      "loss": 5.7421,
      "mean_token_accuracy": 0.18264487236738206,
      "num_tokens": 11100048.0,
      "step": 5945
    },
    {
      "entropy": 5.874642705917358,
      "epoch": 0.5254327092900035,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004978443682954756,
      "loss": 5.7466,
      "mean_token_accuracy": 0.17708369791507722,
      "num_tokens": 11109871.0,
      "step": 5950
    },
    {
      "entropy": 5.954572439193726,
      "epoch": 0.5258742493818439,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004978400173897061,
      "loss": 5.7382,
      "mean_token_accuracy": 0.1706852361559868,
      "num_tokens": 11119450.0,
      "step": 5955
    },
    {
      "entropy": 6.011610317230224,
      "epoch": 0.5263157894736842,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004978356621186275,
      "loss": 5.7024,
      "mean_token_accuracy": 0.19025731831789017,
      "num_tokens": 11128321.0,
      "step": 5960
    },
    {
      "entropy": 5.893124532699585,
      "epoch": 0.5267573295655246,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004978313024823249,
      "loss": 5.7508,
      "mean_token_accuracy": 0.1797892838716507,
      "num_tokens": 11138657.0,
      "step": 5965
    },
    {
      "entropy": 5.876752948760986,
      "epoch": 0.5271988696573648,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004978269384808839,
      "loss": 5.7308,
      "mean_token_accuracy": 0.1822565272450447,
      "num_tokens": 11147704.0,
      "step": 5970
    },
    {
      "entropy": 5.982156419754029,
      "epoch": 0.5276404097492052,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004978225701143898,
      "loss": 5.8386,
      "mean_token_accuracy": 0.1702994167804718,
      "num_tokens": 11156979.0,
      "step": 5975
    },
    {
      "entropy": 5.976283884048462,
      "epoch": 0.5280819498410456,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004978181973829284,
      "loss": 5.7233,
      "mean_token_accuracy": 0.17943737357854844,
      "num_tokens": 11165539.0,
      "step": 5980
    },
    {
      "entropy": 6.0231208324432375,
      "epoch": 0.5285234899328859,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004978138202865851,
      "loss": 5.7906,
      "mean_token_accuracy": 0.16854341477155685,
      "num_tokens": 11175911.0,
      "step": 5985
    },
    {
      "entropy": 5.924521636962891,
      "epoch": 0.5289650300247263,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004978094388254459,
      "loss": 5.8217,
      "mean_token_accuracy": 0.16968544870615004,
      "num_tokens": 11185365.0,
      "step": 5990
    },
    {
      "entropy": 5.926220464706421,
      "epoch": 0.5294065701165666,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004978050529995965,
      "loss": 5.7211,
      "mean_token_accuracy": 0.17357961237430572,
      "num_tokens": 11195318.0,
      "step": 5995
    },
    {
      "entropy": 5.973835372924805,
      "epoch": 0.5298481102084069,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004978006628091228,
      "loss": 5.8353,
      "mean_token_accuracy": 0.17482065111398698,
      "num_tokens": 11205246.0,
      "step": 6000
    },
    {
      "epoch": 0.5298481102084069,
      "eval_entropy": 5.7665525598407195,
      "eval_loss": 5.782026290893555,
      "eval_mean_token_accuracy": 0.18332941638198202,
      "eval_num_tokens": 11205246.0,
      "eval_runtime": 20.2512,
      "eval_samples_per_second": 1743.846,
      "eval_steps_per_second": 218.012,
      "step": 6000
    },
    {
      "entropy": 5.990317106246948,
      "epoch": 0.5302896503002472,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004977962682541108,
      "loss": 5.781,
      "mean_token_accuracy": 0.17299736738204957,
      "num_tokens": 11215183.0,
      "step": 6005
    },
    {
      "entropy": 5.952910041809082,
      "epoch": 0.5307311903920876,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004977918693346467,
      "loss": 5.7518,
      "mean_token_accuracy": 0.17333558350801467,
      "num_tokens": 11224127.0,
      "step": 6010
    },
    {
      "entropy": 5.951988935470581,
      "epoch": 0.531172730483928,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977874660508165,
      "loss": 5.7806,
      "mean_token_accuracy": 0.17980630993843078,
      "num_tokens": 11233014.0,
      "step": 6015
    },
    {
      "entropy": 5.9942632675170895,
      "epoch": 0.5316142705757683,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004977830584027065,
      "loss": 5.7706,
      "mean_token_accuracy": 0.1789461463689804,
      "num_tokens": 11242251.0,
      "step": 6020
    },
    {
      "entropy": 6.00589075088501,
      "epoch": 0.5320558106676087,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004977786463904033,
      "loss": 5.8286,
      "mean_token_accuracy": 0.17321041971445084,
      "num_tokens": 11251947.0,
      "step": 6025
    },
    {
      "entropy": 6.000608110427857,
      "epoch": 0.5324973507594489,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004977742300139929,
      "loss": 5.8648,
      "mean_token_accuracy": 0.16567921191453933,
      "num_tokens": 11261125.0,
      "step": 6030
    },
    {
      "entropy": 6.012487840652466,
      "epoch": 0.5329388908512893,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004977698092735621,
      "loss": 5.739,
      "mean_token_accuracy": 0.17683637589216233,
      "num_tokens": 11270334.0,
      "step": 6035
    },
    {
      "entropy": 5.920803642272949,
      "epoch": 0.5333804309431296,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004977653841691975,
      "loss": 5.7561,
      "mean_token_accuracy": 0.17879996448755264,
      "num_tokens": 11280497.0,
      "step": 6040
    },
    {
      "entropy": 5.9857405662536625,
      "epoch": 0.53382197103497,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977609547009857,
      "loss": 5.6924,
      "mean_token_accuracy": 0.17791507840156556,
      "num_tokens": 11288636.0,
      "step": 6045
    },
    {
      "entropy": 5.947061967849732,
      "epoch": 0.5342635111268104,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977565208690134,
      "loss": 5.8118,
      "mean_token_accuracy": 0.17096064984798431,
      "num_tokens": 11297519.0,
      "step": 6050
    },
    {
      "entropy": 5.937090826034546,
      "epoch": 0.5347050512186506,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004977520826733677,
      "loss": 5.7466,
      "mean_token_accuracy": 0.1841330721974373,
      "num_tokens": 11306280.0,
      "step": 6055
    },
    {
      "entropy": 5.941117191314698,
      "epoch": 0.535146591310491,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004977476401141353,
      "loss": 5.7608,
      "mean_token_accuracy": 0.17097747325897217,
      "num_tokens": 11315366.0,
      "step": 6060
    },
    {
      "entropy": 5.977073001861572,
      "epoch": 0.5355881314023313,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004977431931914032,
      "loss": 5.7688,
      "mean_token_accuracy": 0.1683377668261528,
      "num_tokens": 11324664.0,
      "step": 6065
    },
    {
      "entropy": 6.0049549579620365,
      "epoch": 0.5360296714941717,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004977387419052586,
      "loss": 5.752,
      "mean_token_accuracy": 0.1675189822912216,
      "num_tokens": 11334118.0,
      "step": 6070
    },
    {
      "entropy": 6.066574954986573,
      "epoch": 0.536471211586012,
      "grad_norm": 0.83984375,
      "learning_rate": 0.000497734286255789,
      "loss": 5.8661,
      "mean_token_accuracy": 0.16980284303426743,
      "num_tokens": 11342684.0,
      "step": 6075
    },
    {
      "entropy": 5.942904949188232,
      "epoch": 0.5369127516778524,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004977298262430813,
      "loss": 5.787,
      "mean_token_accuracy": 0.18056662529706954,
      "num_tokens": 11351298.0,
      "step": 6080
    },
    {
      "entropy": 5.964593315124512,
      "epoch": 0.5373542917696927,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004977253618672229,
      "loss": 5.7353,
      "mean_token_accuracy": 0.17694929987192154,
      "num_tokens": 11360264.0,
      "step": 6085
    },
    {
      "entropy": 5.946565723419189,
      "epoch": 0.537795831861533,
      "grad_norm": 0.875,
      "learning_rate": 0.0004977208931283013,
      "loss": 5.7171,
      "mean_token_accuracy": 0.179603773355484,
      "num_tokens": 11369815.0,
      "step": 6090
    },
    {
      "entropy": 5.9731690883636475,
      "epoch": 0.5382373719533734,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977164200264042,
      "loss": 5.7954,
      "mean_token_accuracy": 0.17322832047939302,
      "num_tokens": 11379131.0,
      "step": 6095
    },
    {
      "entropy": 5.961735391616822,
      "epoch": 0.5386789120452137,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004977119425616191,
      "loss": 5.6836,
      "mean_token_accuracy": 0.18143999129533767,
      "num_tokens": 11387692.0,
      "step": 6100
    },
    {
      "entropy": 5.944207382202149,
      "epoch": 0.5391204521370541,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004977074607340336,
      "loss": 5.7364,
      "mean_token_accuracy": 0.17622338831424714,
      "num_tokens": 11396278.0,
      "step": 6105
    },
    {
      "entropy": 5.873842573165893,
      "epoch": 0.5395619922288943,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004977029745437357,
      "loss": 5.6798,
      "mean_token_accuracy": 0.1845750167965889,
      "num_tokens": 11404790.0,
      "step": 6110
    },
    {
      "entropy": 5.9296238899230955,
      "epoch": 0.5400035323207347,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000497698483990813,
      "loss": 5.7827,
      "mean_token_accuracy": 0.17046503573656083,
      "num_tokens": 11414614.0,
      "step": 6115
    },
    {
      "entropy": 5.982711362838745,
      "epoch": 0.5404450724125751,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976939890753539,
      "loss": 5.7841,
      "mean_token_accuracy": 0.17147547155618667,
      "num_tokens": 11423696.0,
      "step": 6120
    },
    {
      "entropy": 5.949193668365479,
      "epoch": 0.5408866125044154,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497689489797446,
      "loss": 5.7296,
      "mean_token_accuracy": 0.1846212074160576,
      "num_tokens": 11432669.0,
      "step": 6125
    },
    {
      "entropy": 5.935843181610108,
      "epoch": 0.5413281525962558,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004976849861571778,
      "loss": 5.7638,
      "mean_token_accuracy": 0.17586904317140578,
      "num_tokens": 11442132.0,
      "step": 6130
    },
    {
      "entropy": 5.958372163772583,
      "epoch": 0.541769692688096,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004976804781546373,
      "loss": 5.7237,
      "mean_token_accuracy": 0.17114574760198592,
      "num_tokens": 11451821.0,
      "step": 6135
    },
    {
      "entropy": 5.9710346221923825,
      "epoch": 0.5422112327799364,
      "grad_norm": 1.0,
      "learning_rate": 0.0004976759657899129,
      "loss": 5.7422,
      "mean_token_accuracy": 0.1799530491232872,
      "num_tokens": 11460242.0,
      "step": 6140
    },
    {
      "entropy": 5.869692325592041,
      "epoch": 0.5426527728717767,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000497671449063093,
      "loss": 5.7729,
      "mean_token_accuracy": 0.18124603480100632,
      "num_tokens": 11469588.0,
      "step": 6145
    },
    {
      "entropy": 6.013301944732666,
      "epoch": 0.5430943129636171,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000497666927974266,
      "loss": 5.8456,
      "mean_token_accuracy": 0.16814137101173401,
      "num_tokens": 11480401.0,
      "step": 6150
    },
    {
      "entropy": 6.017386198043823,
      "epoch": 0.5435358530554575,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004976624025235205,
      "loss": 5.7602,
      "mean_token_accuracy": 0.1720780089497566,
      "num_tokens": 11489117.0,
      "step": 6155
    },
    {
      "entropy": 5.8355712890625,
      "epoch": 0.5439773931472978,
      "grad_norm": 0.875,
      "learning_rate": 0.0004976578727109453,
      "loss": 5.7019,
      "mean_token_accuracy": 0.18260259926319122,
      "num_tokens": 11497612.0,
      "step": 6160
    },
    {
      "entropy": 5.938451480865479,
      "epoch": 0.5444189332391381,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004976533385366289,
      "loss": 5.7813,
      "mean_token_accuracy": 0.17646318674087524,
      "num_tokens": 11507024.0,
      "step": 6165
    },
    {
      "entropy": 5.959741735458374,
      "epoch": 0.5448604733309784,
      "grad_norm": 0.8125,
      "learning_rate": 0.0004976488000006602,
      "loss": 5.7019,
      "mean_token_accuracy": 0.18477845042943955,
      "num_tokens": 11516873.0,
      "step": 6170
    },
    {
      "entropy": 6.025303840637207,
      "epoch": 0.5453020134228188,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004976442571031282,
      "loss": 5.8072,
      "mean_token_accuracy": 0.18425012081861497,
      "num_tokens": 11526128.0,
      "step": 6175
    },
    {
      "entropy": 5.902001094818115,
      "epoch": 0.5457435535146591,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004976397098441218,
      "loss": 5.8113,
      "mean_token_accuracy": 0.18117952942848206,
      "num_tokens": 11535030.0,
      "step": 6180
    },
    {
      "entropy": 5.852960729598999,
      "epoch": 0.5461850936064995,
      "grad_norm": 1.0,
      "learning_rate": 0.0004976351582237301,
      "loss": 5.6982,
      "mean_token_accuracy": 0.1818488821387291,
      "num_tokens": 11543113.0,
      "step": 6185
    },
    {
      "entropy": 5.96119122505188,
      "epoch": 0.5466266336983399,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004976306022420423,
      "loss": 5.7295,
      "mean_token_accuracy": 0.17677829265594483,
      "num_tokens": 11552744.0,
      "step": 6190
    },
    {
      "entropy": 6.032695865631103,
      "epoch": 0.5470681737901801,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004976260418991476,
      "loss": 5.7739,
      "mean_token_accuracy": 0.1760561227798462,
      "num_tokens": 11561274.0,
      "step": 6195
    },
    {
      "entropy": 6.003128910064698,
      "epoch": 0.5475097138820205,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004976214771951354,
      "loss": 5.8,
      "mean_token_accuracy": 0.17939124554395675,
      "num_tokens": 11570485.0,
      "step": 6200
    },
    {
      "entropy": 5.972816276550293,
      "epoch": 0.5479512539738608,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004976169081300949,
      "loss": 5.7707,
      "mean_token_accuracy": 0.17227309495210646,
      "num_tokens": 11579380.0,
      "step": 6205
    },
    {
      "entropy": 5.934311819076538,
      "epoch": 0.5483927940657012,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004976123347041159,
      "loss": 5.75,
      "mean_token_accuracy": 0.1758207708597183,
      "num_tokens": 11588686.0,
      "step": 6210
    },
    {
      "entropy": 5.94596152305603,
      "epoch": 0.5488343341575415,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004976077569172879,
      "loss": 5.6867,
      "mean_token_accuracy": 0.18259273767471312,
      "num_tokens": 11598547.0,
      "step": 6215
    },
    {
      "entropy": 5.933783531188965,
      "epoch": 0.5492758742493818,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004976031747697006,
      "loss": 5.7744,
      "mean_token_accuracy": 0.1780215844511986,
      "num_tokens": 11608407.0,
      "step": 6220
    },
    {
      "entropy": 5.873191547393799,
      "epoch": 0.5497174143412222,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004975985882614435,
      "loss": 5.7298,
      "mean_token_accuracy": 0.1796346575021744,
      "num_tokens": 11618641.0,
      "step": 6225
    },
    {
      "entropy": 5.968663263320923,
      "epoch": 0.5501589544330625,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004975939973926068,
      "loss": 5.7934,
      "mean_token_accuracy": 0.16603399068117142,
      "num_tokens": 11627082.0,
      "step": 6230
    },
    {
      "entropy": 5.934994173049927,
      "epoch": 0.5506004945249029,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004975894021632803,
      "loss": 5.6769,
      "mean_token_accuracy": 0.1807957574725151,
      "num_tokens": 11636045.0,
      "step": 6235
    },
    {
      "entropy": 5.929260015487671,
      "epoch": 0.5510420346167432,
      "grad_norm": 0.9375,
      "learning_rate": 0.000497584802573554,
      "loss": 5.7589,
      "mean_token_accuracy": 0.1748830959200859,
      "num_tokens": 11646516.0,
      "step": 6240
    },
    {
      "entropy": 5.919200325012207,
      "epoch": 0.5514835747085836,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000497580198623518,
      "loss": 5.6835,
      "mean_token_accuracy": 0.18040877282619477,
      "num_tokens": 11655512.0,
      "step": 6245
    },
    {
      "entropy": 5.988286733627319,
      "epoch": 0.5519251148004238,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004975755903132623,
      "loss": 5.8009,
      "mean_token_accuracy": 0.17675041109323503,
      "num_tokens": 11665641.0,
      "step": 6250
    },
    {
      "entropy": 5.927127695083618,
      "epoch": 0.5523666548922642,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004975709776428776,
      "loss": 5.7674,
      "mean_token_accuracy": 0.17123859375715256,
      "num_tokens": 11674585.0,
      "step": 6255
    },
    {
      "entropy": 5.883406162261963,
      "epoch": 0.5528081949841046,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000497566360612454,
      "loss": 5.6318,
      "mean_token_accuracy": 0.18883141577243806,
      "num_tokens": 11683498.0,
      "step": 6260
    },
    {
      "entropy": 5.945862483978272,
      "epoch": 0.5532497350759449,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000497561739222082,
      "loss": 5.6778,
      "mean_token_accuracy": 0.1816217005252838,
      "num_tokens": 11692663.0,
      "step": 6265
    },
    {
      "entropy": 5.911245727539063,
      "epoch": 0.5536912751677853,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004975571134718521,
      "loss": 5.8389,
      "mean_token_accuracy": 0.17130470722913743,
      "num_tokens": 11701906.0,
      "step": 6270
    },
    {
      "entropy": 5.953613424301148,
      "epoch": 0.5541328152596255,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004975524833618548,
      "loss": 5.7533,
      "mean_token_accuracy": 0.18487675338983536,
      "num_tokens": 11710932.0,
      "step": 6275
    },
    {
      "entropy": 5.97131118774414,
      "epoch": 0.5545743553514659,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497547848892181,
      "loss": 5.7541,
      "mean_token_accuracy": 0.18125775456428528,
      "num_tokens": 11719953.0,
      "step": 6280
    },
    {
      "entropy": 5.979850387573242,
      "epoch": 0.5550158954433062,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004975432100629215,
      "loss": 5.6412,
      "mean_token_accuracy": 0.18299894779920578,
      "num_tokens": 11728706.0,
      "step": 6285
    },
    {
      "entropy": 5.969555902481079,
      "epoch": 0.5554574355351466,
      "grad_norm": 1.078125,
      "learning_rate": 0.000497538566874167,
      "loss": 5.8242,
      "mean_token_accuracy": 0.1693728670477867,
      "num_tokens": 11738537.0,
      "step": 6290
    },
    {
      "entropy": 5.960725116729736,
      "epoch": 0.555898975626987,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004975339193260084,
      "loss": 5.6493,
      "mean_token_accuracy": 0.18169963359832764,
      "num_tokens": 11747503.0,
      "step": 6295
    },
    {
      "entropy": 5.951407527923584,
      "epoch": 0.5563405157188273,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004975292674185371,
      "loss": 5.7519,
      "mean_token_accuracy": 0.17961287796497344,
      "num_tokens": 11756501.0,
      "step": 6300
    },
    {
      "entropy": 5.8904026508331295,
      "epoch": 0.5567820558106676,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004975246111518439,
      "loss": 5.7268,
      "mean_token_accuracy": 0.1724578857421875,
      "num_tokens": 11765483.0,
      "step": 6305
    },
    {
      "entropy": 6.040583896636963,
      "epoch": 0.5572235959025079,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004975199505260202,
      "loss": 5.8402,
      "mean_token_accuracy": 0.1671855017542839,
      "num_tokens": 11774941.0,
      "step": 6310
    },
    {
      "entropy": 6.042414903640747,
      "epoch": 0.5576651359943483,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004975152855411571,
      "loss": 5.8157,
      "mean_token_accuracy": 0.16865744739770888,
      "num_tokens": 11785080.0,
      "step": 6315
    },
    {
      "entropy": 5.949933290481567,
      "epoch": 0.5581066760861886,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004975106161973462,
      "loss": 5.707,
      "mean_token_accuracy": 0.17533884942531586,
      "num_tokens": 11793933.0,
      "step": 6320
    },
    {
      "entropy": 5.896859550476075,
      "epoch": 0.558548216178029,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000497505942494679,
      "loss": 5.7346,
      "mean_token_accuracy": 0.17790014147758484,
      "num_tokens": 11803592.0,
      "step": 6325
    },
    {
      "entropy": 5.902767181396484,
      "epoch": 0.5589897562698694,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004975012644332467,
      "loss": 5.6507,
      "mean_token_accuracy": 0.18465393632650376,
      "num_tokens": 11812397.0,
      "step": 6330
    },
    {
      "entropy": 5.903921031951905,
      "epoch": 0.5594312963617096,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974965820131413,
      "loss": 5.7665,
      "mean_token_accuracy": 0.173906309902668,
      "num_tokens": 11821734.0,
      "step": 6335
    },
    {
      "entropy": 5.908196878433228,
      "epoch": 0.55987283645355,
      "grad_norm": 0.80859375,
      "learning_rate": 0.0004974918952344544,
      "loss": 5.6429,
      "mean_token_accuracy": 0.18138528764247894,
      "num_tokens": 11832107.0,
      "step": 6340
    },
    {
      "entropy": 5.960652589797974,
      "epoch": 0.5603143765453903,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004974872040972777,
      "loss": 5.6944,
      "mean_token_accuracy": 0.17450693845748902,
      "num_tokens": 11841983.0,
      "step": 6345
    },
    {
      "entropy": 5.822365188598633,
      "epoch": 0.5607559166372307,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004974825086017033,
      "loss": 5.6839,
      "mean_token_accuracy": 0.18381367623806,
      "num_tokens": 11851836.0,
      "step": 6350
    },
    {
      "entropy": 5.940042352676391,
      "epoch": 0.561197456729071,
      "grad_norm": 0.890625,
      "learning_rate": 0.000497477808747823,
      "loss": 5.7497,
      "mean_token_accuracy": 0.17437803000211716,
      "num_tokens": 11861109.0,
      "step": 6355
    },
    {
      "entropy": 5.960816669464111,
      "epoch": 0.5616389968209113,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004974731045357292,
      "loss": 5.845,
      "mean_token_accuracy": 0.1676926702260971,
      "num_tokens": 11871015.0,
      "step": 6360
    },
    {
      "entropy": 5.906099796295166,
      "epoch": 0.5620805369127517,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004974683959655136,
      "loss": 5.7129,
      "mean_token_accuracy": 0.18293437659740447,
      "num_tokens": 11881306.0,
      "step": 6365
    },
    {
      "entropy": 5.926035451889038,
      "epoch": 0.562522077004592,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004974636830372686,
      "loss": 5.648,
      "mean_token_accuracy": 0.188501039147377,
      "num_tokens": 11890251.0,
      "step": 6370
    },
    {
      "entropy": 5.925776243209839,
      "epoch": 0.5629636170964324,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004974589657510866,
      "loss": 5.7729,
      "mean_token_accuracy": 0.17742838859558105,
      "num_tokens": 11899662.0,
      "step": 6375
    },
    {
      "entropy": 5.8874077796936035,
      "epoch": 0.5634051571882727,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974542441070601,
      "loss": 5.727,
      "mean_token_accuracy": 0.18206923007965087,
      "num_tokens": 11908699.0,
      "step": 6380
    },
    {
      "entropy": 5.9278564453125,
      "epoch": 0.563846697280113,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004974495181052814,
      "loss": 5.7072,
      "mean_token_accuracy": 0.18245860040187836,
      "num_tokens": 11916637.0,
      "step": 6385
    },
    {
      "entropy": 5.957434034347534,
      "epoch": 0.5642882373719533,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004974447877458433,
      "loss": 5.7751,
      "mean_token_accuracy": 0.17324685156345368,
      "num_tokens": 11926063.0,
      "step": 6390
    },
    {
      "entropy": 5.929507160186768,
      "epoch": 0.5647297774637937,
      "grad_norm": 1.125,
      "learning_rate": 0.0004974400530288381,
      "loss": 5.6611,
      "mean_token_accuracy": 0.18200212568044663,
      "num_tokens": 11934861.0,
      "step": 6395
    },
    {
      "entropy": 5.959511756896973,
      "epoch": 0.5651713175556341,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004974353139543589,
      "loss": 5.7622,
      "mean_token_accuracy": 0.1750079497694969,
      "num_tokens": 11944403.0,
      "step": 6400
    },
    {
      "entropy": 5.965158271789551,
      "epoch": 0.5656128576474744,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004974305705224983,
      "loss": 5.8406,
      "mean_token_accuracy": 0.17911516278982162,
      "num_tokens": 11953306.0,
      "step": 6405
    },
    {
      "entropy": 5.934712839126587,
      "epoch": 0.5660543977393148,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004974258227333493,
      "loss": 5.7194,
      "mean_token_accuracy": 0.17788851857185364,
      "num_tokens": 11962727.0,
      "step": 6410
    },
    {
      "entropy": 5.954853057861328,
      "epoch": 0.566495937831155,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004974210705870052,
      "loss": 5.8619,
      "mean_token_accuracy": 0.17795796543359757,
      "num_tokens": 11972375.0,
      "step": 6415
    },
    {
      "entropy": 5.850558567047119,
      "epoch": 0.5669374779229954,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004974163140835586,
      "loss": 5.6381,
      "mean_token_accuracy": 0.18238368183374404,
      "num_tokens": 11981989.0,
      "step": 6420
    },
    {
      "entropy": 5.973133182525634,
      "epoch": 0.5673790180148357,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004974115532231029,
      "loss": 5.6917,
      "mean_token_accuracy": 0.18118394017219544,
      "num_tokens": 11990865.0,
      "step": 6425
    },
    {
      "entropy": 5.942868566513061,
      "epoch": 0.5678205581066761,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004974067880057315,
      "loss": 5.6916,
      "mean_token_accuracy": 0.1895127236843109,
      "num_tokens": 12000697.0,
      "step": 6430
    },
    {
      "entropy": 5.884861660003662,
      "epoch": 0.5682620981985165,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004974020184315375,
      "loss": 5.7739,
      "mean_token_accuracy": 0.17718288153409958,
      "num_tokens": 12010560.0,
      "step": 6435
    },
    {
      "entropy": 5.981086111068725,
      "epoch": 0.5687036382903567,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004973972445006144,
      "loss": 5.7798,
      "mean_token_accuracy": 0.1795758858323097,
      "num_tokens": 12020760.0,
      "step": 6440
    },
    {
      "entropy": 5.845777130126953,
      "epoch": 0.5691451783821971,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004973924662130559,
      "loss": 5.6409,
      "mean_token_accuracy": 0.18322015553712845,
      "num_tokens": 12030467.0,
      "step": 6445
    },
    {
      "entropy": 5.929102373123169,
      "epoch": 0.5695867184740374,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973876835689554,
      "loss": 5.7441,
      "mean_token_accuracy": 0.17533728629350662,
      "num_tokens": 12039897.0,
      "step": 6450
    },
    {
      "entropy": 5.986848783493042,
      "epoch": 0.5700282585658778,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004973828965684066,
      "loss": 5.7667,
      "mean_token_accuracy": 0.1707102909684181,
      "num_tokens": 12048997.0,
      "step": 6455
    },
    {
      "entropy": 5.889280796051025,
      "epoch": 0.5704697986577181,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004973781052115035,
      "loss": 5.6626,
      "mean_token_accuracy": 0.18899978697299957,
      "num_tokens": 12057493.0,
      "step": 6460
    },
    {
      "entropy": 5.862435913085937,
      "epoch": 0.5709113387495585,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004973733094983396,
      "loss": 5.7348,
      "mean_token_accuracy": 0.18229613900184632,
      "num_tokens": 12066329.0,
      "step": 6465
    },
    {
      "entropy": 5.959417295455933,
      "epoch": 0.5713528788413988,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973685094290092,
      "loss": 5.7654,
      "mean_token_accuracy": 0.1740923374891281,
      "num_tokens": 12076022.0,
      "step": 6470
    },
    {
      "entropy": 5.901052379608155,
      "epoch": 0.5717944189332391,
      "grad_norm": 0.85546875,
      "learning_rate": 0.000497363705003606,
      "loss": 5.7036,
      "mean_token_accuracy": 0.17817170470952987,
      "num_tokens": 12085750.0,
      "step": 6475
    },
    {
      "entropy": 5.895412921905518,
      "epoch": 0.5722359590250795,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004973588962222244,
      "loss": 5.753,
      "mean_token_accuracy": 0.17770908027887344,
      "num_tokens": 12094935.0,
      "step": 6480
    },
    {
      "entropy": 5.895012998580933,
      "epoch": 0.5726774991169198,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004973540830849585,
      "loss": 5.6811,
      "mean_token_accuracy": 0.17887165248394013,
      "num_tokens": 12104970.0,
      "step": 6485
    },
    {
      "entropy": 5.959931230545044,
      "epoch": 0.5731190392087602,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004973492655919025,
      "loss": 5.7748,
      "mean_token_accuracy": 0.17862619161605836,
      "num_tokens": 12116195.0,
      "step": 6490
    },
    {
      "entropy": 5.988841438293457,
      "epoch": 0.5735605793006004,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004973444437431508,
      "loss": 5.7261,
      "mean_token_accuracy": 0.18133638501167298,
      "num_tokens": 12125485.0,
      "step": 6495
    },
    {
      "entropy": 5.886534547805786,
      "epoch": 0.5740021193924408,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004973396175387981,
      "loss": 5.7432,
      "mean_token_accuracy": 0.18072778433561326,
      "num_tokens": 12135474.0,
      "step": 6500
    },
    {
      "entropy": 5.9477166652679445,
      "epoch": 0.5744436594842812,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004973347869789385,
      "loss": 5.7713,
      "mean_token_accuracy": 0.17266511768102646,
      "num_tokens": 12144945.0,
      "step": 6505
    },
    {
      "entropy": 5.9193422317504885,
      "epoch": 0.5748851995761215,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004973299520636671,
      "loss": 5.7062,
      "mean_token_accuracy": 0.18106334358453752,
      "num_tokens": 12154893.0,
      "step": 6510
    },
    {
      "entropy": 5.919309759140015,
      "epoch": 0.5753267396679619,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004973251127930782,
      "loss": 5.7894,
      "mean_token_accuracy": 0.17616706043481828,
      "num_tokens": 12164047.0,
      "step": 6515
    },
    {
      "entropy": 5.9936785221099855,
      "epoch": 0.5757682797598022,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004973202691672669,
      "loss": 5.7679,
      "mean_token_accuracy": 0.1752978190779686,
      "num_tokens": 12172965.0,
      "step": 6520
    },
    {
      "entropy": 5.984523725509644,
      "epoch": 0.5762098198516425,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004973154211863279,
      "loss": 5.7988,
      "mean_token_accuracy": 0.17396161705255508,
      "num_tokens": 12183534.0,
      "step": 6525
    },
    {
      "entropy": 5.926711177825927,
      "epoch": 0.5766513599434828,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004973105688503563,
      "loss": 5.7586,
      "mean_token_accuracy": 0.17371885776519774,
      "num_tokens": 12194370.0,
      "step": 6530
    },
    {
      "entropy": 5.94952425956726,
      "epoch": 0.5770929000353232,
      "grad_norm": 0.875,
      "learning_rate": 0.000497305712159447,
      "loss": 5.7548,
      "mean_token_accuracy": 0.17564807534217836,
      "num_tokens": 12203765.0,
      "step": 6535
    },
    {
      "entropy": 5.950043344497681,
      "epoch": 0.5775344401271636,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004973008511136954,
      "loss": 5.6843,
      "mean_token_accuracy": 0.18619055598974227,
      "num_tokens": 12212889.0,
      "step": 6540
    },
    {
      "entropy": 5.9376091957092285,
      "epoch": 0.5779759802190039,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004972959857131963,
      "loss": 5.7568,
      "mean_token_accuracy": 0.17935430109500886,
      "num_tokens": 12222055.0,
      "step": 6545
    },
    {
      "entropy": 5.945558071136475,
      "epoch": 0.5784175203108443,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004972911159580454,
      "loss": 5.7504,
      "mean_token_accuracy": 0.1824483186006546,
      "num_tokens": 12232092.0,
      "step": 6550
    },
    {
      "entropy": 5.967494964599609,
      "epoch": 0.5788590604026845,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000497286241848338,
      "loss": 5.7899,
      "mean_token_accuracy": 0.17304221391677857,
      "num_tokens": 12241411.0,
      "step": 6555
    },
    {
      "entropy": 5.939413595199585,
      "epoch": 0.5793006004945249,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004972813633841696,
      "loss": 5.6636,
      "mean_token_accuracy": 0.17755235135555267,
      "num_tokens": 12249735.0,
      "step": 6560
    },
    {
      "entropy": 5.896793413162231,
      "epoch": 0.5797421405863652,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004972764805656355,
      "loss": 5.6744,
      "mean_token_accuracy": 0.18935427516698838,
      "num_tokens": 12259132.0,
      "step": 6565
    },
    {
      "entropy": 5.91446475982666,
      "epoch": 0.5801836806782056,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004972715933928318,
      "loss": 5.7427,
      "mean_token_accuracy": 0.18480055928230285,
      "num_tokens": 12269401.0,
      "step": 6570
    },
    {
      "entropy": 5.961280679702758,
      "epoch": 0.580625220770046,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004972667018658539,
      "loss": 5.7714,
      "mean_token_accuracy": 0.1721562162041664,
      "num_tokens": 12279163.0,
      "step": 6575
    },
    {
      "entropy": 5.897429990768432,
      "epoch": 0.5810667608618862,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004972618059847976,
      "loss": 5.6763,
      "mean_token_accuracy": 0.18488114923238755,
      "num_tokens": 12287942.0,
      "step": 6580
    },
    {
      "entropy": 5.941003751754761,
      "epoch": 0.5815083009537266,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004972569057497591,
      "loss": 5.675,
      "mean_token_accuracy": 0.17982326596975326,
      "num_tokens": 12296717.0,
      "step": 6585
    },
    {
      "entropy": 5.910008764266967,
      "epoch": 0.5819498410455669,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004972520011608341,
      "loss": 5.6576,
      "mean_token_accuracy": 0.18751820027828217,
      "num_tokens": 12305983.0,
      "step": 6590
    },
    {
      "entropy": 5.849145221710205,
      "epoch": 0.5823913811374073,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004972470922181189,
      "loss": 5.7389,
      "mean_token_accuracy": 0.17131290584802628,
      "num_tokens": 12315338.0,
      "step": 6595
    },
    {
      "entropy": 5.961781024932861,
      "epoch": 0.5828329212292476,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004972421789217094,
      "loss": 5.8171,
      "mean_token_accuracy": 0.1724231258034706,
      "num_tokens": 12324828.0,
      "step": 6600
    },
    {
      "entropy": 6.009534645080566,
      "epoch": 0.583274461321088,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972372612717022,
      "loss": 5.686,
      "mean_token_accuracy": 0.18648577928543092,
      "num_tokens": 12332844.0,
      "step": 6605
    },
    {
      "entropy": 5.874768495559692,
      "epoch": 0.5837160014129283,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004972323392681934,
      "loss": 5.7203,
      "mean_token_accuracy": 0.18115541487932205,
      "num_tokens": 12342062.0,
      "step": 6610
    },
    {
      "entropy": 5.906319189071655,
      "epoch": 0.5841575415047686,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004972274129112794,
      "loss": 5.7949,
      "mean_token_accuracy": 0.17332905381917954,
      "num_tokens": 12352071.0,
      "step": 6615
    },
    {
      "entropy": 6.078907680511475,
      "epoch": 0.584599081596609,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004972224822010567,
      "loss": 5.8644,
      "mean_token_accuracy": 0.1694015458226204,
      "num_tokens": 12362025.0,
      "step": 6620
    },
    {
      "entropy": 5.988093709945678,
      "epoch": 0.5850406216884493,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004972175471376222,
      "loss": 5.8597,
      "mean_token_accuracy": 0.16808632612228394,
      "num_tokens": 12371547.0,
      "step": 6625
    },
    {
      "entropy": 5.961208009719849,
      "epoch": 0.5854821617802897,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004972126077210722,
      "loss": 5.7425,
      "mean_token_accuracy": 0.17402835786342621,
      "num_tokens": 12381385.0,
      "step": 6630
    },
    {
      "entropy": 5.982282543182373,
      "epoch": 0.5859237018721299,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004972076639515035,
      "loss": 5.8134,
      "mean_token_accuracy": 0.17358503490686417,
      "num_tokens": 12391302.0,
      "step": 6635
    },
    {
      "entropy": 5.828851175308228,
      "epoch": 0.5863652419639703,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004972027158290131,
      "loss": 5.6969,
      "mean_token_accuracy": 0.1809057265520096,
      "num_tokens": 12400388.0,
      "step": 6640
    },
    {
      "entropy": 5.867154169082641,
      "epoch": 0.5868067820558107,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000497197763353698,
      "loss": 5.7531,
      "mean_token_accuracy": 0.17146917283535004,
      "num_tokens": 12409695.0,
      "step": 6645
    },
    {
      "entropy": 5.974938201904297,
      "epoch": 0.587248322147651,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000497192806525655,
      "loss": 5.6163,
      "mean_token_accuracy": 0.19104450494050979,
      "num_tokens": 12418728.0,
      "step": 6650
    },
    {
      "entropy": 5.9528045654296875,
      "epoch": 0.5876898622394914,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004971878453449811,
      "loss": 5.7788,
      "mean_token_accuracy": 0.18384409248828887,
      "num_tokens": 12428072.0,
      "step": 6655
    },
    {
      "entropy": 5.993191766738891,
      "epoch": 0.5881314023313317,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000497182879811774,
      "loss": 5.7131,
      "mean_token_accuracy": 0.18061271756887437,
      "num_tokens": 12436624.0,
      "step": 6660
    },
    {
      "entropy": 5.881068277359009,
      "epoch": 0.588572942423172,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004971779099261304,
      "loss": 5.7265,
      "mean_token_accuracy": 0.17819689810276032,
      "num_tokens": 12446370.0,
      "step": 6665
    },
    {
      "entropy": 5.889710187911987,
      "epoch": 0.5890144825150123,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497172935688148,
      "loss": 5.785,
      "mean_token_accuracy": 0.17816216349601746,
      "num_tokens": 12456705.0,
      "step": 6670
    },
    {
      "entropy": 5.939290475845337,
      "epoch": 0.5894560226068527,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004971679570979242,
      "loss": 5.7909,
      "mean_token_accuracy": 0.17827335596084595,
      "num_tokens": 12466150.0,
      "step": 6675
    },
    {
      "entropy": 5.957861614227295,
      "epoch": 0.5898975626986931,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004971629741555565,
      "loss": 5.6405,
      "mean_token_accuracy": 0.18321777880191803,
      "num_tokens": 12475051.0,
      "step": 6680
    },
    {
      "entropy": 5.9179764747619625,
      "epoch": 0.5903391027905334,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004971579868611423,
      "loss": 5.7193,
      "mean_token_accuracy": 0.18429083675146102,
      "num_tokens": 12485427.0,
      "step": 6685
    },
    {
      "entropy": 5.848150014877319,
      "epoch": 0.5907806428823738,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004971529952147797,
      "loss": 5.6583,
      "mean_token_accuracy": 0.18403666615486144,
      "num_tokens": 12494489.0,
      "step": 6690
    },
    {
      "entropy": 5.954574680328369,
      "epoch": 0.591222182974214,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000497147999216566,
      "loss": 5.6894,
      "mean_token_accuracy": 0.1826411232352257,
      "num_tokens": 12503656.0,
      "step": 6695
    },
    {
      "entropy": 5.95741491317749,
      "epoch": 0.5916637230660544,
      "grad_norm": 1.25,
      "learning_rate": 0.0004971429988665996,
      "loss": 5.7271,
      "mean_token_accuracy": 0.17112142443656922,
      "num_tokens": 12512940.0,
      "step": 6700
    },
    {
      "entropy": 5.92096962928772,
      "epoch": 0.5921052631578947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497137994164978,
      "loss": 5.6653,
      "mean_token_accuracy": 0.18557649552822114,
      "num_tokens": 12523186.0,
      "step": 6705
    },
    {
      "entropy": 5.926599788665771,
      "epoch": 0.5925468032497351,
      "grad_norm": 1.0,
      "learning_rate": 0.0004971329851117996,
      "loss": 5.709,
      "mean_token_accuracy": 0.1801753506064415,
      "num_tokens": 12533018.0,
      "step": 6710
    },
    {
      "entropy": 5.996683883666992,
      "epoch": 0.5929883433415755,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004971279717071622,
      "loss": 5.7826,
      "mean_token_accuracy": 0.17882051169872284,
      "num_tokens": 12542889.0,
      "step": 6715
    },
    {
      "entropy": 5.967419385910034,
      "epoch": 0.5934298834334157,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004971229539511643,
      "loss": 5.7956,
      "mean_token_accuracy": 0.17490288019180297,
      "num_tokens": 12553625.0,
      "step": 6720
    },
    {
      "entropy": 5.867678117752075,
      "epoch": 0.5938714235252561,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000497117931843904,
      "loss": 5.7149,
      "mean_token_accuracy": 0.17882569432258605,
      "num_tokens": 12563012.0,
      "step": 6725
    },
    {
      "entropy": 6.035449504852295,
      "epoch": 0.5943129636170964,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004971129053854797,
      "loss": 5.8473,
      "mean_token_accuracy": 0.1676483556628227,
      "num_tokens": 12572892.0,
      "step": 6730
    },
    {
      "entropy": 5.953235340118408,
      "epoch": 0.5947545037089368,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049710787457599,
      "loss": 5.6586,
      "mean_token_accuracy": 0.1844158798456192,
      "num_tokens": 12581550.0,
      "step": 6735
    },
    {
      "entropy": 5.877220201492309,
      "epoch": 0.5951960438007771,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004971028394155332,
      "loss": 5.7398,
      "mean_token_accuracy": 0.17651234865188598,
      "num_tokens": 12590599.0,
      "step": 6740
    },
    {
      "entropy": 5.981046724319458,
      "epoch": 0.5956375838926175,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004970977999042082,
      "loss": 5.7146,
      "mean_token_accuracy": 0.18010410368442537,
      "num_tokens": 12599497.0,
      "step": 6745
    },
    {
      "entropy": 5.942370986938476,
      "epoch": 0.5960791239844578,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004970927560421137,
      "loss": 5.7958,
      "mean_token_accuracy": 0.17954052835702897,
      "num_tokens": 12608273.0,
      "step": 6750
    },
    {
      "entropy": 5.882658863067627,
      "epoch": 0.5965206640762981,
      "grad_norm": 1.125,
      "learning_rate": 0.0004970877078293483,
      "loss": 5.6383,
      "mean_token_accuracy": 0.18409765362739564,
      "num_tokens": 12617287.0,
      "step": 6755
    },
    {
      "entropy": 5.927299499511719,
      "epoch": 0.5969622041681385,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004970826552660111,
      "loss": 5.7166,
      "mean_token_accuracy": 0.17487230598926545,
      "num_tokens": 12626874.0,
      "step": 6760
    },
    {
      "entropy": 5.94651575088501,
      "epoch": 0.5974037442599788,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970775983522009,
      "loss": 5.7682,
      "mean_token_accuracy": 0.17655521929264067,
      "num_tokens": 12635218.0,
      "step": 6765
    },
    {
      "entropy": 5.948738765716553,
      "epoch": 0.5978452843518192,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970725370880168,
      "loss": 5.7085,
      "mean_token_accuracy": 0.17957753241062163,
      "num_tokens": 12644395.0,
      "step": 6770
    },
    {
      "entropy": 5.971686220169067,
      "epoch": 0.5982868244436594,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004970674714735581,
      "loss": 5.7582,
      "mean_token_accuracy": 0.1807936668395996,
      "num_tokens": 12654540.0,
      "step": 6775
    },
    {
      "entropy": 5.892604112625122,
      "epoch": 0.5987283645354998,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004970624015089239,
      "loss": 5.5915,
      "mean_token_accuracy": 0.18906807750463486,
      "num_tokens": 12663546.0,
      "step": 6780
    },
    {
      "entropy": 5.866572380065918,
      "epoch": 0.5991699046273402,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970573271942135,
      "loss": 5.7395,
      "mean_token_accuracy": 0.18058196157217027,
      "num_tokens": 12672353.0,
      "step": 6785
    },
    {
      "entropy": 5.866556930541992,
      "epoch": 0.5996114447191805,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004970522485295265,
      "loss": 5.6094,
      "mean_token_accuracy": 0.19295398741960526,
      "num_tokens": 12680833.0,
      "step": 6790
    },
    {
      "entropy": 5.851094532012939,
      "epoch": 0.6000529848110209,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004970471655149621,
      "loss": 5.5789,
      "mean_token_accuracy": 0.1865946739912033,
      "num_tokens": 12690033.0,
      "step": 6795
    },
    {
      "entropy": 5.816410923004151,
      "epoch": 0.6004945249028611,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049704207815062,
      "loss": 5.6477,
      "mean_token_accuracy": 0.18757187128067015,
      "num_tokens": 12700127.0,
      "step": 6800
    },
    {
      "entropy": 5.935033321380615,
      "epoch": 0.6009360649947015,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004970369864366,
      "loss": 5.6766,
      "mean_token_accuracy": 0.18124886006116867,
      "num_tokens": 12709803.0,
      "step": 6805
    },
    {
      "entropy": 5.913188362121582,
      "epoch": 0.6013776050865418,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004970318903730017,
      "loss": 5.6686,
      "mean_token_accuracy": 0.18383529633283616,
      "num_tokens": 12718876.0,
      "step": 6810
    },
    {
      "entropy": 5.947503089904785,
      "epoch": 0.6018191451783822,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004970267899599249,
      "loss": 5.7802,
      "mean_token_accuracy": 0.17513394802808763,
      "num_tokens": 12728745.0,
      "step": 6815
    },
    {
      "entropy": 5.921718597412109,
      "epoch": 0.6022606852702226,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004970216851974695,
      "loss": 5.7881,
      "mean_token_accuracy": 0.17280806452035904,
      "num_tokens": 12738764.0,
      "step": 6820
    },
    {
      "entropy": 5.870873689651489,
      "epoch": 0.6027022253620629,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004970165760857356,
      "loss": 5.7253,
      "mean_token_accuracy": 0.17291287034749986,
      "num_tokens": 12747888.0,
      "step": 6825
    },
    {
      "entropy": 6.020090627670288,
      "epoch": 0.6031437654539032,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004970114626248232,
      "loss": 5.7377,
      "mean_token_accuracy": 0.18093507140874862,
      "num_tokens": 12757763.0,
      "step": 6830
    },
    {
      "entropy": 5.94991626739502,
      "epoch": 0.6035853055457435,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004970063448148326,
      "loss": 5.77,
      "mean_token_accuracy": 0.17415328770875932,
      "num_tokens": 12767924.0,
      "step": 6835
    },
    {
      "entropy": 5.948456811904907,
      "epoch": 0.6040268456375839,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004970012226558639,
      "loss": 5.7911,
      "mean_token_accuracy": 0.17666855156421662,
      "num_tokens": 12777140.0,
      "step": 6840
    },
    {
      "entropy": 5.975889825820923,
      "epoch": 0.6044683857294242,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004969960961480176,
      "loss": 5.7618,
      "mean_token_accuracy": 0.1819861873984337,
      "num_tokens": 12786731.0,
      "step": 6845
    },
    {
      "entropy": 5.902105522155762,
      "epoch": 0.6049099258212646,
      "grad_norm": 1.0,
      "learning_rate": 0.0004969909652913939,
      "loss": 5.6598,
      "mean_token_accuracy": 0.18601689636707305,
      "num_tokens": 12796839.0,
      "step": 6850
    },
    {
      "entropy": 5.9462145328521725,
      "epoch": 0.605351465913105,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004969858300860935,
      "loss": 5.7836,
      "mean_token_accuracy": 0.17639988511800767,
      "num_tokens": 12805944.0,
      "step": 6855
    },
    {
      "entropy": 5.95946044921875,
      "epoch": 0.6057930060049452,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004969806905322169,
      "loss": 5.7202,
      "mean_token_accuracy": 0.1749048963189125,
      "num_tokens": 12814692.0,
      "step": 6860
    },
    {
      "entropy": 5.877088022232056,
      "epoch": 0.6062345460967856,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004969755466298649,
      "loss": 5.6688,
      "mean_token_accuracy": 0.17678387016057967,
      "num_tokens": 12824506.0,
      "step": 6865
    },
    {
      "entropy": 5.915852737426758,
      "epoch": 0.6066760861886259,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969703983791382,
      "loss": 5.7392,
      "mean_token_accuracy": 0.18122698962688447,
      "num_tokens": 12835157.0,
      "step": 6870
    },
    {
      "entropy": 5.91579122543335,
      "epoch": 0.6071176262804663,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004969652457801376,
      "loss": 5.6357,
      "mean_token_accuracy": 0.1784660905599594,
      "num_tokens": 12844281.0,
      "step": 6875
    },
    {
      "entropy": 5.997165012359619,
      "epoch": 0.6075591663723066,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004969600888329641,
      "loss": 5.7117,
      "mean_token_accuracy": 0.18102120906114577,
      "num_tokens": 12854486.0,
      "step": 6880
    },
    {
      "entropy": 5.928967142105103,
      "epoch": 0.608000706464147,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004969549275377186,
      "loss": 5.701,
      "mean_token_accuracy": 0.17599090039730073,
      "num_tokens": 12863574.0,
      "step": 6885
    },
    {
      "entropy": 5.84666748046875,
      "epoch": 0.6084422465559873,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969497618945024,
      "loss": 5.6549,
      "mean_token_accuracy": 0.18378470838069916,
      "num_tokens": 12872450.0,
      "step": 6890
    },
    {
      "entropy": 5.866508674621582,
      "epoch": 0.6088837866478276,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004969445919034165,
      "loss": 5.6066,
      "mean_token_accuracy": 0.19402503222227097,
      "num_tokens": 12881299.0,
      "step": 6895
    },
    {
      "entropy": 5.8215114116668705,
      "epoch": 0.609325326739668,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004969394175645623,
      "loss": 5.5739,
      "mean_token_accuracy": 0.18549795001745223,
      "num_tokens": 12889943.0,
      "step": 6900
    },
    {
      "entropy": 5.840115118026733,
      "epoch": 0.6097668668315083,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004969342388780411,
      "loss": 5.6779,
      "mean_token_accuracy": 0.18420430421829223,
      "num_tokens": 12899497.0,
      "step": 6905
    },
    {
      "entropy": 5.927986431121826,
      "epoch": 0.6102084069233487,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004969290558439546,
      "loss": 5.6608,
      "mean_token_accuracy": 0.177994304895401,
      "num_tokens": 12908677.0,
      "step": 6910
    },
    {
      "entropy": 5.953958320617676,
      "epoch": 0.6106499470151889,
      "grad_norm": 0.82421875,
      "learning_rate": 0.0004969238684624038,
      "loss": 5.7974,
      "mean_token_accuracy": 0.17282726615667343,
      "num_tokens": 12918468.0,
      "step": 6915
    },
    {
      "entropy": 5.8817911624908445,
      "epoch": 0.6110914871070293,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004969186767334908,
      "loss": 5.677,
      "mean_token_accuracy": 0.18767081946134567,
      "num_tokens": 12928527.0,
      "step": 6920
    },
    {
      "entropy": 5.910523414611816,
      "epoch": 0.6115330271988697,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004969134806573172,
      "loss": 5.6662,
      "mean_token_accuracy": 0.1900987759232521,
      "num_tokens": 12937021.0,
      "step": 6925
    },
    {
      "entropy": 5.804785871505738,
      "epoch": 0.61197456729071,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004969082802339846,
      "loss": 5.5893,
      "mean_token_accuracy": 0.1889987915754318,
      "num_tokens": 12945570.0,
      "step": 6930
    },
    {
      "entropy": 5.893771982192993,
      "epoch": 0.6124161073825504,
      "grad_norm": 0.8828125,
      "learning_rate": 0.000496903075463595,
      "loss": 5.7471,
      "mean_token_accuracy": 0.18079857677221298,
      "num_tokens": 12954594.0,
      "step": 6935
    },
    {
      "entropy": 5.93186502456665,
      "epoch": 0.6128576474743906,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004968978663462504,
      "loss": 5.6933,
      "mean_token_accuracy": 0.18866560161113738,
      "num_tokens": 12963328.0,
      "step": 6940
    },
    {
      "entropy": 5.902009201049805,
      "epoch": 0.613299187566231,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004968926528820529,
      "loss": 5.7045,
      "mean_token_accuracy": 0.1821183055639267,
      "num_tokens": 12972806.0,
      "step": 6945
    },
    {
      "entropy": 5.9072239875793455,
      "epoch": 0.6137407276580713,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004968874350711044,
      "loss": 5.7147,
      "mean_token_accuracy": 0.17737514525651932,
      "num_tokens": 12981940.0,
      "step": 6950
    },
    {
      "entropy": 5.902434349060059,
      "epoch": 0.6141822677499117,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968822129135072,
      "loss": 5.6743,
      "mean_token_accuracy": 0.18073017597198487,
      "num_tokens": 12990501.0,
      "step": 6955
    },
    {
      "entropy": 5.960966110229492,
      "epoch": 0.6146238078417521,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004968769864093637,
      "loss": 5.701,
      "mean_token_accuracy": 0.17968421429395676,
      "num_tokens": 12999734.0,
      "step": 6960
    },
    {
      "entropy": 6.008027362823486,
      "epoch": 0.6150653479335924,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968717555587762,
      "loss": 5.7938,
      "mean_token_accuracy": 0.17155789136886596,
      "num_tokens": 13009323.0,
      "step": 6965
    },
    {
      "entropy": 5.979251575469971,
      "epoch": 0.6155068880254327,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004968665203618472,
      "loss": 5.7889,
      "mean_token_accuracy": 0.18105098307132722,
      "num_tokens": 13018136.0,
      "step": 6970
    },
    {
      "entropy": 5.955159854888916,
      "epoch": 0.615948428117273,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968612808186793,
      "loss": 5.7365,
      "mean_token_accuracy": 0.17707560658454896,
      "num_tokens": 13026494.0,
      "step": 6975
    },
    {
      "entropy": 5.939317274093628,
      "epoch": 0.6163899682091134,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004968560369293751,
      "loss": 5.7283,
      "mean_token_accuracy": 0.17934589833021164,
      "num_tokens": 13036548.0,
      "step": 6980
    },
    {
      "entropy": 5.967477321624756,
      "epoch": 0.6168315083009537,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004968507886940372,
      "loss": 5.7605,
      "mean_token_accuracy": 0.18426348268985748,
      "num_tokens": 13045653.0,
      "step": 6985
    },
    {
      "entropy": 6.00388650894165,
      "epoch": 0.6172730483927941,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004968455361127686,
      "loss": 5.8066,
      "mean_token_accuracy": 0.17793480753898622,
      "num_tokens": 13056204.0,
      "step": 6990
    },
    {
      "entropy": 5.94718656539917,
      "epoch": 0.6177145884846345,
      "grad_norm": 0.828125,
      "learning_rate": 0.0004968402791856721,
      "loss": 5.7655,
      "mean_token_accuracy": 0.1818525716662407,
      "num_tokens": 13067215.0,
      "step": 6995
    },
    {
      "entropy": 5.893466234207153,
      "epoch": 0.6181561285764747,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968350179128507,
      "loss": 5.7287,
      "mean_token_accuracy": 0.17613782733678818,
      "num_tokens": 13076824.0,
      "step": 7000
    },
    {
      "entropy": 5.9018186092376705,
      "epoch": 0.6185976686683151,
      "grad_norm": 0.875,
      "learning_rate": 0.0004968297522944075,
      "loss": 5.6894,
      "mean_token_accuracy": 0.1825672522187233,
      "num_tokens": 13086205.0,
      "step": 7005
    },
    {
      "entropy": 5.870771694183349,
      "epoch": 0.6190392087601554,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004968244823304455,
      "loss": 5.6302,
      "mean_token_accuracy": 0.18527472019195557,
      "num_tokens": 13096381.0,
      "step": 7010
    },
    {
      "entropy": 5.887761640548706,
      "epoch": 0.6194807488519958,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004968192080210683,
      "loss": 5.6947,
      "mean_token_accuracy": 0.18954527229070664,
      "num_tokens": 13106533.0,
      "step": 7015
    },
    {
      "entropy": 5.844252872467041,
      "epoch": 0.619922288943836,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004968139293663787,
      "loss": 5.6596,
      "mean_token_accuracy": 0.18514977842569352,
      "num_tokens": 13116605.0,
      "step": 7020
    },
    {
      "entropy": 5.889920091629028,
      "epoch": 0.6203638290356764,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004968086463664806,
      "loss": 5.6567,
      "mean_token_accuracy": 0.1948334753513336,
      "num_tokens": 13125316.0,
      "step": 7025
    },
    {
      "entropy": 6.018892908096314,
      "epoch": 0.6208053691275168,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496803359021477,
      "loss": 5.8357,
      "mean_token_accuracy": 0.18200814127922058,
      "num_tokens": 13135595.0,
      "step": 7030
    },
    {
      "entropy": 5.931438970565796,
      "epoch": 0.6212469092193571,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004967980673314719,
      "loss": 5.763,
      "mean_token_accuracy": 0.1763605758547783,
      "num_tokens": 13144395.0,
      "step": 7035
    },
    {
      "entropy": 5.873738098144531,
      "epoch": 0.6216884493111975,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004967927712965688,
      "loss": 5.7099,
      "mean_token_accuracy": 0.17793530225753784,
      "num_tokens": 13153906.0,
      "step": 7040
    },
    {
      "entropy": 5.951040601730346,
      "epoch": 0.6221299894030378,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0004967874709168714,
      "loss": 5.6495,
      "mean_token_accuracy": 0.18728768527507783,
      "num_tokens": 13163840.0,
      "step": 7045
    },
    {
      "entropy": 5.889948749542237,
      "epoch": 0.6225715294948782,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004967821661924837,
      "loss": 5.6492,
      "mean_token_accuracy": 0.18374582082033158,
      "num_tokens": 13173224.0,
      "step": 7050
    },
    {
      "entropy": 5.907984256744385,
      "epoch": 0.6230130695867184,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004967768571235094,
      "loss": 5.7068,
      "mean_token_accuracy": 0.1758308783173561,
      "num_tokens": 13182664.0,
      "step": 7055
    },
    {
      "entropy": 5.906645822525024,
      "epoch": 0.6234546096785588,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004967715437100526,
      "loss": 5.5975,
      "mean_token_accuracy": 0.1858988508582115,
      "num_tokens": 13191076.0,
      "step": 7060
    },
    {
      "entropy": 5.988809823989868,
      "epoch": 0.6238961497703992,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967662259522174,
      "loss": 5.8038,
      "mean_token_accuracy": 0.1728333130478859,
      "num_tokens": 13201075.0,
      "step": 7065
    },
    {
      "entropy": 5.931052255630493,
      "epoch": 0.6243376898622395,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000496760903850108,
      "loss": 5.7399,
      "mean_token_accuracy": 0.18440952450037001,
      "num_tokens": 13210602.0,
      "step": 7070
    },
    {
      "entropy": 5.948591327667236,
      "epoch": 0.6247792299540799,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004967555774038286,
      "loss": 5.7262,
      "mean_token_accuracy": 0.17679778188467027,
      "num_tokens": 13219601.0,
      "step": 7075
    },
    {
      "entropy": 5.946039438247681,
      "epoch": 0.6252207700459201,
      "grad_norm": 0.76953125,
      "learning_rate": 0.0004967502466134836,
      "loss": 5.7772,
      "mean_token_accuracy": 0.1736842855811119,
      "num_tokens": 13230675.0,
      "step": 7080
    },
    {
      "entropy": 5.926424598693847,
      "epoch": 0.6256623101377605,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004967449114791773,
      "loss": 5.7228,
      "mean_token_accuracy": 0.18056838661432267,
      "num_tokens": 13240349.0,
      "step": 7085
    },
    {
      "entropy": 5.9506090641021725,
      "epoch": 0.6261038502296008,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004967395720010143,
      "loss": 5.7857,
      "mean_token_accuracy": 0.16999740153551102,
      "num_tokens": 13249867.0,
      "step": 7090
    },
    {
      "entropy": 5.962334060668946,
      "epoch": 0.6265453903214412,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004967342281790992,
      "loss": 5.7031,
      "mean_token_accuracy": 0.18430460393428802,
      "num_tokens": 13259143.0,
      "step": 7095
    },
    {
      "entropy": 5.884312629699707,
      "epoch": 0.6269869304132816,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004967288800135368,
      "loss": 5.6321,
      "mean_token_accuracy": 0.18402725905179979,
      "num_tokens": 13268673.0,
      "step": 7100
    },
    {
      "entropy": 5.897164916992187,
      "epoch": 0.6274284705051218,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004967235275044315,
      "loss": 5.6694,
      "mean_token_accuracy": 0.1823020711541176,
      "num_tokens": 13278035.0,
      "step": 7105
    },
    {
      "entropy": 5.945112085342407,
      "epoch": 0.6278700105969622,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004967181706518886,
      "loss": 5.7944,
      "mean_token_accuracy": 0.16981353759765624,
      "num_tokens": 13286950.0,
      "step": 7110
    },
    {
      "entropy": 5.843374967575073,
      "epoch": 0.6283115506888025,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004967128094560128,
      "loss": 5.6535,
      "mean_token_accuracy": 0.1964161604642868,
      "num_tokens": 13296425.0,
      "step": 7115
    },
    {
      "entropy": 6.004785966873169,
      "epoch": 0.6287530907806429,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004967074439169092,
      "loss": 5.7793,
      "mean_token_accuracy": 0.17232254892587662,
      "num_tokens": 13306094.0,
      "step": 7120
    },
    {
      "entropy": 5.918289661407471,
      "epoch": 0.6291946308724832,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967020740346827,
      "loss": 5.7147,
      "mean_token_accuracy": 0.1865583762526512,
      "num_tokens": 13315181.0,
      "step": 7125
    },
    {
      "entropy": 5.852770471572876,
      "epoch": 0.6296361709643236,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004966966998094387,
      "loss": 5.6237,
      "mean_token_accuracy": 0.1867580607533455,
      "num_tokens": 13324108.0,
      "step": 7130
    },
    {
      "entropy": 5.890590524673462,
      "epoch": 0.630077711056164,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004966913212412825,
      "loss": 5.6933,
      "mean_token_accuracy": 0.18645042479038237,
      "num_tokens": 13333101.0,
      "step": 7135
    },
    {
      "entropy": 5.852170133590699,
      "epoch": 0.6305192511480042,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004966859383303194,
      "loss": 5.6105,
      "mean_token_accuracy": 0.19161757826805115,
      "num_tokens": 13341804.0,
      "step": 7140
    },
    {
      "entropy": 5.885224866867065,
      "epoch": 0.6309607912398446,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004966805510766549,
      "loss": 5.6598,
      "mean_token_accuracy": 0.18454794138669967,
      "num_tokens": 13350911.0,
      "step": 7145
    },
    {
      "entropy": 5.943019104003906,
      "epoch": 0.6314023313316849,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004966751594803944,
      "loss": 5.6698,
      "mean_token_accuracy": 0.18281008452177047,
      "num_tokens": 13360657.0,
      "step": 7150
    },
    {
      "entropy": 5.802535104751587,
      "epoch": 0.6318438714235253,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004966697635416437,
      "loss": 5.6219,
      "mean_token_accuracy": 0.18310205191373824,
      "num_tokens": 13370462.0,
      "step": 7155
    },
    {
      "entropy": 5.953017282485962,
      "epoch": 0.6322854115153655,
      "grad_norm": 1.0,
      "learning_rate": 0.0004966643632605082,
      "loss": 5.7309,
      "mean_token_accuracy": 0.17605062574148178,
      "num_tokens": 13379446.0,
      "step": 7160
    },
    {
      "entropy": 5.932848262786865,
      "epoch": 0.6327269516072059,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000496658958637094,
      "loss": 5.7173,
      "mean_token_accuracy": 0.18260978460311889,
      "num_tokens": 13389238.0,
      "step": 7165
    },
    {
      "entropy": 5.886120319366455,
      "epoch": 0.6331684916990463,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000496653549671507,
      "loss": 5.71,
      "mean_token_accuracy": 0.1812327668070793,
      "num_tokens": 13398613.0,
      "step": 7170
    },
    {
      "entropy": 5.922674942016601,
      "epoch": 0.6336100317908866,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004966481363638529,
      "loss": 5.7151,
      "mean_token_accuracy": 0.1866316482424736,
      "num_tokens": 13407600.0,
      "step": 7175
    },
    {
      "entropy": 5.776447343826294,
      "epoch": 0.634051571882727,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004966427187142378,
      "loss": 5.6108,
      "mean_token_accuracy": 0.18910133689641953,
      "num_tokens": 13415974.0,
      "step": 7180
    },
    {
      "entropy": 5.91736102104187,
      "epoch": 0.6344931119745673,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004966372967227681,
      "loss": 5.6084,
      "mean_token_accuracy": 0.1872192531824112,
      "num_tokens": 13425628.0,
      "step": 7185
    },
    {
      "entropy": 5.85201473236084,
      "epoch": 0.6349346520664076,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004966318703895496,
      "loss": 5.6374,
      "mean_token_accuracy": 0.18410265296697617,
      "num_tokens": 13435573.0,
      "step": 7190
    },
    {
      "entropy": 5.8759300231933596,
      "epoch": 0.6353761921582479,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004966264397146888,
      "loss": 5.6048,
      "mean_token_accuracy": 0.1849142238497734,
      "num_tokens": 13443969.0,
      "step": 7195
    },
    {
      "entropy": 5.881554222106933,
      "epoch": 0.6358177322500883,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004966210046982921,
      "loss": 5.6813,
      "mean_token_accuracy": 0.18429166376590728,
      "num_tokens": 13452823.0,
      "step": 7200
    },
    {
      "entropy": 5.915408277511597,
      "epoch": 0.6362592723419287,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004966155653404661,
      "loss": 5.711,
      "mean_token_accuracy": 0.17995356470346452,
      "num_tokens": 13461090.0,
      "step": 7205
    },
    {
      "entropy": 5.9935754299163815,
      "epoch": 0.636700812433769,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004966101216413172,
      "loss": 5.7096,
      "mean_token_accuracy": 0.16896679252386093,
      "num_tokens": 13471425.0,
      "step": 7210
    },
    {
      "entropy": 5.949064111709594,
      "epoch": 0.6371423525256094,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004966046736009519,
      "loss": 5.7538,
      "mean_token_accuracy": 0.180668406188488,
      "num_tokens": 13479755.0,
      "step": 7215
    },
    {
      "entropy": 5.84395956993103,
      "epoch": 0.6375838926174496,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004965992212194772,
      "loss": 5.6444,
      "mean_token_accuracy": 0.1895698606967926,
      "num_tokens": 13488854.0,
      "step": 7220
    },
    {
      "entropy": 5.914668321609497,
      "epoch": 0.63802543270929,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004965937644969998,
      "loss": 5.7083,
      "mean_token_accuracy": 0.17636025100946426,
      "num_tokens": 13497987.0,
      "step": 7225
    },
    {
      "entropy": 5.91913480758667,
      "epoch": 0.6384669728011303,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004965883034336266,
      "loss": 5.7369,
      "mean_token_accuracy": 0.178534634411335,
      "num_tokens": 13506960.0,
      "step": 7230
    },
    {
      "entropy": 5.8813330173492435,
      "epoch": 0.6389085128929707,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004965828380294645,
      "loss": 5.6579,
      "mean_token_accuracy": 0.18809425383806228,
      "num_tokens": 13515907.0,
      "step": 7235
    },
    {
      "entropy": 5.9257501602172855,
      "epoch": 0.6393500529848111,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004965773682846207,
      "loss": 5.7611,
      "mean_token_accuracy": 0.1841880515217781,
      "num_tokens": 13524876.0,
      "step": 7240
    },
    {
      "entropy": 5.961688566207886,
      "epoch": 0.6397915930766513,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004965718941992021,
      "loss": 5.6709,
      "mean_token_accuracy": 0.18797953724861144,
      "num_tokens": 13532891.0,
      "step": 7245
    },
    {
      "entropy": 5.994545555114746,
      "epoch": 0.6402331331684917,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004965664157733162,
      "loss": 5.7375,
      "mean_token_accuracy": 0.18454429805278777,
      "num_tokens": 13542409.0,
      "step": 7250
    },
    {
      "entropy": 5.864827823638916,
      "epoch": 0.640674673260332,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004965609330070703,
      "loss": 5.7061,
      "mean_token_accuracy": 0.18280179649591446,
      "num_tokens": 13551947.0,
      "step": 7255
    },
    {
      "entropy": 5.94833664894104,
      "epoch": 0.6411162133521724,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004965554459005717,
      "loss": 5.7365,
      "mean_token_accuracy": 0.17946964353322983,
      "num_tokens": 13560812.0,
      "step": 7260
    },
    {
      "entropy": 5.968832635879517,
      "epoch": 0.6415577534440127,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004965499544539278,
      "loss": 5.7258,
      "mean_token_accuracy": 0.18233955651521683,
      "num_tokens": 13570197.0,
      "step": 7265
    },
    {
      "entropy": 5.9128265380859375,
      "epoch": 0.641999293535853,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004965444586672463,
      "loss": 5.6503,
      "mean_token_accuracy": 0.18139110803604125,
      "num_tokens": 13579614.0,
      "step": 7270
    },
    {
      "entropy": 5.899955463409424,
      "epoch": 0.6424408336276934,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000496538958540635,
      "loss": 5.6852,
      "mean_token_accuracy": 0.18292355984449388,
      "num_tokens": 13589511.0,
      "step": 7275
    },
    {
      "entropy": 5.9232399463653564,
      "epoch": 0.6428823737195337,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004965334540742015,
      "loss": 5.6729,
      "mean_token_accuracy": 0.18746213912963866,
      "num_tokens": 13597718.0,
      "step": 7280
    },
    {
      "entropy": 5.885659217834473,
      "epoch": 0.6433239138113741,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004965279452680535,
      "loss": 5.7669,
      "mean_token_accuracy": 0.17772504538297654,
      "num_tokens": 13607599.0,
      "step": 7285
    },
    {
      "entropy": 5.995887422561646,
      "epoch": 0.6437654539032144,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004965224321222991,
      "loss": 5.7932,
      "mean_token_accuracy": 0.1795261546969414,
      "num_tokens": 13616513.0,
      "step": 7290
    },
    {
      "entropy": 5.901009464263916,
      "epoch": 0.6442069939950548,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004965169146370462,
      "loss": 5.6643,
      "mean_token_accuracy": 0.18733725994825362,
      "num_tokens": 13625505.0,
      "step": 7295
    },
    {
      "entropy": 5.942426013946533,
      "epoch": 0.644648534086895,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004965113928124031,
      "loss": 5.7231,
      "mean_token_accuracy": 0.17932885885238647,
      "num_tokens": 13635340.0,
      "step": 7300
    },
    {
      "entropy": 5.905205011367798,
      "epoch": 0.6450900741787354,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004965058666484775,
      "loss": 5.6533,
      "mean_token_accuracy": 0.1771724820137024,
      "num_tokens": 13644886.0,
      "step": 7305
    },
    {
      "entropy": 5.897004556655884,
      "epoch": 0.6455316142705758,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004965003361453782,
      "loss": 5.675,
      "mean_token_accuracy": 0.18789544999599456,
      "num_tokens": 13654093.0,
      "step": 7310
    },
    {
      "entropy": 5.986611986160279,
      "epoch": 0.6459731543624161,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004964948013032132,
      "loss": 5.7582,
      "mean_token_accuracy": 0.1729942262172699,
      "num_tokens": 13663095.0,
      "step": 7315
    },
    {
      "entropy": 5.9798657417297365,
      "epoch": 0.6464146944542565,
      "grad_norm": 0.8125,
      "learning_rate": 0.000496489262122091,
      "loss": 5.78,
      "mean_token_accuracy": 0.17928575426340104,
      "num_tokens": 13673624.0,
      "step": 7320
    },
    {
      "entropy": 5.979729175567627,
      "epoch": 0.6468562345460968,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004964837186021201,
      "loss": 5.6401,
      "mean_token_accuracy": 0.1817734122276306,
      "num_tokens": 13683145.0,
      "step": 7325
    },
    {
      "entropy": 5.907504510879517,
      "epoch": 0.6472977746379371,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964781707434093,
      "loss": 5.6629,
      "mean_token_accuracy": 0.17742054164409637,
      "num_tokens": 13691963.0,
      "step": 7330
    },
    {
      "entropy": 5.77130274772644,
      "epoch": 0.6477393147297774,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004964726185460669,
      "loss": 5.5599,
      "mean_token_accuracy": 0.18250133246183395,
      "num_tokens": 13700926.0,
      "step": 7335
    },
    {
      "entropy": 5.937017917633057,
      "epoch": 0.6481808548216178,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004964670620102019,
      "loss": 5.7943,
      "mean_token_accuracy": 0.17252365052700042,
      "num_tokens": 13709824.0,
      "step": 7340
    },
    {
      "entropy": 5.945567226409912,
      "epoch": 0.6486223949134582,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004964615011359231,
      "loss": 5.6856,
      "mean_token_accuracy": 0.17581720650196075,
      "num_tokens": 13718789.0,
      "step": 7345
    },
    {
      "entropy": 5.929153537750244,
      "epoch": 0.6490639350052985,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004964559359233395,
      "loss": 5.6952,
      "mean_token_accuracy": 0.18001507371664047,
      "num_tokens": 13728180.0,
      "step": 7350
    },
    {
      "entropy": 5.902329540252685,
      "epoch": 0.6495054750971389,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004964503663725601,
      "loss": 5.8018,
      "mean_token_accuracy": 0.17223318219184874,
      "num_tokens": 13737819.0,
      "step": 7355
    },
    {
      "entropy": 5.906650400161743,
      "epoch": 0.6499470151889791,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964447924836938,
      "loss": 5.7688,
      "mean_token_accuracy": 0.17573816925287247,
      "num_tokens": 13747519.0,
      "step": 7360
    },
    {
      "entropy": 5.989020252227784,
      "epoch": 0.6503885552808195,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004964392142568501,
      "loss": 5.7569,
      "mean_token_accuracy": 0.1694555476307869,
      "num_tokens": 13757821.0,
      "step": 7365
    },
    {
      "entropy": 5.8534832954406735,
      "epoch": 0.6508300953726598,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004964336316921381,
      "loss": 5.5698,
      "mean_token_accuracy": 0.1853997752070427,
      "num_tokens": 13766560.0,
      "step": 7370
    },
    {
      "entropy": 5.892873287200928,
      "epoch": 0.6512716354645002,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004964280447896672,
      "loss": 5.6008,
      "mean_token_accuracy": 0.18718067109584807,
      "num_tokens": 13775560.0,
      "step": 7375
    },
    {
      "entropy": 5.887672233581543,
      "epoch": 0.6517131755563406,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004964224535495468,
      "loss": 5.6739,
      "mean_token_accuracy": 0.18187779188156128,
      "num_tokens": 13784391.0,
      "step": 7380
    },
    {
      "entropy": 5.880650043487549,
      "epoch": 0.6521547156481808,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004964168579718864,
      "loss": 5.6567,
      "mean_token_accuracy": 0.18475085198879243,
      "num_tokens": 13794008.0,
      "step": 7385
    },
    {
      "entropy": 5.972517251968384,
      "epoch": 0.6525962557400212,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004964112580567959,
      "loss": 5.7224,
      "mean_token_accuracy": 0.17432813197374344,
      "num_tokens": 13804321.0,
      "step": 7390
    },
    {
      "entropy": 5.9455334663391115,
      "epoch": 0.6530377958318615,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004964056538043845,
      "loss": 5.7159,
      "mean_token_accuracy": 0.17365975230932235,
      "num_tokens": 13813492.0,
      "step": 7395
    },
    {
      "entropy": 5.919697284698486,
      "epoch": 0.6534793359237019,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004964000452147624,
      "loss": 5.7034,
      "mean_token_accuracy": 0.1820142850279808,
      "num_tokens": 13822471.0,
      "step": 7400
    },
    {
      "entropy": 5.901232194900513,
      "epoch": 0.6539208760155422,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004963944322880394,
      "loss": 5.7557,
      "mean_token_accuracy": 0.17409348785877227,
      "num_tokens": 13832789.0,
      "step": 7405
    },
    {
      "entropy": 5.958118200302124,
      "epoch": 0.6543624161073825,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963888150243253,
      "loss": 5.709,
      "mean_token_accuracy": 0.1774753138422966,
      "num_tokens": 13842230.0,
      "step": 7410
    },
    {
      "entropy": 5.988575839996338,
      "epoch": 0.6548039561992229,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004963831934237302,
      "loss": 5.6402,
      "mean_token_accuracy": 0.18648962676525116,
      "num_tokens": 13851542.0,
      "step": 7415
    },
    {
      "entropy": 5.891352653503418,
      "epoch": 0.6552454962910632,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004963775674863643,
      "loss": 5.6725,
      "mean_token_accuracy": 0.1804959684610367,
      "num_tokens": 13860753.0,
      "step": 7420
    },
    {
      "entropy": 5.835614538192749,
      "epoch": 0.6556870363829036,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004963719372123378,
      "loss": 5.6639,
      "mean_token_accuracy": 0.18368293344974518,
      "num_tokens": 13870133.0,
      "step": 7425
    },
    {
      "entropy": 5.902837753295898,
      "epoch": 0.6561285764747439,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004963663026017609,
      "loss": 5.7511,
      "mean_token_accuracy": 0.18147976696491241,
      "num_tokens": 13879086.0,
      "step": 7430
    },
    {
      "entropy": 5.9467963695526125,
      "epoch": 0.6565701165665843,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004963606636547441,
      "loss": 5.6768,
      "mean_token_accuracy": 0.18935618698596954,
      "num_tokens": 13888706.0,
      "step": 7435
    },
    {
      "entropy": 5.905650615692139,
      "epoch": 0.6570116566584245,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004963550203713976,
      "loss": 5.7261,
      "mean_token_accuracy": 0.17851834148168563,
      "num_tokens": 13897455.0,
      "step": 7440
    },
    {
      "entropy": 5.907645559310913,
      "epoch": 0.6574531967502649,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004963493727518323,
      "loss": 5.6984,
      "mean_token_accuracy": 0.18513126969337462,
      "num_tokens": 13907697.0,
      "step": 7445
    },
    {
      "entropy": 5.926703357696534,
      "epoch": 0.6578947368421053,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004963437207961585,
      "loss": 5.7227,
      "mean_token_accuracy": 0.1811746746301651,
      "num_tokens": 13917124.0,
      "step": 7450
    },
    {
      "entropy": 5.971260261535645,
      "epoch": 0.6583362769339456,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004963380645044874,
      "loss": 5.6622,
      "mean_token_accuracy": 0.18359404802322388,
      "num_tokens": 13926434.0,
      "step": 7455
    },
    {
      "entropy": 5.838030576705933,
      "epoch": 0.658777817025786,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004963324038769292,
      "loss": 5.6398,
      "mean_token_accuracy": 0.1903223752975464,
      "num_tokens": 13934883.0,
      "step": 7460
    },
    {
      "entropy": 5.900352668762207,
      "epoch": 0.6592193571176262,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004963267389135953,
      "loss": 5.7429,
      "mean_token_accuracy": 0.1787104442715645,
      "num_tokens": 13944514.0,
      "step": 7465
    },
    {
      "entropy": 5.90740818977356,
      "epoch": 0.6596608972094666,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004963210696145964,
      "loss": 5.6602,
      "mean_token_accuracy": 0.179434372484684,
      "num_tokens": 13953434.0,
      "step": 7470
    },
    {
      "entropy": 5.9329437732696535,
      "epoch": 0.6601024373013069,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004963153959800436,
      "loss": 5.6573,
      "mean_token_accuracy": 0.18745767027139665,
      "num_tokens": 13962493.0,
      "step": 7475
    },
    {
      "entropy": 5.919755315780639,
      "epoch": 0.6605439773931473,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000496309718010048,
      "loss": 5.7087,
      "mean_token_accuracy": 0.1846643343567848,
      "num_tokens": 13971193.0,
      "step": 7480
    },
    {
      "entropy": 5.865570878982544,
      "epoch": 0.6609855174849877,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004963040357047209,
      "loss": 5.693,
      "mean_token_accuracy": 0.18486219942569732,
      "num_tokens": 13981829.0,
      "step": 7485
    },
    {
      "entropy": 5.938862991333008,
      "epoch": 0.661427057576828,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004962983490641737,
      "loss": 5.6524,
      "mean_token_accuracy": 0.17965342551469804,
      "num_tokens": 13990262.0,
      "step": 7490
    },
    {
      "entropy": 5.931566143035889,
      "epoch": 0.6618685976686683,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004962926580885177,
      "loss": 5.711,
      "mean_token_accuracy": 0.17974878549575807,
      "num_tokens": 14000100.0,
      "step": 7495
    },
    {
      "entropy": 5.928611373901367,
      "epoch": 0.6623101377605086,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004962869627778642,
      "loss": 5.7297,
      "mean_token_accuracy": 0.18271702378988267,
      "num_tokens": 14009330.0,
      "step": 7500
    },
    {
      "entropy": 5.877379941940307,
      "epoch": 0.662751677852349,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004962812631323252,
      "loss": 5.7004,
      "mean_token_accuracy": 0.17838508486747742,
      "num_tokens": 14018109.0,
      "step": 7505
    },
    {
      "entropy": 5.857556009292603,
      "epoch": 0.6631932179441893,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000496275559152012,
      "loss": 5.5939,
      "mean_token_accuracy": 0.18382253348827363,
      "num_tokens": 14027318.0,
      "step": 7510
    },
    {
      "entropy": 5.90235743522644,
      "epoch": 0.6636347580360297,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004962698508370365,
      "loss": 5.7087,
      "mean_token_accuracy": 0.18098356127738952,
      "num_tokens": 14037061.0,
      "step": 7515
    },
    {
      "entropy": 5.900296354293824,
      "epoch": 0.6640762981278701,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004962641381875103,
      "loss": 5.6673,
      "mean_token_accuracy": 0.17706697136163713,
      "num_tokens": 14046008.0,
      "step": 7520
    },
    {
      "entropy": 5.904057312011719,
      "epoch": 0.6645178382197103,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004962584212035456,
      "loss": 5.6779,
      "mean_token_accuracy": 0.17278148829936982,
      "num_tokens": 14054535.0,
      "step": 7525
    },
    {
      "entropy": 5.935598373413086,
      "epoch": 0.6649593783115507,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962526998852543,
      "loss": 5.6885,
      "mean_token_accuracy": 0.18802266865968703,
      "num_tokens": 14063680.0,
      "step": 7530
    },
    {
      "entropy": 5.948052835464478,
      "epoch": 0.665400918403391,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004962469742327483,
      "loss": 5.7086,
      "mean_token_accuracy": 0.18357841819524764,
      "num_tokens": 14072206.0,
      "step": 7535
    },
    {
      "entropy": 5.953330135345459,
      "epoch": 0.6658424584952314,
      "grad_norm": 0.90625,
      "learning_rate": 0.00049624124424614,
      "loss": 5.7194,
      "mean_token_accuracy": 0.18233713656663894,
      "num_tokens": 14081943.0,
      "step": 7540
    },
    {
      "entropy": 5.907440185546875,
      "epoch": 0.6662839985870717,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004962355099255415,
      "loss": 5.6778,
      "mean_token_accuracy": 0.18833412528038024,
      "num_tokens": 14091695.0,
      "step": 7545
    },
    {
      "entropy": 5.81530237197876,
      "epoch": 0.666725538678912,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004962297712710653,
      "loss": 5.6031,
      "mean_token_accuracy": 0.1932504266500473,
      "num_tokens": 14100811.0,
      "step": 7550
    },
    {
      "entropy": 5.941954612731934,
      "epoch": 0.6671670787707524,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004962240282828236,
      "loss": 5.7873,
      "mean_token_accuracy": 0.1661756306886673,
      "num_tokens": 14109818.0,
      "step": 7555
    },
    {
      "entropy": 5.935464525222779,
      "epoch": 0.6676086188625927,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000496218280960929,
      "loss": 5.7011,
      "mean_token_accuracy": 0.17778615057468414,
      "num_tokens": 14119811.0,
      "step": 7560
    },
    {
      "entropy": 5.979626893997192,
      "epoch": 0.6680501589544331,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004962125293054941,
      "loss": 5.6657,
      "mean_token_accuracy": 0.18736949414014817,
      "num_tokens": 14128963.0,
      "step": 7565
    },
    {
      "entropy": 5.878252077102661,
      "epoch": 0.6684916990462734,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004962067733166317,
      "loss": 5.7049,
      "mean_token_accuracy": 0.17186809182167054,
      "num_tokens": 14138532.0,
      "step": 7570
    },
    {
      "entropy": 5.826684474945068,
      "epoch": 0.6689332391381138,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004962010129944542,
      "loss": 5.6849,
      "mean_token_accuracy": 0.18159957826137543,
      "num_tokens": 14148107.0,
      "step": 7575
    },
    {
      "entropy": 5.974263906478882,
      "epoch": 0.669374779229954,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004961952483390747,
      "loss": 5.7294,
      "mean_token_accuracy": 0.1763908162713051,
      "num_tokens": 14158823.0,
      "step": 7580
    },
    {
      "entropy": 5.860423851013183,
      "epoch": 0.6698163193217944,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961894793506062,
      "loss": 5.6734,
      "mean_token_accuracy": 0.1888882637023926,
      "num_tokens": 14168147.0,
      "step": 7585
    },
    {
      "entropy": 5.892768669128418,
      "epoch": 0.6702578594136348,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004961837060291615,
      "loss": 5.7062,
      "mean_token_accuracy": 0.17821233719587326,
      "num_tokens": 14177123.0,
      "step": 7590
    },
    {
      "entropy": 5.907606601715088,
      "epoch": 0.6706993995054751,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004961779283748538,
      "loss": 5.6805,
      "mean_token_accuracy": 0.1794887065887451,
      "num_tokens": 14186614.0,
      "step": 7595
    },
    {
      "entropy": 5.9154106140136715,
      "epoch": 0.6711409395973155,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004961721463877963,
      "loss": 5.6989,
      "mean_token_accuracy": 0.18051617294549943,
      "num_tokens": 14195310.0,
      "step": 7600
    },
    {
      "entropy": 5.919904899597168,
      "epoch": 0.6715824796891557,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004961663600681023,
      "loss": 5.6704,
      "mean_token_accuracy": 0.18061379790306092,
      "num_tokens": 14204321.0,
      "step": 7605
    },
    {
      "entropy": 5.88912878036499,
      "epoch": 0.6720240197809961,
      "grad_norm": 1.015625,
      "learning_rate": 0.000496160569415885,
      "loss": 5.7036,
      "mean_token_accuracy": 0.18224740624427796,
      "num_tokens": 14213615.0,
      "step": 7610
    },
    {
      "entropy": 5.953593444824219,
      "epoch": 0.6724655598728364,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004961547744312579,
      "loss": 5.571,
      "mean_token_accuracy": 0.19122107177972794,
      "num_tokens": 14222715.0,
      "step": 7615
    },
    {
      "entropy": 5.953861474990845,
      "epoch": 0.6729070999646768,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004961489751143346,
      "loss": 5.7101,
      "mean_token_accuracy": 0.18252974301576613,
      "num_tokens": 14232117.0,
      "step": 7620
    },
    {
      "entropy": 5.8538073062896725,
      "epoch": 0.6733486400565172,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004961431714652287,
      "loss": 5.5862,
      "mean_token_accuracy": 0.19271744936704635,
      "num_tokens": 14240606.0,
      "step": 7625
    },
    {
      "entropy": 5.886905479431152,
      "epoch": 0.6737901801483575,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004961373634840538,
      "loss": 5.7103,
      "mean_token_accuracy": 0.1752036228775978,
      "num_tokens": 14249844.0,
      "step": 7630
    },
    {
      "entropy": 5.935414505004883,
      "epoch": 0.6742317202401978,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004961315511709237,
      "loss": 5.7062,
      "mean_token_accuracy": 0.18466105908155442,
      "num_tokens": 14259231.0,
      "step": 7635
    },
    {
      "entropy": 5.972374820709229,
      "epoch": 0.6746732603320381,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004961257345259524,
      "loss": 5.7107,
      "mean_token_accuracy": 0.1831063836812973,
      "num_tokens": 14268508.0,
      "step": 7640
    },
    {
      "entropy": 5.89817533493042,
      "epoch": 0.6751148004238785,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004961199135492535,
      "loss": 5.6311,
      "mean_token_accuracy": 0.18840737044811248,
      "num_tokens": 14278057.0,
      "step": 7645
    },
    {
      "entropy": 5.872119998931884,
      "epoch": 0.6755563405157188,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004961140882409414,
      "loss": 5.6891,
      "mean_token_accuracy": 0.1815480887889862,
      "num_tokens": 14287225.0,
      "step": 7650
    },
    {
      "entropy": 5.89968466758728,
      "epoch": 0.6759978806075592,
      "grad_norm": 0.85546875,
      "learning_rate": 0.00049610825860113,
      "loss": 5.655,
      "mean_token_accuracy": 0.1813758134841919,
      "num_tokens": 14296529.0,
      "step": 7655
    },
    {
      "entropy": 5.920412063598633,
      "epoch": 0.6764394206993996,
      "grad_norm": 1.0,
      "learning_rate": 0.0004961024246299336,
      "loss": 5.6277,
      "mean_token_accuracy": 0.18376876413822174,
      "num_tokens": 14305675.0,
      "step": 7660
    },
    {
      "entropy": 5.9062878608703615,
      "epoch": 0.6768809607912398,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004960965863274665,
      "loss": 5.5924,
      "mean_token_accuracy": 0.18598417639732362,
      "num_tokens": 14313632.0,
      "step": 7665
    },
    {
      "entropy": 5.870595502853393,
      "epoch": 0.6773225008830802,
      "grad_norm": 1.0,
      "learning_rate": 0.000496090743693843,
      "loss": 5.7035,
      "mean_token_accuracy": 0.18137972801923752,
      "num_tokens": 14323353.0,
      "step": 7670
    },
    {
      "entropy": 5.984563302993775,
      "epoch": 0.6777640409749205,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960848967291775,
      "loss": 5.7151,
      "mean_token_accuracy": 0.1740511640906334,
      "num_tokens": 14331953.0,
      "step": 7675
    },
    {
      "entropy": 5.88297929763794,
      "epoch": 0.6782055810667609,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004960790454335846,
      "loss": 5.7209,
      "mean_token_accuracy": 0.17738450318574905,
      "num_tokens": 14342456.0,
      "step": 7680
    },
    {
      "entropy": 5.808986568450928,
      "epoch": 0.6786471211586012,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000496073189807179,
      "loss": 5.5301,
      "mean_token_accuracy": 0.19175227284431456,
      "num_tokens": 14351013.0,
      "step": 7685
    },
    {
      "entropy": 5.941276884078979,
      "epoch": 0.6790886612504415,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004960673298500753,
      "loss": 5.8007,
      "mean_token_accuracy": 0.17323903888463973,
      "num_tokens": 14360887.0,
      "step": 7690
    },
    {
      "entropy": 5.945207118988037,
      "epoch": 0.6795302013422819,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004960614655623883,
      "loss": 5.7007,
      "mean_token_accuracy": 0.1831446796655655,
      "num_tokens": 14370510.0,
      "step": 7695
    },
    {
      "entropy": 5.913691139221191,
      "epoch": 0.6799717414341222,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000496055596944233,
      "loss": 5.6944,
      "mean_token_accuracy": 0.18526873290538787,
      "num_tokens": 14379315.0,
      "step": 7700
    },
    {
      "entropy": 5.881440448760986,
      "epoch": 0.6804132815259626,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004960497239957243,
      "loss": 5.6668,
      "mean_token_accuracy": 0.18388139009475707,
      "num_tokens": 14388537.0,
      "step": 7705
    },
    {
      "entropy": 5.987562608718872,
      "epoch": 0.6808548216178029,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960438467169772,
      "loss": 5.7216,
      "mean_token_accuracy": 0.17798287868499757,
      "num_tokens": 14398347.0,
      "step": 7710
    },
    {
      "entropy": 5.966675758361816,
      "epoch": 0.6812963617096433,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960379651081068,
      "loss": 5.6293,
      "mean_token_accuracy": 0.1845902293920517,
      "num_tokens": 14407304.0,
      "step": 7715
    },
    {
      "entropy": 5.903362655639649,
      "epoch": 0.6817379018014835,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004960320791692283,
      "loss": 5.6624,
      "mean_token_accuracy": 0.1791662245988846,
      "num_tokens": 14416641.0,
      "step": 7720
    },
    {
      "entropy": 5.976517868041992,
      "epoch": 0.6821794418933239,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004960261889004572,
      "loss": 5.7595,
      "mean_token_accuracy": 0.17492255866527556,
      "num_tokens": 14426095.0,
      "step": 7725
    },
    {
      "entropy": 5.935105276107788,
      "epoch": 0.6826209819851643,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004960202943019088,
      "loss": 5.6682,
      "mean_token_accuracy": 0.18540456146001816,
      "num_tokens": 14434563.0,
      "step": 7730
    },
    {
      "entropy": 5.794443988800049,
      "epoch": 0.6830625220770046,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960143953736984,
      "loss": 5.598,
      "mean_token_accuracy": 0.18621402233839035,
      "num_tokens": 14444319.0,
      "step": 7735
    },
    {
      "entropy": 5.9014829158782955,
      "epoch": 0.683504062168845,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004960084921159419,
      "loss": 5.6398,
      "mean_token_accuracy": 0.1842109367251396,
      "num_tokens": 14453381.0,
      "step": 7740
    },
    {
      "entropy": 5.819450187683105,
      "epoch": 0.6839456022606852,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004960025845287546,
      "loss": 5.6404,
      "mean_token_accuracy": 0.18381152004003526,
      "num_tokens": 14462660.0,
      "step": 7745
    },
    {
      "entropy": 5.850420904159546,
      "epoch": 0.6843871423525256,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004959966726122523,
      "loss": 5.5589,
      "mean_token_accuracy": 0.1945916920900345,
      "num_tokens": 14471820.0,
      "step": 7750
    },
    {
      "entropy": 5.949693870544434,
      "epoch": 0.6848286824443659,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000495990756366551,
      "loss": 5.7463,
      "mean_token_accuracy": 0.18212738186120986,
      "num_tokens": 14481663.0,
      "step": 7755
    },
    {
      "entropy": 5.947234582901001,
      "epoch": 0.6852702225362063,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004959848357917664,
      "loss": 5.7543,
      "mean_token_accuracy": 0.17507750540971756,
      "num_tokens": 14491515.0,
      "step": 7760
    },
    {
      "entropy": 5.843775367736816,
      "epoch": 0.6857117626280467,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004959789108880145,
      "loss": 5.5817,
      "mean_token_accuracy": 0.1806255117058754,
      "num_tokens": 14500852.0,
      "step": 7765
    },
    {
      "entropy": 5.86192569732666,
      "epoch": 0.686153302719887,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004959729816554115,
      "loss": 5.6308,
      "mean_token_accuracy": 0.19001229852437973,
      "num_tokens": 14509771.0,
      "step": 7770
    },
    {
      "entropy": 5.83790431022644,
      "epoch": 0.6865948428117273,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004959670480940734,
      "loss": 5.6172,
      "mean_token_accuracy": 0.18619886487722398,
      "num_tokens": 14519377.0,
      "step": 7775
    },
    {
      "entropy": 5.864809226989746,
      "epoch": 0.6870363829035676,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004959611102041165,
      "loss": 5.6722,
      "mean_token_accuracy": 0.18079863488674164,
      "num_tokens": 14528998.0,
      "step": 7780
    },
    {
      "entropy": 5.904181909561157,
      "epoch": 0.687477922995408,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004959551679856571,
      "loss": 5.6997,
      "mean_token_accuracy": 0.1849734291434288,
      "num_tokens": 14538153.0,
      "step": 7785
    },
    {
      "entropy": 5.8900541305542,
      "epoch": 0.6879194630872483,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004959492214388117,
      "loss": 5.6349,
      "mean_token_accuracy": 0.18393201678991317,
      "num_tokens": 14547631.0,
      "step": 7790
    },
    {
      "entropy": 5.935344123840332,
      "epoch": 0.6883610031790887,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004959432705636968,
      "loss": 5.5612,
      "mean_token_accuracy": 0.1882770836353302,
      "num_tokens": 14557685.0,
      "step": 7795
    },
    {
      "entropy": 5.903605842590332,
      "epoch": 0.688802543270929,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004959373153604287,
      "loss": 5.7088,
      "mean_token_accuracy": 0.18453071862459183,
      "num_tokens": 14567360.0,
      "step": 7800
    },
    {
      "entropy": 5.804054594039917,
      "epoch": 0.6892440833627693,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004959313558291243,
      "loss": 5.6375,
      "mean_token_accuracy": 0.18478822112083435,
      "num_tokens": 14576916.0,
      "step": 7805
    },
    {
      "entropy": 5.898049783706665,
      "epoch": 0.6896856234546097,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004959253919699004,
      "loss": 5.555,
      "mean_token_accuracy": 0.1911950170993805,
      "num_tokens": 14586716.0,
      "step": 7810
    },
    {
      "entropy": 5.912984180450439,
      "epoch": 0.69012716354645,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004959194237828735,
      "loss": 5.6252,
      "mean_token_accuracy": 0.1866038516163826,
      "num_tokens": 14595690.0,
      "step": 7815
    },
    {
      "entropy": 5.8310645580291744,
      "epoch": 0.6905687036382904,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004959134512681609,
      "loss": 5.7008,
      "mean_token_accuracy": 0.18006411343812942,
      "num_tokens": 14605481.0,
      "step": 7820
    },
    {
      "entropy": 5.893993234634399,
      "epoch": 0.6910102437301306,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004959074744258794,
      "loss": 5.7398,
      "mean_token_accuracy": 0.17599762678146363,
      "num_tokens": 14615242.0,
      "step": 7825
    },
    {
      "entropy": 5.87747049331665,
      "epoch": 0.691451783821971,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004959014932561461,
      "loss": 5.5771,
      "mean_token_accuracy": 0.19491011202335357,
      "num_tokens": 14624343.0,
      "step": 7830
    },
    {
      "entropy": 5.802929401397705,
      "epoch": 0.6918933239138114,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004958955077590782,
      "loss": 5.5422,
      "mean_token_accuracy": 0.19644290953874588,
      "num_tokens": 14633014.0,
      "step": 7835
    },
    {
      "entropy": 5.872480010986328,
      "epoch": 0.6923348640056517,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004958895179347929,
      "loss": 5.6687,
      "mean_token_accuracy": 0.18705726861953736,
      "num_tokens": 14642796.0,
      "step": 7840
    },
    {
      "entropy": 5.943118047714234,
      "epoch": 0.6927764040974921,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004958835237834075,
      "loss": 5.716,
      "mean_token_accuracy": 0.174401131272316,
      "num_tokens": 14652388.0,
      "step": 7845
    },
    {
      "entropy": 5.9830615520477295,
      "epoch": 0.6932179441893324,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004958775253050396,
      "loss": 5.6614,
      "mean_token_accuracy": 0.18240052610635757,
      "num_tokens": 14661219.0,
      "step": 7850
    },
    {
      "entropy": 5.8752015113830565,
      "epoch": 0.6936594842811727,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004958715224998066,
      "loss": 5.6783,
      "mean_token_accuracy": 0.18555933982133865,
      "num_tokens": 14670878.0,
      "step": 7855
    },
    {
      "entropy": 5.82140998840332,
      "epoch": 0.694101024373013,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004958655153678261,
      "loss": 5.7262,
      "mean_token_accuracy": 0.17967288792133332,
      "num_tokens": 14680893.0,
      "step": 7860
    },
    {
      "entropy": 5.9207089900970455,
      "epoch": 0.6945425644648534,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004958595039092156,
      "loss": 5.6324,
      "mean_token_accuracy": 0.18184648156166078,
      "num_tokens": 14689249.0,
      "step": 7865
    },
    {
      "entropy": 5.896468544006348,
      "epoch": 0.6949841045566938,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004958534881240932,
      "loss": 5.6245,
      "mean_token_accuracy": 0.1842005655169487,
      "num_tokens": 14698724.0,
      "step": 7870
    },
    {
      "entropy": 5.853013896942139,
      "epoch": 0.6954256446485341,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004958474680125766,
      "loss": 5.5998,
      "mean_token_accuracy": 0.181177119910717,
      "num_tokens": 14707756.0,
      "step": 7875
    },
    {
      "entropy": 5.86539044380188,
      "epoch": 0.6958671847403745,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004958414435747837,
      "loss": 5.616,
      "mean_token_accuracy": 0.18928422033786774,
      "num_tokens": 14717416.0,
      "step": 7880
    },
    {
      "entropy": 5.847776317596436,
      "epoch": 0.6963087248322147,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004958354148108324,
      "loss": 5.6802,
      "mean_token_accuracy": 0.182321697473526,
      "num_tokens": 14727321.0,
      "step": 7885
    },
    {
      "entropy": 5.7946008205413815,
      "epoch": 0.6967502649240551,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000495829381720841,
      "loss": 5.5477,
      "mean_token_accuracy": 0.19297181218862533,
      "num_tokens": 14736298.0,
      "step": 7890
    },
    {
      "entropy": 5.8987609386444095,
      "epoch": 0.6971918050158954,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004958233443049275,
      "loss": 5.5849,
      "mean_token_accuracy": 0.18445069193840027,
      "num_tokens": 14745893.0,
      "step": 7895
    },
    {
      "entropy": 5.892021942138672,
      "epoch": 0.6976333451077358,
      "grad_norm": 1.0,
      "learning_rate": 0.0004958173025632103,
      "loss": 5.664,
      "mean_token_accuracy": 0.17899592816829682,
      "num_tokens": 14755277.0,
      "step": 7900
    },
    {
      "entropy": 5.957773065567016,
      "epoch": 0.6980748851995762,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004958112564958079,
      "loss": 5.7665,
      "mean_token_accuracy": 0.17907627373933793,
      "num_tokens": 14764369.0,
      "step": 7905
    },
    {
      "entropy": 5.851858806610108,
      "epoch": 0.6985164252914164,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004958052061028385,
      "loss": 5.6058,
      "mean_token_accuracy": 0.1951722025871277,
      "num_tokens": 14772970.0,
      "step": 7910
    },
    {
      "entropy": 5.824829816818237,
      "epoch": 0.6989579653832568,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004957991513844205,
      "loss": 5.6783,
      "mean_token_accuracy": 0.183780537545681,
      "num_tokens": 14782862.0,
      "step": 7915
    },
    {
      "entropy": 5.947744512557984,
      "epoch": 0.6993995054750971,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004957930923406729,
      "loss": 5.6932,
      "mean_token_accuracy": 0.1842297852039337,
      "num_tokens": 14792042.0,
      "step": 7920
    },
    {
      "entropy": 5.92308783531189,
      "epoch": 0.6998410455669375,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004957870289717142,
      "loss": 5.6239,
      "mean_token_accuracy": 0.18697006553411483,
      "num_tokens": 14801032.0,
      "step": 7925
    },
    {
      "entropy": 5.812475728988647,
      "epoch": 0.7002825856587778,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004957809612776631,
      "loss": 5.6368,
      "mean_token_accuracy": 0.18773610293865203,
      "num_tokens": 14810594.0,
      "step": 7930
    },
    {
      "entropy": 5.906245374679566,
      "epoch": 0.7007241257506182,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004957748892586384,
      "loss": 5.7392,
      "mean_token_accuracy": 0.18038935363292694,
      "num_tokens": 14820570.0,
      "step": 7935
    },
    {
      "entropy": 5.768427610397339,
      "epoch": 0.7011656658424585,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004957688129147594,
      "loss": 5.5486,
      "mean_token_accuracy": 0.1953514665365219,
      "num_tokens": 14830078.0,
      "step": 7940
    },
    {
      "entropy": 5.886784267425537,
      "epoch": 0.7016072059342988,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004957627322461448,
      "loss": 5.6258,
      "mean_token_accuracy": 0.1836659014225006,
      "num_tokens": 14839974.0,
      "step": 7945
    },
    {
      "entropy": 5.916618680953979,
      "epoch": 0.7020487460261392,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004957566472529139,
      "loss": 5.7203,
      "mean_token_accuracy": 0.17270487248897554,
      "num_tokens": 14849551.0,
      "step": 7950
    },
    {
      "entropy": 5.903427076339722,
      "epoch": 0.7024902861179795,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004957505579351858,
      "loss": 5.6697,
      "mean_token_accuracy": 0.17774245142936707,
      "num_tokens": 14859011.0,
      "step": 7955
    },
    {
      "entropy": 5.865008449554443,
      "epoch": 0.7029318262098199,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004957444642930798,
      "loss": 5.6071,
      "mean_token_accuracy": 0.18349462747573853,
      "num_tokens": 14867948.0,
      "step": 7960
    },
    {
      "entropy": 5.898176765441894,
      "epoch": 0.7033733663016601,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004957383663267152,
      "loss": 5.656,
      "mean_token_accuracy": 0.18263181746006013,
      "num_tokens": 14877233.0,
      "step": 7965
    },
    {
      "entropy": 5.915711975097656,
      "epoch": 0.7038149063935005,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004957322640362118,
      "loss": 5.6758,
      "mean_token_accuracy": 0.1814686819911003,
      "num_tokens": 14887439.0,
      "step": 7970
    },
    {
      "entropy": 5.960000324249267,
      "epoch": 0.7042564464853409,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004957261574216887,
      "loss": 5.7501,
      "mean_token_accuracy": 0.17440729439258576,
      "num_tokens": 14897064.0,
      "step": 7975
    },
    {
      "entropy": 5.924387741088867,
      "epoch": 0.7046979865771812,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004957200464832656,
      "loss": 5.7028,
      "mean_token_accuracy": 0.17842092514038085,
      "num_tokens": 14907054.0,
      "step": 7980
    },
    {
      "entropy": 5.930465459823608,
      "epoch": 0.7051395266690216,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004957139312210626,
      "loss": 5.6654,
      "mean_token_accuracy": 0.18382232934236525,
      "num_tokens": 14916969.0,
      "step": 7985
    },
    {
      "entropy": 5.894822502136231,
      "epoch": 0.7055810667608619,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000495707811635199,
      "loss": 5.6861,
      "mean_token_accuracy": 0.18471913486719133,
      "num_tokens": 14926404.0,
      "step": 7990
    },
    {
      "entropy": 5.876964044570923,
      "epoch": 0.7060226068527022,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004957016877257949,
      "loss": 5.6544,
      "mean_token_accuracy": 0.182310651242733,
      "num_tokens": 14935693.0,
      "step": 7995
    },
    {
      "entropy": 5.899308919906616,
      "epoch": 0.7064641469445425,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004956955594929704,
      "loss": 5.596,
      "mean_token_accuracy": 0.184686329215765,
      "num_tokens": 14945147.0,
      "step": 8000
    },
    {
      "entropy": 5.859074115753174,
      "epoch": 0.7069056870363829,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004956894269368454,
      "loss": 5.6438,
      "mean_token_accuracy": 0.18151623457670213,
      "num_tokens": 14954022.0,
      "step": 8005
    },
    {
      "entropy": 5.803120088577271,
      "epoch": 0.7073472271282233,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004956832900575399,
      "loss": 5.6509,
      "mean_token_accuracy": 0.18272694200277328,
      "num_tokens": 14962977.0,
      "step": 8010
    },
    {
      "entropy": 5.9160982131958,
      "epoch": 0.7077887672200636,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004956771488551742,
      "loss": 5.6914,
      "mean_token_accuracy": 0.1767708733677864,
      "num_tokens": 14973420.0,
      "step": 8015
    },
    {
      "entropy": 5.975215291976928,
      "epoch": 0.708230307311904,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004956710033298689,
      "loss": 5.6465,
      "mean_token_accuracy": 0.1861080601811409,
      "num_tokens": 14981762.0,
      "step": 8020
    },
    {
      "entropy": 5.950171566009521,
      "epoch": 0.7086718474037442,
      "grad_norm": 0.9375,
      "learning_rate": 0.000495664853481744,
      "loss": 5.7171,
      "mean_token_accuracy": 0.17843022793531418,
      "num_tokens": 14991449.0,
      "step": 8025
    },
    {
      "entropy": 5.917905616760254,
      "epoch": 0.7091133874955846,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004956586993109201,
      "loss": 5.6514,
      "mean_token_accuracy": 0.18210796564817427,
      "num_tokens": 15000802.0,
      "step": 8030
    },
    {
      "entropy": 5.865891695022583,
      "epoch": 0.7095549275874249,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004956525408175179,
      "loss": 5.6694,
      "mean_token_accuracy": 0.18268949687480926,
      "num_tokens": 15010580.0,
      "step": 8035
    },
    {
      "entropy": 5.844978952407837,
      "epoch": 0.7099964676792653,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004956463780016578,
      "loss": 5.7012,
      "mean_token_accuracy": 0.18287778943777083,
      "num_tokens": 15020059.0,
      "step": 8040
    },
    {
      "entropy": 5.900697183609009,
      "epoch": 0.7104380077711057,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004956402108634607,
      "loss": 5.6449,
      "mean_token_accuracy": 0.1812546208500862,
      "num_tokens": 15030268.0,
      "step": 8045
    },
    {
      "entropy": 5.875079154968262,
      "epoch": 0.7108795478629459,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004956340394030473,
      "loss": 5.6166,
      "mean_token_accuracy": 0.18594399243593215,
      "num_tokens": 15038865.0,
      "step": 8050
    },
    {
      "entropy": 5.94498291015625,
      "epoch": 0.7113210879547863,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004956278636205386,
      "loss": 5.7501,
      "mean_token_accuracy": 0.17342514544725418,
      "num_tokens": 15048427.0,
      "step": 8055
    },
    {
      "entropy": 5.866507053375244,
      "epoch": 0.7117626280466266,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004956216835160556,
      "loss": 5.6176,
      "mean_token_accuracy": 0.1918194204568863,
      "num_tokens": 15057366.0,
      "step": 8060
    },
    {
      "entropy": 5.917425632476807,
      "epoch": 0.712204168138467,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004956154990897191,
      "loss": 5.6597,
      "mean_token_accuracy": 0.18279267549514772,
      "num_tokens": 15067098.0,
      "step": 8065
    },
    {
      "entropy": 5.884406900405883,
      "epoch": 0.7126457082303073,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004956093103416505,
      "loss": 5.6655,
      "mean_token_accuracy": 0.18610380887985228,
      "num_tokens": 15075734.0,
      "step": 8070
    },
    {
      "entropy": 5.898903656005859,
      "epoch": 0.7130872483221476,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004956031172719711,
      "loss": 5.7048,
      "mean_token_accuracy": 0.1823004812002182,
      "num_tokens": 15084625.0,
      "step": 8075
    },
    {
      "entropy": 5.792678642272949,
      "epoch": 0.713528788413988,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004955969198808021,
      "loss": 5.5319,
      "mean_token_accuracy": 0.19952139854431153,
      "num_tokens": 15093256.0,
      "step": 8080
    },
    {
      "entropy": 5.816761541366577,
      "epoch": 0.7139703285058283,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004955907181682648,
      "loss": 5.6,
      "mean_token_accuracy": 0.191490775346756,
      "num_tokens": 15103175.0,
      "step": 8085
    },
    {
      "entropy": 5.834416055679322,
      "epoch": 0.7144118685976687,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004955845121344809,
      "loss": 5.6853,
      "mean_token_accuracy": 0.1825338825583458,
      "num_tokens": 15112789.0,
      "step": 8090
    },
    {
      "entropy": 5.914598226547241,
      "epoch": 0.714853408689509,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004955783017795717,
      "loss": 5.7234,
      "mean_token_accuracy": 0.18739004880189897,
      "num_tokens": 15121869.0,
      "step": 8095
    },
    {
      "entropy": 5.868374967575074,
      "epoch": 0.7152949487813494,
      "grad_norm": 0.875,
      "learning_rate": 0.000495572087103659,
      "loss": 5.6229,
      "mean_token_accuracy": 0.18730900734663009,
      "num_tokens": 15131441.0,
      "step": 8100
    },
    {
      "entropy": 5.806985235214233,
      "epoch": 0.7157364888731896,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004955658681068647,
      "loss": 5.6615,
      "mean_token_accuracy": 0.18200474381446838,
      "num_tokens": 15140641.0,
      "step": 8105
    },
    {
      "entropy": 5.958062267303466,
      "epoch": 0.71617802896503,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004955596447893105,
      "loss": 5.6552,
      "mean_token_accuracy": 0.18732800036668779,
      "num_tokens": 15150073.0,
      "step": 8110
    },
    {
      "entropy": 5.803730249404907,
      "epoch": 0.7166195690568704,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004955534171511181,
      "loss": 5.3936,
      "mean_token_accuracy": 0.20334044098854065,
      "num_tokens": 15159879.0,
      "step": 8115
    },
    {
      "entropy": 5.778502416610718,
      "epoch": 0.7170611091487107,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00049554718519241,
      "loss": 5.6052,
      "mean_token_accuracy": 0.19128098487854003,
      "num_tokens": 15168316.0,
      "step": 8120
    },
    {
      "entropy": 5.765700244903565,
      "epoch": 0.7175026492405511,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004955409489133078,
      "loss": 5.6519,
      "mean_token_accuracy": 0.1813172772526741,
      "num_tokens": 15176700.0,
      "step": 8125
    },
    {
      "entropy": 5.927375173568725,
      "epoch": 0.7179441893323913,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004955347083139338,
      "loss": 5.681,
      "mean_token_accuracy": 0.18737523555755614,
      "num_tokens": 15185997.0,
      "step": 8130
    },
    {
      "entropy": 5.912154483795166,
      "epoch": 0.7183857294242317,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004955284633944104,
      "loss": 5.6616,
      "mean_token_accuracy": 0.18576339930295943,
      "num_tokens": 15195341.0,
      "step": 8135
    },
    {
      "entropy": 5.85348105430603,
      "epoch": 0.718827269516072,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004955222141548597,
      "loss": 5.6797,
      "mean_token_accuracy": 0.18169899731874467,
      "num_tokens": 15205501.0,
      "step": 8140
    },
    {
      "entropy": 5.862333488464356,
      "epoch": 0.7192688096079124,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004955159605954043,
      "loss": 5.6344,
      "mean_token_accuracy": 0.18649196475744248,
      "num_tokens": 15214778.0,
      "step": 8145
    },
    {
      "entropy": 5.957325029373169,
      "epoch": 0.7197103496997528,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004955097027161666,
      "loss": 5.6663,
      "mean_token_accuracy": 0.18574378043413162,
      "num_tokens": 15224201.0,
      "step": 8150
    },
    {
      "entropy": 5.847279453277588,
      "epoch": 0.7201518897915931,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004955034405172693,
      "loss": 5.6008,
      "mean_token_accuracy": 0.18820102363824845,
      "num_tokens": 15233472.0,
      "step": 8155
    },
    {
      "entropy": 5.881206130981445,
      "epoch": 0.7205934298834334,
      "grad_norm": 1.015625,
      "learning_rate": 0.000495497173998835,
      "loss": 5.6868,
      "mean_token_accuracy": 0.17715079635381697,
      "num_tokens": 15242323.0,
      "step": 8160
    },
    {
      "entropy": 5.893172788619995,
      "epoch": 0.7210349699752737,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004954909031609864,
      "loss": 5.6699,
      "mean_token_accuracy": 0.18143803179264067,
      "num_tokens": 15252952.0,
      "step": 8165
    },
    {
      "entropy": 5.937598085403442,
      "epoch": 0.7214765100671141,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004954846280038465,
      "loss": 5.6777,
      "mean_token_accuracy": 0.18283233791589737,
      "num_tokens": 15261815.0,
      "step": 8170
    },
    {
      "entropy": 5.916687059402466,
      "epoch": 0.7219180501589544,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004954783485275381,
      "loss": 5.6998,
      "mean_token_accuracy": 0.177165387570858,
      "num_tokens": 15271026.0,
      "step": 8175
    },
    {
      "entropy": 5.846047449111938,
      "epoch": 0.7223595902507948,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004954720647321843,
      "loss": 5.5364,
      "mean_token_accuracy": 0.19162006825208663,
      "num_tokens": 15279795.0,
      "step": 8180
    },
    {
      "entropy": 5.923773765563965,
      "epoch": 0.7228011303426352,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004954657766179081,
      "loss": 5.6822,
      "mean_token_accuracy": 0.1771241009235382,
      "num_tokens": 15289113.0,
      "step": 8185
    },
    {
      "entropy": 5.846442317962646,
      "epoch": 0.7232426704344754,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004954594841848327,
      "loss": 5.5244,
      "mean_token_accuracy": 0.19908362179994582,
      "num_tokens": 15298271.0,
      "step": 8190
    },
    {
      "entropy": 5.895449018478393,
      "epoch": 0.7236842105263158,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004954531874330814,
      "loss": 5.6342,
      "mean_token_accuracy": 0.18278907835483552,
      "num_tokens": 15307677.0,
      "step": 8195
    },
    {
      "entropy": 5.8396324634552,
      "epoch": 0.7241257506181561,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004954468863627774,
      "loss": 5.6438,
      "mean_token_accuracy": 0.18307971209287643,
      "num_tokens": 15316637.0,
      "step": 8200
    },
    {
      "entropy": 5.833258199691772,
      "epoch": 0.7245672907099965,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004954405809740445,
      "loss": 5.63,
      "mean_token_accuracy": 0.17943803071975709,
      "num_tokens": 15325221.0,
      "step": 8205
    },
    {
      "entropy": 5.881443452835083,
      "epoch": 0.7250088308018368,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004954342712670058,
      "loss": 5.6632,
      "mean_token_accuracy": 0.1839708626270294,
      "num_tokens": 15335877.0,
      "step": 8210
    },
    {
      "entropy": 5.872735929489136,
      "epoch": 0.7254503708936771,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004954279572417852,
      "loss": 5.6677,
      "mean_token_accuracy": 0.18076927959918976,
      "num_tokens": 15345096.0,
      "step": 8215
    },
    {
      "entropy": 5.841041040420532,
      "epoch": 0.7258919109855175,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004954216388985063,
      "loss": 5.5006,
      "mean_token_accuracy": 0.19473296403884888,
      "num_tokens": 15353785.0,
      "step": 8220
    },
    {
      "entropy": 5.870520782470703,
      "epoch": 0.7263334510773578,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004954153162372928,
      "loss": 5.6024,
      "mean_token_accuracy": 0.1930699646472931,
      "num_tokens": 15362761.0,
      "step": 8225
    },
    {
      "entropy": 5.888486576080322,
      "epoch": 0.7267749911691982,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004954089892582686,
      "loss": 5.6081,
      "mean_token_accuracy": 0.18426867425441742,
      "num_tokens": 15371951.0,
      "step": 8230
    },
    {
      "entropy": 5.898847579956055,
      "epoch": 0.7272165312610385,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004954026579615577,
      "loss": 5.6633,
      "mean_token_accuracy": 0.18143970519304276,
      "num_tokens": 15380277.0,
      "step": 8235
    },
    {
      "entropy": 5.8413866519927975,
      "epoch": 0.7276580713528789,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004953963223472841,
      "loss": 5.6742,
      "mean_token_accuracy": 0.18348696678876877,
      "num_tokens": 15390951.0,
      "step": 8240
    },
    {
      "entropy": 5.908440494537354,
      "epoch": 0.7280996114447191,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004953899824155718,
      "loss": 5.6227,
      "mean_token_accuracy": 0.18988921642303466,
      "num_tokens": 15399394.0,
      "step": 8245
    },
    {
      "entropy": 5.865427255630493,
      "epoch": 0.7285411515365595,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004953836381665452,
      "loss": 5.5761,
      "mean_token_accuracy": 0.18975492119789122,
      "num_tokens": 15408644.0,
      "step": 8250
    },
    {
      "entropy": 5.946927690505982,
      "epoch": 0.7289826916283999,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004953772896003284,
      "loss": 5.6464,
      "mean_token_accuracy": 0.18333052694797516,
      "num_tokens": 15417398.0,
      "step": 8255
    },
    {
      "entropy": 5.93438024520874,
      "epoch": 0.7294242317202402,
      "grad_norm": 1.0,
      "learning_rate": 0.0004953709367170458,
      "loss": 5.717,
      "mean_token_accuracy": 0.1741793468594551,
      "num_tokens": 15426532.0,
      "step": 8260
    },
    {
      "entropy": 5.8816978931427,
      "epoch": 0.7298657718120806,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000495364579516822,
      "loss": 5.6952,
      "mean_token_accuracy": 0.17792603224515915,
      "num_tokens": 15436371.0,
      "step": 8265
    },
    {
      "entropy": 5.964777851104737,
      "epoch": 0.7303073119039208,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004953582179997813,
      "loss": 5.7434,
      "mean_token_accuracy": 0.18140242248773575,
      "num_tokens": 15446600.0,
      "step": 8270
    },
    {
      "entropy": 5.91185245513916,
      "epoch": 0.7307488519957612,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004953518521660484,
      "loss": 5.5995,
      "mean_token_accuracy": 0.1830684170126915,
      "num_tokens": 15455547.0,
      "step": 8275
    },
    {
      "entropy": 5.856135416030884,
      "epoch": 0.7311903920876015,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004953454820157481,
      "loss": 5.6275,
      "mean_token_accuracy": 0.18180723786354064,
      "num_tokens": 15464543.0,
      "step": 8280
    },
    {
      "entropy": 5.958267164230347,
      "epoch": 0.7316319321794419,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004953391075490051,
      "loss": 5.7714,
      "mean_token_accuracy": 0.16960600912570953,
      "num_tokens": 15473626.0,
      "step": 8285
    },
    {
      "entropy": 5.923289155960083,
      "epoch": 0.7320734722712823,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004953327287659444,
      "loss": 5.689,
      "mean_token_accuracy": 0.17944410294294358,
      "num_tokens": 15484164.0,
      "step": 8290
    },
    {
      "entropy": 6.006247711181641,
      "epoch": 0.7325150123631226,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004953263456666907,
      "loss": 5.695,
      "mean_token_accuracy": 0.17470264732837676,
      "num_tokens": 15493718.0,
      "step": 8295
    },
    {
      "entropy": 5.919910049438476,
      "epoch": 0.7329565524549629,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004953199582513693,
      "loss": 5.6388,
      "mean_token_accuracy": 0.18529484421014786,
      "num_tokens": 15503339.0,
      "step": 8300
    },
    {
      "entropy": 5.870715141296387,
      "epoch": 0.7333980925468032,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004953135665201052,
      "loss": 5.6187,
      "mean_token_accuracy": 0.18870222121477126,
      "num_tokens": 15513314.0,
      "step": 8305
    },
    {
      "entropy": 5.898948812484742,
      "epoch": 0.7338396326386436,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004953071704730237,
      "loss": 5.6969,
      "mean_token_accuracy": 0.17855761349201202,
      "num_tokens": 15522943.0,
      "step": 8310
    },
    {
      "entropy": 5.9014928340911865,
      "epoch": 0.7342811727304839,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004953007701102499,
      "loss": 5.6605,
      "mean_token_accuracy": 0.18277857601642608,
      "num_tokens": 15531975.0,
      "step": 8315
    },
    {
      "entropy": 5.902299976348877,
      "epoch": 0.7347227128223243,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004952943654319094,
      "loss": 5.5941,
      "mean_token_accuracy": 0.1843016892671585,
      "num_tokens": 15541139.0,
      "step": 8320
    },
    {
      "entropy": 5.9181843280792235,
      "epoch": 0.7351642529141647,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004952879564381276,
      "loss": 5.6706,
      "mean_token_accuracy": 0.1839447870850563,
      "num_tokens": 15550604.0,
      "step": 8325
    },
    {
      "entropy": 5.814180707931518,
      "epoch": 0.7356057930060049,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004952815431290298,
      "loss": 5.5866,
      "mean_token_accuracy": 0.18736102283000947,
      "num_tokens": 15558995.0,
      "step": 8330
    },
    {
      "entropy": 5.975278282165528,
      "epoch": 0.7360473330978453,
      "grad_norm": 1.03125,
      "learning_rate": 0.000495275125504742,
      "loss": 5.6574,
      "mean_token_accuracy": 0.18505442291498184,
      "num_tokens": 15568229.0,
      "step": 8335
    },
    {
      "entropy": 5.837584114074707,
      "epoch": 0.7364888731896856,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004952687035653899,
      "loss": 5.6092,
      "mean_token_accuracy": 0.18216995596885682,
      "num_tokens": 15577042.0,
      "step": 8340
    },
    {
      "entropy": 5.812479686737061,
      "epoch": 0.736930413281526,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004952622773110989,
      "loss": 5.52,
      "mean_token_accuracy": 0.19470138996839523,
      "num_tokens": 15585825.0,
      "step": 8345
    },
    {
      "entropy": 5.863739252090454,
      "epoch": 0.7373719533733663,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004952558467419953,
      "loss": 5.5351,
      "mean_token_accuracy": 0.19085845798254014,
      "num_tokens": 15595486.0,
      "step": 8350
    },
    {
      "entropy": 5.866566705703735,
      "epoch": 0.7378134934652066,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000495249411858205,
      "loss": 5.6003,
      "mean_token_accuracy": 0.17631158828735352,
      "num_tokens": 15605312.0,
      "step": 8355
    },
    {
      "entropy": 5.840285158157348,
      "epoch": 0.738255033557047,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004952429726598538,
      "loss": 5.5771,
      "mean_token_accuracy": 0.19373610317707063,
      "num_tokens": 15615185.0,
      "step": 8360
    },
    {
      "entropy": 5.883241939544678,
      "epoch": 0.7386965736488873,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004952365291470682,
      "loss": 5.6903,
      "mean_token_accuracy": 0.18867729008197784,
      "num_tokens": 15625259.0,
      "step": 8365
    },
    {
      "entropy": 5.9134824752807615,
      "epoch": 0.7391381137407277,
      "grad_norm": 0.875,
      "learning_rate": 0.0004952300813199742,
      "loss": 5.6882,
      "mean_token_accuracy": 0.1814059793949127,
      "num_tokens": 15635040.0,
      "step": 8370
    },
    {
      "entropy": 5.942975568771362,
      "epoch": 0.739579653832568,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004952236291786981,
      "loss": 5.6992,
      "mean_token_accuracy": 0.18363112807273865,
      "num_tokens": 15644996.0,
      "step": 8375
    },
    {
      "entropy": 5.800611877441407,
      "epoch": 0.7400211939244083,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004952171727233665,
      "loss": 5.5721,
      "mean_token_accuracy": 0.1871659651398659,
      "num_tokens": 15654598.0,
      "step": 8380
    },
    {
      "entropy": 5.870110893249512,
      "epoch": 0.7404627340162486,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004952107119541055,
      "loss": 5.5809,
      "mean_token_accuracy": 0.1836706057190895,
      "num_tokens": 15664213.0,
      "step": 8385
    },
    {
      "entropy": 5.889188480377197,
      "epoch": 0.740904274108089,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004952042468710421,
      "loss": 5.6605,
      "mean_token_accuracy": 0.1854677751660347,
      "num_tokens": 15673595.0,
      "step": 8390
    },
    {
      "entropy": 5.942821264266968,
      "epoch": 0.7413458141999294,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004951977774743027,
      "loss": 5.7411,
      "mean_token_accuracy": 0.17860967963933944,
      "num_tokens": 15682753.0,
      "step": 8395
    },
    {
      "entropy": 5.832130432128906,
      "epoch": 0.7417873542917697,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004951913037640139,
      "loss": 5.5636,
      "mean_token_accuracy": 0.1966960161924362,
      "num_tokens": 15691508.0,
      "step": 8400
    },
    {
      "entropy": 5.7541731834411625,
      "epoch": 0.7422288943836101,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004951848257403029,
      "loss": 5.531,
      "mean_token_accuracy": 0.18669057488441468,
      "num_tokens": 15701110.0,
      "step": 8405
    },
    {
      "entropy": 5.792100095748902,
      "epoch": 0.7426704344754503,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004951783434032963,
      "loss": 5.4955,
      "mean_token_accuracy": 0.1957699477672577,
      "num_tokens": 15710565.0,
      "step": 8410
    },
    {
      "entropy": 5.883226823806763,
      "epoch": 0.7431119745672907,
      "grad_norm": 1.125,
      "learning_rate": 0.0004951718567531211,
      "loss": 5.6617,
      "mean_token_accuracy": 0.1794003278017044,
      "num_tokens": 15718364.0,
      "step": 8415
    },
    {
      "entropy": 5.934400129318237,
      "epoch": 0.743553514659131,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004951653657899045,
      "loss": 5.7172,
      "mean_token_accuracy": 0.17798707634210587,
      "num_tokens": 15729127.0,
      "step": 8420
    },
    {
      "entropy": 5.944851684570312,
      "epoch": 0.7439950547509714,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004951588705137737,
      "loss": 5.6884,
      "mean_token_accuracy": 0.17930149734020234,
      "num_tokens": 15738491.0,
      "step": 8425
    },
    {
      "entropy": 5.883695507049561,
      "epoch": 0.7444365948428118,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004951523709248557,
      "loss": 5.5617,
      "mean_token_accuracy": 0.19158599972724916,
      "num_tokens": 15748346.0,
      "step": 8430
    },
    {
      "entropy": 5.836772108078003,
      "epoch": 0.744878134934652,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000495145867023278,
      "loss": 5.5833,
      "mean_token_accuracy": 0.18637182414531708,
      "num_tokens": 15757242.0,
      "step": 8435
    },
    {
      "entropy": 5.957921314239502,
      "epoch": 0.7453196750264924,
      "grad_norm": 0.83203125,
      "learning_rate": 0.000495139358809168,
      "loss": 5.6306,
      "mean_token_accuracy": 0.18897896707057954,
      "num_tokens": 15767236.0,
      "step": 8440
    },
    {
      "entropy": 5.909159517288208,
      "epoch": 0.7457612151183327,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004951328462826532,
      "loss": 5.7068,
      "mean_token_accuracy": 0.17895558327436448,
      "num_tokens": 15776977.0,
      "step": 8445
    },
    {
      "entropy": 5.813918399810791,
      "epoch": 0.7462027552101731,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004951263294438611,
      "loss": 5.5944,
      "mean_token_accuracy": 0.17929491102695466,
      "num_tokens": 15787333.0,
      "step": 8450
    },
    {
      "entropy": 5.803722190856933,
      "epoch": 0.7466442953020134,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004951198082929194,
      "loss": 5.4828,
      "mean_token_accuracy": 0.2009884908795357,
      "num_tokens": 15794773.0,
      "step": 8455
    },
    {
      "entropy": 5.792421913146972,
      "epoch": 0.7470858353938538,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004951132828299558,
      "loss": 5.5766,
      "mean_token_accuracy": 0.19700139313936232,
      "num_tokens": 15803803.0,
      "step": 8460
    },
    {
      "entropy": 5.856327486038208,
      "epoch": 0.7475273754856941,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004951067530550982,
      "loss": 5.6327,
      "mean_token_accuracy": 0.18049795925617218,
      "num_tokens": 15814021.0,
      "step": 8465
    },
    {
      "entropy": 5.89320011138916,
      "epoch": 0.7479689155775344,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004951002189684745,
      "loss": 5.6179,
      "mean_token_accuracy": 0.17834028005599975,
      "num_tokens": 15822974.0,
      "step": 8470
    },
    {
      "entropy": 5.7720318794250485,
      "epoch": 0.7484104556693748,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004950936805702127,
      "loss": 5.4998,
      "mean_token_accuracy": 0.19415683150291443,
      "num_tokens": 15832202.0,
      "step": 8475
    },
    {
      "entropy": 5.834465599060058,
      "epoch": 0.7488519957612151,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004950871378604409,
      "loss": 5.6309,
      "mean_token_accuracy": 0.19115645736455916,
      "num_tokens": 15841223.0,
      "step": 8480
    },
    {
      "entropy": 5.873974132537842,
      "epoch": 0.7492935358530555,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004950805908392872,
      "loss": 5.5861,
      "mean_token_accuracy": 0.1882975786924362,
      "num_tokens": 15850260.0,
      "step": 8485
    },
    {
      "entropy": 5.867472314834595,
      "epoch": 0.7497350759448957,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004950740395068799,
      "loss": 5.6677,
      "mean_token_accuracy": 0.1869749143719673,
      "num_tokens": 15859780.0,
      "step": 8490
    },
    {
      "entropy": 5.826950740814209,
      "epoch": 0.7501766160367361,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004950674838633473,
      "loss": 5.5941,
      "mean_token_accuracy": 0.18690303415060044,
      "num_tokens": 15868331.0,
      "step": 8495
    },
    {
      "entropy": 5.867539358139038,
      "epoch": 0.7506181561285765,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004950609239088178,
      "loss": 5.6468,
      "mean_token_accuracy": 0.17849708199501038,
      "num_tokens": 15877359.0,
      "step": 8500
    },
    {
      "entropy": 5.786456060409546,
      "epoch": 0.7510596962204168,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004950543596434201,
      "loss": 5.555,
      "mean_token_accuracy": 0.19861937165260315,
      "num_tokens": 15886976.0,
      "step": 8505
    },
    {
      "entropy": 5.780098247528076,
      "epoch": 0.7515012363122572,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004950477910672825,
      "loss": 5.6306,
      "mean_token_accuracy": 0.18838491588830947,
      "num_tokens": 15896139.0,
      "step": 8510
    },
    {
      "entropy": 5.891390085220337,
      "epoch": 0.7519427764040975,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000495041218180534,
      "loss": 5.6435,
      "mean_token_accuracy": 0.18974340856075286,
      "num_tokens": 15904672.0,
      "step": 8515
    },
    {
      "entropy": 5.787608051300049,
      "epoch": 0.7523843164959378,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000495034640983303,
      "loss": 5.478,
      "mean_token_accuracy": 0.19494728446006776,
      "num_tokens": 15912716.0,
      "step": 8520
    },
    {
      "entropy": 5.833557415008545,
      "epoch": 0.7528258565877781,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004950280594757186,
      "loss": 5.5967,
      "mean_token_accuracy": 0.18717412501573563,
      "num_tokens": 15921974.0,
      "step": 8525
    },
    {
      "entropy": 5.920584440231323,
      "epoch": 0.7532673966796185,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004950214736579097,
      "loss": 5.6112,
      "mean_token_accuracy": 0.18524771481752395,
      "num_tokens": 15930859.0,
      "step": 8530
    },
    {
      "entropy": 5.784576272964477,
      "epoch": 0.7537089367714589,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004950148835300051,
      "loss": 5.5324,
      "mean_token_accuracy": 0.1943311259150505,
      "num_tokens": 15940052.0,
      "step": 8535
    },
    {
      "entropy": 5.823805046081543,
      "epoch": 0.7541504768632992,
      "grad_norm": 1.0,
      "learning_rate": 0.0004950082890921341,
      "loss": 5.6219,
      "mean_token_accuracy": 0.18529213070869446,
      "num_tokens": 15948422.0,
      "step": 8540
    },
    {
      "entropy": 5.894643259048462,
      "epoch": 0.7545920169551396,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004950016903444261,
      "loss": 5.6906,
      "mean_token_accuracy": 0.17810325771570207,
      "num_tokens": 15957433.0,
      "step": 8545
    },
    {
      "entropy": 5.915259504318238,
      "epoch": 0.7550335570469798,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004949950872870099,
      "loss": 5.6259,
      "mean_token_accuracy": 0.18364077657461167,
      "num_tokens": 15966542.0,
      "step": 8550
    },
    {
      "entropy": 5.895453834533692,
      "epoch": 0.7554750971388202,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000494988479920015,
      "loss": 5.7576,
      "mean_token_accuracy": 0.17443610578775406,
      "num_tokens": 15977067.0,
      "step": 8555
    },
    {
      "entropy": 5.903998708724975,
      "epoch": 0.7559166372306605,
      "grad_norm": 0.875,
      "learning_rate": 0.000494981868243571,
      "loss": 5.6459,
      "mean_token_accuracy": 0.18578411638736725,
      "num_tokens": 15986502.0,
      "step": 8560
    },
    {
      "entropy": 5.942930603027344,
      "epoch": 0.7563581773225009,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004949752522578073,
      "loss": 5.7161,
      "mean_token_accuracy": 0.18248875737190245,
      "num_tokens": 15996788.0,
      "step": 8565
    },
    {
      "entropy": 5.891510248184204,
      "epoch": 0.7567997174143413,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004949686319628535,
      "loss": 5.5889,
      "mean_token_accuracy": 0.18840514570474626,
      "num_tokens": 16007212.0,
      "step": 8570
    },
    {
      "entropy": 5.804558944702149,
      "epoch": 0.7572412575061815,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004949620073588394,
      "loss": 5.6435,
      "mean_token_accuracy": 0.18354785442352295,
      "num_tokens": 16017000.0,
      "step": 8575
    },
    {
      "entropy": 5.941242122650147,
      "epoch": 0.7576827975980219,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004949553784458945,
      "loss": 5.7309,
      "mean_token_accuracy": 0.17222158014774322,
      "num_tokens": 16026647.0,
      "step": 8580
    },
    {
      "entropy": 5.841959381103516,
      "epoch": 0.7581243376898622,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004949487452241489,
      "loss": 5.4496,
      "mean_token_accuracy": 0.2036088615655899,
      "num_tokens": 16034773.0,
      "step": 8585
    },
    {
      "entropy": 5.879746627807617,
      "epoch": 0.7585658777817026,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004949421076937325,
      "loss": 5.6027,
      "mean_token_accuracy": 0.18925876021385193,
      "num_tokens": 16043787.0,
      "step": 8590
    },
    {
      "entropy": 5.76485013961792,
      "epoch": 0.7590074178735429,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004949354658547753,
      "loss": 5.5737,
      "mean_token_accuracy": 0.19907401651144027,
      "num_tokens": 16053130.0,
      "step": 8595
    },
    {
      "entropy": 5.870529079437256,
      "epoch": 0.7594489579653833,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004949288197074074,
      "loss": 5.7198,
      "mean_token_accuracy": 0.17960111498832704,
      "num_tokens": 16062529.0,
      "step": 8600
    },
    {
      "entropy": 5.8299237251281735,
      "epoch": 0.7598904980572236,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004949221692517589,
      "loss": 5.5684,
      "mean_token_accuracy": 0.18861594051122665,
      "num_tokens": 16071932.0,
      "step": 8605
    },
    {
      "entropy": 5.880710124969482,
      "epoch": 0.7603320381490639,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004949155144879602,
      "loss": 5.5314,
      "mean_token_accuracy": 0.19198206812143326,
      "num_tokens": 16081077.0,
      "step": 8610
    },
    {
      "entropy": 5.786740303039551,
      "epoch": 0.7607735782409043,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004949088554161417,
      "loss": 5.4939,
      "mean_token_accuracy": 0.19128501862287522,
      "num_tokens": 16089725.0,
      "step": 8615
    },
    {
      "entropy": 5.847886180877685,
      "epoch": 0.7612151183327446,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004949021920364339,
      "loss": 5.6605,
      "mean_token_accuracy": 0.18638603687286376,
      "num_tokens": 16099362.0,
      "step": 8620
    },
    {
      "entropy": 5.94460711479187,
      "epoch": 0.761656658424585,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004948955243489671,
      "loss": 5.675,
      "mean_token_accuracy": 0.18102553784847258,
      "num_tokens": 16108746.0,
      "step": 8625
    },
    {
      "entropy": 5.966106033325195,
      "epoch": 0.7620981985164252,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004948888523538721,
      "loss": 5.8383,
      "mean_token_accuracy": 0.1732643410563469,
      "num_tokens": 16118902.0,
      "step": 8630
    },
    {
      "entropy": 5.881408309936523,
      "epoch": 0.7625397386082656,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004948821760512795,
      "loss": 5.5767,
      "mean_token_accuracy": 0.1854456201195717,
      "num_tokens": 16127781.0,
      "step": 8635
    },
    {
      "entropy": 5.865966653823852,
      "epoch": 0.762981278700106,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004948754954413201,
      "loss": 5.5434,
      "mean_token_accuracy": 0.20540579557418823,
      "num_tokens": 16136004.0,
      "step": 8640
    },
    {
      "entropy": 5.809378910064697,
      "epoch": 0.7634228187919463,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004948688105241248,
      "loss": 5.5865,
      "mean_token_accuracy": 0.1957641437649727,
      "num_tokens": 16145023.0,
      "step": 8645
    },
    {
      "entropy": 5.872867155075073,
      "epoch": 0.7638643588837867,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004948621212998246,
      "loss": 5.6456,
      "mean_token_accuracy": 0.18539468497037886,
      "num_tokens": 16154242.0,
      "step": 8650
    },
    {
      "entropy": 5.940119457244873,
      "epoch": 0.764305898975627,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004948554277685505,
      "loss": 5.619,
      "mean_token_accuracy": 0.18480096608400345,
      "num_tokens": 16163252.0,
      "step": 8655
    },
    {
      "entropy": 5.909817171096802,
      "epoch": 0.7647474390674673,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004948487299304335,
      "loss": 5.552,
      "mean_token_accuracy": 0.1940522775053978,
      "num_tokens": 16171363.0,
      "step": 8660
    },
    {
      "entropy": 5.8778034210205075,
      "epoch": 0.7651889791593076,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494842027785605,
      "loss": 5.7388,
      "mean_token_accuracy": 0.18207193166017532,
      "num_tokens": 16181234.0,
      "step": 8665
    },
    {
      "entropy": 5.895607423782349,
      "epoch": 0.765630519251148,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004948353213341961,
      "loss": 5.6517,
      "mean_token_accuracy": 0.1834381639957428,
      "num_tokens": 16189420.0,
      "step": 8670
    },
    {
      "entropy": 5.9458598613739015,
      "epoch": 0.7660720593429884,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004948286105763385,
      "loss": 5.7867,
      "mean_token_accuracy": 0.17419244572520257,
      "num_tokens": 16199202.0,
      "step": 8675
    },
    {
      "entropy": 5.9134142875671385,
      "epoch": 0.7665135994348287,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004948218955121632,
      "loss": 5.6212,
      "mean_token_accuracy": 0.1843989074230194,
      "num_tokens": 16208481.0,
      "step": 8680
    },
    {
      "entropy": 5.904475688934326,
      "epoch": 0.766955139526669,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004948151761418021,
      "loss": 5.7078,
      "mean_token_accuracy": 0.18244910091161728,
      "num_tokens": 16217282.0,
      "step": 8685
    },
    {
      "entropy": 5.880133628845215,
      "epoch": 0.7673966796185093,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004948084524653868,
      "loss": 5.5712,
      "mean_token_accuracy": 0.1891437828540802,
      "num_tokens": 16226567.0,
      "step": 8690
    },
    {
      "entropy": 5.866838026046753,
      "epoch": 0.7678382197103497,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004948017244830489,
      "loss": 5.7673,
      "mean_token_accuracy": 0.17605815678834916,
      "num_tokens": 16236255.0,
      "step": 8695
    },
    {
      "entropy": 5.869735670089722,
      "epoch": 0.76827975980219,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004947949921949201,
      "loss": 5.5621,
      "mean_token_accuracy": 0.19431234449148177,
      "num_tokens": 16244838.0,
      "step": 8700
    },
    {
      "entropy": 5.7782117366790775,
      "epoch": 0.7687212998940304,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004947882556011325,
      "loss": 5.4373,
      "mean_token_accuracy": 0.20313273072242738,
      "num_tokens": 16254101.0,
      "step": 8705
    },
    {
      "entropy": 5.712455606460571,
      "epoch": 0.7691628399858708,
      "grad_norm": 0.859375,
      "learning_rate": 0.000494781514701818,
      "loss": 5.5341,
      "mean_token_accuracy": 0.1988469660282135,
      "num_tokens": 16263496.0,
      "step": 8710
    },
    {
      "entropy": 5.932302951812744,
      "epoch": 0.769604380077711,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004947747694971085,
      "loss": 5.6368,
      "mean_token_accuracy": 0.19015247076749803,
      "num_tokens": 16271884.0,
      "step": 8715
    },
    {
      "entropy": 5.857388734817505,
      "epoch": 0.7700459201695514,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004947680199871363,
      "loss": 5.5507,
      "mean_token_accuracy": 0.19477736204862595,
      "num_tokens": 16281197.0,
      "step": 8720
    },
    {
      "entropy": 5.808503437042236,
      "epoch": 0.7704874602613917,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004947612661720336,
      "loss": 5.646,
      "mean_token_accuracy": 0.18876955658197403,
      "num_tokens": 16289908.0,
      "step": 8725
    },
    {
      "entropy": 5.825726795196533,
      "epoch": 0.7709290003532321,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004947545080519327,
      "loss": 5.599,
      "mean_token_accuracy": 0.18349761068820952,
      "num_tokens": 16299237.0,
      "step": 8730
    },
    {
      "entropy": 5.9406637191772464,
      "epoch": 0.7713705404450724,
      "grad_norm": 1.046875,
      "learning_rate": 0.000494747745626966,
      "loss": 5.6251,
      "mean_token_accuracy": 0.18967668414115907,
      "num_tokens": 16308175.0,
      "step": 8735
    },
    {
      "entropy": 5.865895175933838,
      "epoch": 0.7718120805369127,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004947409788972659,
      "loss": 5.5776,
      "mean_token_accuracy": 0.19288718998432158,
      "num_tokens": 16317121.0,
      "step": 8740
    },
    {
      "entropy": 5.81405348777771,
      "epoch": 0.7722536206287531,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004947342078629649,
      "loss": 5.5723,
      "mean_token_accuracy": 0.18960059136152269,
      "num_tokens": 16325764.0,
      "step": 8745
    },
    {
      "entropy": 5.863961315155029,
      "epoch": 0.7726951607205934,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004947274325241959,
      "loss": 5.5695,
      "mean_token_accuracy": 0.1862097531557083,
      "num_tokens": 16333823.0,
      "step": 8750
    },
    {
      "entropy": 5.800171661376953,
      "epoch": 0.7731367008124338,
      "grad_norm": 1.0,
      "learning_rate": 0.0004947206528810913,
      "loss": 5.5642,
      "mean_token_accuracy": 0.1920272096991539,
      "num_tokens": 16341918.0,
      "step": 8755
    },
    {
      "entropy": 5.817489528656006,
      "epoch": 0.7735782409042741,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004947138689337843,
      "loss": 5.6122,
      "mean_token_accuracy": 0.18937198519706727,
      "num_tokens": 16351904.0,
      "step": 8760
    },
    {
      "entropy": 5.950567388534546,
      "epoch": 0.7740197809961145,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004947070806824074,
      "loss": 5.62,
      "mean_token_accuracy": 0.18496680557727813,
      "num_tokens": 16360921.0,
      "step": 8765
    },
    {
      "entropy": 5.908766412734986,
      "epoch": 0.7744613210879547,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004947002881270937,
      "loss": 5.6423,
      "mean_token_accuracy": 0.1874388188123703,
      "num_tokens": 16370929.0,
      "step": 8770
    },
    {
      "entropy": 5.890651512145996,
      "epoch": 0.7749028611797951,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004946934912679764,
      "loss": 5.6359,
      "mean_token_accuracy": 0.18902267664670944,
      "num_tokens": 16380099.0,
      "step": 8775
    },
    {
      "entropy": 5.898251152038574,
      "epoch": 0.7753444012716355,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004946866901051885,
      "loss": 5.6891,
      "mean_token_accuracy": 0.18321269005537033,
      "num_tokens": 16389109.0,
      "step": 8780
    },
    {
      "entropy": 5.88666090965271,
      "epoch": 0.7757859413634758,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004946798846388634,
      "loss": 5.7083,
      "mean_token_accuracy": 0.17529224157333373,
      "num_tokens": 16398601.0,
      "step": 8785
    },
    {
      "entropy": 5.8759900569915775,
      "epoch": 0.7762274814553162,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946730748691342,
      "loss": 5.5623,
      "mean_token_accuracy": 0.19695408195257186,
      "num_tokens": 16406950.0,
      "step": 8790
    },
    {
      "entropy": 5.928198385238647,
      "epoch": 0.7766690215471564,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004946662607961344,
      "loss": 5.5653,
      "mean_token_accuracy": 0.1901753857731819,
      "num_tokens": 16414680.0,
      "step": 8795
    },
    {
      "entropy": 5.811094951629639,
      "epoch": 0.7771105616389968,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004946594424199976,
      "loss": 5.63,
      "mean_token_accuracy": 0.18612197637557984,
      "num_tokens": 16424060.0,
      "step": 8800
    },
    {
      "entropy": 5.8213338375091555,
      "epoch": 0.7775521017308371,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004946526197408571,
      "loss": 5.6604,
      "mean_token_accuracy": 0.18676266372203826,
      "num_tokens": 16433691.0,
      "step": 8805
    },
    {
      "entropy": 5.919007301330566,
      "epoch": 0.7779936418226775,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004946457927588467,
      "loss": 5.6535,
      "mean_token_accuracy": 0.18097317218780518,
      "num_tokens": 16443234.0,
      "step": 8810
    },
    {
      "entropy": 5.85936803817749,
      "epoch": 0.7784351819145179,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004946389614741002,
      "loss": 5.6024,
      "mean_token_accuracy": 0.19192105978727342,
      "num_tokens": 16452084.0,
      "step": 8815
    },
    {
      "entropy": 5.853142642974854,
      "epoch": 0.7788767220063582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946321258867513,
      "loss": 5.5928,
      "mean_token_accuracy": 0.1917520835995674,
      "num_tokens": 16460135.0,
      "step": 8820
    },
    {
      "entropy": 5.902470922470092,
      "epoch": 0.7793182620981985,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000494625285996934,
      "loss": 5.7215,
      "mean_token_accuracy": 0.18264151215553284,
      "num_tokens": 16469919.0,
      "step": 8825
    },
    {
      "entropy": 5.948286771774292,
      "epoch": 0.7797598021900388,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004946184418047823,
      "loss": 5.6396,
      "mean_token_accuracy": 0.1841666057705879,
      "num_tokens": 16478473.0,
      "step": 8830
    },
    {
      "entropy": 5.9353759765625,
      "epoch": 0.7802013422818792,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946115933104301,
      "loss": 5.6427,
      "mean_token_accuracy": 0.17996163070201873,
      "num_tokens": 16486878.0,
      "step": 8835
    },
    {
      "entropy": 5.856232929229736,
      "epoch": 0.7806428823737195,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004946047405140119,
      "loss": 5.6813,
      "mean_token_accuracy": 0.1801102951169014,
      "num_tokens": 16496798.0,
      "step": 8840
    },
    {
      "entropy": 5.931548357009888,
      "epoch": 0.7810844224655599,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004945978834156616,
      "loss": 5.5854,
      "mean_token_accuracy": 0.19176071733236313,
      "num_tokens": 16506508.0,
      "step": 8845
    },
    {
      "entropy": 5.789986038208008,
      "epoch": 0.7815259625574003,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004945910220155136,
      "loss": 5.5651,
      "mean_token_accuracy": 0.19404177665710448,
      "num_tokens": 16514979.0,
      "step": 8850
    },
    {
      "entropy": 5.872536373138428,
      "epoch": 0.7819675026492405,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004945841563137025,
      "loss": 5.6417,
      "mean_token_accuracy": 0.1903441235423088,
      "num_tokens": 16524654.0,
      "step": 8855
    },
    {
      "entropy": 5.909209680557251,
      "epoch": 0.7824090427410809,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004945772863103626,
      "loss": 5.7241,
      "mean_token_accuracy": 0.18177213817834853,
      "num_tokens": 16534486.0,
      "step": 8860
    },
    {
      "entropy": 5.937249708175659,
      "epoch": 0.7828505828329212,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004945704120056286,
      "loss": 5.6423,
      "mean_token_accuracy": 0.19361244142055511,
      "num_tokens": 16543718.0,
      "step": 8865
    },
    {
      "entropy": 5.8497367858886715,
      "epoch": 0.7832921229247616,
      "grad_norm": 1.0,
      "learning_rate": 0.000494563533399635,
      "loss": 5.6089,
      "mean_token_accuracy": 0.18692233711481093,
      "num_tokens": 16553159.0,
      "step": 8870
    },
    {
      "entropy": 5.948692417144775,
      "epoch": 0.7837336630166019,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004945566504925167,
      "loss": 5.6642,
      "mean_token_accuracy": 0.1829291120171547,
      "num_tokens": 16562398.0,
      "step": 8875
    },
    {
      "entropy": 5.894802904129028,
      "epoch": 0.7841752031084422,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004945497632844084,
      "loss": 5.6048,
      "mean_token_accuracy": 0.19393257647752762,
      "num_tokens": 16571530.0,
      "step": 8880
    },
    {
      "entropy": 5.789991092681885,
      "epoch": 0.7846167432002826,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004945428717754453,
      "loss": 5.6248,
      "mean_token_accuracy": 0.1875714972615242,
      "num_tokens": 16581074.0,
      "step": 8885
    },
    {
      "entropy": 5.832242250442505,
      "epoch": 0.7850582832921229,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000494535975965762,
      "loss": 5.5046,
      "mean_token_accuracy": 0.1969103291630745,
      "num_tokens": 16589574.0,
      "step": 8890
    },
    {
      "entropy": 5.781494998931885,
      "epoch": 0.7854998233839633,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004945290758554939,
      "loss": 5.5657,
      "mean_token_accuracy": 0.19426276236772538,
      "num_tokens": 16599623.0,
      "step": 8895
    },
    {
      "entropy": 5.818102407455444,
      "epoch": 0.7859413634758036,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000494522171444776,
      "loss": 5.6364,
      "mean_token_accuracy": 0.17526367157697678,
      "num_tokens": 16608559.0,
      "step": 8900
    },
    {
      "entropy": 5.861512613296509,
      "epoch": 0.786382903567644,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004945152627337437,
      "loss": 5.5512,
      "mean_token_accuracy": 0.19917489737272262,
      "num_tokens": 16617148.0,
      "step": 8905
    },
    {
      "entropy": 5.7910514831542965,
      "epoch": 0.7868244436594842,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004945083497225321,
      "loss": 5.5606,
      "mean_token_accuracy": 0.1904294088482857,
      "num_tokens": 16626405.0,
      "step": 8910
    },
    {
      "entropy": 5.840649938583374,
      "epoch": 0.7872659837513246,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004945014324112769,
      "loss": 5.6101,
      "mean_token_accuracy": 0.1881376251578331,
      "num_tokens": 16635116.0,
      "step": 8915
    },
    {
      "entropy": 5.78248348236084,
      "epoch": 0.787707523843165,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004944945108001134,
      "loss": 5.5713,
      "mean_token_accuracy": 0.18788688331842424,
      "num_tokens": 16644381.0,
      "step": 8920
    },
    {
      "entropy": 5.86023359298706,
      "epoch": 0.7881490639350053,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004944875848891773,
      "loss": 5.6207,
      "mean_token_accuracy": 0.19069672822952272,
      "num_tokens": 16654275.0,
      "step": 8925
    },
    {
      "entropy": 5.885168266296387,
      "epoch": 0.7885906040268457,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004944806546786042,
      "loss": 5.7006,
      "mean_token_accuracy": 0.17883567065000533,
      "num_tokens": 16663484.0,
      "step": 8930
    },
    {
      "entropy": 5.799995088577271,
      "epoch": 0.7890321441186859,
      "grad_norm": 0.9375,
      "learning_rate": 0.00049447372016853,
      "loss": 5.64,
      "mean_token_accuracy": 0.18918595016002654,
      "num_tokens": 16672490.0,
      "step": 8935
    },
    {
      "entropy": 5.899666690826416,
      "epoch": 0.7894736842105263,
      "grad_norm": 1.0,
      "learning_rate": 0.0004944667813590904,
      "loss": 5.59,
      "mean_token_accuracy": 0.18535562306642533,
      "num_tokens": 16681242.0,
      "step": 8940
    },
    {
      "entropy": 5.812223339080811,
      "epoch": 0.7899152243023666,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004944598382504214,
      "loss": 5.5646,
      "mean_token_accuracy": 0.1899319499731064,
      "num_tokens": 16690359.0,
      "step": 8945
    },
    {
      "entropy": 5.85008430480957,
      "epoch": 0.790356764394207,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004944528908426589,
      "loss": 5.637,
      "mean_token_accuracy": 0.1899271935224533,
      "num_tokens": 16700774.0,
      "step": 8950
    },
    {
      "entropy": 5.957952642440796,
      "epoch": 0.7907983044860474,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004944459391359391,
      "loss": 5.7111,
      "mean_token_accuracy": 0.1864597260951996,
      "num_tokens": 16710513.0,
      "step": 8955
    },
    {
      "entropy": 5.826300096511841,
      "epoch": 0.7912398445778877,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004944389831303982,
      "loss": 5.6322,
      "mean_token_accuracy": 0.18941447585821153,
      "num_tokens": 16720785.0,
      "step": 8960
    },
    {
      "entropy": 5.864205551147461,
      "epoch": 0.791681384669728,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004944320228261723,
      "loss": 5.6036,
      "mean_token_accuracy": 0.18326730877161027,
      "num_tokens": 16729083.0,
      "step": 8965
    },
    {
      "entropy": 5.831407117843628,
      "epoch": 0.7921229247615683,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000494425058223398,
      "loss": 5.6218,
      "mean_token_accuracy": 0.1872497618198395,
      "num_tokens": 16738313.0,
      "step": 8970
    },
    {
      "entropy": 5.846746397018433,
      "epoch": 0.7925644648534087,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004944180893222117,
      "loss": 5.5538,
      "mean_token_accuracy": 0.19500797092914582,
      "num_tokens": 16747191.0,
      "step": 8975
    },
    {
      "entropy": 5.7963032722473145,
      "epoch": 0.793006004945249,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004944111161227496,
      "loss": 5.5776,
      "mean_token_accuracy": 0.18987296223640443,
      "num_tokens": 16755921.0,
      "step": 8980
    },
    {
      "entropy": 5.910832405090332,
      "epoch": 0.7934475450370894,
      "grad_norm": 1.0,
      "learning_rate": 0.0004944041386251486,
      "loss": 5.7473,
      "mean_token_accuracy": 0.18214513510465621,
      "num_tokens": 16765394.0,
      "step": 8985
    },
    {
      "entropy": 5.920050525665284,
      "epoch": 0.7938890851289298,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004943971568295454,
      "loss": 5.6229,
      "mean_token_accuracy": 0.19057771414518357,
      "num_tokens": 16774824.0,
      "step": 8990
    },
    {
      "entropy": 5.925590467453003,
      "epoch": 0.79433062522077,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004943901707360765,
      "loss": 5.7477,
      "mean_token_accuracy": 0.17744729071855544,
      "num_tokens": 16783945.0,
      "step": 8995
    },
    {
      "entropy": 5.856736326217652,
      "epoch": 0.7947721653126104,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000494383180344879,
      "loss": 5.5277,
      "mean_token_accuracy": 0.19507913142442704,
      "num_tokens": 16792618.0,
      "step": 9000
    },
    {
      "epoch": 0.7947721653126104,
      "eval_entropy": 5.591507935388982,
      "eval_loss": 5.619070529937744,
      "eval_mean_token_accuracy": 0.19495929661475855,
      "eval_num_tokens": 16792618.0,
      "eval_runtime": 20.2421,
      "eval_samples_per_second": 1744.631,
      "eval_steps_per_second": 218.11,
      "step": 9000
    },
    {
      "entropy": 5.825322103500366,
      "epoch": 0.7952137054044507,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004943761856560898,
      "loss": 5.5992,
      "mean_token_accuracy": 0.18942335098981858,
      "num_tokens": 16801210.0,
      "step": 9005
    },
    {
      "entropy": 5.867608642578125,
      "epoch": 0.7956552454962911,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000494369186669846,
      "loss": 5.6397,
      "mean_token_accuracy": 0.18785736858844757,
      "num_tokens": 16810371.0,
      "step": 9010
    },
    {
      "entropy": 5.823724222183228,
      "epoch": 0.7960967855881314,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004943621833862844,
      "loss": 5.545,
      "mean_token_accuracy": 0.19277409613132476,
      "num_tokens": 16818886.0,
      "step": 9015
    },
    {
      "entropy": 5.842554903030395,
      "epoch": 0.7965383256799717,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004943551758055425,
      "loss": 5.6224,
      "mean_token_accuracy": 0.18964377343654631,
      "num_tokens": 16827705.0,
      "step": 9020
    },
    {
      "entropy": 5.8641541481018065,
      "epoch": 0.7969798657718121,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004943481639277574,
      "loss": 5.582,
      "mean_token_accuracy": 0.18107532560825348,
      "num_tokens": 16836313.0,
      "step": 9025
    },
    {
      "entropy": 5.9201160907745365,
      "epoch": 0.7974214058636524,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004943411477530665,
      "loss": 5.6618,
      "mean_token_accuracy": 0.1813226357102394,
      "num_tokens": 16845016.0,
      "step": 9030
    },
    {
      "entropy": 5.87960410118103,
      "epoch": 0.7978629459554928,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004943341272816073,
      "loss": 5.627,
      "mean_token_accuracy": 0.1824415385723114,
      "num_tokens": 16854025.0,
      "step": 9035
    },
    {
      "entropy": 5.828975486755371,
      "epoch": 0.7983044860473331,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004943271025135173,
      "loss": 5.5683,
      "mean_token_accuracy": 0.18993970900774002,
      "num_tokens": 16862859.0,
      "step": 9040
    },
    {
      "entropy": 5.855637502670288,
      "epoch": 0.7987460261391734,
      "grad_norm": 0.875,
      "learning_rate": 0.0004943200734489341,
      "loss": 5.6629,
      "mean_token_accuracy": 0.18977779150009155,
      "num_tokens": 16872226.0,
      "step": 9045
    },
    {
      "entropy": 5.803923225402832,
      "epoch": 0.7991875662310137,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004943130400879953,
      "loss": 5.5409,
      "mean_token_accuracy": 0.1905651032924652,
      "num_tokens": 16881977.0,
      "step": 9050
    },
    {
      "entropy": 5.857625722885132,
      "epoch": 0.7996291063228541,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004943060024308389,
      "loss": 5.6599,
      "mean_token_accuracy": 0.1782623678445816,
      "num_tokens": 16891656.0,
      "step": 9055
    },
    {
      "entropy": 5.903197383880615,
      "epoch": 0.8000706464146945,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004942989604776026,
      "loss": 5.5252,
      "mean_token_accuracy": 0.19363559782505035,
      "num_tokens": 16900944.0,
      "step": 9060
    },
    {
      "entropy": 5.761235761642456,
      "epoch": 0.8005121865065348,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004942919142284244,
      "loss": 5.546,
      "mean_token_accuracy": 0.18910249769687654,
      "num_tokens": 16910154.0,
      "step": 9065
    },
    {
      "entropy": 5.7982673168182375,
      "epoch": 0.8009537265983752,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004942848636834423,
      "loss": 5.6586,
      "mean_token_accuracy": 0.1833169311285019,
      "num_tokens": 16919694.0,
      "step": 9070
    },
    {
      "entropy": 5.882076978683472,
      "epoch": 0.8013952666902154,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004942778088427944,
      "loss": 5.6214,
      "mean_token_accuracy": 0.18945298343896866,
      "num_tokens": 16928530.0,
      "step": 9075
    },
    {
      "entropy": 5.927755212783813,
      "epoch": 0.8018368067820558,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000494270749706619,
      "loss": 5.6551,
      "mean_token_accuracy": 0.18146197944879533,
      "num_tokens": 16938614.0,
      "step": 9080
    },
    {
      "entropy": 5.8563508033752445,
      "epoch": 0.8022783468738961,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004942636862750544,
      "loss": 5.595,
      "mean_token_accuracy": 0.18589434772729874,
      "num_tokens": 16947655.0,
      "step": 9085
    },
    {
      "entropy": 5.8665442943573,
      "epoch": 0.8027198869657365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004942566185482387,
      "loss": 5.6857,
      "mean_token_accuracy": 0.1815369561314583,
      "num_tokens": 16957101.0,
      "step": 9090
    },
    {
      "entropy": 5.843801975250244,
      "epoch": 0.8031614270575769,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004942495465263106,
      "loss": 5.7167,
      "mean_token_accuracy": 0.1784170463681221,
      "num_tokens": 16966421.0,
      "step": 9095
    },
    {
      "entropy": 5.925464344024658,
      "epoch": 0.8036029671494171,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004942424702094085,
      "loss": 5.6041,
      "mean_token_accuracy": 0.19640391767024995,
      "num_tokens": 16975656.0,
      "step": 9100
    },
    {
      "entropy": 5.887157869338989,
      "epoch": 0.8040445072412575,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004942353895976713,
      "loss": 5.6632,
      "mean_token_accuracy": 0.18462532758712769,
      "num_tokens": 16984409.0,
      "step": 9105
    },
    {
      "entropy": 5.816635465621948,
      "epoch": 0.8044860473330978,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004942283046912374,
      "loss": 5.5567,
      "mean_token_accuracy": 0.1862815201282501,
      "num_tokens": 16992728.0,
      "step": 9110
    },
    {
      "entropy": 5.8094785690307615,
      "epoch": 0.8049275874249382,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004942212154902456,
      "loss": 5.4597,
      "mean_token_accuracy": 0.20306771099567414,
      "num_tokens": 17001616.0,
      "step": 9115
    },
    {
      "entropy": 5.818345212936402,
      "epoch": 0.8053691275167785,
      "grad_norm": 1.0,
      "learning_rate": 0.000494214121994835,
      "loss": 5.641,
      "mean_token_accuracy": 0.18777192533016204,
      "num_tokens": 17010655.0,
      "step": 9120
    },
    {
      "entropy": 5.862553977966309,
      "epoch": 0.8058106676086189,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004942070242051444,
      "loss": 5.5987,
      "mean_token_accuracy": 0.18641175031661988,
      "num_tokens": 17019819.0,
      "step": 9125
    },
    {
      "entropy": 5.934750032424927,
      "epoch": 0.8062522077004592,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004941999221213129,
      "loss": 5.6402,
      "mean_token_accuracy": 0.18712895065546037,
      "num_tokens": 17029218.0,
      "step": 9130
    },
    {
      "entropy": 5.9235235214233395,
      "epoch": 0.8066937477922995,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004941928157434796,
      "loss": 5.5558,
      "mean_token_accuracy": 0.1911984696984291,
      "num_tokens": 17037880.0,
      "step": 9135
    },
    {
      "entropy": 5.789757156372071,
      "epoch": 0.8071352878841399,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004941857050717836,
      "loss": 5.6241,
      "mean_token_accuracy": 0.1851594939827919,
      "num_tokens": 17047934.0,
      "step": 9140
    },
    {
      "entropy": 5.932223892211914,
      "epoch": 0.8075768279759802,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004941785901063642,
      "loss": 5.691,
      "mean_token_accuracy": 0.18222017735242843,
      "num_tokens": 17057258.0,
      "step": 9145
    },
    {
      "entropy": 5.885777521133423,
      "epoch": 0.8080183680678206,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004941714708473611,
      "loss": 5.6141,
      "mean_token_accuracy": 0.18538845628499984,
      "num_tokens": 17066387.0,
      "step": 9150
    },
    {
      "entropy": 5.799456453323364,
      "epoch": 0.8084599081596608,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004941643472949134,
      "loss": 5.5675,
      "mean_token_accuracy": 0.1871365562081337,
      "num_tokens": 17075952.0,
      "step": 9155
    },
    {
      "entropy": 5.7912437915802,
      "epoch": 0.8089014482515012,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004941572194491608,
      "loss": 5.4397,
      "mean_token_accuracy": 0.20129949748516082,
      "num_tokens": 17084209.0,
      "step": 9160
    },
    {
      "entropy": 5.819268655776978,
      "epoch": 0.8093429883433416,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000494150087310243,
      "loss": 5.6811,
      "mean_token_accuracy": 0.18660429269075393,
      "num_tokens": 17093395.0,
      "step": 9165
    },
    {
      "entropy": 5.889335632324219,
      "epoch": 0.8097845284351819,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004941429508782995,
      "loss": 5.5633,
      "mean_token_accuracy": 0.19305034577846528,
      "num_tokens": 17102646.0,
      "step": 9170
    },
    {
      "entropy": 5.912970304489136,
      "epoch": 0.8102260685270223,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004941358101534703,
      "loss": 5.577,
      "mean_token_accuracy": 0.1855246216058731,
      "num_tokens": 17111530.0,
      "step": 9175
    },
    {
      "entropy": 5.789304494857788,
      "epoch": 0.8106676086188626,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004941286651358952,
      "loss": 5.5631,
      "mean_token_accuracy": 0.1946306437253952,
      "num_tokens": 17119659.0,
      "step": 9180
    },
    {
      "entropy": 5.7249321937561035,
      "epoch": 0.8111091487107029,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004941215158257142,
      "loss": 5.5494,
      "mean_token_accuracy": 0.19357234835624695,
      "num_tokens": 17127931.0,
      "step": 9185
    },
    {
      "entropy": 5.942633867263794,
      "epoch": 0.8115506888025432,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004941143622230672,
      "loss": 5.5768,
      "mean_token_accuracy": 0.19068775326013565,
      "num_tokens": 17137374.0,
      "step": 9190
    },
    {
      "entropy": 5.92955060005188,
      "epoch": 0.8119922288943836,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004941072043280946,
      "loss": 5.6504,
      "mean_token_accuracy": 0.18410056233406066,
      "num_tokens": 17147257.0,
      "step": 9195
    },
    {
      "entropy": 5.830786895751953,
      "epoch": 0.812433768986224,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004941000421409365,
      "loss": 5.5152,
      "mean_token_accuracy": 0.19687575548887254,
      "num_tokens": 17156233.0,
      "step": 9200
    },
    {
      "entropy": 5.914954757690429,
      "epoch": 0.8128753090780643,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004940928756617331,
      "loss": 5.713,
      "mean_token_accuracy": 0.1751517355442047,
      "num_tokens": 17166364.0,
      "step": 9205
    },
    {
      "entropy": 5.830341339111328,
      "epoch": 0.8133168491699047,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004940857048906248,
      "loss": 5.6207,
      "mean_token_accuracy": 0.1892386794090271,
      "num_tokens": 17175853.0,
      "step": 9210
    },
    {
      "entropy": 5.878078269958496,
      "epoch": 0.8137583892617449,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004940785298277523,
      "loss": 5.7153,
      "mean_token_accuracy": 0.18135832846164704,
      "num_tokens": 17187849.0,
      "step": 9215
    },
    {
      "entropy": 5.920613622665405,
      "epoch": 0.8141999293535853,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000494071350473256,
      "loss": 5.5822,
      "mean_token_accuracy": 0.18615618199110032,
      "num_tokens": 17196646.0,
      "step": 9220
    },
    {
      "entropy": 5.888037824630738,
      "epoch": 0.8146414694454256,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004940641668272765,
      "loss": 5.7214,
      "mean_token_accuracy": 0.18530243486166,
      "num_tokens": 17205919.0,
      "step": 9225
    },
    {
      "entropy": 5.902007102966309,
      "epoch": 0.815083009537266,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004940569788899544,
      "loss": 5.6713,
      "mean_token_accuracy": 0.18742569386959076,
      "num_tokens": 17215085.0,
      "step": 9230
    },
    {
      "entropy": 5.867893171310425,
      "epoch": 0.8155245496291064,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004940497866614309,
      "loss": 5.6533,
      "mean_token_accuracy": 0.18850154876708985,
      "num_tokens": 17223633.0,
      "step": 9235
    },
    {
      "entropy": 5.853533172607422,
      "epoch": 0.8159660897209466,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004940425901418465,
      "loss": 5.5872,
      "mean_token_accuracy": 0.19215673506259917,
      "num_tokens": 17232915.0,
      "step": 9240
    },
    {
      "entropy": 5.933935165405273,
      "epoch": 0.816407629812787,
      "grad_norm": 1.25,
      "learning_rate": 0.0004940353893313423,
      "loss": 5.676,
      "mean_token_accuracy": 0.18245417177677153,
      "num_tokens": 17242392.0,
      "step": 9245
    },
    {
      "entropy": 5.964029026031494,
      "epoch": 0.8168491699046273,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004940281842300596,
      "loss": 5.7735,
      "mean_token_accuracy": 0.17744607329368592,
      "num_tokens": 17252259.0,
      "step": 9250
    },
    {
      "entropy": 5.921383571624756,
      "epoch": 0.8172907099964677,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004940209748381392,
      "loss": 5.6371,
      "mean_token_accuracy": 0.18096522390842437,
      "num_tokens": 17261584.0,
      "step": 9255
    },
    {
      "entropy": 5.890787363052368,
      "epoch": 0.817732250088308,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004940137611557225,
      "loss": 5.5562,
      "mean_token_accuracy": 0.19099668562412261,
      "num_tokens": 17270901.0,
      "step": 9260
    },
    {
      "entropy": 5.876820230484009,
      "epoch": 0.8181737901801484,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004940065431829508,
      "loss": 5.6184,
      "mean_token_accuracy": 0.18632869124412538,
      "num_tokens": 17280024.0,
      "step": 9265
    },
    {
      "entropy": 5.889107370376587,
      "epoch": 0.8186153302719887,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004939993209199654,
      "loss": 5.6784,
      "mean_token_accuracy": 0.1789259731769562,
      "num_tokens": 17290131.0,
      "step": 9270
    },
    {
      "entropy": 5.96183500289917,
      "epoch": 0.819056870363829,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004939920943669079,
      "loss": 5.6668,
      "mean_token_accuracy": 0.18145134747028352,
      "num_tokens": 17299453.0,
      "step": 9275
    },
    {
      "entropy": 5.8522838115692135,
      "epoch": 0.8194984104556694,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004939848635239198,
      "loss": 5.6284,
      "mean_token_accuracy": 0.18608871698379517,
      "num_tokens": 17309344.0,
      "step": 9280
    },
    {
      "entropy": 5.8321044921875,
      "epoch": 0.8199399505475097,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004939776283911429,
      "loss": 5.5445,
      "mean_token_accuracy": 0.1960724487900734,
      "num_tokens": 17318636.0,
      "step": 9285
    },
    {
      "entropy": 5.826576089859008,
      "epoch": 0.8203814906393501,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004939703889687186,
      "loss": 5.6117,
      "mean_token_accuracy": 0.18429741710424424,
      "num_tokens": 17328146.0,
      "step": 9290
    },
    {
      "entropy": 5.860490322113037,
      "epoch": 0.8208230307311903,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004939631452567891,
      "loss": 5.6699,
      "mean_token_accuracy": 0.18432924002408982,
      "num_tokens": 17338542.0,
      "step": 9295
    },
    {
      "entropy": 5.937232255935669,
      "epoch": 0.8212645708230307,
      "grad_norm": 0.9375,
      "learning_rate": 0.000493955897255496,
      "loss": 5.6259,
      "mean_token_accuracy": 0.1889485612511635,
      "num_tokens": 17347577.0,
      "step": 9300
    },
    {
      "entropy": 5.814497280120849,
      "epoch": 0.8217061109148711,
      "grad_norm": 1.0,
      "learning_rate": 0.0004939486449649814,
      "loss": 5.606,
      "mean_token_accuracy": 0.1901299074292183,
      "num_tokens": 17356554.0,
      "step": 9305
    },
    {
      "entropy": 5.827168941497803,
      "epoch": 0.8221476510067114,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004939413883853873,
      "loss": 5.6382,
      "mean_token_accuracy": 0.18767885267734527,
      "num_tokens": 17366432.0,
      "step": 9310
    },
    {
      "entropy": 5.914646291732788,
      "epoch": 0.8225891910985518,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000493934127516856,
      "loss": 5.619,
      "mean_token_accuracy": 0.19058295339345932,
      "num_tokens": 17376556.0,
      "step": 9315
    },
    {
      "entropy": 5.845771455764771,
      "epoch": 0.823030731190392,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004939268623595297,
      "loss": 5.659,
      "mean_token_accuracy": 0.18527737408876419,
      "num_tokens": 17385382.0,
      "step": 9320
    },
    {
      "entropy": 5.861846494674682,
      "epoch": 0.8234722712822324,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004939195929135506,
      "loss": 5.516,
      "mean_token_accuracy": 0.1856905475258827,
      "num_tokens": 17393229.0,
      "step": 9325
    },
    {
      "entropy": 5.8488586902618405,
      "epoch": 0.8239138113740727,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939123191790612,
      "loss": 5.5046,
      "mean_token_accuracy": 0.18935805708169937,
      "num_tokens": 17402439.0,
      "step": 9330
    },
    {
      "entropy": 5.842728233337402,
      "epoch": 0.8243553514659131,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004939050411562041,
      "loss": 5.6861,
      "mean_token_accuracy": 0.17682990729808806,
      "num_tokens": 17412349.0,
      "step": 9335
    },
    {
      "entropy": 5.7745805263519285,
      "epoch": 0.8247968915577535,
      "grad_norm": 1.0,
      "learning_rate": 0.0004938977588451216,
      "loss": 5.4547,
      "mean_token_accuracy": 0.20448464900255203,
      "num_tokens": 17420842.0,
      "step": 9340
    },
    {
      "entropy": 5.804004430770874,
      "epoch": 0.8252384316495938,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004938904722459565,
      "loss": 5.5068,
      "mean_token_accuracy": 0.20089084506034852,
      "num_tokens": 17429877.0,
      "step": 9345
    },
    {
      "entropy": 5.835537624359131,
      "epoch": 0.8256799717414341,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004938831813588515,
      "loss": 5.6559,
      "mean_token_accuracy": 0.17827803045511245,
      "num_tokens": 17439508.0,
      "step": 9350
    },
    {
      "entropy": 5.783529901504517,
      "epoch": 0.8261215118332744,
      "grad_norm": 1.0,
      "learning_rate": 0.0004938758861839495,
      "loss": 5.5674,
      "mean_token_accuracy": 0.19206740260124205,
      "num_tokens": 17449004.0,
      "step": 9355
    },
    {
      "entropy": 5.813489484786987,
      "epoch": 0.8265630519251148,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004938685867213934,
      "loss": 5.5891,
      "mean_token_accuracy": 0.18597745895385742,
      "num_tokens": 17458462.0,
      "step": 9360
    },
    {
      "entropy": 5.885543584823608,
      "epoch": 0.8270045920169551,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004938612829713262,
      "loss": 5.6062,
      "mean_token_accuracy": 0.1844778299331665,
      "num_tokens": 17466808.0,
      "step": 9365
    },
    {
      "entropy": 5.891280174255371,
      "epoch": 0.8274461321087955,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004938539749338909,
      "loss": 5.6734,
      "mean_token_accuracy": 0.18365688621997833,
      "num_tokens": 17476441.0,
      "step": 9370
    },
    {
      "entropy": 5.8662230491638185,
      "epoch": 0.8278876722006359,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004938466626092308,
      "loss": 5.5768,
      "mean_token_accuracy": 0.1952987253665924,
      "num_tokens": 17485393.0,
      "step": 9375
    },
    {
      "entropy": 5.807134485244751,
      "epoch": 0.8283292122924761,
      "grad_norm": 1.015625,
      "learning_rate": 0.000493839345997489,
      "loss": 5.6094,
      "mean_token_accuracy": 0.19124255627393721,
      "num_tokens": 17494904.0,
      "step": 9380
    },
    {
      "entropy": 5.896303987503051,
      "epoch": 0.8287707523843165,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004938320250988087,
      "loss": 5.6033,
      "mean_token_accuracy": 0.18640150576829911,
      "num_tokens": 17505137.0,
      "step": 9385
    },
    {
      "entropy": 5.881055641174316,
      "epoch": 0.8292122924761568,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004938246999133337,
      "loss": 5.6649,
      "mean_token_accuracy": 0.18294226974248887,
      "num_tokens": 17514039.0,
      "step": 9390
    },
    {
      "entropy": 5.825823831558227,
      "epoch": 0.8296538325679972,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004938173704412071,
      "loss": 5.5456,
      "mean_token_accuracy": 0.18992763757705688,
      "num_tokens": 17522423.0,
      "step": 9395
    },
    {
      "entropy": 5.826774454116821,
      "epoch": 0.8300953726598375,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004938100366825728,
      "loss": 5.5187,
      "mean_token_accuracy": 0.19853198379278184,
      "num_tokens": 17531958.0,
      "step": 9400
    },
    {
      "entropy": 5.802187108993531,
      "epoch": 0.8305369127516778,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004938026986375742,
      "loss": 5.5884,
      "mean_token_accuracy": 0.19164701849222182,
      "num_tokens": 17540905.0,
      "step": 9405
    },
    {
      "entropy": 5.846265411376953,
      "epoch": 0.8309784528435182,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004937953563063553,
      "loss": 5.6312,
      "mean_token_accuracy": 0.18226529955863952,
      "num_tokens": 17551007.0,
      "step": 9410
    },
    {
      "entropy": 5.79426941871643,
      "epoch": 0.8314199929353585,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004937880096890598,
      "loss": 5.4573,
      "mean_token_accuracy": 0.20695766806602478,
      "num_tokens": 17559403.0,
      "step": 9415
    },
    {
      "entropy": 5.8758955001831055,
      "epoch": 0.8318615330271989,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004937806587858315,
      "loss": 5.6001,
      "mean_token_accuracy": 0.18914103060960769,
      "num_tokens": 17569191.0,
      "step": 9420
    },
    {
      "entropy": 5.900097465515136,
      "epoch": 0.8323030731190392,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004937733035968147,
      "loss": 5.7011,
      "mean_token_accuracy": 0.17884953022003175,
      "num_tokens": 17578538.0,
      "step": 9425
    },
    {
      "entropy": 5.870207452774048,
      "epoch": 0.8327446132108796,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004937659441221531,
      "loss": 5.5631,
      "mean_token_accuracy": 0.1875049278140068,
      "num_tokens": 17587344.0,
      "step": 9430
    },
    {
      "entropy": 5.817720985412597,
      "epoch": 0.8331861533027198,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004937585803619912,
      "loss": 5.5161,
      "mean_token_accuracy": 0.1912759780883789,
      "num_tokens": 17596033.0,
      "step": 9435
    },
    {
      "entropy": 5.836237287521362,
      "epoch": 0.8336276933945602,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004937512123164731,
      "loss": 5.4889,
      "mean_token_accuracy": 0.20142687410116195,
      "num_tokens": 17605860.0,
      "step": 9440
    },
    {
      "entropy": 5.844070720672607,
      "epoch": 0.8340692334864006,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004937438399857431,
      "loss": 5.6416,
      "mean_token_accuracy": 0.18426611870527268,
      "num_tokens": 17615819.0,
      "step": 9445
    },
    {
      "entropy": 5.843596935272217,
      "epoch": 0.8345107735782409,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004937364633699459,
      "loss": 5.6088,
      "mean_token_accuracy": 0.19166117459535598,
      "num_tokens": 17625478.0,
      "step": 9450
    },
    {
      "entropy": 5.911627244949341,
      "epoch": 0.8349523136700813,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004937290824692255,
      "loss": 5.5726,
      "mean_token_accuracy": 0.19171651750802993,
      "num_tokens": 17635329.0,
      "step": 9455
    },
    {
      "entropy": 5.814493846893311,
      "epoch": 0.8353938537619215,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004937216972837269,
      "loss": 5.5565,
      "mean_token_accuracy": 0.1882588431239128,
      "num_tokens": 17644129.0,
      "step": 9460
    },
    {
      "entropy": 5.733139610290527,
      "epoch": 0.8358353938537619,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004937143078135946,
      "loss": 5.4978,
      "mean_token_accuracy": 0.19633229672908784,
      "num_tokens": 17652836.0,
      "step": 9465
    },
    {
      "entropy": 5.855417060852051,
      "epoch": 0.8362769339456022,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004937069140589735,
      "loss": 5.5625,
      "mean_token_accuracy": 0.18842038065195083,
      "num_tokens": 17661222.0,
      "step": 9470
    },
    {
      "entropy": 5.822893047332764,
      "epoch": 0.8367184740374426,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004936995160200083,
      "loss": 5.4958,
      "mean_token_accuracy": 0.20001437067985534,
      "num_tokens": 17669637.0,
      "step": 9475
    },
    {
      "entropy": 5.933269119262695,
      "epoch": 0.837160014129283,
      "grad_norm": 1.0,
      "learning_rate": 0.0004936921136968439,
      "loss": 5.7061,
      "mean_token_accuracy": 0.1770281821489334,
      "num_tokens": 17678589.0,
      "step": 9480
    },
    {
      "entropy": 5.864948320388794,
      "epoch": 0.8376015542211233,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004936847070896254,
      "loss": 5.6684,
      "mean_token_accuracy": 0.1817604660987854,
      "num_tokens": 17689129.0,
      "step": 9485
    },
    {
      "entropy": 5.9451494216918945,
      "epoch": 0.8380430943129636,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004936772961984979,
      "loss": 5.6541,
      "mean_token_accuracy": 0.18763858824968338,
      "num_tokens": 17698864.0,
      "step": 9490
    },
    {
      "entropy": 5.964946937561035,
      "epoch": 0.8384846344048039,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004936698810236065,
      "loss": 5.7361,
      "mean_token_accuracy": 0.17936290800571442,
      "num_tokens": 17707900.0,
      "step": 9495
    },
    {
      "entropy": 5.836407423019409,
      "epoch": 0.8389261744966443,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004936624615650964,
      "loss": 5.5412,
      "mean_token_accuracy": 0.19406685680150987,
      "num_tokens": 17716983.0,
      "step": 9500
    },
    {
      "entropy": 5.838777542114258,
      "epoch": 0.8393677145884846,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004936550378231133,
      "loss": 5.5538,
      "mean_token_accuracy": 0.18904087096452712,
      "num_tokens": 17726503.0,
      "step": 9505
    },
    {
      "entropy": 5.873986625671387,
      "epoch": 0.839809254680325,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004936476097978023,
      "loss": 5.6409,
      "mean_token_accuracy": 0.18771174252033235,
      "num_tokens": 17735708.0,
      "step": 9510
    },
    {
      "entropy": 5.859618520736694,
      "epoch": 0.8402507947721654,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004936401774893088,
      "loss": 5.5591,
      "mean_token_accuracy": 0.18589181452989578,
      "num_tokens": 17745169.0,
      "step": 9515
    },
    {
      "entropy": 5.937458419799805,
      "epoch": 0.8406923348640056,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004936327408977786,
      "loss": 5.7008,
      "mean_token_accuracy": 0.17849522083997726,
      "num_tokens": 17755630.0,
      "step": 9520
    },
    {
      "entropy": 5.868137788772583,
      "epoch": 0.841133874955846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004936253000233574,
      "loss": 5.5731,
      "mean_token_accuracy": 0.18807740211486818,
      "num_tokens": 17764469.0,
      "step": 9525
    },
    {
      "entropy": 5.833565092086792,
      "epoch": 0.8415754150476863,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004936178548661909,
      "loss": 5.6655,
      "mean_token_accuracy": 0.18485989272594452,
      "num_tokens": 17774977.0,
      "step": 9530
    },
    {
      "entropy": 5.835967254638672,
      "epoch": 0.8420169551395267,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000493610405426425,
      "loss": 5.582,
      "mean_token_accuracy": 0.1799257680773735,
      "num_tokens": 17784250.0,
      "step": 9535
    },
    {
      "entropy": 5.891266679763794,
      "epoch": 0.842458495231367,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004936029517042056,
      "loss": 5.6253,
      "mean_token_accuracy": 0.19154638797044754,
      "num_tokens": 17793858.0,
      "step": 9540
    },
    {
      "entropy": 5.947404050827027,
      "epoch": 0.8429000353232073,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004935954936996787,
      "loss": 5.7057,
      "mean_token_accuracy": 0.1791315421462059,
      "num_tokens": 17804330.0,
      "step": 9545
    },
    {
      "entropy": 5.87277250289917,
      "epoch": 0.8433415754150477,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004935880314129903,
      "loss": 5.6006,
      "mean_token_accuracy": 0.1922000303864479,
      "num_tokens": 17813750.0,
      "step": 9550
    },
    {
      "entropy": 5.731779670715332,
      "epoch": 0.843783115506888,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004935805648442868,
      "loss": 5.5664,
      "mean_token_accuracy": 0.19010436683893203,
      "num_tokens": 17823537.0,
      "step": 9555
    },
    {
      "entropy": 5.836687755584717,
      "epoch": 0.8442246555987284,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004935730939937143,
      "loss": 5.5825,
      "mean_token_accuracy": 0.19426935017108918,
      "num_tokens": 17833298.0,
      "step": 9560
    },
    {
      "entropy": 5.900979709625244,
      "epoch": 0.8446661956905687,
      "grad_norm": 1.0,
      "learning_rate": 0.0004935656188614192,
      "loss": 5.4836,
      "mean_token_accuracy": 0.19837422668933868,
      "num_tokens": 17842339.0,
      "step": 9565
    },
    {
      "entropy": 5.874470138549805,
      "epoch": 0.845107735782409,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004935581394475479,
      "loss": 5.7009,
      "mean_token_accuracy": 0.18750401884317397,
      "num_tokens": 17851827.0,
      "step": 9570
    },
    {
      "entropy": 5.7602321147918705,
      "epoch": 0.8455492758742493,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004935506557522469,
      "loss": 5.5391,
      "mean_token_accuracy": 0.19153136610984803,
      "num_tokens": 17860788.0,
      "step": 9575
    },
    {
      "entropy": 5.866920423507691,
      "epoch": 0.8459908159660897,
      "grad_norm": 0.921875,
      "learning_rate": 0.000493543167775663,
      "loss": 5.5448,
      "mean_token_accuracy": 0.19585472494363784,
      "num_tokens": 17869849.0,
      "step": 9580
    },
    {
      "entropy": 5.86996488571167,
      "epoch": 0.8464323560579301,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004935356755179426,
      "loss": 5.5749,
      "mean_token_accuracy": 0.18391755521297454,
      "num_tokens": 17879431.0,
      "step": 9585
    },
    {
      "entropy": 5.9047904968261715,
      "epoch": 0.8468738961497704,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004935281789792326,
      "loss": 5.6799,
      "mean_token_accuracy": 0.1816534087061882,
      "num_tokens": 17888599.0,
      "step": 9590
    },
    {
      "entropy": 5.820351839065552,
      "epoch": 0.8473154362416108,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004935206781596799,
      "loss": 5.5834,
      "mean_token_accuracy": 0.1976203590631485,
      "num_tokens": 17898662.0,
      "step": 9595
    },
    {
      "entropy": 5.9325127601623535,
      "epoch": 0.847756976333451,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004935131730594315,
      "loss": 5.622,
      "mean_token_accuracy": 0.1850518763065338,
      "num_tokens": 17907810.0,
      "step": 9600
    },
    {
      "entropy": 5.853305768966675,
      "epoch": 0.8481985164252914,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004935056636786342,
      "loss": 5.5828,
      "mean_token_accuracy": 0.18844246566295625,
      "num_tokens": 17917611.0,
      "step": 9605
    },
    {
      "entropy": 5.834033870697022,
      "epoch": 0.8486400565171317,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004934981500174353,
      "loss": 5.6101,
      "mean_token_accuracy": 0.18328014016151428,
      "num_tokens": 17926076.0,
      "step": 9610
    },
    {
      "entropy": 5.88135986328125,
      "epoch": 0.8490815966089721,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004934906320759818,
      "loss": 5.628,
      "mean_token_accuracy": 0.19137745052576066,
      "num_tokens": 17935521.0,
      "step": 9615
    },
    {
      "entropy": 5.8876971244812015,
      "epoch": 0.8495231367008125,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004934831098544211,
      "loss": 5.5865,
      "mean_token_accuracy": 0.19253529310226442,
      "num_tokens": 17944849.0,
      "step": 9620
    },
    {
      "entropy": 5.886591815948487,
      "epoch": 0.8499646767926528,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004934755833529007,
      "loss": 5.5721,
      "mean_token_accuracy": 0.18591910004615783,
      "num_tokens": 17953988.0,
      "step": 9625
    },
    {
      "entropy": 5.807942008972168,
      "epoch": 0.8504062168844931,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004934680525715677,
      "loss": 5.5739,
      "mean_token_accuracy": 0.1885495513677597,
      "num_tokens": 17963319.0,
      "step": 9630
    },
    {
      "entropy": 5.822053098678589,
      "epoch": 0.8508477569763334,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004934605175105698,
      "loss": 5.5384,
      "mean_token_accuracy": 0.20306545943021775,
      "num_tokens": 17972054.0,
      "step": 9635
    },
    {
      "entropy": 5.9353546619415285,
      "epoch": 0.8512892970681738,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004934529781700546,
      "loss": 5.687,
      "mean_token_accuracy": 0.18286622017621995,
      "num_tokens": 17981630.0,
      "step": 9640
    },
    {
      "entropy": 5.850656509399414,
      "epoch": 0.8517308371600141,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004934454345501698,
      "loss": 5.6222,
      "mean_token_accuracy": 0.18302069157361983,
      "num_tokens": 17990987.0,
      "step": 9645
    },
    {
      "entropy": 5.742798328399658,
      "epoch": 0.8521723772518545,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004934378866510633,
      "loss": 5.5217,
      "mean_token_accuracy": 0.19494544416666032,
      "num_tokens": 17999284.0,
      "step": 9650
    },
    {
      "entropy": 5.88429822921753,
      "epoch": 0.8526139173436948,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004934303344728828,
      "loss": 5.5357,
      "mean_token_accuracy": 0.18915798515081406,
      "num_tokens": 18008670.0,
      "step": 9655
    },
    {
      "entropy": 5.821765184402466,
      "epoch": 0.8530554574355351,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004934227780157761,
      "loss": 5.5523,
      "mean_token_accuracy": 0.18990737497806548,
      "num_tokens": 18018374.0,
      "step": 9660
    },
    {
      "entropy": 5.900697135925293,
      "epoch": 0.8534969975273755,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004934152172798916,
      "loss": 5.6485,
      "mean_token_accuracy": 0.187437105178833,
      "num_tokens": 18026943.0,
      "step": 9665
    },
    {
      "entropy": 5.905522871017456,
      "epoch": 0.8539385376192158,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004934076522653772,
      "loss": 5.5986,
      "mean_token_accuracy": 0.18579338490962982,
      "num_tokens": 18036270.0,
      "step": 9670
    },
    {
      "entropy": 5.760067796707153,
      "epoch": 0.8543800777110562,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000493400082972381,
      "loss": 5.5694,
      "mean_token_accuracy": 0.18511977046728134,
      "num_tokens": 18045611.0,
      "step": 9675
    },
    {
      "entropy": 5.7979716777801515,
      "epoch": 0.8548216178028964,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004933925094010516,
      "loss": 5.6022,
      "mean_token_accuracy": 0.18361486494541168,
      "num_tokens": 18055152.0,
      "step": 9680
    },
    {
      "entropy": 5.896060514450073,
      "epoch": 0.8552631578947368,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004933849315515369,
      "loss": 5.5878,
      "mean_token_accuracy": 0.18785662204027176,
      "num_tokens": 18064376.0,
      "step": 9685
    },
    {
      "entropy": 5.780475234985351,
      "epoch": 0.8557046979865772,
      "grad_norm": 0.875,
      "learning_rate": 0.0004933773494239857,
      "loss": 5.5042,
      "mean_token_accuracy": 0.20165061950683594,
      "num_tokens": 18073569.0,
      "step": 9690
    },
    {
      "entropy": 5.783981561660767,
      "epoch": 0.8561462380784175,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004933697630185464,
      "loss": 5.5754,
      "mean_token_accuracy": 0.19270028918981552,
      "num_tokens": 18083264.0,
      "step": 9695
    },
    {
      "entropy": 5.887913942337036,
      "epoch": 0.8565877781702579,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004933621723353677,
      "loss": 5.5802,
      "mean_token_accuracy": 0.18706991970539094,
      "num_tokens": 18093151.0,
      "step": 9700
    },
    {
      "entropy": 5.7990175724029545,
      "epoch": 0.8570293182620982,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000493354577374598,
      "loss": 5.4653,
      "mean_token_accuracy": 0.20603102296590806,
      "num_tokens": 18102087.0,
      "step": 9705
    },
    {
      "entropy": 5.854032421112061,
      "epoch": 0.8574708583539385,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004933469781363865,
      "loss": 5.6354,
      "mean_token_accuracy": 0.18712162971496582,
      "num_tokens": 18111854.0,
      "step": 9710
    },
    {
      "entropy": 5.869991159439087,
      "epoch": 0.8579123984457788,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004933393746208819,
      "loss": 5.6492,
      "mean_token_accuracy": 0.17795441001653672,
      "num_tokens": 18120520.0,
      "step": 9715
    },
    {
      "entropy": 5.8892655849456785,
      "epoch": 0.8583539385376192,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0004933317668282331,
      "loss": 5.7079,
      "mean_token_accuracy": 0.17710036486387254,
      "num_tokens": 18130557.0,
      "step": 9720
    },
    {
      "entropy": 5.832837152481079,
      "epoch": 0.8587954786294596,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004933241547585891,
      "loss": 5.4832,
      "mean_token_accuracy": 0.20000069737434387,
      "num_tokens": 18139199.0,
      "step": 9725
    },
    {
      "entropy": 5.786106538772583,
      "epoch": 0.8592370187212999,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004933165384120991,
      "loss": 5.5505,
      "mean_token_accuracy": 0.19100535064935684,
      "num_tokens": 18150039.0,
      "step": 9730
    },
    {
      "entropy": 5.812739372253418,
      "epoch": 0.8596785588131403,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004933089177889124,
      "loss": 5.5447,
      "mean_token_accuracy": 0.18598742187023162,
      "num_tokens": 18159451.0,
      "step": 9735
    },
    {
      "entropy": 5.868243551254272,
      "epoch": 0.8601200989049805,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004933012928891781,
      "loss": 5.6304,
      "mean_token_accuracy": 0.1877829000353813,
      "num_tokens": 18169896.0,
      "step": 9740
    },
    {
      "entropy": 5.8395103931427,
      "epoch": 0.8605616389968209,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004932936637130456,
      "loss": 5.5366,
      "mean_token_accuracy": 0.199994458258152,
      "num_tokens": 18177764.0,
      "step": 9745
    },
    {
      "entropy": 5.842320156097412,
      "epoch": 0.8610031790886612,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004932860302606644,
      "loss": 5.5981,
      "mean_token_accuracy": 0.19198889136314393,
      "num_tokens": 18186940.0,
      "step": 9750
    },
    {
      "entropy": 5.880097675323486,
      "epoch": 0.8614447191805016,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004932783925321838,
      "loss": 5.4928,
      "mean_token_accuracy": 0.19632400572299957,
      "num_tokens": 18195374.0,
      "step": 9755
    },
    {
      "entropy": 5.78803334236145,
      "epoch": 0.861886259272342,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000493270750527754,
      "loss": 5.5754,
      "mean_token_accuracy": 0.19211004823446273,
      "num_tokens": 18204779.0,
      "step": 9760
    },
    {
      "entropy": 5.795253658294678,
      "epoch": 0.8623277993641822,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004932631042475241,
      "loss": 5.5211,
      "mean_token_accuracy": 0.1943109005689621,
      "num_tokens": 18214099.0,
      "step": 9765
    },
    {
      "entropy": 5.85060133934021,
      "epoch": 0.8627693394560226,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004932554536916441,
      "loss": 5.5957,
      "mean_token_accuracy": 0.1854076012969017,
      "num_tokens": 18223190.0,
      "step": 9770
    },
    {
      "entropy": 5.875854444503784,
      "epoch": 0.8632108795478629,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000493247798860264,
      "loss": 5.5688,
      "mean_token_accuracy": 0.19499116837978364,
      "num_tokens": 18232380.0,
      "step": 9775
    },
    {
      "entropy": 5.912772750854492,
      "epoch": 0.8636524196397033,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004932401397535337,
      "loss": 5.6423,
      "mean_token_accuracy": 0.1843586429953575,
      "num_tokens": 18242621.0,
      "step": 9780
    },
    {
      "entropy": 5.894945049285889,
      "epoch": 0.8640939597315436,
      "grad_norm": 0.875,
      "learning_rate": 0.000493232476371603,
      "loss": 5.6958,
      "mean_token_accuracy": 0.18013483583927153,
      "num_tokens": 18253183.0,
      "step": 9785
    },
    {
      "entropy": 5.782314443588257,
      "epoch": 0.864535499823384,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004932248087146223,
      "loss": 5.5122,
      "mean_token_accuracy": 0.19055208265781404,
      "num_tokens": 18261967.0,
      "step": 9790
    },
    {
      "entropy": 5.764515495300293,
      "epoch": 0.8649770399152243,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004932171367827417,
      "loss": 5.5869,
      "mean_token_accuracy": 0.19173365384340285,
      "num_tokens": 18270902.0,
      "step": 9795
    },
    {
      "entropy": 5.841852235794067,
      "epoch": 0.8654185800070646,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004932094605761115,
      "loss": 5.6048,
      "mean_token_accuracy": 0.19432415813207626,
      "num_tokens": 18279387.0,
      "step": 9800
    },
    {
      "entropy": 5.799977827072143,
      "epoch": 0.865860120098905,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004932017800948821,
      "loss": 5.3995,
      "mean_token_accuracy": 0.20187714397907258,
      "num_tokens": 18288363.0,
      "step": 9805
    },
    {
      "entropy": 5.957182359695435,
      "epoch": 0.8663016601907453,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004931940953392038,
      "loss": 5.7171,
      "mean_token_accuracy": 0.17543497085571289,
      "num_tokens": 18297398.0,
      "step": 9810
    },
    {
      "entropy": 5.918104887008667,
      "epoch": 0.8667432002825857,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004931864063092274,
      "loss": 5.569,
      "mean_token_accuracy": 0.19007169157266618,
      "num_tokens": 18306230.0,
      "step": 9815
    },
    {
      "entropy": 5.856356287002564,
      "epoch": 0.8671847403744259,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004931787130051034,
      "loss": 5.6938,
      "mean_token_accuracy": 0.18917493373155594,
      "num_tokens": 18316395.0,
      "step": 9820
    },
    {
      "entropy": 5.824563980102539,
      "epoch": 0.8676262804662663,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004931710154269824,
      "loss": 5.5723,
      "mean_token_accuracy": 0.19153814613819123,
      "num_tokens": 18326020.0,
      "step": 9825
    },
    {
      "entropy": 5.7984528064727785,
      "epoch": 0.8680678205581067,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004931633135750154,
      "loss": 5.5643,
      "mean_token_accuracy": 0.19000247418880462,
      "num_tokens": 18335320.0,
      "step": 9830
    },
    {
      "entropy": 5.848925161361694,
      "epoch": 0.868509360649947,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004931556074493531,
      "loss": 5.6002,
      "mean_token_accuracy": 0.1890251263976097,
      "num_tokens": 18345137.0,
      "step": 9835
    },
    {
      "entropy": 5.887432813644409,
      "epoch": 0.8689509007417874,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004931478970501465,
      "loss": 5.6235,
      "mean_token_accuracy": 0.18818622529506684,
      "num_tokens": 18354289.0,
      "step": 9840
    },
    {
      "entropy": 5.948161697387695,
      "epoch": 0.8693924408336277,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004931401823775466,
      "loss": 5.7096,
      "mean_token_accuracy": 0.1750674292445183,
      "num_tokens": 18363701.0,
      "step": 9845
    },
    {
      "entropy": 5.916717147827148,
      "epoch": 0.869833980925468,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004931324634317047,
      "loss": 5.5695,
      "mean_token_accuracy": 0.18598162978887559,
      "num_tokens": 18373336.0,
      "step": 9850
    },
    {
      "entropy": 5.889935493469238,
      "epoch": 0.8702755210173083,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004931247402127719,
      "loss": 5.5954,
      "mean_token_accuracy": 0.1866972878575325,
      "num_tokens": 18383233.0,
      "step": 9855
    },
    {
      "entropy": 5.881876516342163,
      "epoch": 0.8707170611091487,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004931170127208995,
      "loss": 5.6465,
      "mean_token_accuracy": 0.18168477565050126,
      "num_tokens": 18393214.0,
      "step": 9860
    },
    {
      "entropy": 5.842750072479248,
      "epoch": 0.8711586012009891,
      "grad_norm": 1.0,
      "learning_rate": 0.0004931092809562388,
      "loss": 5.5281,
      "mean_token_accuracy": 0.188180635869503,
      "num_tokens": 18401773.0,
      "step": 9865
    },
    {
      "entropy": 5.790875864028931,
      "epoch": 0.8716001412928294,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004931015449189414,
      "loss": 5.592,
      "mean_token_accuracy": 0.18443243950605392,
      "num_tokens": 18411748.0,
      "step": 9870
    },
    {
      "entropy": 5.86829023361206,
      "epoch": 0.8720416813846698,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004930938046091587,
      "loss": 5.6401,
      "mean_token_accuracy": 0.185184346139431,
      "num_tokens": 18421399.0,
      "step": 9875
    },
    {
      "entropy": 5.869137191772461,
      "epoch": 0.87248322147651,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004930860600270425,
      "loss": 5.626,
      "mean_token_accuracy": 0.1813107594847679,
      "num_tokens": 18430052.0,
      "step": 9880
    },
    {
      "entropy": 5.8621752738952635,
      "epoch": 0.8729247615683504,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004930783111727443,
      "loss": 5.4581,
      "mean_token_accuracy": 0.20144166499376298,
      "num_tokens": 18438268.0,
      "step": 9885
    },
    {
      "entropy": 5.791285467147827,
      "epoch": 0.8733663016601907,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000493070558046416,
      "loss": 5.5989,
      "mean_token_accuracy": 0.185325887799263,
      "num_tokens": 18446964.0,
      "step": 9890
    },
    {
      "entropy": 5.905385112762451,
      "epoch": 0.8738078417520311,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004930628006482097,
      "loss": 5.5936,
      "mean_token_accuracy": 0.17827080488204955,
      "num_tokens": 18455701.0,
      "step": 9895
    },
    {
      "entropy": 5.879530239105224,
      "epoch": 0.8742493818438715,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004930550389782769,
      "loss": 5.5543,
      "mean_token_accuracy": 0.19134498983621598,
      "num_tokens": 18465635.0,
      "step": 9900
    },
    {
      "entropy": 5.808335256576538,
      "epoch": 0.8746909219357117,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004930472730367701,
      "loss": 5.4903,
      "mean_token_accuracy": 0.1899092212319374,
      "num_tokens": 18474833.0,
      "step": 9905
    },
    {
      "entropy": 5.82249641418457,
      "epoch": 0.8751324620275521,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004930395028238412,
      "loss": 5.5671,
      "mean_token_accuracy": 0.19031829833984376,
      "num_tokens": 18482987.0,
      "step": 9910
    },
    {
      "entropy": 5.788164138793945,
      "epoch": 0.8755740021193924,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004930317283396423,
      "loss": 5.4798,
      "mean_token_accuracy": 0.1948615252971649,
      "num_tokens": 18491274.0,
      "step": 9915
    },
    {
      "entropy": 5.836555004119873,
      "epoch": 0.8760155422112328,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000493023949584326,
      "loss": 5.6372,
      "mean_token_accuracy": 0.1819130912423134,
      "num_tokens": 18500811.0,
      "step": 9920
    },
    {
      "entropy": 5.72904372215271,
      "epoch": 0.8764570823030731,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004930161665580445,
      "loss": 5.4097,
      "mean_token_accuracy": 0.20912078022956848,
      "num_tokens": 18509475.0,
      "step": 9925
    },
    {
      "entropy": 5.906229877471924,
      "epoch": 0.8768986223949135,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004930083792609502,
      "loss": 5.5801,
      "mean_token_accuracy": 0.1860847517848015,
      "num_tokens": 18519410.0,
      "step": 9930
    },
    {
      "entropy": 5.686052227020264,
      "epoch": 0.8773401624867538,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004930005876931958,
      "loss": 5.4132,
      "mean_token_accuracy": 0.20658794939517974,
      "num_tokens": 18529194.0,
      "step": 9935
    },
    {
      "entropy": 5.85654730796814,
      "epoch": 0.8777817025785941,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004929927918549339,
      "loss": 5.6348,
      "mean_token_accuracy": 0.18094472736120223,
      "num_tokens": 18539166.0,
      "step": 9940
    },
    {
      "entropy": 5.871091413497925,
      "epoch": 0.8782232426704345,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004929849917463171,
      "loss": 5.5506,
      "mean_token_accuracy": 0.19216870218515397,
      "num_tokens": 18548166.0,
      "step": 9945
    },
    {
      "entropy": 5.802586507797241,
      "epoch": 0.8786647827622748,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004929771873674984,
      "loss": 5.5709,
      "mean_token_accuracy": 0.18824805915355683,
      "num_tokens": 18557041.0,
      "step": 9950
    },
    {
      "entropy": 5.86892409324646,
      "epoch": 0.8791063228541152,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004929693787186305,
      "loss": 5.5759,
      "mean_token_accuracy": 0.1842259109020233,
      "num_tokens": 18566461.0,
      "step": 9955
    },
    {
      "entropy": 5.824876546859741,
      "epoch": 0.8795478629459554,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004929615657998664,
      "loss": 5.5174,
      "mean_token_accuracy": 0.19374087154865266,
      "num_tokens": 18576334.0,
      "step": 9960
    },
    {
      "entropy": 5.834353828430176,
      "epoch": 0.8799894030377958,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004929537486113593,
      "loss": 5.5218,
      "mean_token_accuracy": 0.19154039919376373,
      "num_tokens": 18584882.0,
      "step": 9965
    },
    {
      "entropy": 5.880415534973144,
      "epoch": 0.8804309431296362,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004929459271532621,
      "loss": 5.5657,
      "mean_token_accuracy": 0.19408414959907533,
      "num_tokens": 18593669.0,
      "step": 9970
    },
    {
      "entropy": 5.8575982570648195,
      "epoch": 0.8808724832214765,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004929381014257282,
      "loss": 5.6155,
      "mean_token_accuracy": 0.19106159508228301,
      "num_tokens": 18603357.0,
      "step": 9975
    },
    {
      "entropy": 5.810539817810058,
      "epoch": 0.8813140233133169,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004929302714289107,
      "loss": 5.5186,
      "mean_token_accuracy": 0.19460396021604537,
      "num_tokens": 18611856.0,
      "step": 9980
    },
    {
      "entropy": 5.78901777267456,
      "epoch": 0.8817555634051572,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004929224371629634,
      "loss": 5.5747,
      "mean_token_accuracy": 0.1909811556339264,
      "num_tokens": 18620500.0,
      "step": 9985
    },
    {
      "entropy": 5.798228359222412,
      "epoch": 0.8821971034969975,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004929145986280394,
      "loss": 5.5013,
      "mean_token_accuracy": 0.19727822244167328,
      "num_tokens": 18628836.0,
      "step": 9990
    },
    {
      "entropy": 5.819610834121704,
      "epoch": 0.8826386435888378,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004929067558242922,
      "loss": 5.4532,
      "mean_token_accuracy": 0.20147583782672882,
      "num_tokens": 18636831.0,
      "step": 9995
    },
    {
      "entropy": 5.819434881210327,
      "epoch": 0.8830801836806782,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004928989087518758,
      "loss": 5.5911,
      "mean_token_accuracy": 0.19338216483592988,
      "num_tokens": 18646327.0,
      "step": 10000
    },
    {
      "entropy": 5.841499662399292,
      "epoch": 0.8835217237725186,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004928910574109435,
      "loss": 5.6318,
      "mean_token_accuracy": 0.18724457770586014,
      "num_tokens": 18656125.0,
      "step": 10005
    },
    {
      "entropy": 5.8710731029510494,
      "epoch": 0.8839632638643589,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004928832018016495,
      "loss": 5.6313,
      "mean_token_accuracy": 0.18780807703733443,
      "num_tokens": 18665853.0,
      "step": 10010
    },
    {
      "entropy": 5.9379126071929935,
      "epoch": 0.8844048039561992,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004928753419241472,
      "loss": 5.6719,
      "mean_token_accuracy": 0.18405981361865997,
      "num_tokens": 18676674.0,
      "step": 10015
    },
    {
      "entropy": 5.857097625732422,
      "epoch": 0.8848463440480395,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004928674777785912,
      "loss": 5.4977,
      "mean_token_accuracy": 0.19990508407354354,
      "num_tokens": 18685535.0,
      "step": 10020
    },
    {
      "entropy": 5.807556390762329,
      "epoch": 0.8852878841398799,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004928596093651351,
      "loss": 5.5442,
      "mean_token_accuracy": 0.18528246730566025,
      "num_tokens": 18694186.0,
      "step": 10025
    },
    {
      "entropy": 5.787950134277343,
      "epoch": 0.8857294242317202,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000492851736683933,
      "loss": 5.5218,
      "mean_token_accuracy": 0.19080675393342972,
      "num_tokens": 18703061.0,
      "step": 10030
    },
    {
      "entropy": 5.785179281234742,
      "epoch": 0.8861709643235606,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004928438597351395,
      "loss": 5.4461,
      "mean_token_accuracy": 0.19986527413129807,
      "num_tokens": 18712466.0,
      "step": 10035
    },
    {
      "entropy": 5.904344892501831,
      "epoch": 0.886612504415401,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004928359785189086,
      "loss": 5.5937,
      "mean_token_accuracy": 0.1826832503080368,
      "num_tokens": 18721522.0,
      "step": 10040
    },
    {
      "entropy": 5.783854866027832,
      "epoch": 0.8870540445072412,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004928280930353948,
      "loss": 5.5113,
      "mean_token_accuracy": 0.18971350640058518,
      "num_tokens": 18730488.0,
      "step": 10045
    },
    {
      "entropy": 5.807558202743531,
      "epoch": 0.8874955845990816,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004928202032847525,
      "loss": 5.4634,
      "mean_token_accuracy": 0.19593009650707244,
      "num_tokens": 18740318.0,
      "step": 10050
    },
    {
      "entropy": 5.844618654251098,
      "epoch": 0.8879371246909219,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004928123092671362,
      "loss": 5.5804,
      "mean_token_accuracy": 0.18753331601619722,
      "num_tokens": 18748731.0,
      "step": 10055
    },
    {
      "entropy": 5.775624227523804,
      "epoch": 0.8883786647827623,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004928044109827008,
      "loss": 5.5029,
      "mean_token_accuracy": 0.19979605823755264,
      "num_tokens": 18757847.0,
      "step": 10060
    },
    {
      "entropy": 5.871121978759765,
      "epoch": 0.8888202048746026,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004927965084316007,
      "loss": 5.5828,
      "mean_token_accuracy": 0.18818929344415664,
      "num_tokens": 18767429.0,
      "step": 10065
    },
    {
      "entropy": 5.83223180770874,
      "epoch": 0.889261744966443,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000492788601613991,
      "loss": 5.5099,
      "mean_token_accuracy": 0.19484381079673768,
      "num_tokens": 18776652.0,
      "step": 10070
    },
    {
      "entropy": 5.853069925308228,
      "epoch": 0.8897032850582833,
      "grad_norm": 1.0,
      "learning_rate": 0.0004927806905300265,
      "loss": 5.5848,
      "mean_token_accuracy": 0.18585562855005264,
      "num_tokens": 18785687.0,
      "step": 10075
    },
    {
      "entropy": 5.8013709545135494,
      "epoch": 0.8901448251501236,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000492772775179862,
      "loss": 5.5981,
      "mean_token_accuracy": 0.19164907187223434,
      "num_tokens": 18794924.0,
      "step": 10080
    },
    {
      "entropy": 5.740886402130127,
      "epoch": 0.890586365241964,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004927648555636528,
      "loss": 5.4636,
      "mean_token_accuracy": 0.20413612872362136,
      "num_tokens": 18803359.0,
      "step": 10085
    },
    {
      "entropy": 5.848564958572387,
      "epoch": 0.8910279053338043,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004927569316815539,
      "loss": 5.6846,
      "mean_token_accuracy": 0.18487467914819716,
      "num_tokens": 18813496.0,
      "step": 10090
    },
    {
      "entropy": 5.871120262145996,
      "epoch": 0.8914694454256447,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004927490035337205,
      "loss": 5.5264,
      "mean_token_accuracy": 0.1974919855594635,
      "num_tokens": 18823274.0,
      "step": 10095
    },
    {
      "entropy": 5.836826467514038,
      "epoch": 0.8919109855174849,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000492741071120308,
      "loss": 5.5709,
      "mean_token_accuracy": 0.1899980917572975,
      "num_tokens": 18832436.0,
      "step": 10100
    },
    {
      "entropy": 5.861043167114258,
      "epoch": 0.8923525256093253,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004927331344414717,
      "loss": 5.6186,
      "mean_token_accuracy": 0.19028650522232055,
      "num_tokens": 18841923.0,
      "step": 10105
    },
    {
      "entropy": 5.8604639053344725,
      "epoch": 0.8927940657011657,
      "grad_norm": 1.0,
      "learning_rate": 0.0004927251934973672,
      "loss": 5.6378,
      "mean_token_accuracy": 0.19036390632390976,
      "num_tokens": 18850957.0,
      "step": 10110
    },
    {
      "entropy": 5.9429018020629885,
      "epoch": 0.893235605793006,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004927172482881499,
      "loss": 5.6574,
      "mean_token_accuracy": 0.18523878306150438,
      "num_tokens": 18859622.0,
      "step": 10115
    },
    {
      "entropy": 5.913838958740234,
      "epoch": 0.8936771458848464,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004927092988139755,
      "loss": 5.5954,
      "mean_token_accuracy": 0.18543685376644134,
      "num_tokens": 18868555.0,
      "step": 10120
    },
    {
      "entropy": 5.873710680007934,
      "epoch": 0.8941186859766866,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004927013450749999,
      "loss": 5.6223,
      "mean_token_accuracy": 0.18576771765947342,
      "num_tokens": 18877445.0,
      "step": 10125
    },
    {
      "entropy": 5.796116781234741,
      "epoch": 0.894560226068527,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004926933870713787,
      "loss": 5.4899,
      "mean_token_accuracy": 0.19211601465940475,
      "num_tokens": 18886386.0,
      "step": 10130
    },
    {
      "entropy": 5.885198497772217,
      "epoch": 0.8950017661603673,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004926854248032678,
      "loss": 5.5841,
      "mean_token_accuracy": 0.18588477224111558,
      "num_tokens": 18895016.0,
      "step": 10135
    },
    {
      "entropy": 5.77524151802063,
      "epoch": 0.8954433062522077,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004926774582708234,
      "loss": 5.4651,
      "mean_token_accuracy": 0.19851129949092866,
      "num_tokens": 18904126.0,
      "step": 10140
    },
    {
      "entropy": 5.775547122955322,
      "epoch": 0.8958848463440481,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004926694874742012,
      "loss": 5.5271,
      "mean_token_accuracy": 0.1910506159067154,
      "num_tokens": 18913064.0,
      "step": 10145
    },
    {
      "entropy": 5.818687582015992,
      "epoch": 0.8963263864358884,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004926615124135577,
      "loss": 5.6335,
      "mean_token_accuracy": 0.18564965426921845,
      "num_tokens": 18921189.0,
      "step": 10150
    },
    {
      "entropy": 5.848613119125366,
      "epoch": 0.8967679265277287,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004926535330890488,
      "loss": 5.5471,
      "mean_token_accuracy": 0.19497582465410232,
      "num_tokens": 18930818.0,
      "step": 10155
    },
    {
      "entropy": 5.931802415847779,
      "epoch": 0.897209466619569,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004926455495008311,
      "loss": 5.5703,
      "mean_token_accuracy": 0.18524253219366074,
      "num_tokens": 18940376.0,
      "step": 10160
    },
    {
      "entropy": 5.858433055877685,
      "epoch": 0.8976510067114094,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004926375616490608,
      "loss": 5.5997,
      "mean_token_accuracy": 0.19515008479356766,
      "num_tokens": 18949418.0,
      "step": 10165
    },
    {
      "entropy": 5.832569408416748,
      "epoch": 0.8980925468032497,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004926295695338945,
      "loss": 5.5395,
      "mean_token_accuracy": 0.19587756246328353,
      "num_tokens": 18959315.0,
      "step": 10170
    },
    {
      "entropy": 5.773690319061279,
      "epoch": 0.8985340868950901,
      "grad_norm": 0.81640625,
      "learning_rate": 0.0004926215731554887,
      "loss": 5.4795,
      "mean_token_accuracy": 0.19844437688589095,
      "num_tokens": 18968336.0,
      "step": 10175
    },
    {
      "entropy": 5.825376367568969,
      "epoch": 0.8989756269869305,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000492613572514,
      "loss": 5.5255,
      "mean_token_accuracy": 0.19697312414646148,
      "num_tokens": 18977123.0,
      "step": 10180
    },
    {
      "entropy": 5.857279682159424,
      "epoch": 0.8994171670787707,
      "grad_norm": 1.0,
      "learning_rate": 0.0004926055676095851,
      "loss": 5.6546,
      "mean_token_accuracy": 0.1828354775905609,
      "num_tokens": 18986150.0,
      "step": 10185
    },
    {
      "entropy": 5.920010948181153,
      "epoch": 0.8998587071706111,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004925975584424012,
      "loss": 5.6333,
      "mean_token_accuracy": 0.18982790857553483,
      "num_tokens": 18996010.0,
      "step": 10190
    },
    {
      "entropy": 5.860252571105957,
      "epoch": 0.9003002472624514,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004925895450126046,
      "loss": 5.6349,
      "mean_token_accuracy": 0.1805057168006897,
      "num_tokens": 19005100.0,
      "step": 10195
    },
    {
      "entropy": 5.828482389450073,
      "epoch": 0.9007417873542918,
      "grad_norm": 1.0,
      "learning_rate": 0.0004925815273203526,
      "loss": 5.5176,
      "mean_token_accuracy": 0.1949925735592842,
      "num_tokens": 19013844.0,
      "step": 10200
    },
    {
      "entropy": 5.925581741333008,
      "epoch": 0.901183327446132,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004925735053658022,
      "loss": 5.6748,
      "mean_token_accuracy": 0.18633326888084412,
      "num_tokens": 19024218.0,
      "step": 10205
    },
    {
      "entropy": 5.870132875442505,
      "epoch": 0.9016248675379724,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004925654791491106,
      "loss": 5.5969,
      "mean_token_accuracy": 0.18883185982704162,
      "num_tokens": 19034725.0,
      "step": 10210
    },
    {
      "entropy": 5.887659358978271,
      "epoch": 0.9020664076298128,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004925574486704351,
      "loss": 5.6185,
      "mean_token_accuracy": 0.18139611631631852,
      "num_tokens": 19044597.0,
      "step": 10215
    },
    {
      "entropy": 5.978512763977051,
      "epoch": 0.9025079477216531,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004925494139299329,
      "loss": 5.7472,
      "mean_token_accuracy": 0.17511816322803497,
      "num_tokens": 19055124.0,
      "step": 10220
    },
    {
      "entropy": 5.832484722137451,
      "epoch": 0.9029494878134935,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004925413749277613,
      "loss": 5.4823,
      "mean_token_accuracy": 0.1957518771290779,
      "num_tokens": 19064516.0,
      "step": 10225
    },
    {
      "entropy": 5.787093019485473,
      "epoch": 0.9033910279053338,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004925333316640779,
      "loss": 5.5421,
      "mean_token_accuracy": 0.19380549490451812,
      "num_tokens": 19072926.0,
      "step": 10230
    },
    {
      "entropy": 5.859374189376831,
      "epoch": 0.9038325679971742,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004925252841390403,
      "loss": 5.5995,
      "mean_token_accuracy": 0.18605128526687623,
      "num_tokens": 19082313.0,
      "step": 10235
    },
    {
      "entropy": 5.799882936477661,
      "epoch": 0.9042741080890144,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004925172323528061,
      "loss": 5.565,
      "mean_token_accuracy": 0.19741909354925155,
      "num_tokens": 19091891.0,
      "step": 10240
    },
    {
      "entropy": 5.858547878265381,
      "epoch": 0.9047156481808548,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000492509176305533,
      "loss": 5.5876,
      "mean_token_accuracy": 0.1897084191441536,
      "num_tokens": 19100311.0,
      "step": 10245
    },
    {
      "entropy": 5.900732564926147,
      "epoch": 0.9051571882726952,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004925011159973788,
      "loss": 5.5921,
      "mean_token_accuracy": 0.18747150152921677,
      "num_tokens": 19110463.0,
      "step": 10250
    },
    {
      "entropy": 5.870510864257812,
      "epoch": 0.9055987283645355,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004924930514285015,
      "loss": 5.5646,
      "mean_token_accuracy": 0.19496434628963472,
      "num_tokens": 19120723.0,
      "step": 10255
    },
    {
      "entropy": 5.783599853515625,
      "epoch": 0.9060402684563759,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000492484982599059,
      "loss": 5.5296,
      "mean_token_accuracy": 0.19487620741128922,
      "num_tokens": 19130259.0,
      "step": 10260
    },
    {
      "entropy": 5.849772882461548,
      "epoch": 0.9064818085482161,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004924769095092094,
      "loss": 5.6082,
      "mean_token_accuracy": 0.18738152384757994,
      "num_tokens": 19140256.0,
      "step": 10265
    },
    {
      "entropy": 5.847735500335693,
      "epoch": 0.9069233486400565,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004924688321591109,
      "loss": 5.6108,
      "mean_token_accuracy": 0.19002716541290282,
      "num_tokens": 19149165.0,
      "step": 10270
    },
    {
      "entropy": 5.806421995162964,
      "epoch": 0.9073648887318968,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004924607505489215,
      "loss": 5.5302,
      "mean_token_accuracy": 0.1930050030350685,
      "num_tokens": 19159254.0,
      "step": 10275
    },
    {
      "entropy": 5.856638717651367,
      "epoch": 0.9078064288237372,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004924526646787998,
      "loss": 5.5815,
      "mean_token_accuracy": 0.18753056079149247,
      "num_tokens": 19168166.0,
      "step": 10280
    },
    {
      "entropy": 5.875579071044922,
      "epoch": 0.9082479689155776,
      "grad_norm": 0.984375,
      "learning_rate": 0.000492444574548904,
      "loss": 5.5756,
      "mean_token_accuracy": 0.18910295367240906,
      "num_tokens": 19176887.0,
      "step": 10285
    },
    {
      "entropy": 5.816029071807861,
      "epoch": 0.9086895090074179,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004924364801593927,
      "loss": 5.5055,
      "mean_token_accuracy": 0.19390737116336823,
      "num_tokens": 19185510.0,
      "step": 10290
    },
    {
      "entropy": 5.730973148345948,
      "epoch": 0.9091310490992582,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004924283815104243,
      "loss": 5.539,
      "mean_token_accuracy": 0.2019898921251297,
      "num_tokens": 19195459.0,
      "step": 10295
    },
    {
      "entropy": 5.832513618469238,
      "epoch": 0.9095725891910985,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004924202786021578,
      "loss": 5.5633,
      "mean_token_accuracy": 0.19126464575529098,
      "num_tokens": 19204590.0,
      "step": 10300
    },
    {
      "entropy": 5.8629861831665036,
      "epoch": 0.9100141292829389,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004924121714347515,
      "loss": 5.509,
      "mean_token_accuracy": 0.1883266180753708,
      "num_tokens": 19213302.0,
      "step": 10305
    },
    {
      "entropy": 5.921752691268921,
      "epoch": 0.9104556693747792,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004924040600083644,
      "loss": 5.6284,
      "mean_token_accuracy": 0.18644503206014634,
      "num_tokens": 19223172.0,
      "step": 10310
    },
    {
      "entropy": 5.79363374710083,
      "epoch": 0.9108972094666196,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004923959443231554,
      "loss": 5.512,
      "mean_token_accuracy": 0.19097822457551955,
      "num_tokens": 19232234.0,
      "step": 10315
    },
    {
      "entropy": 5.706864833831787,
      "epoch": 0.91133874955846,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004923878243792836,
      "loss": 5.5101,
      "mean_token_accuracy": 0.19105002135038376,
      "num_tokens": 19241070.0,
      "step": 10320
    },
    {
      "entropy": 5.867473459243774,
      "epoch": 0.9117802896503002,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004923797001769079,
      "loss": 5.6053,
      "mean_token_accuracy": 0.1921307310461998,
      "num_tokens": 19251094.0,
      "step": 10325
    },
    {
      "entropy": 5.951294708251953,
      "epoch": 0.9122218297421406,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004923715717161875,
      "loss": 5.712,
      "mean_token_accuracy": 0.17849336862564086,
      "num_tokens": 19260588.0,
      "step": 10330
    },
    {
      "entropy": 5.920758962631226,
      "epoch": 0.9126633698339809,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004923634389972817,
      "loss": 5.7312,
      "mean_token_accuracy": 0.18118082880973815,
      "num_tokens": 19270687.0,
      "step": 10335
    },
    {
      "entropy": 5.824967288970948,
      "epoch": 0.9131049099258213,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004923553020203497,
      "loss": 5.5033,
      "mean_token_accuracy": 0.19493811577558517,
      "num_tokens": 19278739.0,
      "step": 10340
    },
    {
      "entropy": 5.960403347015381,
      "epoch": 0.9135464500176615,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004923471607855511,
      "loss": 5.674,
      "mean_token_accuracy": 0.17853326499462127,
      "num_tokens": 19289184.0,
      "step": 10345
    },
    {
      "entropy": 5.817495632171631,
      "epoch": 0.9139879901095019,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000492339015293045,
      "loss": 5.5119,
      "mean_token_accuracy": 0.20601145029067994,
      "num_tokens": 19298390.0,
      "step": 10350
    },
    {
      "entropy": 5.710802125930786,
      "epoch": 0.9144295302013423,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004923308655429913,
      "loss": 5.4672,
      "mean_token_accuracy": 0.2000226393342018,
      "num_tokens": 19307227.0,
      "step": 10355
    },
    {
      "entropy": 5.861149406433105,
      "epoch": 0.9148710702931826,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004923227115355495,
      "loss": 5.4825,
      "mean_token_accuracy": 0.18927655071020127,
      "num_tokens": 19317701.0,
      "step": 10360
    },
    {
      "entropy": 5.870515537261963,
      "epoch": 0.915312610385023,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004923145532708794,
      "loss": 5.609,
      "mean_token_accuracy": 0.18576886802911757,
      "num_tokens": 19326983.0,
      "step": 10365
    },
    {
      "entropy": 5.838311290740966,
      "epoch": 0.9157541504768633,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004923063907491408,
      "loss": 5.6094,
      "mean_token_accuracy": 0.1861104741692543,
      "num_tokens": 19337157.0,
      "step": 10370
    },
    {
      "entropy": 5.907456350326538,
      "epoch": 0.9161956905687036,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004922982239704936,
      "loss": 5.706,
      "mean_token_accuracy": 0.1902213379740715,
      "num_tokens": 19346845.0,
      "step": 10375
    },
    {
      "entropy": 5.897551155090332,
      "epoch": 0.9166372306605439,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004922900529350978,
      "loss": 5.5666,
      "mean_token_accuracy": 0.19248107224702835,
      "num_tokens": 19356905.0,
      "step": 10380
    },
    {
      "entropy": 5.88671841621399,
      "epoch": 0.9170787707523843,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004922818776431134,
      "loss": 5.5663,
      "mean_token_accuracy": 0.18569775819778442,
      "num_tokens": 19366595.0,
      "step": 10385
    },
    {
      "entropy": 5.902255821228027,
      "epoch": 0.9175203108442247,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004922736980947005,
      "loss": 5.6216,
      "mean_token_accuracy": 0.185403411090374,
      "num_tokens": 19377213.0,
      "step": 10390
    },
    {
      "entropy": 5.812246799468994,
      "epoch": 0.917961850936065,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004922655142900196,
      "loss": 5.4878,
      "mean_token_accuracy": 0.19387887716293334,
      "num_tokens": 19385510.0,
      "step": 10395
    },
    {
      "entropy": 5.769075489044189,
      "epoch": 0.9184033910279054,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004922573262292308,
      "loss": 5.533,
      "mean_token_accuracy": 0.18867291808128356,
      "num_tokens": 19394505.0,
      "step": 10400
    },
    {
      "entropy": 5.840726375579834,
      "epoch": 0.9188449311197456,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004922491339124945,
      "loss": 5.4133,
      "mean_token_accuracy": 0.2043693780899048,
      "num_tokens": 19403140.0,
      "step": 10405
    },
    {
      "entropy": 5.870825862884521,
      "epoch": 0.919286471211586,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004922409373399713,
      "loss": 5.6429,
      "mean_token_accuracy": 0.18787774443626404,
      "num_tokens": 19412709.0,
      "step": 10410
    },
    {
      "entropy": 5.760507488250733,
      "epoch": 0.9197280113034263,
      "grad_norm": 0.84375,
      "learning_rate": 0.0004922327365118216,
      "loss": 5.4797,
      "mean_token_accuracy": 0.19664435982704162,
      "num_tokens": 19422442.0,
      "step": 10415
    },
    {
      "entropy": 5.799489831924438,
      "epoch": 0.9201695513952667,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004922245314282062,
      "loss": 5.5869,
      "mean_token_accuracy": 0.1881786435842514,
      "num_tokens": 19432153.0,
      "step": 10420
    },
    {
      "entropy": 5.823181200027466,
      "epoch": 0.9206110914871071,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004922163220892858,
      "loss": 5.4102,
      "mean_token_accuracy": 0.20729821771383286,
      "num_tokens": 19440811.0,
      "step": 10425
    },
    {
      "entropy": 5.832983493804932,
      "epoch": 0.9210526315789473,
      "grad_norm": 0.86328125,
      "learning_rate": 0.000492208108495221,
      "loss": 5.5976,
      "mean_token_accuracy": 0.18627535849809645,
      "num_tokens": 19451330.0,
      "step": 10430
    },
    {
      "entropy": 5.754718780517578,
      "epoch": 0.9214941716707877,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000492199890646173,
      "loss": 5.4342,
      "mean_token_accuracy": 0.2057944655418396,
      "num_tokens": 19459689.0,
      "step": 10435
    },
    {
      "entropy": 5.9080911636352536,
      "epoch": 0.921935711762628,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004921916685423028,
      "loss": 5.5012,
      "mean_token_accuracy": 0.19368696510791777,
      "num_tokens": 19468110.0,
      "step": 10440
    },
    {
      "entropy": 5.8278931140899655,
      "epoch": 0.9223772518544684,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004921834421837711,
      "loss": 5.5555,
      "mean_token_accuracy": 0.18965283781290054,
      "num_tokens": 19477190.0,
      "step": 10445
    },
    {
      "entropy": 5.878892230987549,
      "epoch": 0.9228187919463087,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004921752115707394,
      "loss": 5.6379,
      "mean_token_accuracy": 0.1885254517197609,
      "num_tokens": 19485757.0,
      "step": 10450
    },
    {
      "entropy": 5.852978515625,
      "epoch": 0.9232603320381491,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004921669767033688,
      "loss": 5.5819,
      "mean_token_accuracy": 0.18951420485973358,
      "num_tokens": 19494846.0,
      "step": 10455
    },
    {
      "entropy": 5.862094497680664,
      "epoch": 0.9237018721299894,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004921587375818207,
      "loss": 5.58,
      "mean_token_accuracy": 0.1880530074238777,
      "num_tokens": 19504307.0,
      "step": 10460
    },
    {
      "entropy": 5.860325145721435,
      "epoch": 0.9241434122218297,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004921504942062564,
      "loss": 5.5183,
      "mean_token_accuracy": 0.18791094273328782,
      "num_tokens": 19512741.0,
      "step": 10465
    },
    {
      "entropy": 5.765936851501465,
      "epoch": 0.9245849523136701,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004921422465768374,
      "loss": 5.43,
      "mean_token_accuracy": 0.1966376557946205,
      "num_tokens": 19521823.0,
      "step": 10470
    },
    {
      "entropy": 5.794504547119141,
      "epoch": 0.9250264924055104,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004921339946937253,
      "loss": 5.6214,
      "mean_token_accuracy": 0.1903781697154045,
      "num_tokens": 19531048.0,
      "step": 10475
    },
    {
      "entropy": 5.863368797302246,
      "epoch": 0.9254680324973508,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004921257385570818,
      "loss": 5.5655,
      "mean_token_accuracy": 0.19053512662649155,
      "num_tokens": 19539543.0,
      "step": 10480
    },
    {
      "entropy": 5.94217700958252,
      "epoch": 0.925909572589191,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004921174781670685,
      "loss": 5.6751,
      "mean_token_accuracy": 0.18787076473236083,
      "num_tokens": 19549242.0,
      "step": 10485
    },
    {
      "entropy": 5.89439754486084,
      "epoch": 0.9263511126810314,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004921092135238473,
      "loss": 5.5648,
      "mean_token_accuracy": 0.18893613815307617,
      "num_tokens": 19558259.0,
      "step": 10490
    },
    {
      "entropy": 5.848619270324707,
      "epoch": 0.9267926527728718,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00049210094462758,
      "loss": 5.5645,
      "mean_token_accuracy": 0.18988049328327178,
      "num_tokens": 19567327.0,
      "step": 10495
    },
    {
      "entropy": 5.851853847503662,
      "epoch": 0.9272341928647121,
      "grad_norm": 1.0,
      "learning_rate": 0.0004920926714784288,
      "loss": 5.5915,
      "mean_token_accuracy": 0.18649715185165405,
      "num_tokens": 19576696.0,
      "step": 10500
    },
    {
      "entropy": 5.78373966217041,
      "epoch": 0.9276757329565525,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004920843940765555,
      "loss": 5.4707,
      "mean_token_accuracy": 0.20074461102485658,
      "num_tokens": 19585708.0,
      "step": 10505
    },
    {
      "entropy": 5.900629997253418,
      "epoch": 0.9281172730483928,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004920761124221224,
      "loss": 5.6463,
      "mean_token_accuracy": 0.19012684673070906,
      "num_tokens": 19595199.0,
      "step": 10510
    },
    {
      "entropy": 5.890163230895996,
      "epoch": 0.9285588131402331,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004920678265152919,
      "loss": 5.5683,
      "mean_token_accuracy": 0.1849339336156845,
      "num_tokens": 19604781.0,
      "step": 10515
    },
    {
      "entropy": 5.79541654586792,
      "epoch": 0.9290003532320734,
      "grad_norm": 0.984375,
      "learning_rate": 0.000492059536356226,
      "loss": 5.5189,
      "mean_token_accuracy": 0.19828343540430068,
      "num_tokens": 19614203.0,
      "step": 10520
    },
    {
      "entropy": 5.859884405136109,
      "epoch": 0.9294418933239138,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004920512419450872,
      "loss": 5.5467,
      "mean_token_accuracy": 0.1924946814775467,
      "num_tokens": 19623664.0,
      "step": 10525
    },
    {
      "entropy": 5.855048704147339,
      "epoch": 0.9298834334157542,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000492042943282038,
      "loss": 5.5473,
      "mean_token_accuracy": 0.19281800091266632,
      "num_tokens": 19632960.0,
      "step": 10530
    },
    {
      "entropy": 5.839187288284302,
      "epoch": 0.9303249735075945,
      "grad_norm": 0.890625,
      "learning_rate": 0.000492034640367241,
      "loss": 5.6609,
      "mean_token_accuracy": 0.18112767189741136,
      "num_tokens": 19643087.0,
      "step": 10535
    },
    {
      "entropy": 5.922775077819824,
      "epoch": 0.9307665135994349,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000492026333200859,
      "loss": 5.6156,
      "mean_token_accuracy": 0.18796712905168533,
      "num_tokens": 19653015.0,
      "step": 10540
    },
    {
      "entropy": 5.935885286331176,
      "epoch": 0.9312080536912751,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004920180217830543,
      "loss": 5.6198,
      "mean_token_accuracy": 0.18706526309251786,
      "num_tokens": 19661569.0,
      "step": 10545
    },
    {
      "entropy": 5.706429433822632,
      "epoch": 0.9316495937831155,
      "grad_norm": 1.0,
      "learning_rate": 0.00049200970611399,
      "loss": 5.4372,
      "mean_token_accuracy": 0.20040739327669144,
      "num_tokens": 19669559.0,
      "step": 10550
    },
    {
      "entropy": 5.738175296783448,
      "epoch": 0.9320911338749558,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004920013861938291,
      "loss": 5.4618,
      "mean_token_accuracy": 0.20482187420129777,
      "num_tokens": 19678766.0,
      "step": 10555
    },
    {
      "entropy": 5.83841872215271,
      "epoch": 0.9325326739667962,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004919930620227345,
      "loss": 5.6172,
      "mean_token_accuracy": 0.18378598093986512,
      "num_tokens": 19688422.0,
      "step": 10560
    },
    {
      "entropy": 5.947232913970947,
      "epoch": 0.9329742140586366,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004919847336008692,
      "loss": 5.6382,
      "mean_token_accuracy": 0.1858677551150322,
      "num_tokens": 19697206.0,
      "step": 10565
    },
    {
      "entropy": 5.870072269439698,
      "epoch": 0.9334157541504768,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004919764009283963,
      "loss": 5.6012,
      "mean_token_accuracy": 0.1842290610074997,
      "num_tokens": 19706065.0,
      "step": 10570
    },
    {
      "entropy": 5.8446617603302,
      "epoch": 0.9338572942423172,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004919680640054792,
      "loss": 5.5312,
      "mean_token_accuracy": 0.19687247425317764,
      "num_tokens": 19715402.0,
      "step": 10575
    },
    {
      "entropy": 5.799044561386109,
      "epoch": 0.9342988343341575,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000491959722832281,
      "loss": 5.4632,
      "mean_token_accuracy": 0.1981448546051979,
      "num_tokens": 19724328.0,
      "step": 10580
    },
    {
      "entropy": 5.792357778549194,
      "epoch": 0.9347403744259979,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004919513774089653,
      "loss": 5.581,
      "mean_token_accuracy": 0.19021185338497162,
      "num_tokens": 19734087.0,
      "step": 10585
    },
    {
      "entropy": 5.932400131225586,
      "epoch": 0.9351819145178382,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004919430277356955,
      "loss": 5.5321,
      "mean_token_accuracy": 0.19613374769687653,
      "num_tokens": 19743010.0,
      "step": 10590
    },
    {
      "entropy": 5.843912744522095,
      "epoch": 0.9356234546096786,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004919346738126351,
      "loss": 5.4913,
      "mean_token_accuracy": 0.19796763360500336,
      "num_tokens": 19752212.0,
      "step": 10595
    },
    {
      "entropy": 5.823946142196656,
      "epoch": 0.9360649947015189,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004919263156399478,
      "loss": 5.588,
      "mean_token_accuracy": 0.19262956976890563,
      "num_tokens": 19762496.0,
      "step": 10600
    },
    {
      "entropy": 5.8475781917572025,
      "epoch": 0.9365065347933592,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004919179532177976,
      "loss": 5.5768,
      "mean_token_accuracy": 0.1893441617488861,
      "num_tokens": 19772665.0,
      "step": 10605
    },
    {
      "entropy": 5.775518417358398,
      "epoch": 0.9369480748851996,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004919095865463479,
      "loss": 5.419,
      "mean_token_accuracy": 0.20027465969324112,
      "num_tokens": 19781639.0,
      "step": 10610
    },
    {
      "entropy": 5.844583129882812,
      "epoch": 0.9373896149770399,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004919012156257627,
      "loss": 5.5176,
      "mean_token_accuracy": 0.19134962409734727,
      "num_tokens": 19791475.0,
      "step": 10615
    },
    {
      "entropy": 5.819388580322266,
      "epoch": 0.9378311550688803,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004918928404562061,
      "loss": 5.5021,
      "mean_token_accuracy": 0.19283542335033416,
      "num_tokens": 19800967.0,
      "step": 10620
    },
    {
      "entropy": 5.782877779006958,
      "epoch": 0.9382726951607206,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004918844610378421,
      "loss": 5.6631,
      "mean_token_accuracy": 0.18687088489532472,
      "num_tokens": 19810303.0,
      "step": 10625
    },
    {
      "entropy": 5.933231925964355,
      "epoch": 0.9387142352525609,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004918760773708348,
      "loss": 5.5558,
      "mean_token_accuracy": 0.18964565843343734,
      "num_tokens": 19819049.0,
      "step": 10630
    },
    {
      "entropy": 5.9463951587677,
      "epoch": 0.9391557753444013,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004918676894553485,
      "loss": 5.5939,
      "mean_token_accuracy": 0.19474973082542418,
      "num_tokens": 19828253.0,
      "step": 10635
    },
    {
      "entropy": 5.834896755218506,
      "epoch": 0.9395973154362416,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004918592972915476,
      "loss": 5.5447,
      "mean_token_accuracy": 0.1950840249657631,
      "num_tokens": 19837673.0,
      "step": 10640
    },
    {
      "entropy": 5.758587121963501,
      "epoch": 0.940038855528082,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004918509008795963,
      "loss": 5.5817,
      "mean_token_accuracy": 0.18384605646133423,
      "num_tokens": 19846289.0,
      "step": 10645
    },
    {
      "entropy": 5.782199668884277,
      "epoch": 0.9404803956199222,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004918425002196594,
      "loss": 5.5483,
      "mean_token_accuracy": 0.19021508246660232,
      "num_tokens": 19855384.0,
      "step": 10650
    },
    {
      "entropy": 5.815890741348267,
      "epoch": 0.9409219357117626,
      "grad_norm": 1.625,
      "learning_rate": 0.000491834095311901,
      "loss": 5.4442,
      "mean_token_accuracy": 0.20328225791454316,
      "num_tokens": 19864688.0,
      "step": 10655
    },
    {
      "entropy": 5.7918627738952635,
      "epoch": 0.941363475803603,
      "grad_norm": 1.09375,
      "learning_rate": 0.000491825686156486,
      "loss": 5.4835,
      "mean_token_accuracy": 0.1945149391889572,
      "num_tokens": 19873692.0,
      "step": 10660
    },
    {
      "entropy": 5.855662870407104,
      "epoch": 0.9418050158954433,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004918172727535792,
      "loss": 5.5903,
      "mean_token_accuracy": 0.17907644957304,
      "num_tokens": 19882063.0,
      "step": 10665
    },
    {
      "entropy": 5.889585542678833,
      "epoch": 0.9422465559872837,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004918088551033454,
      "loss": 5.5299,
      "mean_token_accuracy": 0.19539738744497298,
      "num_tokens": 19891727.0,
      "step": 10670
    },
    {
      "entropy": 5.788696193695069,
      "epoch": 0.942688096079124,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004918004332059493,
      "loss": 5.5827,
      "mean_token_accuracy": 0.1965092420578003,
      "num_tokens": 19901851.0,
      "step": 10675
    },
    {
      "entropy": 5.855720949172974,
      "epoch": 0.9431296361709643,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004917920070615561,
      "loss": 5.5253,
      "mean_token_accuracy": 0.1925203323364258,
      "num_tokens": 19911092.0,
      "step": 10680
    },
    {
      "entropy": 5.8097868919372555,
      "epoch": 0.9435711762628046,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004917835766703307,
      "loss": 5.493,
      "mean_token_accuracy": 0.2018080472946167,
      "num_tokens": 19920842.0,
      "step": 10685
    },
    {
      "entropy": 5.9051590919494625,
      "epoch": 0.944012716354645,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004917751420324384,
      "loss": 5.5925,
      "mean_token_accuracy": 0.18529444336891174,
      "num_tokens": 19930833.0,
      "step": 10690
    },
    {
      "entropy": 5.908380222320557,
      "epoch": 0.9444542564464854,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004917667031480444,
      "loss": 5.6317,
      "mean_token_accuracy": 0.19101576805114745,
      "num_tokens": 19940514.0,
      "step": 10695
    },
    {
      "entropy": 5.79524040222168,
      "epoch": 0.9448957965383257,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004917582600173138,
      "loss": 5.4725,
      "mean_token_accuracy": 0.19828383177518843,
      "num_tokens": 19948949.0,
      "step": 10700
    },
    {
      "entropy": 5.715598058700562,
      "epoch": 0.9453373366301661,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004917498126404123,
      "loss": 5.4482,
      "mean_token_accuracy": 0.19933301210403442,
      "num_tokens": 19958410.0,
      "step": 10705
    },
    {
      "entropy": 5.811657333374024,
      "epoch": 0.9457788767220063,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004917413610175052,
      "loss": 5.4453,
      "mean_token_accuracy": 0.2015089213848114,
      "num_tokens": 19968160.0,
      "step": 10710
    },
    {
      "entropy": 5.925013494491577,
      "epoch": 0.9462204168138467,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004917329051487582,
      "loss": 5.6116,
      "mean_token_accuracy": 0.1815849557518959,
      "num_tokens": 19976811.0,
      "step": 10715
    },
    {
      "entropy": 5.927849149703979,
      "epoch": 0.946661956905687,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004917244450343368,
      "loss": 5.6883,
      "mean_token_accuracy": 0.17899896651506425,
      "num_tokens": 19986326.0,
      "step": 10720
    },
    {
      "entropy": 5.78308892250061,
      "epoch": 0.9471034969975274,
      "grad_norm": 0.875,
      "learning_rate": 0.0004917159806744067,
      "loss": 5.4072,
      "mean_token_accuracy": 0.19397269636392594,
      "num_tokens": 19996255.0,
      "step": 10725
    },
    {
      "entropy": 5.804378938674927,
      "epoch": 0.9475450370893678,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004917075120691338,
      "loss": 5.6274,
      "mean_token_accuracy": 0.18804312944412233,
      "num_tokens": 20004805.0,
      "step": 10730
    },
    {
      "entropy": 5.79609866142273,
      "epoch": 0.947986577181208,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004916990392186841,
      "loss": 5.5205,
      "mean_token_accuracy": 0.19648486524820327,
      "num_tokens": 20014189.0,
      "step": 10735
    },
    {
      "entropy": 5.874725818634033,
      "epoch": 0.9484281172730484,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004916905621232233,
      "loss": 5.5106,
      "mean_token_accuracy": 0.18644982278347016,
      "num_tokens": 20023261.0,
      "step": 10740
    },
    {
      "entropy": 5.811865997314453,
      "epoch": 0.9488696573648887,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004916820807829177,
      "loss": 5.4675,
      "mean_token_accuracy": 0.20290163904428482,
      "num_tokens": 20032433.0,
      "step": 10745
    },
    {
      "entropy": 5.813940620422363,
      "epoch": 0.9493111974567291,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004916735951979335,
      "loss": 5.5916,
      "mean_token_accuracy": 0.18647033274173735,
      "num_tokens": 20041455.0,
      "step": 10750
    },
    {
      "entropy": 5.860266447067261,
      "epoch": 0.9497527375485694,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004916651053684366,
      "loss": 5.6197,
      "mean_token_accuracy": 0.19127440452575684,
      "num_tokens": 20050088.0,
      "step": 10755
    },
    {
      "entropy": 5.873831081390381,
      "epoch": 0.9501942776404098,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004916566112945936,
      "loss": 5.5936,
      "mean_token_accuracy": 0.1924014210700989,
      "num_tokens": 20059114.0,
      "step": 10760
    },
    {
      "entropy": 5.8611561298370365,
      "epoch": 0.9506358177322501,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004916481129765708,
      "loss": 5.5789,
      "mean_token_accuracy": 0.19046310633420943,
      "num_tokens": 20068692.0,
      "step": 10765
    },
    {
      "entropy": 5.838329076766968,
      "epoch": 0.9510773578240904,
      "grad_norm": 2.1875,
      "learning_rate": 0.0004916396104145347,
      "loss": 5.6301,
      "mean_token_accuracy": 0.18845141381025315,
      "num_tokens": 20078380.0,
      "step": 10770
    },
    {
      "entropy": 5.733320808410644,
      "epoch": 0.9515188979159308,
      "grad_norm": 1.0,
      "learning_rate": 0.0004916311036086518,
      "loss": 5.449,
      "mean_token_accuracy": 0.20709470957517623,
      "num_tokens": 20087849.0,
      "step": 10775
    },
    {
      "entropy": 5.778975915908814,
      "epoch": 0.9519604380077711,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004916225925590889,
      "loss": 5.3979,
      "mean_token_accuracy": 0.2057466834783554,
      "num_tokens": 20097253.0,
      "step": 10780
    },
    {
      "entropy": 5.8431253910064695,
      "epoch": 0.9524019780996115,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004916140772660124,
      "loss": 5.5891,
      "mean_token_accuracy": 0.1924793854355812,
      "num_tokens": 20105931.0,
      "step": 10785
    },
    {
      "entropy": 5.866446781158447,
      "epoch": 0.9528435181914517,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004916055577295895,
      "loss": 5.474,
      "mean_token_accuracy": 0.20115301311016082,
      "num_tokens": 20115546.0,
      "step": 10790
    },
    {
      "entropy": 5.872866916656494,
      "epoch": 0.9532850582832921,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004915970339499868,
      "loss": 5.5021,
      "mean_token_accuracy": 0.1942845806479454,
      "num_tokens": 20125009.0,
      "step": 10795
    },
    {
      "entropy": 5.719097089767456,
      "epoch": 0.9537265983751325,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004915885059273714,
      "loss": 5.4817,
      "mean_token_accuracy": 0.19818040132522582,
      "num_tokens": 20134007.0,
      "step": 10800
    },
    {
      "entropy": 5.820677709579468,
      "epoch": 0.9541681384669728,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004915799736619104,
      "loss": 5.4953,
      "mean_token_accuracy": 0.19351319819688798,
      "num_tokens": 20142617.0,
      "step": 10805
    },
    {
      "entropy": 5.808750438690185,
      "epoch": 0.9546096785588132,
      "grad_norm": 1.0,
      "learning_rate": 0.000491571437153771,
      "loss": 5.5193,
      "mean_token_accuracy": 0.19794072806835175,
      "num_tokens": 20151732.0,
      "step": 10810
    },
    {
      "entropy": 5.856479501724243,
      "epoch": 0.9550512186506535,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004915628964031202,
      "loss": 5.6163,
      "mean_token_accuracy": 0.1860255166888237,
      "num_tokens": 20161733.0,
      "step": 10815
    },
    {
      "entropy": 5.874025154113769,
      "epoch": 0.9554927587424938,
      "grad_norm": 1.0,
      "learning_rate": 0.0004915543514101253,
      "loss": 5.5447,
      "mean_token_accuracy": 0.18483669310808182,
      "num_tokens": 20171022.0,
      "step": 10820
    },
    {
      "entropy": 5.85857367515564,
      "epoch": 0.9559342988343341,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004915458021749541,
      "loss": 5.5463,
      "mean_token_accuracy": 0.18598238825798036,
      "num_tokens": 20180809.0,
      "step": 10825
    },
    {
      "entropy": 5.783239555358887,
      "epoch": 0.9563758389261745,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004915372486977736,
      "loss": 5.486,
      "mean_token_accuracy": 0.19979268461465835,
      "num_tokens": 20189614.0,
      "step": 10830
    },
    {
      "entropy": 5.841081809997559,
      "epoch": 0.9568173790180149,
      "grad_norm": 0.875,
      "learning_rate": 0.0004915286909787516,
      "loss": 5.6297,
      "mean_token_accuracy": 0.18066627979278566,
      "num_tokens": 20199136.0,
      "step": 10835
    },
    {
      "entropy": 5.846323442459107,
      "epoch": 0.9572589191098552,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004915201290180556,
      "loss": 5.529,
      "mean_token_accuracy": 0.19071774035692216,
      "num_tokens": 20208768.0,
      "step": 10840
    },
    {
      "entropy": 5.7795311450958256,
      "epoch": 0.9577004592016956,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004915115628158535,
      "loss": 5.512,
      "mean_token_accuracy": 0.19033466279506683,
      "num_tokens": 20218109.0,
      "step": 10845
    },
    {
      "entropy": 5.84051480293274,
      "epoch": 0.9581419992935358,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000491502992372313,
      "loss": 5.5144,
      "mean_token_accuracy": 0.19057008475065232,
      "num_tokens": 20226768.0,
      "step": 10850
    },
    {
      "entropy": 5.81554651260376,
      "epoch": 0.9585835393853762,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000491494417687602,
      "loss": 5.471,
      "mean_token_accuracy": 0.19765634089708328,
      "num_tokens": 20236519.0,
      "step": 10855
    },
    {
      "entropy": 5.762387752532959,
      "epoch": 0.9590250794772165,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004914858387618886,
      "loss": 5.4908,
      "mean_token_accuracy": 0.19534918218851088,
      "num_tokens": 20245611.0,
      "step": 10860
    },
    {
      "entropy": 5.750170135498047,
      "epoch": 0.9594666195690569,
      "grad_norm": 0.83984375,
      "learning_rate": 0.0004914772555953406,
      "loss": 5.406,
      "mean_token_accuracy": 0.20583200305700303,
      "num_tokens": 20255089.0,
      "step": 10865
    },
    {
      "entropy": 5.817626190185547,
      "epoch": 0.9599081596608973,
      "grad_norm": 1.0,
      "learning_rate": 0.0004914686681881264,
      "loss": 5.5574,
      "mean_token_accuracy": 0.1936030000448227,
      "num_tokens": 20264794.0,
      "step": 10870
    },
    {
      "entropy": 5.796624183654785,
      "epoch": 0.9603496997527375,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000491460076540414,
      "loss": 5.416,
      "mean_token_accuracy": 0.20647903829813002,
      "num_tokens": 20274088.0,
      "step": 10875
    },
    {
      "entropy": 5.8205211639404295,
      "epoch": 0.9607912398445779,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004914514806523718,
      "loss": 5.6139,
      "mean_token_accuracy": 0.18427006900310516,
      "num_tokens": 20283837.0,
      "step": 10880
    },
    {
      "entropy": 5.822281312942505,
      "epoch": 0.9612327799364182,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004914428805241683,
      "loss": 5.5071,
      "mean_token_accuracy": 0.1921115681529045,
      "num_tokens": 20293287.0,
      "step": 10885
    },
    {
      "entropy": 5.7551898002624515,
      "epoch": 0.9616743200282586,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004914342761559718,
      "loss": 5.4649,
      "mean_token_accuracy": 0.20005710870027543,
      "num_tokens": 20302262.0,
      "step": 10890
    },
    {
      "entropy": 5.847769927978516,
      "epoch": 0.9621158601200989,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004914256675479509,
      "loss": 5.561,
      "mean_token_accuracy": 0.18721564412117003,
      "num_tokens": 20312475.0,
      "step": 10895
    },
    {
      "entropy": 5.871733570098877,
      "epoch": 0.9625574002119393,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004914170547002742,
      "loss": 5.5751,
      "mean_token_accuracy": 0.18982745260000228,
      "num_tokens": 20322243.0,
      "step": 10900
    },
    {
      "entropy": 5.778389406204224,
      "epoch": 0.9629989403037796,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004914084376131106,
      "loss": 5.4688,
      "mean_token_accuracy": 0.19254059046506883,
      "num_tokens": 20331299.0,
      "step": 10905
    },
    {
      "entropy": 5.7243060111999515,
      "epoch": 0.9634404803956199,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004913998162866288,
      "loss": 5.4259,
      "mean_token_accuracy": 0.20277778804302216,
      "num_tokens": 20340800.0,
      "step": 10910
    },
    {
      "entropy": 5.820883941650391,
      "epoch": 0.9638820204874603,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004913911907209976,
      "loss": 5.5237,
      "mean_token_accuracy": 0.19660518169403077,
      "num_tokens": 20350292.0,
      "step": 10915
    },
    {
      "entropy": 5.881887340545655,
      "epoch": 0.9643235605793006,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004913825609163861,
      "loss": 5.6065,
      "mean_token_accuracy": 0.1878852590918541,
      "num_tokens": 20360109.0,
      "step": 10920
    },
    {
      "entropy": 5.729526853561401,
      "epoch": 0.964765100671141,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004913739268729633,
      "loss": 5.4341,
      "mean_token_accuracy": 0.20365360230207444,
      "num_tokens": 20368917.0,
      "step": 10925
    },
    {
      "entropy": 5.871054363250733,
      "epoch": 0.9652066407629812,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004913652885908983,
      "loss": 5.6194,
      "mean_token_accuracy": 0.18444482535123824,
      "num_tokens": 20378580.0,
      "step": 10930
    },
    {
      "entropy": 5.847923469543457,
      "epoch": 0.9656481808548216,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004913566460703603,
      "loss": 5.4444,
      "mean_token_accuracy": 0.19596558958292007,
      "num_tokens": 20387497.0,
      "step": 10935
    },
    {
      "entropy": 5.832335758209228,
      "epoch": 0.966089720946662,
      "grad_norm": 1.0,
      "learning_rate": 0.0004913479993115187,
      "loss": 5.5996,
      "mean_token_accuracy": 0.18531107306480407,
      "num_tokens": 20396586.0,
      "step": 10940
    },
    {
      "entropy": 5.873624277114868,
      "epoch": 0.9665312610385023,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004913393483145428,
      "loss": 5.687,
      "mean_token_accuracy": 0.1777090162038803,
      "num_tokens": 20406183.0,
      "step": 10945
    },
    {
      "entropy": 5.987033367156982,
      "epoch": 0.9669728011303427,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004913306930796022,
      "loss": 5.6099,
      "mean_token_accuracy": 0.1890963390469551,
      "num_tokens": 20415716.0,
      "step": 10950
    },
    {
      "entropy": 5.862414169311523,
      "epoch": 0.967414341222183,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004913220336068662,
      "loss": 5.5392,
      "mean_token_accuracy": 0.1910262554883957,
      "num_tokens": 20424715.0,
      "step": 10955
    },
    {
      "entropy": 5.779024362564087,
      "epoch": 0.9678558813140233,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004913133698965048,
      "loss": 5.5316,
      "mean_token_accuracy": 0.19780020266771317,
      "num_tokens": 20434686.0,
      "step": 10960
    },
    {
      "entropy": 5.89177713394165,
      "epoch": 0.9682974214058636,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004913047019486874,
      "loss": 5.5132,
      "mean_token_accuracy": 0.19592755883932114,
      "num_tokens": 20443619.0,
      "step": 10965
    },
    {
      "entropy": 5.808563947677612,
      "epoch": 0.968738961497704,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004912960297635839,
      "loss": 5.4652,
      "mean_token_accuracy": 0.20202988535165786,
      "num_tokens": 20453540.0,
      "step": 10970
    },
    {
      "entropy": 5.836717319488526,
      "epoch": 0.9691805015895444,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004912873533413641,
      "loss": 5.5787,
      "mean_token_accuracy": 0.1865977704524994,
      "num_tokens": 20463395.0,
      "step": 10975
    },
    {
      "entropy": 5.80081467628479,
      "epoch": 0.9696220416813847,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004912786726821982,
      "loss": 5.5233,
      "mean_token_accuracy": 0.19126374423503875,
      "num_tokens": 20472302.0,
      "step": 10980
    },
    {
      "entropy": 5.828465366363526,
      "epoch": 0.970063581773225,
      "grad_norm": 0.90234375,
      "learning_rate": 0.000491269987786256,
      "loss": 5.5348,
      "mean_token_accuracy": 0.18925092965364457,
      "num_tokens": 20482077.0,
      "step": 10985
    },
    {
      "entropy": 5.882504272460937,
      "epoch": 0.9705051218650653,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004912612986537078,
      "loss": 5.5443,
      "mean_token_accuracy": 0.19295522719621658,
      "num_tokens": 20492177.0,
      "step": 10990
    },
    {
      "entropy": 5.842564868927002,
      "epoch": 0.9709466619569057,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004912526052847238,
      "loss": 5.5354,
      "mean_token_accuracy": 0.19674329757690429,
      "num_tokens": 20502063.0,
      "step": 10995
    },
    {
      "entropy": 5.8237231254577635,
      "epoch": 0.971388202048746,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004912439076794742,
      "loss": 5.501,
      "mean_token_accuracy": 0.1882395938038826,
      "num_tokens": 20510631.0,
      "step": 11000
    },
    {
      "entropy": 5.805966806411743,
      "epoch": 0.9718297421405864,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004912352058381293,
      "loss": 5.5267,
      "mean_token_accuracy": 0.19073240756988524,
      "num_tokens": 20519978.0,
      "step": 11005
    },
    {
      "entropy": 5.838520431518555,
      "epoch": 0.9722712822324268,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004912264997608599,
      "loss": 5.5247,
      "mean_token_accuracy": 0.19575220346450806,
      "num_tokens": 20529346.0,
      "step": 11010
    },
    {
      "entropy": 5.845974016189575,
      "epoch": 0.972712822324267,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004912177894478364,
      "loss": 5.4883,
      "mean_token_accuracy": 0.18457019031047822,
      "num_tokens": 20537619.0,
      "step": 11015
    },
    {
      "entropy": 5.80692048072815,
      "epoch": 0.9731543624161074,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004912090748992291,
      "loss": 5.5325,
      "mean_token_accuracy": 0.19319151490926742,
      "num_tokens": 20547011.0,
      "step": 11020
    },
    {
      "entropy": 5.868833398818969,
      "epoch": 0.9735959025079477,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004912003561152092,
      "loss": 5.4505,
      "mean_token_accuracy": 0.19301554709672927,
      "num_tokens": 20556425.0,
      "step": 11025
    },
    {
      "entropy": 5.82353720664978,
      "epoch": 0.9740374425997881,
      "grad_norm": 1.0,
      "learning_rate": 0.0004911916330959472,
      "loss": 5.4817,
      "mean_token_accuracy": 0.19372252076864244,
      "num_tokens": 20565330.0,
      "step": 11030
    },
    {
      "entropy": 5.782631540298462,
      "epoch": 0.9744789826916284,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004911829058416141,
      "loss": 5.5746,
      "mean_token_accuracy": 0.20310136675834656,
      "num_tokens": 20575320.0,
      "step": 11035
    },
    {
      "entropy": 5.832729387283325,
      "epoch": 0.9749205227834687,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004911741743523808,
      "loss": 5.4792,
      "mean_token_accuracy": 0.20184270590543746,
      "num_tokens": 20585105.0,
      "step": 11040
    },
    {
      "entropy": 5.805664777755737,
      "epoch": 0.9753620628753091,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004911654386284184,
      "loss": 5.5221,
      "mean_token_accuracy": 0.19096221029758453,
      "num_tokens": 20593889.0,
      "step": 11045
    },
    {
      "entropy": 5.816440153121948,
      "epoch": 0.9758036029671494,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000491156698669898,
      "loss": 5.5806,
      "mean_token_accuracy": 0.19517292082309723,
      "num_tokens": 20602322.0,
      "step": 11050
    },
    {
      "entropy": 5.882116222381592,
      "epoch": 0.9762451430589898,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004911479544769908,
      "loss": 5.646,
      "mean_token_accuracy": 0.1793898180127144,
      "num_tokens": 20612077.0,
      "step": 11055
    },
    {
      "entropy": 5.936902141571045,
      "epoch": 0.9766866831508301,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000491139206049868,
      "loss": 5.6115,
      "mean_token_accuracy": 0.19139361679553984,
      "num_tokens": 20620683.0,
      "step": 11060
    },
    {
      "entropy": 5.883061647415161,
      "epoch": 0.9771282232426705,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004911304533887011,
      "loss": 5.6407,
      "mean_token_accuracy": 0.18760707080364228,
      "num_tokens": 20631124.0,
      "step": 11065
    },
    {
      "entropy": 5.846213531494141,
      "epoch": 0.9775697633345107,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004911216964936616,
      "loss": 5.5012,
      "mean_token_accuracy": 0.19065711200237273,
      "num_tokens": 20640515.0,
      "step": 11070
    },
    {
      "entropy": 5.830524349212647,
      "epoch": 0.9780113034263511,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004911129353649211,
      "loss": 5.4786,
      "mean_token_accuracy": 0.19359829723834993,
      "num_tokens": 20649819.0,
      "step": 11075
    },
    {
      "entropy": 5.820531988143921,
      "epoch": 0.9784528435181915,
      "grad_norm": 1.03125,
      "learning_rate": 0.000491104170002651,
      "loss": 5.5104,
      "mean_token_accuracy": 0.18779035657644272,
      "num_tokens": 20657831.0,
      "step": 11080
    },
    {
      "entropy": 5.872558355331421,
      "epoch": 0.9788943836100318,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000491095400407023,
      "loss": 5.6304,
      "mean_token_accuracy": 0.189630950987339,
      "num_tokens": 20666807.0,
      "step": 11085
    },
    {
      "entropy": 5.868529367446899,
      "epoch": 0.9793359237018722,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004910866265782091,
      "loss": 5.4759,
      "mean_token_accuracy": 0.19473941773176193,
      "num_tokens": 20674859.0,
      "step": 11090
    },
    {
      "entropy": 5.881205654144287,
      "epoch": 0.9797774637937124,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004910778485163811,
      "loss": 5.5587,
      "mean_token_accuracy": 0.19576413780450821,
      "num_tokens": 20684060.0,
      "step": 11095
    },
    {
      "entropy": 5.822815370559693,
      "epoch": 0.9802190038855528,
      "grad_norm": 1.015625,
      "learning_rate": 0.000491069066221711,
      "loss": 5.4876,
      "mean_token_accuracy": 0.1994558945298195,
      "num_tokens": 20693446.0,
      "step": 11100
    },
    {
      "entropy": 5.830358982086182,
      "epoch": 0.9806605439773931,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004910602796943707,
      "loss": 5.6416,
      "mean_token_accuracy": 0.18342330753803254,
      "num_tokens": 20702635.0,
      "step": 11105
    },
    {
      "entropy": 5.9078997611999515,
      "epoch": 0.9811020840692335,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004910514889345323,
      "loss": 5.5593,
      "mean_token_accuracy": 0.19043246358633042,
      "num_tokens": 20712256.0,
      "step": 11110
    },
    {
      "entropy": 5.8011908531188965,
      "epoch": 0.9815436241610739,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004910426939423682,
      "loss": 5.5167,
      "mean_token_accuracy": 0.19414657205343247,
      "num_tokens": 20720175.0,
      "step": 11115
    },
    {
      "entropy": 5.811932897567749,
      "epoch": 0.9819851642529142,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004910338947180506,
      "loss": 5.5335,
      "mean_token_accuracy": 0.1913396969437599,
      "num_tokens": 20729849.0,
      "step": 11120
    },
    {
      "entropy": 5.838641405105591,
      "epoch": 0.9824267043447545,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004910250912617519,
      "loss": 5.4708,
      "mean_token_accuracy": 0.19795780032873153,
      "num_tokens": 20739129.0,
      "step": 11125
    },
    {
      "entropy": 5.8044229507446286,
      "epoch": 0.9828682444365948,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004910162835736445,
      "loss": 5.4401,
      "mean_token_accuracy": 0.20202791690826416,
      "num_tokens": 20748632.0,
      "step": 11130
    },
    {
      "entropy": 5.840196371078491,
      "epoch": 0.9833097845284352,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000491007471653901,
      "loss": 5.4914,
      "mean_token_accuracy": 0.1948569595813751,
      "num_tokens": 20758535.0,
      "step": 11135
    },
    {
      "entropy": 5.8420037746429445,
      "epoch": 0.9837513246202755,
      "grad_norm": 0.875,
      "learning_rate": 0.0004909986555026939,
      "loss": 5.5285,
      "mean_token_accuracy": 0.19545526206493377,
      "num_tokens": 20768982.0,
      "step": 11140
    },
    {
      "entropy": 5.8689741611480715,
      "epoch": 0.9841928647121159,
      "grad_norm": 0.984375,
      "learning_rate": 0.000490989835120196,
      "loss": 5.5847,
      "mean_token_accuracy": 0.19294222444295883,
      "num_tokens": 20778936.0,
      "step": 11145
    },
    {
      "entropy": 5.844592666625976,
      "epoch": 0.9846344048039563,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004909810105065801,
      "loss": 5.5818,
      "mean_token_accuracy": 0.18521394282579423,
      "num_tokens": 20788618.0,
      "step": 11150
    },
    {
      "entropy": 5.882683420181275,
      "epoch": 0.9850759448957965,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004909721816620191,
      "loss": 5.5572,
      "mean_token_accuracy": 0.19761741161346436,
      "num_tokens": 20797721.0,
      "step": 11155
    },
    {
      "entropy": 5.8583558082580565,
      "epoch": 0.9855174849876369,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004909633485866857,
      "loss": 5.4983,
      "mean_token_accuracy": 0.19559048116207123,
      "num_tokens": 20807305.0,
      "step": 11160
    },
    {
      "entropy": 5.772975111007691,
      "epoch": 0.9859590250794772,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004909545112807533,
      "loss": 5.4632,
      "mean_token_accuracy": 0.2001844346523285,
      "num_tokens": 20816984.0,
      "step": 11165
    },
    {
      "entropy": 5.901240539550781,
      "epoch": 0.9864005651713176,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004909456697443948,
      "loss": 5.6369,
      "mean_token_accuracy": 0.18369455635547638,
      "num_tokens": 20825738.0,
      "step": 11170
    },
    {
      "entropy": 5.8502623558044435,
      "epoch": 0.9868421052631579,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004909368239777835,
      "loss": 5.5822,
      "mean_token_accuracy": 0.1874794155359268,
      "num_tokens": 20834135.0,
      "step": 11175
    },
    {
      "entropy": 5.9332075119018555,
      "epoch": 0.9872836453549982,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004909279739810926,
      "loss": 5.6398,
      "mean_token_accuracy": 0.18591250330209733,
      "num_tokens": 20843689.0,
      "step": 11180
    },
    {
      "entropy": 5.897461938858032,
      "epoch": 0.9877251854468386,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004909191197544954,
      "loss": 5.5174,
      "mean_token_accuracy": 0.19147263169288636,
      "num_tokens": 20851724.0,
      "step": 11185
    },
    {
      "entropy": 5.9224241256713865,
      "epoch": 0.9881667255386789,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004909102612981655,
      "loss": 5.6748,
      "mean_token_accuracy": 0.18432777673006057,
      "num_tokens": 20862325.0,
      "step": 11190
    },
    {
      "entropy": 5.858700799942016,
      "epoch": 0.9886082656305193,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004909013986122763,
      "loss": 5.5076,
      "mean_token_accuracy": 0.19612646251916885,
      "num_tokens": 20871046.0,
      "step": 11195
    },
    {
      "entropy": 5.917420291900635,
      "epoch": 0.9890498057223596,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004908925316970016,
      "loss": 5.5363,
      "mean_token_accuracy": 0.18652904480695726,
      "num_tokens": 20881605.0,
      "step": 11200
    },
    {
      "entropy": 5.817625427246094,
      "epoch": 0.9894913458142,
      "grad_norm": 0.921875,
      "learning_rate": 0.000490883660552515,
      "loss": 5.4602,
      "mean_token_accuracy": 0.1957890585064888,
      "num_tokens": 20891170.0,
      "step": 11205
    },
    {
      "entropy": 5.833710765838623,
      "epoch": 0.9899328859060402,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004908747851789902,
      "loss": 5.4702,
      "mean_token_accuracy": 0.1861294686794281,
      "num_tokens": 20900307.0,
      "step": 11210
    },
    {
      "entropy": 5.817179775238037,
      "epoch": 0.9903744259978806,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908659055766011,
      "loss": 5.6567,
      "mean_token_accuracy": 0.18575167357921601,
      "num_tokens": 20910623.0,
      "step": 11215
    },
    {
      "entropy": 5.938866853713989,
      "epoch": 0.990815966089721,
      "grad_norm": 1.0,
      "learning_rate": 0.0004908570217455218,
      "loss": 5.5463,
      "mean_token_accuracy": 0.19014012664556504,
      "num_tokens": 20919744.0,
      "step": 11220
    },
    {
      "entropy": 5.842549800872803,
      "epoch": 0.9912575061815613,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004908481336859262,
      "loss": 5.6128,
      "mean_token_accuracy": 0.18795954138040544,
      "num_tokens": 20928701.0,
      "step": 11225
    },
    {
      "entropy": 5.8330910205841064,
      "epoch": 0.9916990462734017,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004908392413979884,
      "loss": 5.5843,
      "mean_token_accuracy": 0.180023954808712,
      "num_tokens": 20937786.0,
      "step": 11230
    },
    {
      "entropy": 5.864147567749024,
      "epoch": 0.9921405863652419,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004908303448818825,
      "loss": 5.4946,
      "mean_token_accuracy": 0.19863616079092025,
      "num_tokens": 20947214.0,
      "step": 11235
    },
    {
      "entropy": 5.832936477661133,
      "epoch": 0.9925821264570823,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000490821444137783,
      "loss": 5.6014,
      "mean_token_accuracy": 0.1971711367368698,
      "num_tokens": 20956041.0,
      "step": 11240
    },
    {
      "entropy": 5.9129539966583256,
      "epoch": 0.9930236665489226,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004908125391658641,
      "loss": 5.569,
      "mean_token_accuracy": 0.19148291647434235,
      "num_tokens": 20966162.0,
      "step": 11245
    },
    {
      "entropy": 5.822826194763183,
      "epoch": 0.993465206640763,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004908036299663003,
      "loss": 5.5614,
      "mean_token_accuracy": 0.1926749274134636,
      "num_tokens": 20976025.0,
      "step": 11250
    },
    {
      "entropy": 5.8316937446594235,
      "epoch": 0.9939067467326034,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004907947165392662,
      "loss": 5.5893,
      "mean_token_accuracy": 0.18729829490184785,
      "num_tokens": 20985606.0,
      "step": 11255
    },
    {
      "entropy": 5.878600978851319,
      "epoch": 0.9943482868244437,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004907857988849363,
      "loss": 5.5732,
      "mean_token_accuracy": 0.1873452693223953,
      "num_tokens": 20994308.0,
      "step": 11260
    },
    {
      "entropy": 5.794317722320557,
      "epoch": 0.994789826916284,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004907768770034853,
      "loss": 5.5024,
      "mean_token_accuracy": 0.19150446206331254,
      "num_tokens": 21003783.0,
      "step": 11265
    },
    {
      "entropy": 5.8252458572387695,
      "epoch": 0.9952313670081243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000490767950895088,
      "loss": 5.4881,
      "mean_token_accuracy": 0.19247478395700454,
      "num_tokens": 21012870.0,
      "step": 11270
    },
    {
      "entropy": 5.8911336898803714,
      "epoch": 0.9956729070999647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004907590205599192,
      "loss": 5.5473,
      "mean_token_accuracy": 0.19090282768011094,
      "num_tokens": 21021987.0,
      "step": 11275
    },
    {
      "entropy": 5.820564222335816,
      "epoch": 0.996114447191805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000490750085998154,
      "loss": 5.5425,
      "mean_token_accuracy": 0.18923987150192262,
      "num_tokens": 21030977.0,
      "step": 11280
    },
    {
      "entropy": 5.850178575515747,
      "epoch": 0.9965559872836454,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004907411472099672,
      "loss": 5.4942,
      "mean_token_accuracy": 0.19618253856897355,
      "num_tokens": 21039259.0,
      "step": 11285
    },
    {
      "entropy": 5.877530908584594,
      "epoch": 0.9969975273754857,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004907322041955341,
      "loss": 5.5794,
      "mean_token_accuracy": 0.18979037553071976,
      "num_tokens": 21049321.0,
      "step": 11290
    },
    {
      "entropy": 5.921963882446289,
      "epoch": 0.997439067467326,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004907232569550298,
      "loss": 5.651,
      "mean_token_accuracy": 0.18208762407302856,
      "num_tokens": 21059469.0,
      "step": 11295
    },
    {
      "entropy": 5.900750780105591,
      "epoch": 0.9978806075591664,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004907143054886297,
      "loss": 5.5703,
      "mean_token_accuracy": 0.1895820155739784,
      "num_tokens": 21068694.0,
      "step": 11300
    },
    {
      "entropy": 5.857248401641845,
      "epoch": 0.9983221476510067,
      "grad_norm": 1.0,
      "learning_rate": 0.000490705349796509,
      "loss": 5.5068,
      "mean_token_accuracy": 0.19134427458047867,
      "num_tokens": 21077232.0,
      "step": 11305
    },
    {
      "entropy": 5.867048740386963,
      "epoch": 0.9987636877428471,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004906963898788431,
      "loss": 5.5268,
      "mean_token_accuracy": 0.1941692277789116,
      "num_tokens": 21086517.0,
      "step": 11310
    },
    {
      "entropy": 5.850531101226807,
      "epoch": 0.9992052278346873,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004906874257358075,
      "loss": 5.5828,
      "mean_token_accuracy": 0.19759476333856582,
      "num_tokens": 21096969.0,
      "step": 11315
    },
    {
      "entropy": 5.858910846710205,
      "epoch": 0.9996467679265277,
      "grad_norm": 1.03125,
      "learning_rate": 0.000490678457367578,
      "loss": 5.5086,
      "mean_token_accuracy": 0.18713059276342392,
      "num_tokens": 21106554.0,
      "step": 11320
    },
    {
      "entropy": 5.785353326797486,
      "epoch": 1.000088308018368,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004906694847743302,
      "loss": 5.5205,
      "mean_token_accuracy": 0.18986949622631072,
      "num_tokens": 21114829.0,
      "step": 11325
    },
    {
      "entropy": 5.84618034362793,
      "epoch": 1.0005298481102085,
      "grad_norm": 1.0,
      "learning_rate": 0.0004906605079562399,
      "loss": 5.458,
      "mean_token_accuracy": 0.19853014200925828,
      "num_tokens": 21123568.0,
      "step": 11330
    },
    {
      "entropy": 5.854660415649414,
      "epoch": 1.0009713882020488,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004906515269134827,
      "loss": 5.5118,
      "mean_token_accuracy": 0.19205471575260163,
      "num_tokens": 21132811.0,
      "step": 11335
    },
    {
      "entropy": 5.89120774269104,
      "epoch": 1.001412928293889,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004906425416462349,
      "loss": 5.5509,
      "mean_token_accuracy": 0.18671243637800217,
      "num_tokens": 21141549.0,
      "step": 11340
    },
    {
      "entropy": 5.838399982452392,
      "epoch": 1.0018544683857293,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004906335521546723,
      "loss": 5.4523,
      "mean_token_accuracy": 0.19941389858722686,
      "num_tokens": 21150603.0,
      "step": 11345
    },
    {
      "entropy": 5.707959413528442,
      "epoch": 1.0022960084775698,
      "grad_norm": 1.0,
      "learning_rate": 0.0004906245584389712,
      "loss": 5.2676,
      "mean_token_accuracy": 0.21225778609514237,
      "num_tokens": 21159842.0,
      "step": 11350
    },
    {
      "entropy": 5.822812271118164,
      "epoch": 1.00273754856941,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004906155604993075,
      "loss": 5.4184,
      "mean_token_accuracy": 0.2034035414457321,
      "num_tokens": 21169139.0,
      "step": 11355
    },
    {
      "entropy": 5.784867143630981,
      "epoch": 1.0031790886612504,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004906065583358576,
      "loss": 5.4127,
      "mean_token_accuracy": 0.19405371248722075,
      "num_tokens": 21178059.0,
      "step": 11360
    },
    {
      "entropy": 5.803128242492676,
      "epoch": 1.0036206287530909,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004905975519487978,
      "loss": 5.4032,
      "mean_token_accuracy": 0.20225947499275207,
      "num_tokens": 21187966.0,
      "step": 11365
    },
    {
      "entropy": 5.765253305435181,
      "epoch": 1.0040621688449312,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004905885413383046,
      "loss": 5.4632,
      "mean_token_accuracy": 0.19135601818561554,
      "num_tokens": 21197460.0,
      "step": 11370
    },
    {
      "entropy": 5.851017570495605,
      "epoch": 1.0045037089367714,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004905795265045546,
      "loss": 5.4939,
      "mean_token_accuracy": 0.1959708407521248,
      "num_tokens": 21206977.0,
      "step": 11375
    },
    {
      "entropy": 5.869827938079834,
      "epoch": 1.0049452490286117,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004905705074477242,
      "loss": 5.4267,
      "mean_token_accuracy": 0.19144175946712494,
      "num_tokens": 21216046.0,
      "step": 11380
    },
    {
      "entropy": 5.856464099884033,
      "epoch": 1.0053867891204522,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004905614841679901,
      "loss": 5.4792,
      "mean_token_accuracy": 0.19193035364151,
      "num_tokens": 21225612.0,
      "step": 11385
    },
    {
      "entropy": 5.785376119613647,
      "epoch": 1.0058283292122925,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004905524566655292,
      "loss": 5.3684,
      "mean_token_accuracy": 0.19887097030878068,
      "num_tokens": 21233641.0,
      "step": 11390
    },
    {
      "entropy": 5.765918016433716,
      "epoch": 1.0062698693041328,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004905434249405183,
      "loss": 5.4281,
      "mean_token_accuracy": 0.19341274499893188,
      "num_tokens": 21243444.0,
      "step": 11395
    },
    {
      "entropy": 5.7887389183044435,
      "epoch": 1.0067114093959733,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004905343889931342,
      "loss": 5.3977,
      "mean_token_accuracy": 0.20267544835805892,
      "num_tokens": 21253473.0,
      "step": 11400
    },
    {
      "entropy": 5.797057723999023,
      "epoch": 1.0071529494878135,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000490525348823554,
      "loss": 5.3677,
      "mean_token_accuracy": 0.21220226287841798,
      "num_tokens": 21262049.0,
      "step": 11405
    },
    {
      "entropy": 5.783495759963989,
      "epoch": 1.0075944895796538,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004905163044319549,
      "loss": 5.4093,
      "mean_token_accuracy": 0.19663029164075851,
      "num_tokens": 21270909.0,
      "step": 11410
    },
    {
      "entropy": 5.816624593734741,
      "epoch": 1.008036029671494,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004905072558185139,
      "loss": 5.5011,
      "mean_token_accuracy": 0.1927421882748604,
      "num_tokens": 21280503.0,
      "step": 11415
    },
    {
      "entropy": 5.871208047866821,
      "epoch": 1.0084775697633346,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004904982029834083,
      "loss": 5.4831,
      "mean_token_accuracy": 0.19088730365037918,
      "num_tokens": 21289957.0,
      "step": 11420
    },
    {
      "entropy": 5.831331872940064,
      "epoch": 1.0089191098551749,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004904891459268154,
      "loss": 5.4584,
      "mean_token_accuracy": 0.19292692244052886,
      "num_tokens": 21298799.0,
      "step": 11425
    },
    {
      "entropy": 5.818262195587158,
      "epoch": 1.0093606499470151,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004904800846489128,
      "loss": 5.4662,
      "mean_token_accuracy": 0.19724620282649993,
      "num_tokens": 21308251.0,
      "step": 11430
    },
    {
      "entropy": 5.905819416046143,
      "epoch": 1.0098021900388556,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004904710191498779,
      "loss": 5.5004,
      "mean_token_accuracy": 0.1947304055094719,
      "num_tokens": 21317705.0,
      "step": 11435
    },
    {
      "entropy": 5.811056852340698,
      "epoch": 1.010243730130696,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004904619494298882,
      "loss": 5.4791,
      "mean_token_accuracy": 0.1861782342195511,
      "num_tokens": 21326867.0,
      "step": 11440
    },
    {
      "entropy": 5.7974344253540036,
      "epoch": 1.0106852702225362,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004904528754891215,
      "loss": 5.4851,
      "mean_token_accuracy": 0.2002352073788643,
      "num_tokens": 21336054.0,
      "step": 11445
    },
    {
      "entropy": 5.823584175109863,
      "epoch": 1.0111268103143765,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004904437973277555,
      "loss": 5.4017,
      "mean_token_accuracy": 0.1982896387577057,
      "num_tokens": 21344457.0,
      "step": 11450
    },
    {
      "entropy": 5.870646190643311,
      "epoch": 1.011568350406217,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004904347149459681,
      "loss": 5.4993,
      "mean_token_accuracy": 0.19316635876893998,
      "num_tokens": 21353639.0,
      "step": 11455
    },
    {
      "entropy": 5.709807586669922,
      "epoch": 1.0120098904980572,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004904256283439371,
      "loss": 5.2969,
      "mean_token_accuracy": 0.20653853714466094,
      "num_tokens": 21362005.0,
      "step": 11460
    },
    {
      "entropy": 5.858596324920654,
      "epoch": 1.0124514305898975,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004904165375218405,
      "loss": 5.4774,
      "mean_token_accuracy": 0.19295329749584197,
      "num_tokens": 21371917.0,
      "step": 11465
    },
    {
      "entropy": 5.855436897277832,
      "epoch": 1.012892970681738,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004904074424798565,
      "loss": 5.4778,
      "mean_token_accuracy": 0.19918979704380035,
      "num_tokens": 21381500.0,
      "step": 11470
    },
    {
      "entropy": 5.782000780105591,
      "epoch": 1.0133345107735783,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004903983432181632,
      "loss": 5.3096,
      "mean_token_accuracy": 0.2015272855758667,
      "num_tokens": 21391280.0,
      "step": 11475
    },
    {
      "entropy": 5.815475749969482,
      "epoch": 1.0137760508654186,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004903892397369388,
      "loss": 5.4567,
      "mean_token_accuracy": 0.19736944139003754,
      "num_tokens": 21400371.0,
      "step": 11480
    },
    {
      "entropy": 5.67401762008667,
      "epoch": 1.0142175909572588,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004903801320363618,
      "loss": 5.4771,
      "mean_token_accuracy": 0.1985686182975769,
      "num_tokens": 21410493.0,
      "step": 11485
    },
    {
      "entropy": 5.90243935585022,
      "epoch": 1.0146591310490993,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004903710201166105,
      "loss": 5.5491,
      "mean_token_accuracy": 0.1836334154009819,
      "num_tokens": 21420516.0,
      "step": 11490
    },
    {
      "entropy": 5.916108655929565,
      "epoch": 1.0151006711409396,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004903619039778634,
      "loss": 5.4953,
      "mean_token_accuracy": 0.18604948073625566,
      "num_tokens": 21429614.0,
      "step": 11495
    },
    {
      "entropy": 5.828366661071778,
      "epoch": 1.0155422112327799,
      "grad_norm": 1.046875,
      "learning_rate": 0.000490352783620299,
      "loss": 5.435,
      "mean_token_accuracy": 0.19844983667135238,
      "num_tokens": 21438595.0,
      "step": 11500
    },
    {
      "entropy": 5.836382675170898,
      "epoch": 1.0159837513246204,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004903436590440961,
      "loss": 5.4777,
      "mean_token_accuracy": 0.1989193946123123,
      "num_tokens": 21447708.0,
      "step": 11505
    },
    {
      "entropy": 5.802752923965454,
      "epoch": 1.0164252914164607,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004903345302494334,
      "loss": 5.4294,
      "mean_token_accuracy": 0.1908747971057892,
      "num_tokens": 21457210.0,
      "step": 11510
    },
    {
      "entropy": 5.840627908706665,
      "epoch": 1.016866831508301,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004903253972364897,
      "loss": 5.4614,
      "mean_token_accuracy": 0.19405657798051834,
      "num_tokens": 21466247.0,
      "step": 11515
    },
    {
      "entropy": 5.7858726501464846,
      "epoch": 1.0173083716001412,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004903162600054439,
      "loss": 5.4048,
      "mean_token_accuracy": 0.20191211253404617,
      "num_tokens": 21475634.0,
      "step": 11520
    },
    {
      "entropy": 5.86365647315979,
      "epoch": 1.0177499116919817,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004903071185564751,
      "loss": 5.522,
      "mean_token_accuracy": 0.19560528099536895,
      "num_tokens": 21485523.0,
      "step": 11525
    },
    {
      "entropy": 5.8727446556091305,
      "epoch": 1.018191451783822,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004902979728897623,
      "loss": 5.472,
      "mean_token_accuracy": 0.19813337475061416,
      "num_tokens": 21494909.0,
      "step": 11530
    },
    {
      "entropy": 5.864111089706421,
      "epoch": 1.0186329918756623,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004902888230054846,
      "loss": 5.5114,
      "mean_token_accuracy": 0.1915932461619377,
      "num_tokens": 21504413.0,
      "step": 11535
    },
    {
      "entropy": 5.755148887634277,
      "epoch": 1.0190745319675028,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902796689038213,
      "loss": 5.3802,
      "mean_token_accuracy": 0.20285746902227403,
      "num_tokens": 21514068.0,
      "step": 11540
    },
    {
      "entropy": 5.793791389465332,
      "epoch": 1.019516072059343,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004902705105849517,
      "loss": 5.3969,
      "mean_token_accuracy": 0.20098552852869034,
      "num_tokens": 21523536.0,
      "step": 11545
    },
    {
      "entropy": 5.786781597137451,
      "epoch": 1.0199576121511833,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004902613480490553,
      "loss": 5.4692,
      "mean_token_accuracy": 0.19177822917699813,
      "num_tokens": 21532554.0,
      "step": 11550
    },
    {
      "entropy": 5.76055417060852,
      "epoch": 1.0203991522430236,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004902521812963116,
      "loss": 5.3321,
      "mean_token_accuracy": 0.20827980637550353,
      "num_tokens": 21542508.0,
      "step": 11555
    },
    {
      "entropy": 5.842724275588989,
      "epoch": 1.020840692334864,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004902430103268998,
      "loss": 5.4541,
      "mean_token_accuracy": 0.18653589189052583,
      "num_tokens": 21551954.0,
      "step": 11560
    },
    {
      "entropy": 5.823522233963013,
      "epoch": 1.0212822324267044,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004902338351410002,
      "loss": 5.3807,
      "mean_token_accuracy": 0.20463838130235673,
      "num_tokens": 21561676.0,
      "step": 11565
    },
    {
      "entropy": 5.840857124328613,
      "epoch": 1.0217237725185446,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000490224655738792,
      "loss": 5.5141,
      "mean_token_accuracy": 0.18839401602745057,
      "num_tokens": 21571916.0,
      "step": 11570
    },
    {
      "entropy": 5.841112852096558,
      "epoch": 1.0221653126103851,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004902154721204552,
      "loss": 5.4459,
      "mean_token_accuracy": 0.20426111817359924,
      "num_tokens": 21581892.0,
      "step": 11575
    },
    {
      "entropy": 5.79775037765503,
      "epoch": 1.0226068527022254,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004902062842861696,
      "loss": 5.4042,
      "mean_token_accuracy": 0.194574473798275,
      "num_tokens": 21591390.0,
      "step": 11580
    },
    {
      "entropy": 5.82957501411438,
      "epoch": 1.0230483927940657,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004901970922361154,
      "loss": 5.473,
      "mean_token_accuracy": 0.19892579168081284,
      "num_tokens": 21600766.0,
      "step": 11585
    },
    {
      "entropy": 5.746204280853272,
      "epoch": 1.023489932885906,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004901878959704726,
      "loss": 5.4019,
      "mean_token_accuracy": 0.20576784610748292,
      "num_tokens": 21609264.0,
      "step": 11590
    },
    {
      "entropy": 5.889167833328247,
      "epoch": 1.0239314729777464,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004901786954894212,
      "loss": 5.4956,
      "mean_token_accuracy": 0.1914335384964943,
      "num_tokens": 21618409.0,
      "step": 11595
    },
    {
      "entropy": 5.795125818252563,
      "epoch": 1.0243730130695867,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004901694907931416,
      "loss": 5.3851,
      "mean_token_accuracy": 0.19335469454526902,
      "num_tokens": 21628158.0,
      "step": 11600
    },
    {
      "entropy": 5.816567993164062,
      "epoch": 1.024814553161427,
      "grad_norm": 0.953125,
      "learning_rate": 0.000490160281881814,
      "loss": 5.4942,
      "mean_token_accuracy": 0.19107363224029542,
      "num_tokens": 21637859.0,
      "step": 11605
    },
    {
      "entropy": 5.850227165222168,
      "epoch": 1.0252560932532675,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004901510687556188,
      "loss": 5.5057,
      "mean_token_accuracy": 0.18931938260793685,
      "num_tokens": 21647389.0,
      "step": 11610
    },
    {
      "entropy": 5.873205041885376,
      "epoch": 1.0256976333451078,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004901418514147367,
      "loss": 5.4553,
      "mean_token_accuracy": 0.19083526283502578,
      "num_tokens": 21655857.0,
      "step": 11615
    },
    {
      "entropy": 5.807099008560181,
      "epoch": 1.026139173436948,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000490132629859348,
      "loss": 5.4184,
      "mean_token_accuracy": 0.20058273077011107,
      "num_tokens": 21666150.0,
      "step": 11620
    },
    {
      "entropy": 5.834448575973511,
      "epoch": 1.0265807135287883,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004901234040896334,
      "loss": 5.4394,
      "mean_token_accuracy": 0.19685986191034316,
      "num_tokens": 21675662.0,
      "step": 11625
    },
    {
      "entropy": 5.859313535690307,
      "epoch": 1.0270222536206288,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004901141741057737,
      "loss": 5.4975,
      "mean_token_accuracy": 0.187775319814682,
      "num_tokens": 21685721.0,
      "step": 11630
    },
    {
      "entropy": 5.846322774887085,
      "epoch": 1.027463793712469,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004901049399079495,
      "loss": 5.4729,
      "mean_token_accuracy": 0.1979638710618019,
      "num_tokens": 21694507.0,
      "step": 11635
    },
    {
      "entropy": 5.857589817047119,
      "epoch": 1.0279053338043094,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004900957014963421,
      "loss": 5.4899,
      "mean_token_accuracy": 0.20053462982177733,
      "num_tokens": 21704092.0,
      "step": 11640
    },
    {
      "entropy": 5.760619688034057,
      "epoch": 1.0283468738961499,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004900864588711321,
      "loss": 5.3866,
      "mean_token_accuracy": 0.20713716447353364,
      "num_tokens": 21713313.0,
      "step": 11645
    },
    {
      "entropy": 5.814091444015503,
      "epoch": 1.0287884139879901,
      "grad_norm": 1.0,
      "learning_rate": 0.0004900772120325008,
      "loss": 5.4371,
      "mean_token_accuracy": 0.2021331310272217,
      "num_tokens": 21721622.0,
      "step": 11650
    },
    {
      "entropy": 5.838172197341919,
      "epoch": 1.0292299540798304,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004900679609806291,
      "loss": 5.413,
      "mean_token_accuracy": 0.2008321091532707,
      "num_tokens": 21730914.0,
      "step": 11655
    },
    {
      "entropy": 5.813170003890991,
      "epoch": 1.0296714941716707,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004900587057156985,
      "loss": 5.3781,
      "mean_token_accuracy": 0.20447985976934432,
      "num_tokens": 21740124.0,
      "step": 11660
    },
    {
      "entropy": 5.815191602706909,
      "epoch": 1.0301130342635112,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004900494462378902,
      "loss": 5.4655,
      "mean_token_accuracy": 0.19556693136692047,
      "num_tokens": 21750013.0,
      "step": 11665
    },
    {
      "entropy": 5.862353706359864,
      "epoch": 1.0305545743553515,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004900401825473854,
      "loss": 5.4884,
      "mean_token_accuracy": 0.19777150452136993,
      "num_tokens": 21758842.0,
      "step": 11670
    },
    {
      "entropy": 5.808915328979492,
      "epoch": 1.0309961144471917,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004900309146443658,
      "loss": 5.3693,
      "mean_token_accuracy": 0.20384867042303084,
      "num_tokens": 21767002.0,
      "step": 11675
    },
    {
      "entropy": 5.862798595428467,
      "epoch": 1.0314376545390322,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004900216425290128,
      "loss": 5.5014,
      "mean_token_accuracy": 0.19282409995794297,
      "num_tokens": 21775372.0,
      "step": 11680
    },
    {
      "entropy": 5.848205852508545,
      "epoch": 1.0318791946308725,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004900123662015082,
      "loss": 5.4869,
      "mean_token_accuracy": 0.1917622983455658,
      "num_tokens": 21784541.0,
      "step": 11685
    },
    {
      "entropy": 5.81615629196167,
      "epoch": 1.0323207347227128,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004900030856620336,
      "loss": 5.388,
      "mean_token_accuracy": 0.20575797706842422,
      "num_tokens": 21794525.0,
      "step": 11690
    },
    {
      "entropy": 5.841868925094604,
      "epoch": 1.032762274814553,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004899938009107708,
      "loss": 5.4658,
      "mean_token_accuracy": 0.20023476779460908,
      "num_tokens": 21804469.0,
      "step": 11695
    },
    {
      "entropy": 5.763099622726441,
      "epoch": 1.0332038149063936,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004899845119479019,
      "loss": 5.4179,
      "mean_token_accuracy": 0.20264574736356736,
      "num_tokens": 21813986.0,
      "step": 11700
    },
    {
      "entropy": 5.846244621276855,
      "epoch": 1.0336453549982338,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004899752187736085,
      "loss": 5.418,
      "mean_token_accuracy": 0.20018214583396912,
      "num_tokens": 21824083.0,
      "step": 11705
    },
    {
      "entropy": 5.750829124450684,
      "epoch": 1.0340868950900741,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004899659213880728,
      "loss": 5.3799,
      "mean_token_accuracy": 0.2069662794470787,
      "num_tokens": 21833526.0,
      "step": 11710
    },
    {
      "entropy": 5.792848777770996,
      "epoch": 1.0345284351819146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004899566197914771,
      "loss": 5.3703,
      "mean_token_accuracy": 0.2027594715356827,
      "num_tokens": 21842537.0,
      "step": 11715
    },
    {
      "entropy": 5.841008043289184,
      "epoch": 1.034969975273755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004899473139840033,
      "loss": 5.446,
      "mean_token_accuracy": 0.194618758559227,
      "num_tokens": 21851604.0,
      "step": 11720
    },
    {
      "entropy": 5.825651741027832,
      "epoch": 1.0354115153655952,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004899380039658339,
      "loss": 5.4928,
      "mean_token_accuracy": 0.19981683492660524,
      "num_tokens": 21861480.0,
      "step": 11725
    },
    {
      "entropy": 5.839599514007569,
      "epoch": 1.0358530554574354,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004899286897371513,
      "loss": 5.4713,
      "mean_token_accuracy": 0.1941080316901207,
      "num_tokens": 21870901.0,
      "step": 11730
    },
    {
      "entropy": 5.751971244812012,
      "epoch": 1.036294595549276,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000489919371298138,
      "loss": 5.3487,
      "mean_token_accuracy": 0.20988212525844574,
      "num_tokens": 21880491.0,
      "step": 11735
    },
    {
      "entropy": 5.80057053565979,
      "epoch": 1.0367361356411162,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004899100486489763,
      "loss": 5.4267,
      "mean_token_accuracy": 0.20379792600870134,
      "num_tokens": 21889833.0,
      "step": 11740
    },
    {
      "entropy": 5.779327392578125,
      "epoch": 1.0371776757329565,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004899007217898491,
      "loss": 5.4792,
      "mean_token_accuracy": 0.19909710586071014,
      "num_tokens": 21899222.0,
      "step": 11745
    },
    {
      "entropy": 5.846719264984131,
      "epoch": 1.037619215824797,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004898913907209389,
      "loss": 5.4592,
      "mean_token_accuracy": 0.196945820748806,
      "num_tokens": 21908848.0,
      "step": 11750
    },
    {
      "entropy": 5.823145055770874,
      "epoch": 1.0380607559166373,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004898820554424285,
      "loss": 5.4181,
      "mean_token_accuracy": 0.19045039564371108,
      "num_tokens": 21917867.0,
      "step": 11755
    },
    {
      "entropy": 5.874126863479614,
      "epoch": 1.0385022960084775,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000489872715954501,
      "loss": 5.4908,
      "mean_token_accuracy": 0.1879080832004547,
      "num_tokens": 21927476.0,
      "step": 11760
    },
    {
      "entropy": 5.924701070785522,
      "epoch": 1.0389438361003178,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004898633722573391,
      "loss": 5.5758,
      "mean_token_accuracy": 0.18321400731801987,
      "num_tokens": 21937068.0,
      "step": 11765
    },
    {
      "entropy": 5.808666515350342,
      "epoch": 1.0393853761921583,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000489854024351126,
      "loss": 5.4365,
      "mean_token_accuracy": 0.1953706905245781,
      "num_tokens": 21946144.0,
      "step": 11770
    },
    {
      "entropy": 5.7899901390075685,
      "epoch": 1.0398269162839986,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004898446722360446,
      "loss": 5.3694,
      "mean_token_accuracy": 0.2072113871574402,
      "num_tokens": 21955346.0,
      "step": 11775
    },
    {
      "entropy": 5.744083833694458,
      "epoch": 1.0402684563758389,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004898353159122784,
      "loss": 5.4414,
      "mean_token_accuracy": 0.20349046885967254,
      "num_tokens": 21965034.0,
      "step": 11780
    },
    {
      "entropy": 5.783421277999878,
      "epoch": 1.0407099964676794,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004898259553800106,
      "loss": 5.4154,
      "mean_token_accuracy": 0.19425492584705353,
      "num_tokens": 21975167.0,
      "step": 11785
    },
    {
      "entropy": 5.900732803344726,
      "epoch": 1.0411515365595196,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004898165906394244,
      "loss": 5.421,
      "mean_token_accuracy": 0.19738931059837342,
      "num_tokens": 21984372.0,
      "step": 11790
    },
    {
      "entropy": 5.803021955490112,
      "epoch": 1.04159307665136,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004898072216907033,
      "loss": 5.4264,
      "mean_token_accuracy": 0.1924193799495697,
      "num_tokens": 21993513.0,
      "step": 11795
    },
    {
      "entropy": 5.823993492126465,
      "epoch": 1.0420346167432002,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004897978485340311,
      "loss": 5.4697,
      "mean_token_accuracy": 0.19322787672281266,
      "num_tokens": 22002593.0,
      "step": 11800
    },
    {
      "entropy": 5.913989877700805,
      "epoch": 1.0424761568350407,
      "grad_norm": 0.953125,
      "learning_rate": 0.000489788471169591,
      "loss": 5.512,
      "mean_token_accuracy": 0.19518644213676453,
      "num_tokens": 22012296.0,
      "step": 11805
    },
    {
      "entropy": 5.894937324523926,
      "epoch": 1.042917696926881,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004897790895975671,
      "loss": 5.5292,
      "mean_token_accuracy": 0.19130600541830062,
      "num_tokens": 22021267.0,
      "step": 11810
    },
    {
      "entropy": 5.827191638946533,
      "epoch": 1.0433592370187212,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004897697038181428,
      "loss": 5.4924,
      "mean_token_accuracy": 0.19480232298374175,
      "num_tokens": 22030878.0,
      "step": 11815
    },
    {
      "entropy": 5.788354349136353,
      "epoch": 1.0438007771105617,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004897603138315022,
      "loss": 5.4657,
      "mean_token_accuracy": 0.19767487049102783,
      "num_tokens": 22040245.0,
      "step": 11820
    },
    {
      "entropy": 5.861823177337646,
      "epoch": 1.044242317202402,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004897509196378293,
      "loss": 5.4704,
      "mean_token_accuracy": 0.19682453721761703,
      "num_tokens": 22050169.0,
      "step": 11825
    },
    {
      "entropy": 5.900480031967163,
      "epoch": 1.0446838572942423,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000489741521237308,
      "loss": 5.4994,
      "mean_token_accuracy": 0.19175503700971602,
      "num_tokens": 22059168.0,
      "step": 11830
    },
    {
      "entropy": 5.776277828216553,
      "epoch": 1.0451253973860826,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004897321186301223,
      "loss": 5.4242,
      "mean_token_accuracy": 0.19125735759735107,
      "num_tokens": 22068326.0,
      "step": 11835
    },
    {
      "entropy": 5.794267559051514,
      "epoch": 1.045566937477923,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004897227118164566,
      "loss": 5.452,
      "mean_token_accuracy": 0.19609965533018112,
      "num_tokens": 22078026.0,
      "step": 11840
    },
    {
      "entropy": 5.873622989654541,
      "epoch": 1.0460084775697633,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000489713300796495,
      "loss": 5.4966,
      "mean_token_accuracy": 0.19001780599355697,
      "num_tokens": 22088230.0,
      "step": 11845
    },
    {
      "entropy": 5.760268783569336,
      "epoch": 1.0464500176616036,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004897038855704221,
      "loss": 5.3219,
      "mean_token_accuracy": 0.20597454011440278,
      "num_tokens": 22096667.0,
      "step": 11850
    },
    {
      "entropy": 5.770038747787476,
      "epoch": 1.0468915577534441,
      "grad_norm": 1.0,
      "learning_rate": 0.0004896944661384222,
      "loss": 5.4385,
      "mean_token_accuracy": 0.19639641493558885,
      "num_tokens": 22106871.0,
      "step": 11855
    },
    {
      "entropy": 5.861229133605957,
      "epoch": 1.0473330978452844,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004896850425006797,
      "loss": 5.3875,
      "mean_token_accuracy": 0.19390991032123567,
      "num_tokens": 22114954.0,
      "step": 11860
    },
    {
      "entropy": 5.826381063461303,
      "epoch": 1.0477746379371247,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004896756146573794,
      "loss": 5.4207,
      "mean_token_accuracy": 0.19825548231601714,
      "num_tokens": 22125293.0,
      "step": 11865
    },
    {
      "entropy": 5.866196918487549,
      "epoch": 1.048216178028965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004896661826087059,
      "loss": 5.4628,
      "mean_token_accuracy": 0.1978798970580101,
      "num_tokens": 22133731.0,
      "step": 11870
    },
    {
      "entropy": 5.7895458221435545,
      "epoch": 1.0486577181208054,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004896567463548439,
      "loss": 5.4534,
      "mean_token_accuracy": 0.18647223711013794,
      "num_tokens": 22143007.0,
      "step": 11875
    },
    {
      "entropy": 5.852534818649292,
      "epoch": 1.0490992582126457,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004896473058959783,
      "loss": 5.4701,
      "mean_token_accuracy": 0.1962908148765564,
      "num_tokens": 22152044.0,
      "step": 11880
    },
    {
      "entropy": 5.888934659957886,
      "epoch": 1.049540798304486,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004896378612322942,
      "loss": 5.5378,
      "mean_token_accuracy": 0.19280146360397338,
      "num_tokens": 22161473.0,
      "step": 11885
    },
    {
      "entropy": 5.837153434753418,
      "epoch": 1.0499823383963265,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004896284123639763,
      "loss": 5.4335,
      "mean_token_accuracy": 0.19596049040555955,
      "num_tokens": 22171704.0,
      "step": 11890
    },
    {
      "entropy": 5.830124616622925,
      "epoch": 1.0504238784881668,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048961895929121,
      "loss": 5.4361,
      "mean_token_accuracy": 0.19259973168373107,
      "num_tokens": 22180593.0,
      "step": 11895
    },
    {
      "entropy": 5.813496732711792,
      "epoch": 1.050865418580007,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004896095020141802,
      "loss": 5.4278,
      "mean_token_accuracy": 0.2036044180393219,
      "num_tokens": 22189552.0,
      "step": 11900
    },
    {
      "entropy": 5.837124395370483,
      "epoch": 1.0513069586718473,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004896000405330724,
      "loss": 5.4551,
      "mean_token_accuracy": 0.19098978042602538,
      "num_tokens": 22199314.0,
      "step": 11905
    },
    {
      "entropy": 5.787033367156982,
      "epoch": 1.0517484987636878,
      "grad_norm": 1.0,
      "learning_rate": 0.000489590574848072,
      "loss": 5.3893,
      "mean_token_accuracy": 0.21002008467912675,
      "num_tokens": 22208128.0,
      "step": 11910
    },
    {
      "entropy": 5.725189113616944,
      "epoch": 1.052190038855528,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000489581104959364,
      "loss": 5.3562,
      "mean_token_accuracy": 0.20892212986946107,
      "num_tokens": 22217700.0,
      "step": 11915
    },
    {
      "entropy": 5.841599559783935,
      "epoch": 1.0526315789473684,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004895716308671343,
      "loss": 5.4628,
      "mean_token_accuracy": 0.18936695754528046,
      "num_tokens": 22225941.0,
      "step": 11920
    },
    {
      "entropy": 5.795419073104858,
      "epoch": 1.0530731190392089,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004895621525715685,
      "loss": 5.4098,
      "mean_token_accuracy": 0.1998206689953804,
      "num_tokens": 22234935.0,
      "step": 11925
    },
    {
      "entropy": 5.824654388427734,
      "epoch": 1.0535146591310491,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004895526700728521,
      "loss": 5.4884,
      "mean_token_accuracy": 0.19739733785390853,
      "num_tokens": 22244087.0,
      "step": 11930
    },
    {
      "entropy": 5.804003524780273,
      "epoch": 1.0539561992228894,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004895431833711708,
      "loss": 5.4452,
      "mean_token_accuracy": 0.19215094894170762,
      "num_tokens": 22254277.0,
      "step": 11935
    },
    {
      "entropy": 5.788774204254151,
      "epoch": 1.0543977393147297,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004895336924667107,
      "loss": 5.4065,
      "mean_token_accuracy": 0.2060361012816429,
      "num_tokens": 22263150.0,
      "step": 11940
    },
    {
      "entropy": 5.789840364456177,
      "epoch": 1.0548392794065702,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004895241973596576,
      "loss": 5.437,
      "mean_token_accuracy": 0.198403038084507,
      "num_tokens": 22271973.0,
      "step": 11945
    },
    {
      "entropy": 5.766195297241211,
      "epoch": 1.0552808194984105,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004895146980501973,
      "loss": 5.4104,
      "mean_token_accuracy": 0.20491848587989808,
      "num_tokens": 22280637.0,
      "step": 11950
    },
    {
      "entropy": 5.802606105804443,
      "epoch": 1.0557223595902507,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004895051945385163,
      "loss": 5.426,
      "mean_token_accuracy": 0.19119401425123214,
      "num_tokens": 22290601.0,
      "step": 11955
    },
    {
      "entropy": 5.851685810089111,
      "epoch": 1.0561638996820912,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004894956868248004,
      "loss": 5.454,
      "mean_token_accuracy": 0.19932712465524674,
      "num_tokens": 22300212.0,
      "step": 11960
    },
    {
      "entropy": 5.875369596481323,
      "epoch": 1.0566054397739315,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004894861749092359,
      "loss": 5.4377,
      "mean_token_accuracy": 0.19410779625177382,
      "num_tokens": 22309171.0,
      "step": 11965
    },
    {
      "entropy": 5.8332763671875,
      "epoch": 1.0570469798657718,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004894766587920094,
      "loss": 5.4529,
      "mean_token_accuracy": 0.20589271783828736,
      "num_tokens": 22318457.0,
      "step": 11970
    },
    {
      "entropy": 5.868374919891357,
      "epoch": 1.057488519957612,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000489467138473307,
      "loss": 5.4526,
      "mean_token_accuracy": 0.19353863447904587,
      "num_tokens": 22327077.0,
      "step": 11975
    },
    {
      "entropy": 5.820459318161011,
      "epoch": 1.0579300600494526,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004894576139533154,
      "loss": 5.443,
      "mean_token_accuracy": 0.19651559889316558,
      "num_tokens": 22336259.0,
      "step": 11980
    },
    {
      "entropy": 5.801056051254273,
      "epoch": 1.0583716001412928,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000489448085232221,
      "loss": 5.4227,
      "mean_token_accuracy": 0.20796411633491516,
      "num_tokens": 22344978.0,
      "step": 11985
    },
    {
      "entropy": 5.804754590988159,
      "epoch": 1.058813140233133,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004894385523102107,
      "loss": 5.4315,
      "mean_token_accuracy": 0.20169849544763566,
      "num_tokens": 22353950.0,
      "step": 11990
    },
    {
      "entropy": 5.834704208374023,
      "epoch": 1.0592546803249736,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004894290151874711,
      "loss": 5.3831,
      "mean_token_accuracy": 0.19994827657938002,
      "num_tokens": 22363100.0,
      "step": 11995
    },
    {
      "entropy": 5.760499382019043,
      "epoch": 1.0596962204168139,
      "grad_norm": 0.89453125,
      "learning_rate": 0.000489419473864189,
      "loss": 5.3758,
      "mean_token_accuracy": 0.20482891499996186,
      "num_tokens": 22371979.0,
      "step": 12000
    },
    {
      "epoch": 1.0596962204168139,
      "eval_entropy": 5.5627229945980945,
      "eval_loss": 5.526999473571777,
      "eval_mean_token_accuracy": 0.20152225287742956,
      "eval_num_tokens": 22371979.0,
      "eval_runtime": 20.0722,
      "eval_samples_per_second": 1759.395,
      "eval_steps_per_second": 219.956,
      "step": 12000
    },
    {
      "entropy": 5.76117901802063,
      "epoch": 1.0601377605086542,
      "grad_norm": 1.0,
      "learning_rate": 0.0004894099283405515,
      "loss": 5.398,
      "mean_token_accuracy": 0.1994665563106537,
      "num_tokens": 22380987.0,
      "step": 12005
    },
    {
      "entropy": 5.820873498916626,
      "epoch": 1.0605793006004944,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004894003786167452,
      "loss": 5.4924,
      "mean_token_accuracy": 0.19171537905931474,
      "num_tokens": 22389367.0,
      "step": 12010
    },
    {
      "entropy": 5.860400533676147,
      "epoch": 1.061020840692335,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004893908246929576,
      "loss": 5.529,
      "mean_token_accuracy": 0.19122651368379592,
      "num_tokens": 22399158.0,
      "step": 12015
    },
    {
      "entropy": 5.865876626968384,
      "epoch": 1.0614623807841752,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004893812665693755,
      "loss": 5.5236,
      "mean_token_accuracy": 0.19261441379785538,
      "num_tokens": 22408245.0,
      "step": 12020
    },
    {
      "entropy": 5.787412929534912,
      "epoch": 1.0619039208760155,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004893717042461865,
      "loss": 5.4741,
      "mean_token_accuracy": 0.1920081853866577,
      "num_tokens": 22417014.0,
      "step": 12025
    },
    {
      "entropy": 5.860195589065552,
      "epoch": 1.062345460967856,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004893621377235777,
      "loss": 5.4167,
      "mean_token_accuracy": 0.1942832052707672,
      "num_tokens": 22426504.0,
      "step": 12030
    },
    {
      "entropy": 5.927928256988525,
      "epoch": 1.0627870010596963,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004893525670017364,
      "loss": 5.5063,
      "mean_token_accuracy": 0.18941380232572555,
      "num_tokens": 22435859.0,
      "step": 12035
    },
    {
      "entropy": 5.7999958992004395,
      "epoch": 1.0632285411515365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004893429920808503,
      "loss": 5.4235,
      "mean_token_accuracy": 0.2001870334148407,
      "num_tokens": 22445112.0,
      "step": 12040
    },
    {
      "entropy": 5.772587394714355,
      "epoch": 1.0636700812433768,
      "grad_norm": 0.953125,
      "learning_rate": 0.000489333412961107,
      "loss": 5.4947,
      "mean_token_accuracy": 0.19539366364479066,
      "num_tokens": 22454602.0,
      "step": 12045
    },
    {
      "entropy": 5.907477617263794,
      "epoch": 1.0641116213352173,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004893238296426939,
      "loss": 5.5175,
      "mean_token_accuracy": 0.19273462295532226,
      "num_tokens": 22463918.0,
      "step": 12050
    },
    {
      "entropy": 5.788310766220093,
      "epoch": 1.0645531614270576,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004893142421257989,
      "loss": 5.3967,
      "mean_token_accuracy": 0.1981455311179161,
      "num_tokens": 22473059.0,
      "step": 12055
    },
    {
      "entropy": 5.795585441589355,
      "epoch": 1.0649947015188979,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004893046504106099,
      "loss": 5.4705,
      "mean_token_accuracy": 0.192001049220562,
      "num_tokens": 22482615.0,
      "step": 12060
    },
    {
      "entropy": 5.797793292999268,
      "epoch": 1.0654362416107384,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004892950544973145,
      "loss": 5.3531,
      "mean_token_accuracy": 0.2064661979675293,
      "num_tokens": 22492141.0,
      "step": 12065
    },
    {
      "entropy": 5.835184335708618,
      "epoch": 1.0658777817025786,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004892854543861009,
      "loss": 5.3335,
      "mean_token_accuracy": 0.20551654249429702,
      "num_tokens": 22502084.0,
      "step": 12070
    },
    {
      "entropy": 5.767578411102295,
      "epoch": 1.066319321794419,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004892758500771572,
      "loss": 5.3795,
      "mean_token_accuracy": 0.19579511880874634,
      "num_tokens": 22510565.0,
      "step": 12075
    },
    {
      "entropy": 5.784418725967408,
      "epoch": 1.0667608618862592,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004892662415706714,
      "loss": 5.4789,
      "mean_token_accuracy": 0.19152628928422927,
      "num_tokens": 22520089.0,
      "step": 12080
    },
    {
      "entropy": 5.802230167388916,
      "epoch": 1.0672024019780997,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004892566288668319,
      "loss": 5.408,
      "mean_token_accuracy": 0.1952907308936119,
      "num_tokens": 22529008.0,
      "step": 12085
    },
    {
      "entropy": 5.837222623825073,
      "epoch": 1.06764394206994,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004892470119658267,
      "loss": 5.4007,
      "mean_token_accuracy": 0.20102533251047133,
      "num_tokens": 22538040.0,
      "step": 12090
    },
    {
      "entropy": 5.809632682800293,
      "epoch": 1.0680854821617802,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004892373908678445,
      "loss": 5.453,
      "mean_token_accuracy": 0.19296614080667496,
      "num_tokens": 22548786.0,
      "step": 12095
    },
    {
      "entropy": 5.81814136505127,
      "epoch": 1.0685270222536207,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004892277655730736,
      "loss": 5.4443,
      "mean_token_accuracy": 0.19625641107559205,
      "num_tokens": 22559841.0,
      "step": 12100
    },
    {
      "entropy": 5.818791198730469,
      "epoch": 1.068968562345461,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004892181360817027,
      "loss": 5.3932,
      "mean_token_accuracy": 0.20388062298297882,
      "num_tokens": 22569338.0,
      "step": 12105
    },
    {
      "entropy": 5.7399450778961185,
      "epoch": 1.0694101024373013,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004892085023939204,
      "loss": 5.4412,
      "mean_token_accuracy": 0.1898152142763138,
      "num_tokens": 22578184.0,
      "step": 12110
    },
    {
      "entropy": 5.8836915493011475,
      "epoch": 1.0698516425291416,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004891988645099151,
      "loss": 5.5404,
      "mean_token_accuracy": 0.18727049976587296,
      "num_tokens": 22587007.0,
      "step": 12115
    },
    {
      "entropy": 5.89788327217102,
      "epoch": 1.070293182620982,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004891892224298761,
      "loss": 5.4614,
      "mean_token_accuracy": 0.19142129123210908,
      "num_tokens": 22596898.0,
      "step": 12120
    },
    {
      "entropy": 5.7848447322845455,
      "epoch": 1.0707347227128223,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000489179576153992,
      "loss": 5.4116,
      "mean_token_accuracy": 0.20285348743200302,
      "num_tokens": 22605627.0,
      "step": 12125
    },
    {
      "entropy": 5.75311050415039,
      "epoch": 1.0711762628046626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004891699256824518,
      "loss": 5.3397,
      "mean_token_accuracy": 0.2017929509282112,
      "num_tokens": 22615269.0,
      "step": 12130
    },
    {
      "entropy": 5.791806364059449,
      "epoch": 1.071617802896503,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004891602710154446,
      "loss": 5.3771,
      "mean_token_accuracy": 0.2027752161026001,
      "num_tokens": 22625048.0,
      "step": 12135
    },
    {
      "entropy": 5.756436395645141,
      "epoch": 1.0720593429883434,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004891506121531593,
      "loss": 5.3711,
      "mean_token_accuracy": 0.20226978957653047,
      "num_tokens": 22634146.0,
      "step": 12140
    },
    {
      "entropy": 5.775600814819336,
      "epoch": 1.0725008830801837,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004891409490957855,
      "loss": 5.4698,
      "mean_token_accuracy": 0.19833989888429643,
      "num_tokens": 22642915.0,
      "step": 12145
    },
    {
      "entropy": 5.828043651580811,
      "epoch": 1.072942423172024,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004891312818435122,
      "loss": 5.4121,
      "mean_token_accuracy": 0.1949666276574135,
      "num_tokens": 22651546.0,
      "step": 12150
    },
    {
      "entropy": 5.841364526748658,
      "epoch": 1.0733839632638644,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000489121610396529,
      "loss": 5.4101,
      "mean_token_accuracy": 0.2037117898464203,
      "num_tokens": 22660712.0,
      "step": 12155
    },
    {
      "entropy": 5.8018913745880125,
      "epoch": 1.0738255033557047,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004891119347550251,
      "loss": 5.4237,
      "mean_token_accuracy": 0.20187718719244002,
      "num_tokens": 22669428.0,
      "step": 12160
    },
    {
      "entropy": 5.8284533500671385,
      "epoch": 1.074267043447545,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004891022549191902,
      "loss": 5.4883,
      "mean_token_accuracy": 0.1896545797586441,
      "num_tokens": 22679126.0,
      "step": 12165
    },
    {
      "entropy": 5.762653779983521,
      "epoch": 1.0747085835393855,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004890925708892138,
      "loss": 5.3786,
      "mean_token_accuracy": 0.2046400487422943,
      "num_tokens": 22688640.0,
      "step": 12170
    },
    {
      "entropy": 5.825605392456055,
      "epoch": 1.0751501236312258,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000489082882665286,
      "loss": 5.371,
      "mean_token_accuracy": 0.2033366784453392,
      "num_tokens": 22698158.0,
      "step": 12175
    },
    {
      "entropy": 5.729478168487549,
      "epoch": 1.075591663723066,
      "grad_norm": 0.984375,
      "learning_rate": 0.000489073190247596,
      "loss": 5.3131,
      "mean_token_accuracy": 0.2104171559214592,
      "num_tokens": 22707649.0,
      "step": 12180
    },
    {
      "entropy": 5.794720220565796,
      "epoch": 1.0760332038149063,
      "grad_norm": 1.0,
      "learning_rate": 0.000489063493636334,
      "loss": 5.4907,
      "mean_token_accuracy": 0.19447026252746583,
      "num_tokens": 22717123.0,
      "step": 12185
    },
    {
      "entropy": 5.749238109588623,
      "epoch": 1.0764747439067468,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004890537928316901,
      "loss": 5.4214,
      "mean_token_accuracy": 0.19723988473415374,
      "num_tokens": 22725921.0,
      "step": 12190
    },
    {
      "entropy": 5.85083327293396,
      "epoch": 1.076916283998587,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004890440878338538,
      "loss": 5.5186,
      "mean_token_accuracy": 0.19249518662691117,
      "num_tokens": 22735798.0,
      "step": 12195
    },
    {
      "entropy": 5.897446632385254,
      "epoch": 1.0773578240904274,
      "grad_norm": 1.0,
      "learning_rate": 0.0004890343786430158,
      "loss": 5.5028,
      "mean_token_accuracy": 0.19175300002098083,
      "num_tokens": 22745583.0,
      "step": 12200
    },
    {
      "entropy": 5.774647665023804,
      "epoch": 1.0777993641822678,
      "grad_norm": 1.03125,
      "learning_rate": 0.000489024665259366,
      "loss": 5.3356,
      "mean_token_accuracy": 0.2094974547624588,
      "num_tokens": 22754693.0,
      "step": 12205
    },
    {
      "entropy": 5.683815670013428,
      "epoch": 1.0782409042741081,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004890149476830947,
      "loss": 5.3428,
      "mean_token_accuracy": 0.206331467628479,
      "num_tokens": 22763416.0,
      "step": 12210
    },
    {
      "entropy": 5.8208612442016605,
      "epoch": 1.0786824443659484,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004890052259143924,
      "loss": 5.424,
      "mean_token_accuracy": 0.20380556285381318,
      "num_tokens": 22772411.0,
      "step": 12215
    },
    {
      "entropy": 5.7585056781768795,
      "epoch": 1.0791239844577887,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004889954999534493,
      "loss": 5.3263,
      "mean_token_accuracy": 0.2080024152994156,
      "num_tokens": 22780148.0,
      "step": 12220
    },
    {
      "entropy": 5.727291059494019,
      "epoch": 1.0795655245496292,
      "grad_norm": 1.0,
      "learning_rate": 0.0004889857698004561,
      "loss": 5.4002,
      "mean_token_accuracy": 0.19845105707645416,
      "num_tokens": 22789453.0,
      "step": 12225
    },
    {
      "entropy": 5.865949726104736,
      "epoch": 1.0800070646414694,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004889760354556033,
      "loss": 5.4346,
      "mean_token_accuracy": 0.20274429768323898,
      "num_tokens": 22798800.0,
      "step": 12230
    },
    {
      "entropy": 5.830954694747925,
      "epoch": 1.0804486047333097,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004889662969190817,
      "loss": 5.5088,
      "mean_token_accuracy": 0.19512779414653778,
      "num_tokens": 22807766.0,
      "step": 12235
    },
    {
      "entropy": 5.779366540908813,
      "epoch": 1.0808901448251502,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004889565541910822,
      "loss": 5.3753,
      "mean_token_accuracy": 0.1985454335808754,
      "num_tokens": 22817316.0,
      "step": 12240
    },
    {
      "entropy": 5.798041868209839,
      "epoch": 1.0813316849169905,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004889468072717953,
      "loss": 5.4642,
      "mean_token_accuracy": 0.2018450379371643,
      "num_tokens": 22827082.0,
      "step": 12245
    },
    {
      "entropy": 5.798942756652832,
      "epoch": 1.0817732250088308,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004889370561614122,
      "loss": 5.4404,
      "mean_token_accuracy": 0.1997752696275711,
      "num_tokens": 22836114.0,
      "step": 12250
    },
    {
      "entropy": 5.87620096206665,
      "epoch": 1.082214765100671,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004889273008601238,
      "loss": 5.5317,
      "mean_token_accuracy": 0.19290950894355774,
      "num_tokens": 22845653.0,
      "step": 12255
    },
    {
      "entropy": 5.8952672481536865,
      "epoch": 1.0826563051925115,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004889175413681213,
      "loss": 5.4682,
      "mean_token_accuracy": 0.19856464862823486,
      "num_tokens": 22855472.0,
      "step": 12260
    },
    {
      "entropy": 5.781057119369507,
      "epoch": 1.0830978452843518,
      "grad_norm": 1.0,
      "learning_rate": 0.0004889077776855957,
      "loss": 5.3743,
      "mean_token_accuracy": 0.2052726164460182,
      "num_tokens": 22864223.0,
      "step": 12265
    },
    {
      "entropy": 5.784585189819336,
      "epoch": 1.083539385376192,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004888980098127385,
      "loss": 5.4608,
      "mean_token_accuracy": 0.19553003907203675,
      "num_tokens": 22872868.0,
      "step": 12270
    },
    {
      "entropy": 5.870864582061768,
      "epoch": 1.0839809254680326,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004888882377497409,
      "loss": 5.4855,
      "mean_token_accuracy": 0.1898757666349411,
      "num_tokens": 22881867.0,
      "step": 12275
    },
    {
      "entropy": 5.829045009613037,
      "epoch": 1.0844224655598729,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004888784614967943,
      "loss": 5.4357,
      "mean_token_accuracy": 0.19417086243629456,
      "num_tokens": 22890904.0,
      "step": 12280
    },
    {
      "entropy": 5.743711328506469,
      "epoch": 1.0848640056517131,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004888686810540905,
      "loss": 5.3738,
      "mean_token_accuracy": 0.20171961784362794,
      "num_tokens": 22900128.0,
      "step": 12285
    },
    {
      "entropy": 5.797252750396728,
      "epoch": 1.0853055457435534,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004888588964218208,
      "loss": 5.4636,
      "mean_token_accuracy": 0.19641826003789903,
      "num_tokens": 22910101.0,
      "step": 12290
    },
    {
      "entropy": 5.815429210662842,
      "epoch": 1.085747085835394,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004888491076001768,
      "loss": 5.4332,
      "mean_token_accuracy": 0.19740083068609238,
      "num_tokens": 22919372.0,
      "step": 12295
    },
    {
      "entropy": 5.834107255935669,
      "epoch": 1.0861886259272342,
      "grad_norm": 1.0,
      "learning_rate": 0.0004888393145893505,
      "loss": 5.4362,
      "mean_token_accuracy": 0.19701868295669556,
      "num_tokens": 22929053.0,
      "step": 12300
    },
    {
      "entropy": 5.711127853393554,
      "epoch": 1.0866301660190745,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004888295173895335,
      "loss": 5.3759,
      "mean_token_accuracy": 0.19772911816835403,
      "num_tokens": 22938173.0,
      "step": 12305
    },
    {
      "entropy": 5.830015850067139,
      "epoch": 1.087071706110915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004888197160009181,
      "loss": 5.5007,
      "mean_token_accuracy": 0.19715835601091386,
      "num_tokens": 22947076.0,
      "step": 12310
    },
    {
      "entropy": 5.806154012680054,
      "epoch": 1.0875132462027552,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000488809910423696,
      "loss": 5.422,
      "mean_token_accuracy": 0.20476145893335343,
      "num_tokens": 22956267.0,
      "step": 12315
    },
    {
      "entropy": 5.898472595214844,
      "epoch": 1.0879547862945955,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004888001006580592,
      "loss": 5.5187,
      "mean_token_accuracy": 0.18518722355365752,
      "num_tokens": 22966751.0,
      "step": 12320
    },
    {
      "entropy": 5.761488008499145,
      "epoch": 1.0883963263864358,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004887902867042002,
      "loss": 5.4251,
      "mean_token_accuracy": 0.20033389180898667,
      "num_tokens": 22975862.0,
      "step": 12325
    },
    {
      "entropy": 5.790075588226318,
      "epoch": 1.0888378664782763,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004887804685623111,
      "loss": 5.3893,
      "mean_token_accuracy": 0.19841696918010712,
      "num_tokens": 22985085.0,
      "step": 12330
    },
    {
      "entropy": 5.807966661453247,
      "epoch": 1.0892794065701166,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004887706462325841,
      "loss": 5.5217,
      "mean_token_accuracy": 0.19649834632873536,
      "num_tokens": 22993718.0,
      "step": 12335
    },
    {
      "entropy": 5.7946618556976315,
      "epoch": 1.0897209466619568,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004887608197152117,
      "loss": 5.3899,
      "mean_token_accuracy": 0.20841073095798493,
      "num_tokens": 23002611.0,
      "step": 12340
    },
    {
      "entropy": 5.859386825561524,
      "epoch": 1.0901624867537973,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004887509890103864,
      "loss": 5.5398,
      "mean_token_accuracy": 0.1920286536216736,
      "num_tokens": 23012333.0,
      "step": 12345
    },
    {
      "entropy": 5.856261110305786,
      "epoch": 1.0906040268456376,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004887411541183009,
      "loss": 5.4803,
      "mean_token_accuracy": 0.20197425931692123,
      "num_tokens": 23020998.0,
      "step": 12350
    },
    {
      "entropy": 5.842471599578857,
      "epoch": 1.091045566937478,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004887313150391477,
      "loss": 5.4516,
      "mean_token_accuracy": 0.19660209268331527,
      "num_tokens": 23029520.0,
      "step": 12355
    },
    {
      "entropy": 5.792888975143432,
      "epoch": 1.0914871070293182,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004887214717731196,
      "loss": 5.5197,
      "mean_token_accuracy": 0.1904631733894348,
      "num_tokens": 23039224.0,
      "step": 12360
    },
    {
      "entropy": 5.844955587387085,
      "epoch": 1.0919286471211587,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004887116243204094,
      "loss": 5.4067,
      "mean_token_accuracy": 0.1981590136885643,
      "num_tokens": 23047851.0,
      "step": 12365
    },
    {
      "entropy": 5.739581966400147,
      "epoch": 1.092370187212999,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048870177268121,
      "loss": 5.4264,
      "mean_token_accuracy": 0.19422738552093505,
      "num_tokens": 23057242.0,
      "step": 12370
    },
    {
      "entropy": 5.841652679443359,
      "epoch": 1.0928117273048392,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004886919168557145,
      "loss": 5.5422,
      "mean_token_accuracy": 0.19310810416936874,
      "num_tokens": 23066873.0,
      "step": 12375
    },
    {
      "entropy": 5.895767641067505,
      "epoch": 1.0932532673966797,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004886820568441158,
      "loss": 5.5146,
      "mean_token_accuracy": 0.1886664405465126,
      "num_tokens": 23076448.0,
      "step": 12380
    },
    {
      "entropy": 5.869763708114624,
      "epoch": 1.09369480748852,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004886721926466072,
      "loss": 5.4382,
      "mean_token_accuracy": 0.1931287795305252,
      "num_tokens": 23084759.0,
      "step": 12385
    },
    {
      "entropy": 5.808056640625,
      "epoch": 1.0941363475803603,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004886623242633817,
      "loss": 5.3883,
      "mean_token_accuracy": 0.200452321767807,
      "num_tokens": 23093236.0,
      "step": 12390
    },
    {
      "entropy": 5.792661809921265,
      "epoch": 1.0945778876722005,
      "grad_norm": 1.0,
      "learning_rate": 0.000488652451694633,
      "loss": 5.4178,
      "mean_token_accuracy": 0.20028265118598937,
      "num_tokens": 23101621.0,
      "step": 12395
    },
    {
      "entropy": 5.79798321723938,
      "epoch": 1.095019427764041,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004886425749405542,
      "loss": 5.4712,
      "mean_token_accuracy": 0.19113578498363495,
      "num_tokens": 23110882.0,
      "step": 12400
    },
    {
      "entropy": 5.850175237655639,
      "epoch": 1.0954609678558813,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004886326940013388,
      "loss": 5.4552,
      "mean_token_accuracy": 0.19971661865711213,
      "num_tokens": 23120669.0,
      "step": 12405
    },
    {
      "entropy": 5.820252561569214,
      "epoch": 1.0959025079477216,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004886228088771805,
      "loss": 5.3615,
      "mean_token_accuracy": 0.19930000305175782,
      "num_tokens": 23129524.0,
      "step": 12410
    },
    {
      "entropy": 5.752618837356567,
      "epoch": 1.096344048039562,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004886129195682729,
      "loss": 5.4056,
      "mean_token_accuracy": 0.20056315809488295,
      "num_tokens": 23139144.0,
      "step": 12415
    },
    {
      "entropy": 5.736331605911255,
      "epoch": 1.0967855881314024,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004886030260748096,
      "loss": 5.3601,
      "mean_token_accuracy": 0.199568709731102,
      "num_tokens": 23148600.0,
      "step": 12420
    },
    {
      "entropy": 5.717127895355224,
      "epoch": 1.0972271282232426,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004885931283969845,
      "loss": 5.3429,
      "mean_token_accuracy": 0.20439921617507933,
      "num_tokens": 23156831.0,
      "step": 12425
    },
    {
      "entropy": 5.88601131439209,
      "epoch": 1.097668668315083,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004885832265349915,
      "loss": 5.5168,
      "mean_token_accuracy": 0.18720456659793855,
      "num_tokens": 23166453.0,
      "step": 12430
    },
    {
      "entropy": 5.818928289413452,
      "epoch": 1.0981102084069234,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004885733204890246,
      "loss": 5.3952,
      "mean_token_accuracy": 0.19314671605825423,
      "num_tokens": 23175832.0,
      "step": 12435
    },
    {
      "entropy": 5.794879484176636,
      "epoch": 1.0985517484987637,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004885634102592778,
      "loss": 5.3893,
      "mean_token_accuracy": 0.19797425717115402,
      "num_tokens": 23185349.0,
      "step": 12440
    },
    {
      "entropy": 5.81860728263855,
      "epoch": 1.098993288590604,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004885534958459454,
      "loss": 5.5292,
      "mean_token_accuracy": 0.2007902055978775,
      "num_tokens": 23195391.0,
      "step": 12445
    },
    {
      "entropy": 5.7743274688720705,
      "epoch": 1.0994348286824445,
      "grad_norm": 2.59375,
      "learning_rate": 0.0004885435772492213,
      "loss": 5.5042,
      "mean_token_accuracy": 0.1974121704697609,
      "num_tokens": 23204689.0,
      "step": 12450
    },
    {
      "entropy": 5.77637882232666,
      "epoch": 1.0998763687742847,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004885336544693,
      "loss": 5.4345,
      "mean_token_accuracy": 0.19952351003885269,
      "num_tokens": 23213182.0,
      "step": 12455
    },
    {
      "entropy": 5.736325311660766,
      "epoch": 1.100317908866125,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004885237275063758,
      "loss": 5.4312,
      "mean_token_accuracy": 0.19424163848161696,
      "num_tokens": 23222058.0,
      "step": 12460
    },
    {
      "entropy": 5.875259733200073,
      "epoch": 1.1007594489579653,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004885137963606433,
      "loss": 5.4669,
      "mean_token_accuracy": 0.2015385299921036,
      "num_tokens": 23231747.0,
      "step": 12465
    },
    {
      "entropy": 5.866737127304077,
      "epoch": 1.1012009890498058,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004885038610322969,
      "loss": 5.4063,
      "mean_token_accuracy": 0.19817569553852082,
      "num_tokens": 23240402.0,
      "step": 12470
    },
    {
      "entropy": 5.813427448272705,
      "epoch": 1.101642529141646,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004884939215215314,
      "loss": 5.4635,
      "mean_token_accuracy": 0.20315919071435928,
      "num_tokens": 23249344.0,
      "step": 12475
    },
    {
      "entropy": 5.760272884368897,
      "epoch": 1.1020840692334863,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004884839778285413,
      "loss": 5.4649,
      "mean_token_accuracy": 0.19652874320745467,
      "num_tokens": 23258157.0,
      "step": 12480
    },
    {
      "entropy": 5.8123131275177,
      "epoch": 1.1025256093253268,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004884740299535216,
      "loss": 5.4448,
      "mean_token_accuracy": 0.18557992577552795,
      "num_tokens": 23267798.0,
      "step": 12485
    },
    {
      "entropy": 5.865637493133545,
      "epoch": 1.1029671494171671,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000488464077896667,
      "loss": 5.485,
      "mean_token_accuracy": 0.19723096042871474,
      "num_tokens": 23276429.0,
      "step": 12490
    },
    {
      "entropy": 5.79170560836792,
      "epoch": 1.1034086895090074,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004884541216581724,
      "loss": 5.3605,
      "mean_token_accuracy": 0.20017392039299012,
      "num_tokens": 23285878.0,
      "step": 12495
    },
    {
      "entropy": 5.742550230026245,
      "epoch": 1.1038502296008477,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000488444161238233,
      "loss": 5.4062,
      "mean_token_accuracy": 0.20699722319841385,
      "num_tokens": 23295683.0,
      "step": 12500
    },
    {
      "entropy": 5.8427427291870115,
      "epoch": 1.1042917696926882,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004884341966370439,
      "loss": 5.4578,
      "mean_token_accuracy": 0.19892906546592712,
      "num_tokens": 23305539.0,
      "step": 12505
    },
    {
      "entropy": 5.772446918487549,
      "epoch": 1.1047333097845284,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004884242278548003,
      "loss": 5.4352,
      "mean_token_accuracy": 0.19846121519804,
      "num_tokens": 23314889.0,
      "step": 12510
    },
    {
      "entropy": 5.848331785202026,
      "epoch": 1.1051748498763687,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004884142548916975,
      "loss": 5.4896,
      "mean_token_accuracy": 0.1966036155819893,
      "num_tokens": 23324425.0,
      "step": 12515
    },
    {
      "entropy": 5.777439594268799,
      "epoch": 1.1056163899682092,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004884042777479308,
      "loss": 5.3814,
      "mean_token_accuracy": 0.19838157445192336,
      "num_tokens": 23332994.0,
      "step": 12520
    },
    {
      "entropy": 5.768653202056885,
      "epoch": 1.1060579300600495,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004883942964236957,
      "loss": 5.3175,
      "mean_token_accuracy": 0.20157167315483093,
      "num_tokens": 23342108.0,
      "step": 12525
    },
    {
      "entropy": 5.787688446044922,
      "epoch": 1.1064994701518898,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004883843109191877,
      "loss": 5.3548,
      "mean_token_accuracy": 0.2093011975288391,
      "num_tokens": 23351439.0,
      "step": 12530
    },
    {
      "entropy": 5.833836698532105,
      "epoch": 1.10694101024373,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004883743212346021,
      "loss": 5.4974,
      "mean_token_accuracy": 0.19144345372915267,
      "num_tokens": 23360591.0,
      "step": 12535
    },
    {
      "entropy": 5.824067211151123,
      "epoch": 1.1073825503355705,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004883643273701353,
      "loss": 5.4278,
      "mean_token_accuracy": 0.1985139235854149,
      "num_tokens": 23369854.0,
      "step": 12540
    },
    {
      "entropy": 5.834714078903199,
      "epoch": 1.1078240904274108,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004883543293259826,
      "loss": 5.5058,
      "mean_token_accuracy": 0.19206878244876863,
      "num_tokens": 23378071.0,
      "step": 12545
    },
    {
      "entropy": 5.8821605205535885,
      "epoch": 1.108265630519251,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004883443271023397,
      "loss": 5.4892,
      "mean_token_accuracy": 0.1873974084854126,
      "num_tokens": 23387083.0,
      "step": 12550
    },
    {
      "entropy": 5.770686149597168,
      "epoch": 1.1087071706110916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000488334320699403,
      "loss": 5.3106,
      "mean_token_accuracy": 0.2036597326397896,
      "num_tokens": 23395008.0,
      "step": 12555
    },
    {
      "entropy": 5.762484693527222,
      "epoch": 1.1091487107029319,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004883243101173681,
      "loss": 5.4334,
      "mean_token_accuracy": 0.19761996120214462,
      "num_tokens": 23404193.0,
      "step": 12560
    },
    {
      "entropy": 5.816957092285156,
      "epoch": 1.1095902507947721,
      "grad_norm": 1.0,
      "learning_rate": 0.0004883142953564315,
      "loss": 5.4231,
      "mean_token_accuracy": 0.19985931515693664,
      "num_tokens": 23413353.0,
      "step": 12565
    },
    {
      "entropy": 5.853426456451416,
      "epoch": 1.1100317908866124,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000488304276416789,
      "loss": 5.4938,
      "mean_token_accuracy": 0.19763486087322235,
      "num_tokens": 23422797.0,
      "step": 12570
    },
    {
      "entropy": 5.828439140319825,
      "epoch": 1.110473330978453,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004882942532986372,
      "loss": 5.5575,
      "mean_token_accuracy": 0.19543415904045106,
      "num_tokens": 23432705.0,
      "step": 12575
    },
    {
      "entropy": 5.825581216812134,
      "epoch": 1.1109148710702932,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004882842260021721,
      "loss": 5.4109,
      "mean_token_accuracy": 0.19751026034355162,
      "num_tokens": 23441613.0,
      "step": 12580
    },
    {
      "entropy": 5.7938634872436525,
      "epoch": 1.1113564111621335,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048827419452759045,
      "loss": 5.4806,
      "mean_token_accuracy": 0.19247215688228608,
      "num_tokens": 23450975.0,
      "step": 12585
    },
    {
      "entropy": 5.847980928421021,
      "epoch": 1.111797951253974,
      "grad_norm": 1.0,
      "learning_rate": 0.00048826415887508857,
      "loss": 5.4778,
      "mean_token_accuracy": 0.19287350475788118,
      "num_tokens": 23460646.0,
      "step": 12590
    },
    {
      "entropy": 5.9207230567932125,
      "epoch": 1.1122394913458142,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048825411904486303,
      "loss": 5.5497,
      "mean_token_accuracy": 0.18543048352003097,
      "num_tokens": 23470454.0,
      "step": 12595
    },
    {
      "entropy": 5.771830892562866,
      "epoch": 1.1126810314376545,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004882440750371108,
      "loss": 5.4194,
      "mean_token_accuracy": 0.20379505902528763,
      "num_tokens": 23480013.0,
      "step": 12600
    },
    {
      "entropy": 5.859670925140381,
      "epoch": 1.1131225715294948,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048823402685202824,
      "loss": 5.4247,
      "mean_token_accuracy": 0.19428976625204086,
      "num_tokens": 23489485.0,
      "step": 12605
    },
    {
      "entropy": 5.772221231460572,
      "epoch": 1.1135641116213353,
      "grad_norm": 1.0,
      "learning_rate": 0.0004882239744898124,
      "loss": 5.4156,
      "mean_token_accuracy": 0.2106514871120453,
      "num_tokens": 23498147.0,
      "step": 12610
    },
    {
      "entropy": 5.678761100769043,
      "epoch": 1.1140056517131756,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004882139179506602,
      "loss": 5.2947,
      "mean_token_accuracy": 0.20896689891815184,
      "num_tokens": 23506614.0,
      "step": 12615
    },
    {
      "entropy": 5.903944253921509,
      "epoch": 1.1144471918050158,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004882038572347685,
      "loss": 5.5749,
      "mean_token_accuracy": 0.19001982361078262,
      "num_tokens": 23514927.0,
      "step": 12620
    },
    {
      "entropy": 5.88766074180603,
      "epoch": 1.1148887318968563,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048819379234233464,
      "loss": 5.5857,
      "mean_token_accuracy": 0.1913965031504631,
      "num_tokens": 23525298.0,
      "step": 12625
    },
    {
      "entropy": 5.938671970367432,
      "epoch": 1.1153302719886966,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004881837232735556,
      "loss": 5.486,
      "mean_token_accuracy": 0.19408829957246782,
      "num_tokens": 23533995.0,
      "step": 12630
    },
    {
      "entropy": 5.736942863464355,
      "epoch": 1.1157718120805369,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004881736500286286,
      "loss": 5.4361,
      "mean_token_accuracy": 0.2004469931125641,
      "num_tokens": 23543900.0,
      "step": 12635
    },
    {
      "entropy": 5.816015386581421,
      "epoch": 1.1162133521723772,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048816357260775115,
      "loss": 5.5347,
      "mean_token_accuracy": 0.19486338496208191,
      "num_tokens": 23553790.0,
      "step": 12640
    },
    {
      "entropy": 5.809583520889282,
      "epoch": 1.1166548922642177,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004881534910111204,
      "loss": 5.4425,
      "mean_token_accuracy": 0.20176970213651657,
      "num_tokens": 23563473.0,
      "step": 12645
    },
    {
      "entropy": 5.83355016708374,
      "epoch": 1.117096432356058,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048814340523893406,
      "loss": 5.3513,
      "mean_token_accuracy": 0.21166206598281861,
      "num_tokens": 23572495.0,
      "step": 12650
    },
    {
      "entropy": 5.870338773727417,
      "epoch": 1.1175379724478982,
      "grad_norm": 0.875,
      "learning_rate": 0.00048813331529138965,
      "loss": 5.525,
      "mean_token_accuracy": 0.18290655612945556,
      "num_tokens": 23582945.0,
      "step": 12655
    },
    {
      "entropy": 5.860627746582031,
      "epoch": 1.1179795125397387,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004881232211686848,
      "loss": 5.5324,
      "mean_token_accuracy": 0.19012165814638138,
      "num_tokens": 23592370.0,
      "step": 12660
    },
    {
      "entropy": 5.8514732837677,
      "epoch": 1.118421052631579,
      "grad_norm": 1.0,
      "learning_rate": 0.0004881131228710172,
      "loss": 5.4836,
      "mean_token_accuracy": 0.18865697979927062,
      "num_tokens": 23601589.0,
      "step": 12665
    },
    {
      "entropy": 5.815238523483276,
      "epoch": 1.1188625927234193,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004881030203985848,
      "loss": 5.3981,
      "mean_token_accuracy": 0.19706527441740035,
      "num_tokens": 23611456.0,
      "step": 12670
    },
    {
      "entropy": 5.828798818588257,
      "epoch": 1.1193041328152595,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048809291375158544,
      "loss": 5.4252,
      "mean_token_accuracy": 0.19195756763219834,
      "num_tokens": 23620362.0,
      "step": 12675
    },
    {
      "entropy": 5.8671183586120605,
      "epoch": 1.1197456729071,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048808280293021703,
      "loss": 5.4716,
      "mean_token_accuracy": 0.19548938572406768,
      "num_tokens": 23629647.0,
      "step": 12680
    },
    {
      "entropy": 5.781561756134034,
      "epoch": 1.1201872129989403,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048807268793467774,
      "loss": 5.3719,
      "mean_token_accuracy": 0.19756822139024735,
      "num_tokens": 23639201.0,
      "step": 12685
    },
    {
      "entropy": 5.771619558334351,
      "epoch": 1.1206287530907806,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048806256876516565,
      "loss": 5.3465,
      "mean_token_accuracy": 0.20281566083431243,
      "num_tokens": 23648166.0,
      "step": 12690
    },
    {
      "entropy": 5.842432594299316,
      "epoch": 1.121070293182621,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048805244542187906,
      "loss": 5.5464,
      "mean_token_accuracy": 0.19111703783273698,
      "num_tokens": 23658478.0,
      "step": 12695
    },
    {
      "entropy": 5.863788509368897,
      "epoch": 1.1215118332744614,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004880423179050161,
      "loss": 5.3773,
      "mean_token_accuracy": 0.20170405209064485,
      "num_tokens": 23667934.0,
      "step": 12700
    },
    {
      "entropy": 5.767390775680542,
      "epoch": 1.1219533733663016,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004880321862147755,
      "loss": 5.4125,
      "mean_token_accuracy": 0.1967923253774643,
      "num_tokens": 23676164.0,
      "step": 12705
    },
    {
      "entropy": 5.8280364036560055,
      "epoch": 1.122394913458142,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048802205035135537,
      "loss": 5.5141,
      "mean_token_accuracy": 0.1940636858344078,
      "num_tokens": 23686223.0,
      "step": 12710
    },
    {
      "entropy": 5.829761028289795,
      "epoch": 1.1228364535499824,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048801191031495444,
      "loss": 5.5021,
      "mean_token_accuracy": 0.19006700962781906,
      "num_tokens": 23695043.0,
      "step": 12715
    },
    {
      "entropy": 5.852439260482788,
      "epoch": 1.1232779936418227,
      "grad_norm": 1.0,
      "learning_rate": 0.0004880017661057714,
      "loss": 5.5161,
      "mean_token_accuracy": 0.18853760063648223,
      "num_tokens": 23704248.0,
      "step": 12720
    },
    {
      "entropy": 5.858014678955078,
      "epoch": 1.123719533733663,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004879916177240049,
      "loss": 5.3975,
      "mean_token_accuracy": 0.19637110084295273,
      "num_tokens": 23713724.0,
      "step": 12725
    },
    {
      "entropy": 5.9444338321685795,
      "epoch": 1.1241610738255035,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004879814651698537,
      "loss": 5.63,
      "mean_token_accuracy": 0.17654719054698945,
      "num_tokens": 23723039.0,
      "step": 12730
    },
    {
      "entropy": 5.799656867980957,
      "epoch": 1.1246026139173437,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048797130844351673,
      "loss": 5.3791,
      "mean_token_accuracy": 0.19843216091394425,
      "num_tokens": 23732024.0,
      "step": 12735
    },
    {
      "entropy": 5.818072175979614,
      "epoch": 1.125044154009184,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048796114754519294,
      "loss": 5.5055,
      "mean_token_accuracy": 0.1969237744808197,
      "num_tokens": 23740739.0,
      "step": 12740
    },
    {
      "entropy": 5.825812911987304,
      "epoch": 1.1254856941010243,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048795098247508145,
      "loss": 5.4227,
      "mean_token_accuracy": 0.1959856018424034,
      "num_tokens": 23750405.0,
      "step": 12745
    },
    {
      "entropy": 5.781939506530762,
      "epoch": 1.1259272341928648,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004879408132333813,
      "loss": 5.4088,
      "mean_token_accuracy": 0.19907349944114686,
      "num_tokens": 23759318.0,
      "step": 12750
    },
    {
      "entropy": 5.84839825630188,
      "epoch": 1.126368774284705,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048793063982029166,
      "loss": 5.5572,
      "mean_token_accuracy": 0.18858904838562013,
      "num_tokens": 23769202.0,
      "step": 12755
    },
    {
      "entropy": 5.808833837509155,
      "epoch": 1.1268103143765453,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000487920462236012,
      "loss": 5.3966,
      "mean_token_accuracy": 0.20316065698862076,
      "num_tokens": 23778478.0,
      "step": 12760
    },
    {
      "entropy": 5.830659866333008,
      "epoch": 1.1272518544683856,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004879102804807415,
      "loss": 5.5015,
      "mean_token_accuracy": 0.19673565477132798,
      "num_tokens": 23788044.0,
      "step": 12765
    },
    {
      "entropy": 5.824640274047852,
      "epoch": 1.127693394560226,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048790009455467967,
      "loss": 5.4895,
      "mean_token_accuracy": 0.18861141949892044,
      "num_tokens": 23796351.0,
      "step": 12770
    },
    {
      "entropy": 5.808917093276977,
      "epoch": 1.1281349346520664,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004878899044580261,
      "loss": 5.4903,
      "mean_token_accuracy": 0.19645975679159164,
      "num_tokens": 23805553.0,
      "step": 12775
    },
    {
      "entropy": 5.785381555557251,
      "epoch": 1.1285764747439067,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048787971019098044,
      "loss": 5.3402,
      "mean_token_accuracy": 0.2021778866648674,
      "num_tokens": 23813742.0,
      "step": 12780
    },
    {
      "entropy": 5.764051151275635,
      "epoch": 1.1290180148357472,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004878695117537423,
      "loss": 5.4249,
      "mean_token_accuracy": 0.1989069864153862,
      "num_tokens": 23822135.0,
      "step": 12785
    },
    {
      "entropy": 5.809045600891113,
      "epoch": 1.1294595549275874,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004878593091465114,
      "loss": 5.4892,
      "mean_token_accuracy": 0.19658098220825196,
      "num_tokens": 23832021.0,
      "step": 12790
    },
    {
      "entropy": 5.8431620597839355,
      "epoch": 1.1299010950194277,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004878491023694877,
      "loss": 5.437,
      "mean_token_accuracy": 0.1971711277961731,
      "num_tokens": 23842432.0,
      "step": 12795
    },
    {
      "entropy": 5.868091726303101,
      "epoch": 1.1303426351112682,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004878388914228713,
      "loss": 5.5085,
      "mean_token_accuracy": 0.1874347910284996,
      "num_tokens": 23852172.0,
      "step": 12800
    },
    {
      "entropy": 5.765841627120972,
      "epoch": 1.1307841752031085,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004878286763068619,
      "loss": 5.3804,
      "mean_token_accuracy": 0.20340241342782975,
      "num_tokens": 23861178.0,
      "step": 12805
    },
    {
      "entropy": 5.834781455993652,
      "epoch": 1.1312257152949488,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004878184570216598,
      "loss": 5.4792,
      "mean_token_accuracy": 0.19326757192611693,
      "num_tokens": 23869998.0,
      "step": 12810
    },
    {
      "entropy": 5.782759332656861,
      "epoch": 1.131667255386789,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048780823356746523,
      "loss": 5.4056,
      "mean_token_accuracy": 0.1978372737765312,
      "num_tokens": 23880404.0,
      "step": 12815
    },
    {
      "entropy": 5.795777750015259,
      "epoch": 1.1321087954786295,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004877980059444784,
      "loss": 5.3639,
      "mean_token_accuracy": 0.20314182341098785,
      "num_tokens": 23889643.0,
      "step": 12820
    },
    {
      "entropy": 5.888383579254151,
      "epoch": 1.1325503355704698,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004877877741528996,
      "loss": 5.437,
      "mean_token_accuracy": 0.19794621467590331,
      "num_tokens": 23899264.0,
      "step": 12825
    },
    {
      "entropy": 5.832461547851563,
      "epoch": 1.13299187566231,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004877775381929294,
      "loss": 5.5,
      "mean_token_accuracy": 0.19860878586769104,
      "num_tokens": 23908685.0,
      "step": 12830
    },
    {
      "entropy": 5.782173013687133,
      "epoch": 1.1334334157541504,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004877672980647682,
      "loss": 5.4654,
      "mean_token_accuracy": 0.1986728385090828,
      "num_tokens": 23917838.0,
      "step": 12835
    },
    {
      "entropy": 5.813170957565307,
      "epoch": 1.1338749558459909,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048775705376861663,
      "loss": 5.4574,
      "mean_token_accuracy": 0.1970720797777176,
      "num_tokens": 23927498.0,
      "step": 12840
    },
    {
      "entropy": 5.8912781238555905,
      "epoch": 1.1343164959378311,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004877468053046753,
      "loss": 5.479,
      "mean_token_accuracy": 0.20567873418331145,
      "num_tokens": 23936205.0,
      "step": 12845
    },
    {
      "entropy": 5.78642258644104,
      "epoch": 1.1347580360296714,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048773655267314516,
      "loss": 5.3761,
      "mean_token_accuracy": 0.2015999123454094,
      "num_tokens": 23945897.0,
      "step": 12850
    },
    {
      "entropy": 5.814234733581543,
      "epoch": 1.135199576121512,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004877262958742268,
      "loss": 5.4265,
      "mean_token_accuracy": 0.1943811669945717,
      "num_tokens": 23954873.0,
      "step": 12855
    },
    {
      "entropy": 5.825142288208008,
      "epoch": 1.1356411162133522,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048771603490812144,
      "loss": 5.5165,
      "mean_token_accuracy": 0.19840817600488664,
      "num_tokens": 23964631.0,
      "step": 12860
    },
    {
      "entropy": 5.790509223937988,
      "epoch": 1.1360826563051925,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004877057697750299,
      "loss": 5.4067,
      "mean_token_accuracy": 0.19818204790353774,
      "num_tokens": 23973935.0,
      "step": 12865
    },
    {
      "entropy": 5.870797109603882,
      "epoch": 1.136524196397033,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004876955004751533,
      "loss": 5.5083,
      "mean_token_accuracy": 0.19151266813278198,
      "num_tokens": 23984340.0,
      "step": 12870
    },
    {
      "entropy": 5.942030000686645,
      "epoch": 1.1369657364888732,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048768522700869284,
      "loss": 5.5811,
      "mean_token_accuracy": 0.18916677087545394,
      "num_tokens": 23994387.0,
      "step": 12875
    },
    {
      "entropy": 5.837050342559815,
      "epoch": 1.1374072765807135,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048767494937584964,
      "loss": 5.3994,
      "mean_token_accuracy": 0.20068531930446626,
      "num_tokens": 24003391.0,
      "step": 12880
    },
    {
      "entropy": 5.788455772399902,
      "epoch": 1.1378488166725538,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048766466757682527,
      "loss": 5.3702,
      "mean_token_accuracy": 0.2022908478975296,
      "num_tokens": 24013501.0,
      "step": 12885
    },
    {
      "entropy": 5.749296617507935,
      "epoch": 1.1382903567643943,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004876543816118209,
      "loss": 5.3519,
      "mean_token_accuracy": 0.20435864776372908,
      "num_tokens": 24022675.0,
      "step": 12890
    },
    {
      "entropy": 5.877848482131958,
      "epoch": 1.1387318968562345,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048764409148103815,
      "loss": 5.5236,
      "mean_token_accuracy": 0.18736222982406617,
      "num_tokens": 24032542.0,
      "step": 12895
    },
    {
      "entropy": 5.7748109817504885,
      "epoch": 1.1391734369480748,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004876337971846787,
      "loss": 5.4126,
      "mean_token_accuracy": 0.19638270884752274,
      "num_tokens": 24041342.0,
      "step": 12900
    },
    {
      "entropy": 5.769016361236572,
      "epoch": 1.139614977039915,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000487623498722944,
      "loss": 5.4094,
      "mean_token_accuracy": 0.2004293456673622,
      "num_tokens": 24051429.0,
      "step": 12905
    },
    {
      "entropy": 5.872424840927124,
      "epoch": 1.1400565171317556,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004876131960960358,
      "loss": 5.434,
      "mean_token_accuracy": 0.19928459078073502,
      "num_tokens": 24061074.0,
      "step": 12910
    },
    {
      "entropy": 5.837236070632935,
      "epoch": 1.1404980572235959,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048760288930415615,
      "loss": 5.45,
      "mean_token_accuracy": 0.1986504539847374,
      "num_tokens": 24069622.0,
      "step": 12915
    },
    {
      "entropy": 5.816545200347901,
      "epoch": 1.1409395973154361,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048759257834750673,
      "loss": 5.4509,
      "mean_token_accuracy": 0.19788366854190825,
      "num_tokens": 24078393.0,
      "step": 12920
    },
    {
      "entropy": 5.840530014038086,
      "epoch": 1.1413811374072766,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004875822632262897,
      "loss": 5.4457,
      "mean_token_accuracy": 0.19962743520736695,
      "num_tokens": 24089157.0,
      "step": 12925
    },
    {
      "entropy": 5.794300603866577,
      "epoch": 1.141822677499117,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048757194394070697,
      "loss": 5.3888,
      "mean_token_accuracy": 0.2025819480419159,
      "num_tokens": 24098022.0,
      "step": 12930
    },
    {
      "entropy": 5.736756801605225,
      "epoch": 1.1422642175909572,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048756162049096077,
      "loss": 5.3845,
      "mean_token_accuracy": 0.19877981543540954,
      "num_tokens": 24107629.0,
      "step": 12935
    },
    {
      "entropy": 5.840106248855591,
      "epoch": 1.1427057576827977,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048755129287725334,
      "loss": 5.4387,
      "mean_token_accuracy": 0.19928207397460937,
      "num_tokens": 24116903.0,
      "step": 12940
    },
    {
      "entropy": 5.804965496063232,
      "epoch": 1.143147297774638,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048754096109978684,
      "loss": 5.393,
      "mean_token_accuracy": 0.19839546978473663,
      "num_tokens": 24126433.0,
      "step": 12945
    },
    {
      "entropy": 5.881588792800903,
      "epoch": 1.1435888378664782,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004875306251587639,
      "loss": 5.5899,
      "mean_token_accuracy": 0.18981537818908692,
      "num_tokens": 24136603.0,
      "step": 12950
    },
    {
      "entropy": 5.862384891510009,
      "epoch": 1.1440303779583185,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004875202850543869,
      "loss": 5.478,
      "mean_token_accuracy": 0.19205156415700914,
      "num_tokens": 24145656.0,
      "step": 12955
    },
    {
      "entropy": 5.814039945602417,
      "epoch": 1.144471918050159,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048750994078685837,
      "loss": 5.5157,
      "mean_token_accuracy": 0.19038341641426088,
      "num_tokens": 24155044.0,
      "step": 12960
    },
    {
      "entropy": 5.813987731933594,
      "epoch": 1.1449134581419993,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004874995923563809,
      "loss": 5.428,
      "mean_token_accuracy": 0.19938230216503144,
      "num_tokens": 24163987.0,
      "step": 12965
    },
    {
      "entropy": 5.818346261978149,
      "epoch": 1.1453549982338396,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004874892397631573,
      "loss": 5.3913,
      "mean_token_accuracy": 0.19698755741119384,
      "num_tokens": 24173542.0,
      "step": 12970
    },
    {
      "entropy": 5.792830610275269,
      "epoch": 1.1457965383256798,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004874788830073904,
      "loss": 5.3938,
      "mean_token_accuracy": 0.19177045077085494,
      "num_tokens": 24182785.0,
      "step": 12975
    },
    {
      "entropy": 5.849016237258911,
      "epoch": 1.1462380784175203,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000487468522089283,
      "loss": 5.4595,
      "mean_token_accuracy": 0.19423977583646773,
      "num_tokens": 24193477.0,
      "step": 12980
    },
    {
      "entropy": 5.897625732421875,
      "epoch": 1.1466796185093606,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048745815700903806,
      "loss": 5.4289,
      "mean_token_accuracy": 0.19301562160253524,
      "num_tokens": 24202289.0,
      "step": 12985
    },
    {
      "entropy": 5.756063318252563,
      "epoch": 1.147121158601201,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048744778776685865,
      "loss": 5.3365,
      "mean_token_accuracy": 0.2100023940205574,
      "num_tokens": 24212199.0,
      "step": 12990
    },
    {
      "entropy": 5.833570289611816,
      "epoch": 1.1475626986930414,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048743741436294794,
      "loss": 5.4837,
      "mean_token_accuracy": 0.1986352488398552,
      "num_tokens": 24220984.0,
      "step": 12995
    },
    {
      "entropy": 5.813465881347656,
      "epoch": 1.1480042387848817,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004874270367975091,
      "loss": 5.4207,
      "mean_token_accuracy": 0.19989872574806214,
      "num_tokens": 24229599.0,
      "step": 13000
    },
    {
      "entropy": 5.848421335220337,
      "epoch": 1.148445778876722,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004874166550707454,
      "loss": 5.4741,
      "mean_token_accuracy": 0.19258294403553008,
      "num_tokens": 24238721.0,
      "step": 13005
    },
    {
      "entropy": 5.874206924438477,
      "epoch": 1.1488873189685624,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00048740626918286026,
      "loss": 5.4957,
      "mean_token_accuracy": 0.18595235794782639,
      "num_tokens": 24248431.0,
      "step": 13010
    },
    {
      "entropy": 5.847845363616943,
      "epoch": 1.1493288590604027,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048739587913405706,
      "loss": 5.4279,
      "mean_token_accuracy": 0.2011121317744255,
      "num_tokens": 24257770.0,
      "step": 13015
    },
    {
      "entropy": 5.872161769866944,
      "epoch": 1.149770399152243,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048738548492453944,
      "loss": 5.4185,
      "mean_token_accuracy": 0.20037979483604432,
      "num_tokens": 24266402.0,
      "step": 13020
    },
    {
      "entropy": 5.807389497756958,
      "epoch": 1.1502119392440833,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004873750865545109,
      "loss": 5.3918,
      "mean_token_accuracy": 0.200962033867836,
      "num_tokens": 24275164.0,
      "step": 13025
    },
    {
      "entropy": 5.817873430252075,
      "epoch": 1.1506534793359238,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048736468402417525,
      "loss": 5.4917,
      "mean_token_accuracy": 0.1918845921754837,
      "num_tokens": 24285650.0,
      "step": 13030
    },
    {
      "entropy": 5.835181093215942,
      "epoch": 1.151095019427764,
      "grad_norm": 1.0,
      "learning_rate": 0.00048735427733373617,
      "loss": 5.4529,
      "mean_token_accuracy": 0.1928752064704895,
      "num_tokens": 24294745.0,
      "step": 13035
    },
    {
      "entropy": 5.873867845535278,
      "epoch": 1.1515365595196043,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048734386648339767,
      "loss": 5.4198,
      "mean_token_accuracy": 0.20396704971790314,
      "num_tokens": 24303837.0,
      "step": 13040
    },
    {
      "entropy": 5.818719005584716,
      "epoch": 1.1519780996114446,
      "grad_norm": 1.0,
      "learning_rate": 0.0004873334514733634,
      "loss": 5.5055,
      "mean_token_accuracy": 0.19916922450065613,
      "num_tokens": 24313417.0,
      "step": 13045
    },
    {
      "entropy": 5.8238574981689455,
      "epoch": 1.152419639703285,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004873230323038378,
      "loss": 5.4209,
      "mean_token_accuracy": 0.20040738433599473,
      "num_tokens": 24322439.0,
      "step": 13050
    },
    {
      "entropy": 5.865116310119629,
      "epoch": 1.1528611797951254,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004873126089750246,
      "loss": 5.497,
      "mean_token_accuracy": 0.19367455840110778,
      "num_tokens": 24332509.0,
      "step": 13055
    },
    {
      "entropy": 5.831319856643677,
      "epoch": 1.1533027198869656,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004873021814871282,
      "loss": 5.5122,
      "mean_token_accuracy": 0.18654235303401948,
      "num_tokens": 24342252.0,
      "step": 13060
    },
    {
      "entropy": 5.872332239151001,
      "epoch": 1.1537442599788061,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004872917498403528,
      "loss": 5.4796,
      "mean_token_accuracy": 0.20358186960220337,
      "num_tokens": 24352459.0,
      "step": 13065
    },
    {
      "entropy": 5.791362237930298,
      "epoch": 1.1541858000706464,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004872813140349028,
      "loss": 5.3531,
      "mean_token_accuracy": 0.20301172137260437,
      "num_tokens": 24361788.0,
      "step": 13070
    },
    {
      "entropy": 5.795202779769897,
      "epoch": 1.1546273401624867,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004872708740709825,
      "loss": 5.4713,
      "mean_token_accuracy": 0.19351743459701537,
      "num_tokens": 24372117.0,
      "step": 13075
    },
    {
      "entropy": 5.746895122528076,
      "epoch": 1.1550688802543272,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048726042994879656,
      "loss": 5.4098,
      "mean_token_accuracy": 0.199046029150486,
      "num_tokens": 24380833.0,
      "step": 13080
    },
    {
      "entropy": 5.847612810134888,
      "epoch": 1.1555104203461675,
      "grad_norm": 1.0,
      "learning_rate": 0.00048724998166854944,
      "loss": 5.4902,
      "mean_token_accuracy": 0.19934733361005783,
      "num_tokens": 24390258.0,
      "step": 13085
    },
    {
      "entropy": 5.8716787815094,
      "epoch": 1.1559519604380077,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000487239529230446,
      "loss": 5.5235,
      "mean_token_accuracy": 0.19019153714179993,
      "num_tokens": 24399359.0,
      "step": 13090
    },
    {
      "entropy": 5.849550867080689,
      "epoch": 1.156393500529848,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048722907263469084,
      "loss": 5.4572,
      "mean_token_accuracy": 0.20486650615930557,
      "num_tokens": 24408012.0,
      "step": 13095
    },
    {
      "entropy": 5.8003052234649655,
      "epoch": 1.1568350406216885,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004872186118814889,
      "loss": 5.4607,
      "mean_token_accuracy": 0.19436182975769042,
      "num_tokens": 24416804.0,
      "step": 13100
    },
    {
      "entropy": 5.683313798904419,
      "epoch": 1.1572765807135288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048720814697104497,
      "loss": 5.2753,
      "mean_token_accuracy": 0.2063358947634697,
      "num_tokens": 24425401.0,
      "step": 13105
    },
    {
      "entropy": 5.722922849655151,
      "epoch": 1.157718120805369,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048719767790356415,
      "loss": 5.3285,
      "mean_token_accuracy": 0.2067764699459076,
      "num_tokens": 24434473.0,
      "step": 13110
    },
    {
      "entropy": 5.92434253692627,
      "epoch": 1.1581596608972093,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048718720467925156,
      "loss": 5.5765,
      "mean_token_accuracy": 0.17944076806306838,
      "num_tokens": 24443154.0,
      "step": 13115
    },
    {
      "entropy": 5.810152959823609,
      "epoch": 1.1586012009890498,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004871767272983122,
      "loss": 5.399,
      "mean_token_accuracy": 0.20420801043510436,
      "num_tokens": 24451308.0,
      "step": 13120
    },
    {
      "entropy": 5.8501222133636475,
      "epoch": 1.1590427410808901,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004871662457609515,
      "loss": 5.4117,
      "mean_token_accuracy": 0.20110041052103042,
      "num_tokens": 24459997.0,
      "step": 13125
    },
    {
      "entropy": 5.81738691329956,
      "epoch": 1.1594842811727304,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004871557600673747,
      "loss": 5.4513,
      "mean_token_accuracy": 0.19508708715438844,
      "num_tokens": 24469448.0,
      "step": 13130
    },
    {
      "entropy": 5.780637216567993,
      "epoch": 1.1599258212645709,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048714527021778714,
      "loss": 5.4361,
      "mean_token_accuracy": 0.2001718685030937,
      "num_tokens": 24478304.0,
      "step": 13135
    },
    {
      "entropy": 5.796525287628174,
      "epoch": 1.1603673613564112,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004871347762123945,
      "loss": 5.3226,
      "mean_token_accuracy": 0.2066611498594284,
      "num_tokens": 24487300.0,
      "step": 13140
    },
    {
      "entropy": 5.790245914459229,
      "epoch": 1.1608089014482514,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048712427805140224,
      "loss": 5.4006,
      "mean_token_accuracy": 0.1998492196202278,
      "num_tokens": 24495893.0,
      "step": 13145
    },
    {
      "entropy": 5.792077970504761,
      "epoch": 1.161250441540092,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004871137757350159,
      "loss": 5.3971,
      "mean_token_accuracy": 0.19801437556743623,
      "num_tokens": 24504603.0,
      "step": 13150
    },
    {
      "entropy": 5.737908554077149,
      "epoch": 1.1616919816319322,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048710326926344153,
      "loss": 5.3665,
      "mean_token_accuracy": 0.2058656930923462,
      "num_tokens": 24513148.0,
      "step": 13155
    },
    {
      "entropy": 5.87685341835022,
      "epoch": 1.1621335217237725,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048709275863688455,
      "loss": 5.532,
      "mean_token_accuracy": 0.18852106034755706,
      "num_tokens": 24524006.0,
      "step": 13160
    },
    {
      "entropy": 5.790403509140015,
      "epoch": 1.1625750618156128,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048708224385555116,
      "loss": 5.3681,
      "mean_token_accuracy": 0.20684442520141602,
      "num_tokens": 24532468.0,
      "step": 13165
    },
    {
      "entropy": 5.831372356414795,
      "epoch": 1.1630166019074533,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004870717249196472,
      "loss": 5.4618,
      "mean_token_accuracy": 0.1951098173856735,
      "num_tokens": 24540887.0,
      "step": 13170
    },
    {
      "entropy": 5.873029279708862,
      "epoch": 1.1634581419992935,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004870612018293788,
      "loss": 5.4052,
      "mean_token_accuracy": 0.19676822125911714,
      "num_tokens": 24550565.0,
      "step": 13175
    },
    {
      "entropy": 5.828689861297607,
      "epoch": 1.1638996820911338,
      "grad_norm": 0.9375,
      "learning_rate": 0.000487050674584952,
      "loss": 5.3904,
      "mean_token_accuracy": 0.20156201124191284,
      "num_tokens": 24560283.0,
      "step": 13180
    },
    {
      "entropy": 5.754884147644043,
      "epoch": 1.164341222182974,
      "grad_norm": 1.109375,
      "learning_rate": 0.000487040143186573,
      "loss": 5.3399,
      "mean_token_accuracy": 0.20346760600805283,
      "num_tokens": 24569527.0,
      "step": 13185
    },
    {
      "entropy": 5.808266115188599,
      "epoch": 1.1647827622748146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048702960763444833,
      "loss": 5.4317,
      "mean_token_accuracy": 0.19581593424081803,
      "num_tokens": 24578702.0,
      "step": 13190
    },
    {
      "entropy": 5.844560813903809,
      "epoch": 1.1652243023666549,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004870190679287842,
      "loss": 5.4059,
      "mean_token_accuracy": 0.20308234989643098,
      "num_tokens": 24588252.0,
      "step": 13195
    },
    {
      "entropy": 5.775644016265869,
      "epoch": 1.1656658424584951,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004870085240697871,
      "loss": 5.3954,
      "mean_token_accuracy": 0.19660432189702987,
      "num_tokens": 24596989.0,
      "step": 13200
    },
    {
      "entropy": 5.801691913604737,
      "epoch": 1.1661073825503356,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048699797605766365,
      "loss": 5.512,
      "mean_token_accuracy": 0.19366351664066314,
      "num_tokens": 24607051.0,
      "step": 13205
    },
    {
      "entropy": 5.912935543060303,
      "epoch": 1.166548922642176,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048698742389262036,
      "loss": 5.5146,
      "mean_token_accuracy": 0.19013242274522782,
      "num_tokens": 24616561.0,
      "step": 13210
    },
    {
      "entropy": 5.820238876342773,
      "epoch": 1.1669904627340162,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048697686757486397,
      "loss": 5.4595,
      "mean_token_accuracy": 0.1996126264333725,
      "num_tokens": 24625942.0,
      "step": 13215
    },
    {
      "entropy": 5.82216625213623,
      "epoch": 1.1674320028258567,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004869663071046013,
      "loss": 5.4288,
      "mean_token_accuracy": 0.19304953664541244,
      "num_tokens": 24635646.0,
      "step": 13220
    },
    {
      "entropy": 5.787486600875854,
      "epoch": 1.167873542917697,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048695574248203935,
      "loss": 5.4145,
      "mean_token_accuracy": 0.20401907563209534,
      "num_tokens": 24645147.0,
      "step": 13225
    },
    {
      "entropy": 5.8801429748535154,
      "epoch": 1.1683150830095372,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004869451737073848,
      "loss": 5.4355,
      "mean_token_accuracy": 0.19647326022386552,
      "num_tokens": 24654796.0,
      "step": 13230
    },
    {
      "entropy": 5.8170005798339846,
      "epoch": 1.1687566231013775,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004869346007808449,
      "loss": 5.4041,
      "mean_token_accuracy": 0.1959619030356407,
      "num_tokens": 24663765.0,
      "step": 13235
    },
    {
      "entropy": 5.8346545696258545,
      "epoch": 1.169198163193218,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048692402370262674,
      "loss": 5.4977,
      "mean_token_accuracy": 0.18846199810504913,
      "num_tokens": 24673758.0,
      "step": 13240
    },
    {
      "entropy": 5.835272121429443,
      "epoch": 1.1696397032850583,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048691344247293747,
      "loss": 5.3756,
      "mean_token_accuracy": 0.19807292222976686,
      "num_tokens": 24682751.0,
      "step": 13245
    },
    {
      "entropy": 5.814353513717651,
      "epoch": 1.1700812433768986,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048690285709198435,
      "loss": 5.466,
      "mean_token_accuracy": 0.19401862025260924,
      "num_tokens": 24691375.0,
      "step": 13250
    },
    {
      "entropy": 5.8145843029022215,
      "epoch": 1.1705227834687388,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004868922675599747,
      "loss": 5.4288,
      "mean_token_accuracy": 0.20230161547660827,
      "num_tokens": 24699852.0,
      "step": 13255
    },
    {
      "entropy": 5.785641479492187,
      "epoch": 1.1709643235605793,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004868816738771162,
      "loss": 5.3133,
      "mean_token_accuracy": 0.2019198477268219,
      "num_tokens": 24708857.0,
      "step": 13260
    },
    {
      "entropy": 5.882981061935425,
      "epoch": 1.1714058636524196,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004868710760436161,
      "loss": 5.5256,
      "mean_token_accuracy": 0.18815020024776458,
      "num_tokens": 24718262.0,
      "step": 13265
    },
    {
      "entropy": 5.868311262130737,
      "epoch": 1.1718474037442599,
      "grad_norm": 0.921875,
      "learning_rate": 0.00048686047405968217,
      "loss": 5.4504,
      "mean_token_accuracy": 0.20123834460973739,
      "num_tokens": 24727458.0,
      "step": 13270
    },
    {
      "entropy": 5.921502304077149,
      "epoch": 1.1722889438361004,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048684986792552206,
      "loss": 5.5357,
      "mean_token_accuracy": 0.1842239499092102,
      "num_tokens": 24737297.0,
      "step": 13275
    },
    {
      "entropy": 5.888692378997803,
      "epoch": 1.1727304839279407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048683925764134346,
      "loss": 5.4004,
      "mean_token_accuracy": 0.20313832759857178,
      "num_tokens": 24746939.0,
      "step": 13280
    },
    {
      "entropy": 5.844958543777466,
      "epoch": 1.173172024019781,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004868286432073543,
      "loss": 5.5455,
      "mean_token_accuracy": 0.19567487835884095,
      "num_tokens": 24756907.0,
      "step": 13285
    },
    {
      "entropy": 5.803437566757202,
      "epoch": 1.1736135641116214,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048681802462376253,
      "loss": 5.3591,
      "mean_token_accuracy": 0.20153522938489915,
      "num_tokens": 24765997.0,
      "step": 13290
    },
    {
      "entropy": 5.88543848991394,
      "epoch": 1.1740551042034617,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048680740189077605,
      "loss": 5.4523,
      "mean_token_accuracy": 0.1859741896390915,
      "num_tokens": 24775090.0,
      "step": 13295
    },
    {
      "entropy": 5.79713716506958,
      "epoch": 1.174496644295302,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048679677500860314,
      "loss": 5.4017,
      "mean_token_accuracy": 0.1957128718495369,
      "num_tokens": 24784709.0,
      "step": 13300
    },
    {
      "entropy": 5.894897937774658,
      "epoch": 1.1749381843871423,
      "grad_norm": 1.0,
      "learning_rate": 0.0004867861439774519,
      "loss": 5.4576,
      "mean_token_accuracy": 0.2024698406457901,
      "num_tokens": 24794473.0,
      "step": 13305
    },
    {
      "entropy": 5.848655700683594,
      "epoch": 1.1753797244789828,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048677550879753037,
      "loss": 5.5414,
      "mean_token_accuracy": 0.1884506344795227,
      "num_tokens": 24804718.0,
      "step": 13310
    },
    {
      "entropy": 5.8445494174957275,
      "epoch": 1.175821264570823,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048676486946904715,
      "loss": 5.4439,
      "mean_token_accuracy": 0.19045979976654054,
      "num_tokens": 24813632.0,
      "step": 13315
    },
    {
      "entropy": 5.774673843383789,
      "epoch": 1.1762628046626633,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004867542259922105,
      "loss": 5.353,
      "mean_token_accuracy": 0.20659865587949752,
      "num_tokens": 24823395.0,
      "step": 13320
    },
    {
      "entropy": 5.863358068466186,
      "epoch": 1.1767043447545038,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004867435783672291,
      "loss": 5.4796,
      "mean_token_accuracy": 0.1941860318183899,
      "num_tokens": 24832869.0,
      "step": 13325
    },
    {
      "entropy": 5.850523662567139,
      "epoch": 1.177145884846344,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004867329265943114,
      "loss": 5.4138,
      "mean_token_accuracy": 0.19960284531116484,
      "num_tokens": 24841706.0,
      "step": 13330
    },
    {
      "entropy": 5.869169807434082,
      "epoch": 1.1775874249381844,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004867222706736661,
      "loss": 5.4558,
      "mean_token_accuracy": 0.1981481581926346,
      "num_tokens": 24851149.0,
      "step": 13335
    },
    {
      "entropy": 5.851470041275024,
      "epoch": 1.1780289650300246,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004867116106055019,
      "loss": 5.4275,
      "mean_token_accuracy": 0.19862915128469466,
      "num_tokens": 24860002.0,
      "step": 13340
    },
    {
      "entropy": 5.807071018218994,
      "epoch": 1.1784705051218651,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048670094639002766,
      "loss": 5.4879,
      "mean_token_accuracy": 0.19347885251045227,
      "num_tokens": 24869457.0,
      "step": 13345
    },
    {
      "entropy": 5.80014934539795,
      "epoch": 1.1789120452137054,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004866902780274523,
      "loss": 5.4097,
      "mean_token_accuracy": 0.1995387315750122,
      "num_tokens": 24879527.0,
      "step": 13350
    },
    {
      "entropy": 5.80926661491394,
      "epoch": 1.1793535853055457,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004866796055179847,
      "loss": 5.43,
      "mean_token_accuracy": 0.2028554856777191,
      "num_tokens": 24889401.0,
      "step": 13355
    },
    {
      "entropy": 5.846627616882325,
      "epoch": 1.1797951253973862,
      "grad_norm": 1.0,
      "learning_rate": 0.0004866689288618342,
      "loss": 5.4328,
      "mean_token_accuracy": 0.2031128704547882,
      "num_tokens": 24897911.0,
      "step": 13360
    },
    {
      "entropy": 5.832061147689819,
      "epoch": 1.1802366654892265,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004866582480592096,
      "loss": 5.4038,
      "mean_token_accuracy": 0.20462207198143006,
      "num_tokens": 24906969.0,
      "step": 13365
    },
    {
      "entropy": 5.841529798507691,
      "epoch": 1.1806782055810667,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004866475631103203,
      "loss": 5.4327,
      "mean_token_accuracy": 0.20104020237922668,
      "num_tokens": 24915713.0,
      "step": 13370
    },
    {
      "entropy": 5.822256994247437,
      "epoch": 1.181119745672907,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048663687401537573,
      "loss": 5.4291,
      "mean_token_accuracy": 0.19380810260772705,
      "num_tokens": 24925484.0,
      "step": 13375
    },
    {
      "entropy": 5.839131879806518,
      "epoch": 1.1815612857647475,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048662618077458507,
      "loss": 5.4626,
      "mean_token_accuracy": 0.19699358493089675,
      "num_tokens": 24934751.0,
      "step": 13380
    },
    {
      "entropy": 5.82229175567627,
      "epoch": 1.1820028258565878,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004866154833881579,
      "loss": 5.5052,
      "mean_token_accuracy": 0.19295311272144317,
      "num_tokens": 24945059.0,
      "step": 13385
    },
    {
      "entropy": 5.831136798858642,
      "epoch": 1.182444365948428,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004866047818563038,
      "loss": 5.4574,
      "mean_token_accuracy": 0.19463883936405182,
      "num_tokens": 24954678.0,
      "step": 13390
    },
    {
      "entropy": 5.803921413421631,
      "epoch": 1.1828859060402686,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004865940761792323,
      "loss": 5.4535,
      "mean_token_accuracy": 0.19376982003450394,
      "num_tokens": 24964093.0,
      "step": 13395
    },
    {
      "entropy": 5.877642822265625,
      "epoch": 1.1833274461321088,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004865833663571533,
      "loss": 5.3945,
      "mean_token_accuracy": 0.205782550573349,
      "num_tokens": 24973903.0,
      "step": 13400
    },
    {
      "entropy": 5.778797292709351,
      "epoch": 1.183768986223949,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048657265239027635,
      "loss": 5.4001,
      "mean_token_accuracy": 0.20267367660999297,
      "num_tokens": 24982458.0,
      "step": 13405
    },
    {
      "entropy": 5.778594207763672,
      "epoch": 1.1842105263157894,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004865619342788116,
      "loss": 5.4687,
      "mean_token_accuracy": 0.1938798740506172,
      "num_tokens": 24992233.0,
      "step": 13410
    },
    {
      "entropy": 5.855686664581299,
      "epoch": 1.1846520664076299,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048655121202296873,
      "loss": 5.3952,
      "mean_token_accuracy": 0.19196417480707167,
      "num_tokens": 25001855.0,
      "step": 13415
    },
    {
      "entropy": 5.785927772521973,
      "epoch": 1.1850936064994702,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048654048562295807,
      "loss": 5.2949,
      "mean_token_accuracy": 0.2096828281879425,
      "num_tokens": 25010950.0,
      "step": 13420
    },
    {
      "entropy": 5.814057970046997,
      "epoch": 1.1855351465913104,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004865297550789895,
      "loss": 5.4715,
      "mean_token_accuracy": 0.18954818695783615,
      "num_tokens": 25020816.0,
      "step": 13425
    },
    {
      "entropy": 5.9006274223327635,
      "epoch": 1.185976686683151,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004865190203912734,
      "loss": 5.4941,
      "mean_token_accuracy": 0.19649789482355118,
      "num_tokens": 25030236.0,
      "step": 13430
    },
    {
      "entropy": 5.812299203872681,
      "epoch": 1.1864182267749912,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048650828156001995,
      "loss": 5.3908,
      "mean_token_accuracy": 0.19811263233423232,
      "num_tokens": 25040236.0,
      "step": 13435
    },
    {
      "entropy": 5.826809358596802,
      "epoch": 1.1868597668668315,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004864975385854396,
      "loss": 5.5173,
      "mean_token_accuracy": 0.19030961990356446,
      "num_tokens": 25049835.0,
      "step": 13440
    },
    {
      "entropy": 5.840577840805054,
      "epoch": 1.1873013069586718,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048648679146774275,
      "loss": 5.4732,
      "mean_token_accuracy": 0.19010272473096848,
      "num_tokens": 25058947.0,
      "step": 13445
    },
    {
      "entropy": 5.888551378250122,
      "epoch": 1.1877428470505123,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004864760402071399,
      "loss": 5.4357,
      "mean_token_accuracy": 0.1927935943007469,
      "num_tokens": 25069128.0,
      "step": 13450
    },
    {
      "entropy": 5.83781099319458,
      "epoch": 1.1881843871423525,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048646528480384175,
      "loss": 5.4931,
      "mean_token_accuracy": 0.18901222944259644,
      "num_tokens": 25079717.0,
      "step": 13455
    },
    {
      "entropy": 5.816900825500488,
      "epoch": 1.1886259272341928,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004864545252580588,
      "loss": 5.4085,
      "mean_token_accuracy": 0.2013822078704834,
      "num_tokens": 25088656.0,
      "step": 13460
    },
    {
      "entropy": 5.830460453033448,
      "epoch": 1.1890674673260333,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048644376157000213,
      "loss": 5.4242,
      "mean_token_accuracy": 0.19458205997943878,
      "num_tokens": 25098066.0,
      "step": 13465
    },
    {
      "entropy": 5.811070537567138,
      "epoch": 1.1895090074178736,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004864329937398823,
      "loss": 5.4383,
      "mean_token_accuracy": 0.19935468584299088,
      "num_tokens": 25107530.0,
      "step": 13470
    },
    {
      "entropy": 5.857369756698608,
      "epoch": 1.1899505475097139,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048642222176791043,
      "loss": 5.38,
      "mean_token_accuracy": 0.20258511751890182,
      "num_tokens": 25116226.0,
      "step": 13475
    },
    {
      "entropy": 5.8830688953399655,
      "epoch": 1.1903920876015541,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004864114456542975,
      "loss": 5.5573,
      "mean_token_accuracy": 0.18713556379079818,
      "num_tokens": 25125619.0,
      "step": 13480
    },
    {
      "entropy": 5.823305940628051,
      "epoch": 1.1908336276933946,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004864006653992546,
      "loss": 5.4095,
      "mean_token_accuracy": 0.20192137360572815,
      "num_tokens": 25134782.0,
      "step": 13485
    },
    {
      "entropy": 5.832695245742798,
      "epoch": 1.191275167785235,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048638988100299283,
      "loss": 5.4645,
      "mean_token_accuracy": 0.20090868026018144,
      "num_tokens": 25144709.0,
      "step": 13490
    },
    {
      "entropy": 5.8342750549316404,
      "epoch": 1.1917167078770752,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004863790924657236,
      "loss": 5.3343,
      "mean_token_accuracy": 0.1985476642847061,
      "num_tokens": 25153771.0,
      "step": 13495
    },
    {
      "entropy": 5.809533023834229,
      "epoch": 1.1921582479689157,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000486368299787658,
      "loss": 5.4937,
      "mean_token_accuracy": 0.19303596913814544,
      "num_tokens": 25163217.0,
      "step": 13500
    },
    {
      "entropy": 5.884714937210083,
      "epoch": 1.192599788060756,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048635750296900784,
      "loss": 5.4969,
      "mean_token_accuracy": 0.19041768461465836,
      "num_tokens": 25172762.0,
      "step": 13505
    },
    {
      "entropy": 5.850723314285278,
      "epoch": 1.1930413281525962,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004863467020099842,
      "loss": 5.4249,
      "mean_token_accuracy": 0.19885963052511216,
      "num_tokens": 25182349.0,
      "step": 13510
    },
    {
      "entropy": 5.875653553009033,
      "epoch": 1.1934828682444365,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048633589691079895,
      "loss": 5.4968,
      "mean_token_accuracy": 0.19478487372398376,
      "num_tokens": 25191971.0,
      "step": 13515
    },
    {
      "entropy": 5.709768342971802,
      "epoch": 1.193924408336277,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004863250876716637,
      "loss": 5.2704,
      "mean_token_accuracy": 0.20902594178915024,
      "num_tokens": 25200476.0,
      "step": 13520
    },
    {
      "entropy": 5.825209093093872,
      "epoch": 1.1943659484281173,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004863142742927902,
      "loss": 5.3754,
      "mean_token_accuracy": 0.19799334704875945,
      "num_tokens": 25209748.0,
      "step": 13525
    },
    {
      "entropy": 5.873579788208008,
      "epoch": 1.1948074885199576,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004863034567743901,
      "loss": 5.4729,
      "mean_token_accuracy": 0.19570437520742417,
      "num_tokens": 25220140.0,
      "step": 13530
    },
    {
      "entropy": 5.757551956176758,
      "epoch": 1.195249028611798,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004862926351166756,
      "loss": 5.3545,
      "mean_token_accuracy": 0.20685475468635559,
      "num_tokens": 25228424.0,
      "step": 13535
    },
    {
      "entropy": 5.827530431747436,
      "epoch": 1.1956905687036383,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004862818093198585,
      "loss": 5.4502,
      "mean_token_accuracy": 0.19304365217685698,
      "num_tokens": 25237448.0,
      "step": 13540
    },
    {
      "entropy": 5.762707901000977,
      "epoch": 1.1961321087954786,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004862709793841509,
      "loss": 5.3744,
      "mean_token_accuracy": 0.20597589612007142,
      "num_tokens": 25247124.0,
      "step": 13545
    },
    {
      "entropy": 5.749617958068848,
      "epoch": 1.1965736488873189,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048626014530976495,
      "loss": 5.4288,
      "mean_token_accuracy": 0.19768060892820358,
      "num_tokens": 25257363.0,
      "step": 13550
    },
    {
      "entropy": 5.8215718269348145,
      "epoch": 1.1970151889791594,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048624930709691296,
      "loss": 5.4882,
      "mean_token_accuracy": 0.19138520658016206,
      "num_tokens": 25266837.0,
      "step": 13555
    },
    {
      "entropy": 5.813835334777832,
      "epoch": 1.1974567290709996,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048623846474580716,
      "loss": 5.522,
      "mean_token_accuracy": 0.18580807745456696,
      "num_tokens": 25275767.0,
      "step": 13560
    },
    {
      "entropy": 5.833759355545044,
      "epoch": 1.19789826916284,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048622761825665994,
      "loss": 5.4664,
      "mean_token_accuracy": 0.19401149749755858,
      "num_tokens": 25285212.0,
      "step": 13565
    },
    {
      "entropy": 5.826594257354737,
      "epoch": 1.1983398092546804,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004862167676296839,
      "loss": 5.3651,
      "mean_token_accuracy": 0.1992584228515625,
      "num_tokens": 25294698.0,
      "step": 13570
    },
    {
      "entropy": 5.805987977981568,
      "epoch": 1.1987813493465207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004862059128650915,
      "loss": 5.4243,
      "mean_token_accuracy": 0.20547815412282944,
      "num_tokens": 25304343.0,
      "step": 13575
    },
    {
      "entropy": 5.828811454772949,
      "epoch": 1.199222889438361,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048619505396309525,
      "loss": 5.3728,
      "mean_token_accuracy": 0.19790565669536592,
      "num_tokens": 25312542.0,
      "step": 13580
    },
    {
      "entropy": 5.837662649154663,
      "epoch": 1.1996644295302012,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004861841909239081,
      "loss": 5.468,
      "mean_token_accuracy": 0.1954321324825287,
      "num_tokens": 25322665.0,
      "step": 13585
    },
    {
      "entropy": 5.810902881622314,
      "epoch": 1.2001059696220417,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048617332374774276,
      "loss": 5.4302,
      "mean_token_accuracy": 0.19678546935319902,
      "num_tokens": 25332234.0,
      "step": 13590
    },
    {
      "entropy": 5.852847099304199,
      "epoch": 1.200547509713882,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048616245243481216,
      "loss": 5.4885,
      "mean_token_accuracy": 0.19284352660179138,
      "num_tokens": 25341462.0,
      "step": 13595
    },
    {
      "entropy": 5.844420433044434,
      "epoch": 1.2009890498057223,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048615157698532913,
      "loss": 5.4582,
      "mean_token_accuracy": 0.19950661957263946,
      "num_tokens": 25351771.0,
      "step": 13600
    },
    {
      "entropy": 5.7953503131866455,
      "epoch": 1.2014305898975628,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048614069739950683,
      "loss": 5.4097,
      "mean_token_accuracy": 0.1965176656842232,
      "num_tokens": 25360108.0,
      "step": 13605
    },
    {
      "entropy": 5.77955436706543,
      "epoch": 1.201872129989403,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004861298136775583,
      "loss": 5.4063,
      "mean_token_accuracy": 0.20105797052383423,
      "num_tokens": 25369635.0,
      "step": 13610
    },
    {
      "entropy": 5.78814663887024,
      "epoch": 1.2023136700812433,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004861189258196969,
      "loss": 5.3736,
      "mean_token_accuracy": 0.20672035217285156,
      "num_tokens": 25379623.0,
      "step": 13615
    },
    {
      "entropy": 5.868280029296875,
      "epoch": 1.2027552101730836,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004861080338261357,
      "loss": 5.4723,
      "mean_token_accuracy": 0.1960707888007164,
      "num_tokens": 25388591.0,
      "step": 13620
    },
    {
      "entropy": 5.8137904644012455,
      "epoch": 1.2031967502649241,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004860971376970882,
      "loss": 5.3815,
      "mean_token_accuracy": 0.20670562088489533,
      "num_tokens": 25397865.0,
      "step": 13625
    },
    {
      "entropy": 5.714167308807373,
      "epoch": 1.2036382903567644,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048608623743276794,
      "loss": 5.3254,
      "mean_token_accuracy": 0.204473876953125,
      "num_tokens": 25407143.0,
      "step": 13630
    },
    {
      "entropy": 5.8470540046691895,
      "epoch": 1.2040798304486047,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048607533303338823,
      "loss": 5.4608,
      "mean_token_accuracy": 0.19450926780700684,
      "num_tokens": 25415721.0,
      "step": 13635
    },
    {
      "entropy": 5.823157882690429,
      "epoch": 1.2045213705404452,
      "grad_norm": 1.0,
      "learning_rate": 0.00048606442449916284,
      "loss": 5.4912,
      "mean_token_accuracy": 0.20159126520156861,
      "num_tokens": 25425319.0,
      "step": 13640
    },
    {
      "entropy": 5.821358633041382,
      "epoch": 1.2049629106322854,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048605351183030537,
      "loss": 5.4381,
      "mean_token_accuracy": 0.20114302784204482,
      "num_tokens": 25435477.0,
      "step": 13645
    },
    {
      "entropy": 5.8079132556915285,
      "epoch": 1.2054044507241257,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004860425950270296,
      "loss": 5.4326,
      "mean_token_accuracy": 0.19563105404376985,
      "num_tokens": 25445633.0,
      "step": 13650
    },
    {
      "entropy": 5.82715630531311,
      "epoch": 1.205845990815966,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048603167408954953,
      "loss": 5.3841,
      "mean_token_accuracy": 0.20574440360069274,
      "num_tokens": 25455228.0,
      "step": 13655
    },
    {
      "entropy": 5.825229167938232,
      "epoch": 1.2062875309078065,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048602074901807894,
      "loss": 5.3951,
      "mean_token_accuracy": 0.19768795669078826,
      "num_tokens": 25463992.0,
      "step": 13660
    },
    {
      "entropy": 5.783743762969971,
      "epoch": 1.2067290709996468,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048600981981283185,
      "loss": 5.3778,
      "mean_token_accuracy": 0.20332661867141724,
      "num_tokens": 25472985.0,
      "step": 13665
    },
    {
      "entropy": 5.781201410293579,
      "epoch": 1.207170611091487,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004859988864740225,
      "loss": 5.3969,
      "mean_token_accuracy": 0.1971193641424179,
      "num_tokens": 25482165.0,
      "step": 13670
    },
    {
      "entropy": 5.787679672241211,
      "epoch": 1.2076121511833275,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004859879490018649,
      "loss": 5.4879,
      "mean_token_accuracy": 0.20230044424533844,
      "num_tokens": 25490643.0,
      "step": 13675
    },
    {
      "entropy": 5.849478387832642,
      "epoch": 1.2080536912751678,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004859770073965733,
      "loss": 5.3876,
      "mean_token_accuracy": 0.19996831715106964,
      "num_tokens": 25499954.0,
      "step": 13680
    },
    {
      "entropy": 5.858248329162597,
      "epoch": 1.208495231367008,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048596606165836223,
      "loss": 5.4277,
      "mean_token_accuracy": 0.19958026111125945,
      "num_tokens": 25509512.0,
      "step": 13685
    },
    {
      "entropy": 5.812077331542969,
      "epoch": 1.2089367714588484,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048595511178744594,
      "loss": 5.3457,
      "mean_token_accuracy": 0.198810675740242,
      "num_tokens": 25518909.0,
      "step": 13690
    },
    {
      "entropy": 5.803182697296142,
      "epoch": 1.2093783115506889,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000485944157784039,
      "loss": 5.4479,
      "mean_token_accuracy": 0.1966306149959564,
      "num_tokens": 25528221.0,
      "step": 13695
    },
    {
      "entropy": 5.84412260055542,
      "epoch": 1.2098198516425291,
      "grad_norm": 1.03125,
      "learning_rate": 0.000485933199648356,
      "loss": 5.4186,
      "mean_token_accuracy": 0.19448938667774202,
      "num_tokens": 25537428.0,
      "step": 13700
    },
    {
      "entropy": 5.828204774856568,
      "epoch": 1.2102613917343694,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048592223738061155,
      "loss": 5.4051,
      "mean_token_accuracy": 0.19483884423971176,
      "num_tokens": 25547608.0,
      "step": 13705
    },
    {
      "entropy": 5.83042197227478,
      "epoch": 1.21070293182621,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048591127098102045,
      "loss": 5.4949,
      "mean_token_accuracy": 0.1958155959844589,
      "num_tokens": 25557482.0,
      "step": 13710
    },
    {
      "entropy": 5.7836699962615965,
      "epoch": 1.2111444719180502,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004859003004497974,
      "loss": 5.3566,
      "mean_token_accuracy": 0.2017431765794754,
      "num_tokens": 25566967.0,
      "step": 13715
    },
    {
      "entropy": 5.862409543991089,
      "epoch": 1.2115860120098905,
      "grad_norm": 1.0,
      "learning_rate": 0.00048588932578715754,
      "loss": 5.3814,
      "mean_token_accuracy": 0.20105107128620148,
      "num_tokens": 25576114.0,
      "step": 13720
    },
    {
      "entropy": 5.899003696441651,
      "epoch": 1.2120275521017307,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048587834699331567,
      "loss": 5.6072,
      "mean_token_accuracy": 0.18123642206192017,
      "num_tokens": 25585953.0,
      "step": 13725
    },
    {
      "entropy": 5.890837287902832,
      "epoch": 1.2124690921935712,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000485867364068487,
      "loss": 5.5367,
      "mean_token_accuracy": 0.1876230001449585,
      "num_tokens": 25595445.0,
      "step": 13730
    },
    {
      "entropy": 5.896729278564453,
      "epoch": 1.2129106322854115,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048585637701288646,
      "loss": 5.4138,
      "mean_token_accuracy": 0.19927503615617753,
      "num_tokens": 25605883.0,
      "step": 13735
    },
    {
      "entropy": 5.8271716117858885,
      "epoch": 1.2133521723772518,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004858453858267295,
      "loss": 5.459,
      "mean_token_accuracy": 0.19328728169202805,
      "num_tokens": 25615653.0,
      "step": 13740
    },
    {
      "entropy": 5.766813802719116,
      "epoch": 1.2137937124690923,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004858343905102314,
      "loss": 5.3864,
      "mean_token_accuracy": 0.20176917165517808,
      "num_tokens": 25625043.0,
      "step": 13745
    },
    {
      "entropy": 5.84102349281311,
      "epoch": 1.2142352525609326,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048582339106360736,
      "loss": 5.4138,
      "mean_token_accuracy": 0.20222438126802444,
      "num_tokens": 25633642.0,
      "step": 13750
    },
    {
      "entropy": 5.866560459136963,
      "epoch": 1.2146767926527728,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048581238748707304,
      "loss": 5.4686,
      "mean_token_accuracy": 0.1934058651328087,
      "num_tokens": 25642154.0,
      "step": 13755
    },
    {
      "entropy": 5.808893013000488,
      "epoch": 1.2151183327446131,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000485801379780844,
      "loss": 5.3959,
      "mean_token_accuracy": 0.19736346006393432,
      "num_tokens": 25650250.0,
      "step": 13760
    },
    {
      "entropy": 5.797326946258545,
      "epoch": 1.2155598728364536,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048579036794513584,
      "loss": 5.4163,
      "mean_token_accuracy": 0.20288915634155275,
      "num_tokens": 25659789.0,
      "step": 13765
    },
    {
      "entropy": 5.815957260131836,
      "epoch": 1.216001412928294,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048577935198016413,
      "loss": 5.4441,
      "mean_token_accuracy": 0.19575237184762956,
      "num_tokens": 25668658.0,
      "step": 13770
    },
    {
      "entropy": 5.835956859588623,
      "epoch": 1.2164429530201342,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004857683318861449,
      "loss": 5.45,
      "mean_token_accuracy": 0.19296166449785232,
      "num_tokens": 25679082.0,
      "step": 13775
    },
    {
      "entropy": 5.858726739883423,
      "epoch": 1.2168844931119747,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048575730766329384,
      "loss": 5.4297,
      "mean_token_accuracy": 0.1935584530234337,
      "num_tokens": 25688408.0,
      "step": 13780
    },
    {
      "entropy": 5.808913040161133,
      "epoch": 1.217326033203815,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048574627931182704,
      "loss": 5.4032,
      "mean_token_accuracy": 0.20316696912050247,
      "num_tokens": 25697415.0,
      "step": 13785
    },
    {
      "entropy": 5.828490257263184,
      "epoch": 1.2177675732956552,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048573524683196047,
      "loss": 5.4758,
      "mean_token_accuracy": 0.1910461187362671,
      "num_tokens": 25706330.0,
      "step": 13790
    },
    {
      "entropy": 5.843984746932984,
      "epoch": 1.2182091133874955,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004857242102239102,
      "loss": 5.4232,
      "mean_token_accuracy": 0.1978504702448845,
      "num_tokens": 25716198.0,
      "step": 13795
    },
    {
      "entropy": 5.848089075088501,
      "epoch": 1.218650653479336,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004857131694878926,
      "loss": 5.4156,
      "mean_token_accuracy": 0.19115940928459169,
      "num_tokens": 25726486.0,
      "step": 13800
    },
    {
      "entropy": 5.811126375198365,
      "epoch": 1.2190921935711763,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048570212462412376,
      "loss": 5.3525,
      "mean_token_accuracy": 0.20375996381044387,
      "num_tokens": 25735559.0,
      "step": 13805
    },
    {
      "entropy": 5.820326900482177,
      "epoch": 1.2195337336630165,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004856910756328202,
      "loss": 5.4127,
      "mean_token_accuracy": 0.19793334752321243,
      "num_tokens": 25744781.0,
      "step": 13810
    },
    {
      "entropy": 5.76948356628418,
      "epoch": 1.219975273754857,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048568002251419835,
      "loss": 5.4159,
      "mean_token_accuracy": 0.20813079923391342,
      "num_tokens": 25753984.0,
      "step": 13815
    },
    {
      "entropy": 5.809472942352295,
      "epoch": 1.2204168138466973,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004856689652684746,
      "loss": 5.4952,
      "mean_token_accuracy": 0.19576571583747865,
      "num_tokens": 25763378.0,
      "step": 13820
    },
    {
      "entropy": 5.833141756057739,
      "epoch": 1.2208583539385376,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004856579038958656,
      "loss": 5.4169,
      "mean_token_accuracy": 0.19592608362436295,
      "num_tokens": 25771907.0,
      "step": 13825
    },
    {
      "entropy": 5.880925941467285,
      "epoch": 1.2212998940303779,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048564683839658805,
      "loss": 5.4578,
      "mean_token_accuracy": 0.20164652019739152,
      "num_tokens": 25780464.0,
      "step": 13830
    },
    {
      "entropy": 5.850726366043091,
      "epoch": 1.2217414341222184,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048563576877085887,
      "loss": 5.468,
      "mean_token_accuracy": 0.20249118357896806,
      "num_tokens": 25790300.0,
      "step": 13835
    },
    {
      "entropy": 5.788461065292358,
      "epoch": 1.2221829742140586,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004856246950188947,
      "loss": 5.3905,
      "mean_token_accuracy": 0.19526033848524094,
      "num_tokens": 25799103.0,
      "step": 13840
    },
    {
      "entropy": 5.857665348052978,
      "epoch": 1.222624514305899,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048561361714091257,
      "loss": 5.4535,
      "mean_token_accuracy": 0.19842367470264435,
      "num_tokens": 25808506.0,
      "step": 13845
    },
    {
      "entropy": 5.91711916923523,
      "epoch": 1.2230660543977394,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048560253513712945,
      "loss": 5.4255,
      "mean_token_accuracy": 0.19719596207141876,
      "num_tokens": 25817965.0,
      "step": 13850
    },
    {
      "entropy": 5.78755373954773,
      "epoch": 1.2235075944895797,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004855914490077624,
      "loss": 5.4066,
      "mean_token_accuracy": 0.19725599884986877,
      "num_tokens": 25827874.0,
      "step": 13855
    },
    {
      "entropy": 5.878895664215088,
      "epoch": 1.22394913458142,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048558035875302864,
      "loss": 5.5216,
      "mean_token_accuracy": 0.19557360112667083,
      "num_tokens": 25836840.0,
      "step": 13860
    },
    {
      "entropy": 5.84828290939331,
      "epoch": 1.2243906746732602,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004855692643731455,
      "loss": 5.4176,
      "mean_token_accuracy": 0.19867638498544693,
      "num_tokens": 25846344.0,
      "step": 13865
    },
    {
      "entropy": 5.772578763961792,
      "epoch": 1.2248322147651007,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048555816586833006,
      "loss": 5.3914,
      "mean_token_accuracy": 0.2070172682404518,
      "num_tokens": 25855336.0,
      "step": 13870
    },
    {
      "entropy": 5.882452440261841,
      "epoch": 1.225273754856941,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048554706323880007,
      "loss": 5.4558,
      "mean_token_accuracy": 0.1981833428144455,
      "num_tokens": 25864402.0,
      "step": 13875
    },
    {
      "entropy": 5.807854795455933,
      "epoch": 1.2257152949487813,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004855359564847728,
      "loss": 5.4173,
      "mean_token_accuracy": 0.20308926999568938,
      "num_tokens": 25873326.0,
      "step": 13880
    },
    {
      "entropy": 5.80874810218811,
      "epoch": 1.2261568350406218,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048552484560646585,
      "loss": 5.5079,
      "mean_token_accuracy": 0.18121581822633742,
      "num_tokens": 25882521.0,
      "step": 13885
    },
    {
      "entropy": 5.800187349319458,
      "epoch": 1.226598375132462,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048551373060409694,
      "loss": 5.3209,
      "mean_token_accuracy": 0.21059587299823762,
      "num_tokens": 25891087.0,
      "step": 13890
    },
    {
      "entropy": 5.84920916557312,
      "epoch": 1.2270399152243023,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048550261147788375,
      "loss": 5.3967,
      "mean_token_accuracy": 0.19528005123138428,
      "num_tokens": 25900757.0,
      "step": 13895
    },
    {
      "entropy": 5.8095996379852295,
      "epoch": 1.2274814553161426,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048549148822804415,
      "loss": 5.3965,
      "mean_token_accuracy": 0.19915473014116286,
      "num_tokens": 25909801.0,
      "step": 13900
    },
    {
      "entropy": 5.852218198776245,
      "epoch": 1.227922995407983,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000485480360854796,
      "loss": 5.523,
      "mean_token_accuracy": 0.18142039179801941,
      "num_tokens": 25919142.0,
      "step": 13905
    },
    {
      "entropy": 5.858808708190918,
      "epoch": 1.2283645354998234,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004854692293583573,
      "loss": 5.419,
      "mean_token_accuracy": 0.20021142810583115,
      "num_tokens": 25927560.0,
      "step": 13910
    },
    {
      "entropy": 5.9525049209594725,
      "epoch": 1.2288060755916637,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048545809373894605,
      "loss": 5.5223,
      "mean_token_accuracy": 0.1962352693080902,
      "num_tokens": 25937184.0,
      "step": 13915
    },
    {
      "entropy": 5.9449995994567875,
      "epoch": 1.2292476156835042,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048544695399678044,
      "loss": 5.533,
      "mean_token_accuracy": 0.18230526745319367,
      "num_tokens": 25946641.0,
      "step": 13920
    },
    {
      "entropy": 5.928172445297241,
      "epoch": 1.2296891557753444,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004854358101320787,
      "loss": 5.59,
      "mean_token_accuracy": 0.1917898952960968,
      "num_tokens": 25955910.0,
      "step": 13925
    },
    {
      "entropy": 5.818526697158814,
      "epoch": 1.2301306958671847,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048542466214505914,
      "loss": 5.4504,
      "mean_token_accuracy": 0.20166580975055695,
      "num_tokens": 25966580.0,
      "step": 13930
    },
    {
      "entropy": 5.803793811798096,
      "epoch": 1.230572235959025,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048541351003594015,
      "loss": 5.3308,
      "mean_token_accuracy": 0.2071426510810852,
      "num_tokens": 25975815.0,
      "step": 13935
    },
    {
      "entropy": 5.809789323806763,
      "epoch": 1.2310137760508655,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004854023538049401,
      "loss": 5.4045,
      "mean_token_accuracy": 0.20217425525188445,
      "num_tokens": 25984784.0,
      "step": 13940
    },
    {
      "entropy": 5.766007328033448,
      "epoch": 1.2314553161427058,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048539119345227774,
      "loss": 5.3754,
      "mean_token_accuracy": 0.20669020116329193,
      "num_tokens": 25994125.0,
      "step": 13945
    },
    {
      "entropy": 5.859079170227051,
      "epoch": 1.231896856234546,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004853800289781714,
      "loss": 5.3981,
      "mean_token_accuracy": 0.20159138441085817,
      "num_tokens": 26002876.0,
      "step": 13950
    },
    {
      "entropy": 5.79793529510498,
      "epoch": 1.2323383963263865,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048536886038284,
      "loss": 5.4347,
      "mean_token_accuracy": 0.2025657996535301,
      "num_tokens": 26012591.0,
      "step": 13955
    },
    {
      "entropy": 5.913652467727661,
      "epoch": 1.2327799364182268,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048535768766650225,
      "loss": 5.5556,
      "mean_token_accuracy": 0.1827597662806511,
      "num_tokens": 26022107.0,
      "step": 13960
    },
    {
      "entropy": 5.863278150558472,
      "epoch": 1.233221476510067,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000485346510829377,
      "loss": 5.5235,
      "mean_token_accuracy": 0.20024748295545577,
      "num_tokens": 26031035.0,
      "step": 13965
    },
    {
      "entropy": 5.81165657043457,
      "epoch": 1.2336630166019074,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048533532987168334,
      "loss": 5.3259,
      "mean_token_accuracy": 0.21129488497972487,
      "num_tokens": 26039980.0,
      "step": 13970
    },
    {
      "entropy": 5.906415891647339,
      "epoch": 1.2341045566937479,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048532414479364016,
      "loss": 5.5538,
      "mean_token_accuracy": 0.18444538563489915,
      "num_tokens": 26049389.0,
      "step": 13975
    },
    {
      "entropy": 5.827001476287842,
      "epoch": 1.2345460967855881,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004853129555954666,
      "loss": 5.383,
      "mean_token_accuracy": 0.20637040138244628,
      "num_tokens": 26058868.0,
      "step": 13980
    },
    {
      "entropy": 5.871132802963257,
      "epoch": 1.2349876368774284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004853017622773818,
      "loss": 5.4582,
      "mean_token_accuracy": 0.19470107704401016,
      "num_tokens": 26068025.0,
      "step": 13985
    },
    {
      "entropy": 5.895386600494385,
      "epoch": 1.235429176969269,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004852905648396051,
      "loss": 5.4353,
      "mean_token_accuracy": 0.1970718115568161,
      "num_tokens": 26076410.0,
      "step": 13990
    },
    {
      "entropy": 5.795963191986084,
      "epoch": 1.2358707170611092,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048527936328235583,
      "loss": 5.3736,
      "mean_token_accuracy": 0.20549770444631577,
      "num_tokens": 26085624.0,
      "step": 13995
    },
    {
      "entropy": 5.846537828445435,
      "epoch": 1.2363122571529495,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004852681576058534,
      "loss": 5.4878,
      "mean_token_accuracy": 0.19413530230522155,
      "num_tokens": 26094898.0,
      "step": 14000
    },
    {
      "entropy": 5.866560268402099,
      "epoch": 1.2367537972447897,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004852569478103174,
      "loss": 5.4536,
      "mean_token_accuracy": 0.1948859751224518,
      "num_tokens": 26105041.0,
      "step": 14005
    },
    {
      "entropy": 5.862928771972657,
      "epoch": 1.2371953373366302,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048524573389596736,
      "loss": 5.4019,
      "mean_token_accuracy": 0.20501069873571395,
      "num_tokens": 26114140.0,
      "step": 14010
    },
    {
      "entropy": 5.724464416503906,
      "epoch": 1.2376368774284705,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048523451586302295,
      "loss": 5.2575,
      "mean_token_accuracy": 0.2127821758389473,
      "num_tokens": 26122611.0,
      "step": 14015
    },
    {
      "entropy": 5.726232576370239,
      "epoch": 1.2380784175203108,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000485223293711704,
      "loss": 5.3401,
      "mean_token_accuracy": 0.20355012714862825,
      "num_tokens": 26131548.0,
      "step": 14020
    },
    {
      "entropy": 5.867557430267334,
      "epoch": 1.2385199576121513,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048521206744223024,
      "loss": 5.3474,
      "mean_token_accuracy": 0.1960148885846138,
      "num_tokens": 26139806.0,
      "step": 14025
    },
    {
      "entropy": 5.912551498413086,
      "epoch": 1.2389614977039916,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048520083705482165,
      "loss": 5.4706,
      "mean_token_accuracy": 0.191637521982193,
      "num_tokens": 26148960.0,
      "step": 14030
    },
    {
      "entropy": 5.831079769134521,
      "epoch": 1.2394030377958318,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004851896025496982,
      "loss": 5.4644,
      "mean_token_accuracy": 0.19882910400629045,
      "num_tokens": 26158498.0,
      "step": 14035
    },
    {
      "entropy": 5.810053777694702,
      "epoch": 1.239844577887672,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048517836392707994,
      "loss": 5.3223,
      "mean_token_accuracy": 0.20472958981990813,
      "num_tokens": 26167319.0,
      "step": 14040
    },
    {
      "entropy": 5.78435640335083,
      "epoch": 1.2402861179795126,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048516712118718713,
      "loss": 5.3455,
      "mean_token_accuracy": 0.2067430794239044,
      "num_tokens": 26176059.0,
      "step": 14045
    },
    {
      "entropy": 5.787687063217163,
      "epoch": 1.2407276580713529,
      "grad_norm": 1.0,
      "learning_rate": 0.0004851558743302399,
      "loss": 5.3743,
      "mean_token_accuracy": 0.20237334966659545,
      "num_tokens": 26185102.0,
      "step": 14050
    },
    {
      "entropy": 5.801526641845703,
      "epoch": 1.2411691981631932,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004851446233564586,
      "loss": 5.3667,
      "mean_token_accuracy": 0.20010706633329392,
      "num_tokens": 26194873.0,
      "step": 14055
    },
    {
      "entropy": 5.849619340896607,
      "epoch": 1.2416107382550337,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004851333682660637,
      "loss": 5.4191,
      "mean_token_accuracy": 0.19707386046648026,
      "num_tokens": 26205362.0,
      "step": 14060
    },
    {
      "entropy": 5.77091875076294,
      "epoch": 1.242052278346874,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048512210905927555,
      "loss": 5.3298,
      "mean_token_accuracy": 0.20205124616622924,
      "num_tokens": 26214266.0,
      "step": 14065
    },
    {
      "entropy": 5.8740599155426025,
      "epoch": 1.2424938184387142,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004851108457363149,
      "loss": 5.491,
      "mean_token_accuracy": 0.1976918712258339,
      "num_tokens": 26223069.0,
      "step": 14070
    },
    {
      "entropy": 5.791577863693237,
      "epoch": 1.2429353585305545,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004850995782974022,
      "loss": 5.4352,
      "mean_token_accuracy": 0.1926703244447708,
      "num_tokens": 26232547.0,
      "step": 14075
    },
    {
      "entropy": 5.906997108459473,
      "epoch": 1.243376898622395,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004850883067427582,
      "loss": 5.4107,
      "mean_token_accuracy": 0.1996164157986641,
      "num_tokens": 26241014.0,
      "step": 14080
    },
    {
      "entropy": 5.804035615921021,
      "epoch": 1.2438184387142353,
      "grad_norm": 0.890625,
      "learning_rate": 0.00048507703107260384,
      "loss": 5.4503,
      "mean_token_accuracy": 0.20347474366426468,
      "num_tokens": 26250784.0,
      "step": 14085
    },
    {
      "entropy": 5.815296602249146,
      "epoch": 1.2442599788060755,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048506575128715984,
      "loss": 5.3905,
      "mean_token_accuracy": 0.20030366480350495,
      "num_tokens": 26259769.0,
      "step": 14090
    },
    {
      "entropy": 5.847494602203369,
      "epoch": 1.244701518897916,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004850544673866474,
      "loss": 5.4149,
      "mean_token_accuracy": 0.20297379940748214,
      "num_tokens": 26269038.0,
      "step": 14095
    },
    {
      "entropy": 5.837285900115967,
      "epoch": 1.2451430589897563,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004850431793712873,
      "loss": 5.3808,
      "mean_token_accuracy": 0.19335044771432877,
      "num_tokens": 26278657.0,
      "step": 14100
    },
    {
      "entropy": 5.765714168548584,
      "epoch": 1.2455845990815966,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048503188724130075,
      "loss": 5.3864,
      "mean_token_accuracy": 0.20418491065502167,
      "num_tokens": 26287458.0,
      "step": 14105
    },
    {
      "entropy": 5.852335691452026,
      "epoch": 1.2460261391734369,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000485020590996909,
      "loss": 5.4659,
      "mean_token_accuracy": 0.19273151755332946,
      "num_tokens": 26296324.0,
      "step": 14110
    },
    {
      "entropy": 5.834042358398437,
      "epoch": 1.2464676792652774,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048500929063833334,
      "loss": 5.3778,
      "mean_token_accuracy": 0.20434997975826263,
      "num_tokens": 26304799.0,
      "step": 14115
    },
    {
      "entropy": 5.866523027420044,
      "epoch": 1.2469092193571176,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048499798616579505,
      "loss": 5.4835,
      "mean_token_accuracy": 0.19575916081666947,
      "num_tokens": 26314223.0,
      "step": 14120
    },
    {
      "entropy": 5.9105627059936525,
      "epoch": 1.247350759448958,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048498667757951574,
      "loss": 5.4739,
      "mean_token_accuracy": 0.18705244809389115,
      "num_tokens": 26323599.0,
      "step": 14125
    },
    {
      "entropy": 5.8942108154296875,
      "epoch": 1.2477922995407984,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004849753648797168,
      "loss": 5.5336,
      "mean_token_accuracy": 0.18914407938718797,
      "num_tokens": 26331929.0,
      "step": 14130
    },
    {
      "entropy": 5.80018048286438,
      "epoch": 1.2482338396326387,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004849640480666199,
      "loss": 5.4078,
      "mean_token_accuracy": 0.19961749017238617,
      "num_tokens": 26340929.0,
      "step": 14135
    },
    {
      "entropy": 5.8703961849212645,
      "epoch": 1.248675379724479,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004849527271404467,
      "loss": 5.3918,
      "mean_token_accuracy": 0.19771150201559068,
      "num_tokens": 26350393.0,
      "step": 14140
    },
    {
      "entropy": 5.902312231063843,
      "epoch": 1.2491169198163192,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048494140210141897,
      "loss": 5.5819,
      "mean_token_accuracy": 0.18880074769258498,
      "num_tokens": 26360709.0,
      "step": 14145
    },
    {
      "entropy": 5.790923738479615,
      "epoch": 1.2495584599081597,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048493007294975855,
      "loss": 5.4037,
      "mean_token_accuracy": 0.19863016456365584,
      "num_tokens": 26369447.0,
      "step": 14150
    },
    {
      "entropy": 5.875974655151367,
      "epoch": 1.25,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048491873968568746,
      "loss": 5.3604,
      "mean_token_accuracy": 0.21115702241659165,
      "num_tokens": 26379281.0,
      "step": 14155
    },
    {
      "entropy": 5.7076949119567875,
      "epoch": 1.2504415400918403,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004849074023094276,
      "loss": 5.3159,
      "mean_token_accuracy": 0.21273223906755448,
      "num_tokens": 26387664.0,
      "step": 14160
    },
    {
      "entropy": 5.753626775741577,
      "epoch": 1.2508830801836806,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00048489606082120106,
      "loss": 5.3486,
      "mean_token_accuracy": 0.20376153737306596,
      "num_tokens": 26397485.0,
      "step": 14165
    },
    {
      "entropy": 5.865854978561401,
      "epoch": 1.251324620275521,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004848847152212301,
      "loss": 5.3105,
      "mean_token_accuracy": 0.20952945202589035,
      "num_tokens": 26406307.0,
      "step": 14170
    },
    {
      "entropy": 5.885340976715088,
      "epoch": 1.2517661603673613,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000484873365509737,
      "loss": 5.5197,
      "mean_token_accuracy": 0.1915242224931717,
      "num_tokens": 26416155.0,
      "step": 14175
    },
    {
      "entropy": 5.837534999847412,
      "epoch": 1.2522077004592016,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048486201168694397,
      "loss": 5.4173,
      "mean_token_accuracy": 0.19935398697853088,
      "num_tokens": 26424147.0,
      "step": 14180
    },
    {
      "entropy": 5.829942417144776,
      "epoch": 1.252649240551042,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048485065375307353,
      "loss": 5.417,
      "mean_token_accuracy": 0.19690395146608353,
      "num_tokens": 26433249.0,
      "step": 14185
    },
    {
      "entropy": 5.879388236999512,
      "epoch": 1.2530907806428824,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048483929170834805,
      "loss": 5.4617,
      "mean_token_accuracy": 0.1976969927549362,
      "num_tokens": 26442346.0,
      "step": 14190
    },
    {
      "entropy": 5.861908054351806,
      "epoch": 1.2535323207347226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048482792555299027,
      "loss": 5.4262,
      "mean_token_accuracy": 0.19566735327243806,
      "num_tokens": 26450885.0,
      "step": 14195
    },
    {
      "entropy": 5.791489171981811,
      "epoch": 1.2539738608265631,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004848165552872227,
      "loss": 5.4648,
      "mean_token_accuracy": 0.20064617842435836,
      "num_tokens": 26459827.0,
      "step": 14200
    },
    {
      "entropy": 5.7747917652130125,
      "epoch": 1.2544154009184034,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004848051809112682,
      "loss": 5.4551,
      "mean_token_accuracy": 0.1935664877295494,
      "num_tokens": 26468357.0,
      "step": 14205
    },
    {
      "entropy": 5.862593698501587,
      "epoch": 1.2548569410102437,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048479380242534955,
      "loss": 5.5092,
      "mean_token_accuracy": 0.19768529385328293,
      "num_tokens": 26477396.0,
      "step": 14210
    },
    {
      "entropy": 5.893388271331787,
      "epoch": 1.2552984811020842,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004847824198296896,
      "loss": 5.481,
      "mean_token_accuracy": 0.19868139773607255,
      "num_tokens": 26486468.0,
      "step": 14215
    },
    {
      "entropy": 5.891446828842163,
      "epoch": 1.2557400211939245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004847710331245113,
      "loss": 5.4485,
      "mean_token_accuracy": 0.19702454507350922,
      "num_tokens": 26496157.0,
      "step": 14220
    },
    {
      "entropy": 5.759126615524292,
      "epoch": 1.2561815612857647,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048475964231003776,
      "loss": 5.3202,
      "mean_token_accuracy": 0.2099949225783348,
      "num_tokens": 26504573.0,
      "step": 14225
    },
    {
      "entropy": 5.724131107330322,
      "epoch": 1.256623101377605,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004847482473864922,
      "loss": 5.2861,
      "mean_token_accuracy": 0.2155895709991455,
      "num_tokens": 26513441.0,
      "step": 14230
    },
    {
      "entropy": 5.8794779777526855,
      "epoch": 1.2570646414694453,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004847368483540977,
      "loss": 5.4995,
      "mean_token_accuracy": 0.1886053577065468,
      "num_tokens": 26522911.0,
      "step": 14235
    },
    {
      "entropy": 5.779993534088135,
      "epoch": 1.2575061815612858,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048472544521307765,
      "loss": 5.3569,
      "mean_token_accuracy": 0.20827997028827666,
      "num_tokens": 26531388.0,
      "step": 14240
    },
    {
      "entropy": 5.836145830154419,
      "epoch": 1.257947721653126,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004847140379636554,
      "loss": 5.3593,
      "mean_token_accuracy": 0.1983882948756218,
      "num_tokens": 26540500.0,
      "step": 14245
    },
    {
      "entropy": 5.845391941070557,
      "epoch": 1.2583892617449663,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004847026266060545,
      "loss": 5.3709,
      "mean_token_accuracy": 0.198971189558506,
      "num_tokens": 26550041.0,
      "step": 14250
    },
    {
      "entropy": 5.791977787017823,
      "epoch": 1.2588308018368068,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004846912111404983,
      "loss": 5.3313,
      "mean_token_accuracy": 0.20202200710773469,
      "num_tokens": 26559594.0,
      "step": 14255
    },
    {
      "entropy": 5.762013292312622,
      "epoch": 1.2592723419286471,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048467979156721055,
      "loss": 5.3989,
      "mean_token_accuracy": 0.20650634914636612,
      "num_tokens": 26569143.0,
      "step": 14260
    },
    {
      "entropy": 5.770469379425049,
      "epoch": 1.2597138820204874,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048466836788641496,
      "loss": 5.3815,
      "mean_token_accuracy": 0.20351646989583969,
      "num_tokens": 26578040.0,
      "step": 14265
    },
    {
      "entropy": 5.801880645751953,
      "epoch": 1.260155422112328,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048465694009833525,
      "loss": 5.4462,
      "mean_token_accuracy": 0.20380374491214753,
      "num_tokens": 26588283.0,
      "step": 14270
    },
    {
      "entropy": 5.890912818908691,
      "epoch": 1.2605969622041682,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048464550820319537,
      "loss": 5.4648,
      "mean_token_accuracy": 0.19421533048152922,
      "num_tokens": 26597815.0,
      "step": 14275
    },
    {
      "entropy": 5.805597448348999,
      "epoch": 1.2610385022960084,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004846340722012192,
      "loss": 5.4091,
      "mean_token_accuracy": 0.19901443123817444,
      "num_tokens": 26606752.0,
      "step": 14280
    },
    {
      "entropy": 5.821068000793457,
      "epoch": 1.261480042387849,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004846226320926308,
      "loss": 5.4364,
      "mean_token_accuracy": 0.1993491008877754,
      "num_tokens": 26615312.0,
      "step": 14285
    },
    {
      "entropy": 5.857545471191406,
      "epoch": 1.2619215824796892,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004846111878776542,
      "loss": 5.4277,
      "mean_token_accuracy": 0.19909490197896956,
      "num_tokens": 26624642.0,
      "step": 14290
    },
    {
      "entropy": 5.769836473464966,
      "epoch": 1.2623631225715295,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004845997395565137,
      "loss": 5.266,
      "mean_token_accuracy": 0.21139654517173767,
      "num_tokens": 26633250.0,
      "step": 14295
    },
    {
      "entropy": 5.746859359741211,
      "epoch": 1.2628046626633698,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004845882871294335,
      "loss": 5.4249,
      "mean_token_accuracy": 0.1992125391960144,
      "num_tokens": 26642422.0,
      "step": 14300
    },
    {
      "entropy": 5.830981349945068,
      "epoch": 1.26324620275521,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004845768305966379,
      "loss": 5.4677,
      "mean_token_accuracy": 0.19054748564958573,
      "num_tokens": 26651978.0,
      "step": 14305
    },
    {
      "entropy": 5.855709075927734,
      "epoch": 1.2636877428470505,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004845653699583515,
      "loss": 5.4805,
      "mean_token_accuracy": 0.19336138367652894,
      "num_tokens": 26662075.0,
      "step": 14310
    },
    {
      "entropy": 5.833240413665772,
      "epoch": 1.2641292829388908,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048455390521479855,
      "loss": 5.3843,
      "mean_token_accuracy": 0.20019437968730927,
      "num_tokens": 26670813.0,
      "step": 14315
    },
    {
      "entropy": 5.8653655529022215,
      "epoch": 1.264570823030731,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048454243636620383,
      "loss": 5.4179,
      "mean_token_accuracy": 0.19402478486299515,
      "num_tokens": 26679867.0,
      "step": 14320
    },
    {
      "entropy": 5.770459222793579,
      "epoch": 1.2650123631225716,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000484530963412792,
      "loss": 5.4271,
      "mean_token_accuracy": 0.19590177685022353,
      "num_tokens": 26688523.0,
      "step": 14325
    },
    {
      "entropy": 5.84591908454895,
      "epoch": 1.2654539032144119,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048451948635478765,
      "loss": 5.4876,
      "mean_token_accuracy": 0.19447792023420335,
      "num_tokens": 26697922.0,
      "step": 14330
    },
    {
      "entropy": 5.927173662185669,
      "epoch": 1.2658954433062521,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004845080051924158,
      "loss": 5.4424,
      "mean_token_accuracy": 0.18839099556207656,
      "num_tokens": 26707058.0,
      "step": 14335
    },
    {
      "entropy": 5.855650472640991,
      "epoch": 1.2663369833980926,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004844965199259013,
      "loss": 5.4508,
      "mean_token_accuracy": 0.19742105156183243,
      "num_tokens": 26715873.0,
      "step": 14340
    },
    {
      "entropy": 5.800871706008911,
      "epoch": 1.266778523489933,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004844850305554692,
      "loss": 5.4298,
      "mean_token_accuracy": 0.2012576714158058,
      "num_tokens": 26726253.0,
      "step": 14345
    },
    {
      "entropy": 5.825993347167969,
      "epoch": 1.2672200635817732,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048447353708134426,
      "loss": 5.4285,
      "mean_token_accuracy": 0.19938729107379913,
      "num_tokens": 26737077.0,
      "step": 14350
    },
    {
      "entropy": 5.819729804992676,
      "epoch": 1.2676616036736137,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000484462039503752,
      "loss": 5.4525,
      "mean_token_accuracy": 0.19648347347974776,
      "num_tokens": 26746603.0,
      "step": 14355
    },
    {
      "entropy": 5.778798151016235,
      "epoch": 1.268103143765454,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048445053782291753,
      "loss": 5.4264,
      "mean_token_accuracy": 0.20044244974851608,
      "num_tokens": 26756174.0,
      "step": 14360
    },
    {
      "entropy": 5.824472713470459,
      "epoch": 1.2685446838572942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048443903203906614,
      "loss": 5.4238,
      "mean_token_accuracy": 0.2045733779668808,
      "num_tokens": 26765124.0,
      "step": 14365
    },
    {
      "entropy": 5.766734218597412,
      "epoch": 1.2689862239491345,
      "grad_norm": 1.0,
      "learning_rate": 0.0004844275221524232,
      "loss": 5.3989,
      "mean_token_accuracy": 0.20548185259103774,
      "num_tokens": 26774438.0,
      "step": 14370
    },
    {
      "entropy": 5.907923460006714,
      "epoch": 1.2694277640409748,
      "grad_norm": 1.0,
      "learning_rate": 0.00048441600816321424,
      "loss": 5.4406,
      "mean_token_accuracy": 0.2006345048546791,
      "num_tokens": 26783613.0,
      "step": 14375
    },
    {
      "entropy": 5.8193682670593265,
      "epoch": 1.2698693041328153,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004844044900716647,
      "loss": 5.2824,
      "mean_token_accuracy": 0.21384110301733017,
      "num_tokens": 26792136.0,
      "step": 14380
    },
    {
      "entropy": 5.701772880554199,
      "epoch": 1.2703108442246556,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004843929678780004,
      "loss": 5.3074,
      "mean_token_accuracy": 0.21431955248117446,
      "num_tokens": 26801481.0,
      "step": 14385
    },
    {
      "entropy": 5.767381381988526,
      "epoch": 1.2707523843164958,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004843814415824469,
      "loss": 5.4005,
      "mean_token_accuracy": 0.20618583559989928,
      "num_tokens": 26810503.0,
      "step": 14390
    },
    {
      "entropy": 5.903650236129761,
      "epoch": 1.2711939244083363,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048436991118523003,
      "loss": 5.4147,
      "mean_token_accuracy": 0.20471629798412322,
      "num_tokens": 26820007.0,
      "step": 14395
    },
    {
      "entropy": 5.845257520675659,
      "epoch": 1.2716354645001766,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048435837668657565,
      "loss": 5.3517,
      "mean_token_accuracy": 0.20296451300382615,
      "num_tokens": 26829088.0,
      "step": 14400
    },
    {
      "entropy": 5.81347689628601,
      "epoch": 1.272077004592017,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004843468380867098,
      "loss": 5.4452,
      "mean_token_accuracy": 0.20157098770141602,
      "num_tokens": 26838619.0,
      "step": 14405
    },
    {
      "entropy": 5.819590997695923,
      "epoch": 1.2725185446838574,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004843352953858585,
      "loss": 5.4645,
      "mean_token_accuracy": 0.1941637635231018,
      "num_tokens": 26848728.0,
      "step": 14410
    },
    {
      "entropy": 5.848436975479126,
      "epoch": 1.2729600847756977,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004843237485842477,
      "loss": 5.3689,
      "mean_token_accuracy": 0.2138514906167984,
      "num_tokens": 26859293.0,
      "step": 14415
    },
    {
      "entropy": 5.810101270675659,
      "epoch": 1.273401624867538,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048431219768210374,
      "loss": 5.4086,
      "mean_token_accuracy": 0.20079730898141862,
      "num_tokens": 26868818.0,
      "step": 14420
    },
    {
      "entropy": 5.862733459472656,
      "epoch": 1.2738431649593784,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004843006426796529,
      "loss": 5.5097,
      "mean_token_accuracy": 0.18794319778680801,
      "num_tokens": 26879245.0,
      "step": 14425
    },
    {
      "entropy": 5.844163131713867,
      "epoch": 1.2742847050512187,
      "grad_norm": 1.0,
      "learning_rate": 0.0004842890835771215,
      "loss": 5.386,
      "mean_token_accuracy": 0.20195904672145842,
      "num_tokens": 26888074.0,
      "step": 14430
    },
    {
      "entropy": 5.851729869842529,
      "epoch": 1.274726245143059,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000484277520374736,
      "loss": 5.4255,
      "mean_token_accuracy": 0.19620058536529542,
      "num_tokens": 26897324.0,
      "step": 14435
    },
    {
      "entropy": 5.910194635391235,
      "epoch": 1.2751677852348993,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048426595307272285,
      "loss": 5.527,
      "mean_token_accuracy": 0.19152487367391585,
      "num_tokens": 26906797.0,
      "step": 14440
    },
    {
      "entropy": 5.864139080047607,
      "epoch": 1.2756093253267395,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048425438167130875,
      "loss": 5.4448,
      "mean_token_accuracy": 0.1987533837556839,
      "num_tokens": 26915737.0,
      "step": 14445
    },
    {
      "entropy": 5.765225744247436,
      "epoch": 1.27605086541858,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00048424280617072026,
      "loss": 5.412,
      "mean_token_accuracy": 0.19996086806058883,
      "num_tokens": 26925364.0,
      "step": 14450
    },
    {
      "entropy": 5.831092691421508,
      "epoch": 1.2764924055104203,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00048423122657118424,
      "loss": 5.4144,
      "mean_token_accuracy": 0.2001582309603691,
      "num_tokens": 26935188.0,
      "step": 14455
    },
    {
      "entropy": 5.8068177700042725,
      "epoch": 1.2769339456022606,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048421964287292745,
      "loss": 5.3796,
      "mean_token_accuracy": 0.19895507395267487,
      "num_tokens": 26944606.0,
      "step": 14460
    },
    {
      "entropy": 5.887812280654908,
      "epoch": 1.277375485694101,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004842080550761769,
      "loss": 5.5068,
      "mean_token_accuracy": 0.1923213228583336,
      "num_tokens": 26955147.0,
      "step": 14465
    },
    {
      "entropy": 5.839098834991455,
      "epoch": 1.2778170257859414,
      "grad_norm": 1.125,
      "learning_rate": 0.0004841964631811595,
      "loss": 5.4364,
      "mean_token_accuracy": 0.20264236479997635,
      "num_tokens": 26964243.0,
      "step": 14470
    },
    {
      "entropy": 5.819336223602295,
      "epoch": 1.2782585658777816,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048418486718810244,
      "loss": 5.4033,
      "mean_token_accuracy": 0.20000549256801606,
      "num_tokens": 26973284.0,
      "step": 14475
    },
    {
      "entropy": 5.841389942169189,
      "epoch": 1.2787001059696221,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004841732670972327,
      "loss": 5.5313,
      "mean_token_accuracy": 0.19234911054372789,
      "num_tokens": 26982757.0,
      "step": 14480
    },
    {
      "entropy": 5.873554134368897,
      "epoch": 1.2791416460614624,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004841616629087776,
      "loss": 5.4393,
      "mean_token_accuracy": 0.19488273859024047,
      "num_tokens": 26991784.0,
      "step": 14485
    },
    {
      "entropy": 5.883336687088013,
      "epoch": 1.2795831861533027,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048415005462296455,
      "loss": 5.4549,
      "mean_token_accuracy": 0.18665586411952972,
      "num_tokens": 27000700.0,
      "step": 14490
    },
    {
      "entropy": 5.838480710983276,
      "epoch": 1.2800247262451432,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004841384422400209,
      "loss": 5.4159,
      "mean_token_accuracy": 0.2099954754114151,
      "num_tokens": 27010089.0,
      "step": 14495
    },
    {
      "entropy": 5.810708618164062,
      "epoch": 1.2804662663369835,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004841268257601741,
      "loss": 5.4194,
      "mean_token_accuracy": 0.20429598540067673,
      "num_tokens": 27018961.0,
      "step": 14500
    },
    {
      "entropy": 5.803251171112061,
      "epoch": 1.2809078064288237,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004841152051836517,
      "loss": 5.4793,
      "mean_token_accuracy": 0.19944901019334793,
      "num_tokens": 27027905.0,
      "step": 14505
    },
    {
      "entropy": 5.829074668884277,
      "epoch": 1.281349346520664,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048410358051068135,
      "loss": 5.3932,
      "mean_token_accuracy": 0.19561435729265214,
      "num_tokens": 27036969.0,
      "step": 14510
    },
    {
      "entropy": 5.755847215652466,
      "epoch": 1.2817908866125043,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004840919517414908,
      "loss": 5.3428,
      "mean_token_accuracy": 0.20895552933216094,
      "num_tokens": 27046654.0,
      "step": 14515
    },
    {
      "entropy": 5.8790534973144535,
      "epoch": 1.2822324267043448,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048408031887630775,
      "loss": 5.4448,
      "mean_token_accuracy": 0.19693035781383514,
      "num_tokens": 27058492.0,
      "step": 14520
    },
    {
      "entropy": 5.867789602279663,
      "epoch": 1.282673966796185,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004840686819153604,
      "loss": 5.4529,
      "mean_token_accuracy": 0.19810487031936647,
      "num_tokens": 27068863.0,
      "step": 14525
    },
    {
      "entropy": 5.802721881866455,
      "epoch": 1.2831155068880253,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004840570408588762,
      "loss": 5.4063,
      "mean_token_accuracy": 0.20112985819578172,
      "num_tokens": 27078793.0,
      "step": 14530
    },
    {
      "entropy": 5.76865701675415,
      "epoch": 1.2835570469798658,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004840453957070836,
      "loss": 5.2506,
      "mean_token_accuracy": 0.21420191824436188,
      "num_tokens": 27087093.0,
      "step": 14535
    },
    {
      "entropy": 5.771711683273315,
      "epoch": 1.283998587071706,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048403374646021057,
      "loss": 5.3317,
      "mean_token_accuracy": 0.20668103098869323,
      "num_tokens": 27096601.0,
      "step": 14540
    },
    {
      "entropy": 5.813027095794678,
      "epoch": 1.2844401271635464,
      "grad_norm": 1.0,
      "learning_rate": 0.00048402209311848533,
      "loss": 5.3975,
      "mean_token_accuracy": 0.20317106544971467,
      "num_tokens": 27105041.0,
      "step": 14545
    },
    {
      "entropy": 5.838200426101684,
      "epoch": 1.2848816672553869,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048401043568213614,
      "loss": 5.4626,
      "mean_token_accuracy": 0.19362209141254424,
      "num_tokens": 27114252.0,
      "step": 14550
    },
    {
      "entropy": 5.802340459823609,
      "epoch": 1.2853232073472272,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004839987741513914,
      "loss": 5.3686,
      "mean_token_accuracy": 0.19881146401166916,
      "num_tokens": 27122480.0,
      "step": 14555
    },
    {
      "entropy": 5.820025205612183,
      "epoch": 1.2857647474390674,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048398710852647954,
      "loss": 5.4209,
      "mean_token_accuracy": 0.19590677171945572,
      "num_tokens": 27132322.0,
      "step": 14560
    },
    {
      "entropy": 5.753691530227661,
      "epoch": 1.286206287530908,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048397543880762905,
      "loss": 5.3121,
      "mean_token_accuracy": 0.21430548578500747,
      "num_tokens": 27140782.0,
      "step": 14565
    },
    {
      "entropy": 5.774447584152222,
      "epoch": 1.2866478276227482,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048396376499506866,
      "loss": 5.3487,
      "mean_token_accuracy": 0.2051307216286659,
      "num_tokens": 27150058.0,
      "step": 14570
    },
    {
      "entropy": 5.7292139530181885,
      "epoch": 1.2870893677145885,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048395208708902686,
      "loss": 5.3925,
      "mean_token_accuracy": 0.20420772433280945,
      "num_tokens": 27159574.0,
      "step": 14575
    },
    {
      "entropy": 5.819472742080689,
      "epoch": 1.2875309078064288,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048394040508973257,
      "loss": 5.3937,
      "mean_token_accuracy": 0.20418492704629898,
      "num_tokens": 27168025.0,
      "step": 14580
    },
    {
      "entropy": 5.773526811599732,
      "epoch": 1.287972447898269,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004839287189974145,
      "loss": 5.3705,
      "mean_token_accuracy": 0.203872212767601,
      "num_tokens": 27177108.0,
      "step": 14585
    },
    {
      "entropy": 5.801486015319824,
      "epoch": 1.2884139879901095,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048391702881230177,
      "loss": 5.4562,
      "mean_token_accuracy": 0.19423876255750655,
      "num_tokens": 27187486.0,
      "step": 14590
    },
    {
      "entropy": 5.919055414199829,
      "epoch": 1.2888555280819498,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004839053345346232,
      "loss": 5.4422,
      "mean_token_accuracy": 0.19504236727952956,
      "num_tokens": 27195962.0,
      "step": 14595
    },
    {
      "entropy": 5.866123962402344,
      "epoch": 1.28929706817379,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004838936361646079,
      "loss": 5.443,
      "mean_token_accuracy": 0.20055250972509384,
      "num_tokens": 27206354.0,
      "step": 14600
    },
    {
      "entropy": 5.732411909103393,
      "epoch": 1.2897386082656306,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004838819337024852,
      "loss": 5.3143,
      "mean_token_accuracy": 0.20928743332624436,
      "num_tokens": 27215326.0,
      "step": 14605
    },
    {
      "entropy": 5.786622333526611,
      "epoch": 1.2901801483574709,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048387022714848406,
      "loss": 5.4083,
      "mean_token_accuracy": 0.20155924260616304,
      "num_tokens": 27224253.0,
      "step": 14610
    },
    {
      "entropy": 5.81986436843872,
      "epoch": 1.2906216884493111,
      "grad_norm": 0.9375,
      "learning_rate": 0.000483858516502834,
      "loss": 5.4091,
      "mean_token_accuracy": 0.201575630903244,
      "num_tokens": 27233007.0,
      "step": 14615
    },
    {
      "entropy": 5.809497880935669,
      "epoch": 1.2910632285411516,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004838468017657644,
      "loss": 5.3373,
      "mean_token_accuracy": 0.20555283278226852,
      "num_tokens": 27242127.0,
      "step": 14620
    },
    {
      "entropy": 5.830899858474732,
      "epoch": 1.291504768632992,
      "grad_norm": 0.90625,
      "learning_rate": 0.00048383508293750474,
      "loss": 5.4886,
      "mean_token_accuracy": 0.18978028148412704,
      "num_tokens": 27252753.0,
      "step": 14625
    },
    {
      "entropy": 5.904895639419555,
      "epoch": 1.2919463087248322,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004838233600182846,
      "loss": 5.4698,
      "mean_token_accuracy": 0.19388115257024766,
      "num_tokens": 27262498.0,
      "step": 14630
    },
    {
      "entropy": 5.9419198513031,
      "epoch": 1.2923878488166727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004838116330083336,
      "loss": 5.4736,
      "mean_token_accuracy": 0.19612549543380736,
      "num_tokens": 27271524.0,
      "step": 14635
    },
    {
      "entropy": 5.708930015563965,
      "epoch": 1.292829388908513,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004837999019078814,
      "loss": 5.3106,
      "mean_token_accuracy": 0.2147142231464386,
      "num_tokens": 27280649.0,
      "step": 14640
    },
    {
      "entropy": 5.815315055847168,
      "epoch": 1.2932709290003532,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004837881667171579,
      "loss": 5.4314,
      "mean_token_accuracy": 0.19222755432128907,
      "num_tokens": 27289593.0,
      "step": 14645
    },
    {
      "entropy": 5.841908502578735,
      "epoch": 1.2937124690921935,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048377642743639297,
      "loss": 5.4191,
      "mean_token_accuracy": 0.20031570494174958,
      "num_tokens": 27298070.0,
      "step": 14650
    },
    {
      "entropy": 5.7707356929779055,
      "epoch": 1.2941540091840338,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004837646840658166,
      "loss": 5.3396,
      "mean_token_accuracy": 0.20702026784420013,
      "num_tokens": 27306738.0,
      "step": 14655
    },
    {
      "entropy": 5.829840850830078,
      "epoch": 1.2945955492758743,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048375293660565875,
      "loss": 5.3435,
      "mean_token_accuracy": 0.2051900789141655,
      "num_tokens": 27315243.0,
      "step": 14660
    },
    {
      "entropy": 5.793229389190674,
      "epoch": 1.2950370893677146,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004837411850561496,
      "loss": 5.3472,
      "mean_token_accuracy": 0.20633992403745652,
      "num_tokens": 27325327.0,
      "step": 14665
    },
    {
      "entropy": 5.82656078338623,
      "epoch": 1.2954786294595548,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048372942941751935,
      "loss": 5.3592,
      "mean_token_accuracy": 0.19919373393058776,
      "num_tokens": 27333436.0,
      "step": 14670
    },
    {
      "entropy": 5.870306777954101,
      "epoch": 1.2959201695513953,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004837176696899984,
      "loss": 5.4502,
      "mean_token_accuracy": 0.19145621359348297,
      "num_tokens": 27342689.0,
      "step": 14675
    },
    {
      "entropy": 5.885797643661499,
      "epoch": 1.2963617096432356,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048370590587381684,
      "loss": 5.4521,
      "mean_token_accuracy": 0.20088817924261093,
      "num_tokens": 27352347.0,
      "step": 14680
    },
    {
      "entropy": 5.858293867111206,
      "epoch": 1.2968032497350759,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048369413796920535,
      "loss": 5.4544,
      "mean_token_accuracy": 0.1977135181427002,
      "num_tokens": 27362153.0,
      "step": 14685
    },
    {
      "entropy": 5.7391194820404055,
      "epoch": 1.2972447898269164,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004836823659763944,
      "loss": 5.3834,
      "mean_token_accuracy": 0.19741292893886567,
      "num_tokens": 27371126.0,
      "step": 14690
    },
    {
      "entropy": 5.774929523468018,
      "epoch": 1.2976863299187567,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048367058989561457,
      "loss": 5.3248,
      "mean_token_accuracy": 0.20115848630666733,
      "num_tokens": 27380727.0,
      "step": 14695
    },
    {
      "entropy": 5.8485064029693605,
      "epoch": 1.298127870010597,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004836588097270966,
      "loss": 5.3699,
      "mean_token_accuracy": 0.2064995586872101,
      "num_tokens": 27389970.0,
      "step": 14700
    },
    {
      "entropy": 5.849681043624878,
      "epoch": 1.2985694101024374,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048364702547107113,
      "loss": 5.3321,
      "mean_token_accuracy": 0.19991912543773652,
      "num_tokens": 27399177.0,
      "step": 14705
    },
    {
      "entropy": 5.836047792434693,
      "epoch": 1.2990109501942777,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004836352371277692,
      "loss": 5.3987,
      "mean_token_accuracy": 0.2044302687048912,
      "num_tokens": 27408776.0,
      "step": 14710
    },
    {
      "entropy": 5.81238169670105,
      "epoch": 1.299452490286118,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004836234446974216,
      "loss": 5.457,
      "mean_token_accuracy": 0.196170312166214,
      "num_tokens": 27418708.0,
      "step": 14715
    },
    {
      "entropy": 5.8453374862670895,
      "epoch": 1.2998940303779583,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004836116481802593,
      "loss": 5.3727,
      "mean_token_accuracy": 0.202563051879406,
      "num_tokens": 27428027.0,
      "step": 14720
    },
    {
      "entropy": 5.795147848129273,
      "epoch": 1.3003355704697985,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048359984757651356,
      "loss": 5.3831,
      "mean_token_accuracy": 0.19861516654491423,
      "num_tokens": 27438214.0,
      "step": 14725
    },
    {
      "entropy": 5.820910692214966,
      "epoch": 1.300777110561639,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048358804288641536,
      "loss": 5.3191,
      "mean_token_accuracy": 0.20154762715101243,
      "num_tokens": 27447413.0,
      "step": 14730
    },
    {
      "entropy": 5.79666838645935,
      "epoch": 1.3012186506534793,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048357623411019607,
      "loss": 5.3446,
      "mean_token_accuracy": 0.21003749072551728,
      "num_tokens": 27456316.0,
      "step": 14735
    },
    {
      "entropy": 5.7720255851745605,
      "epoch": 1.3016601907453196,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000483564421248087,
      "loss": 5.4006,
      "mean_token_accuracy": 0.20428019016981125,
      "num_tokens": 27465346.0,
      "step": 14740
    },
    {
      "entropy": 5.762890577316284,
      "epoch": 1.30210173083716,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048355260430031947,
      "loss": 5.2803,
      "mean_token_accuracy": 0.20367016047239303,
      "num_tokens": 27473367.0,
      "step": 14745
    },
    {
      "entropy": 5.87225456237793,
      "epoch": 1.3025432709290004,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000483540783267125,
      "loss": 5.4654,
      "mean_token_accuracy": 0.20448017865419388,
      "num_tokens": 27482187.0,
      "step": 14750
    },
    {
      "entropy": 5.7717162609100345,
      "epoch": 1.3029848110208406,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048352895814873526,
      "loss": 5.3673,
      "mean_token_accuracy": 0.20130963772535324,
      "num_tokens": 27492406.0,
      "step": 14755
    },
    {
      "entropy": 5.817604494094849,
      "epoch": 1.3034263511126811,
      "grad_norm": 1.0,
      "learning_rate": 0.0004835171289453818,
      "loss": 5.4517,
      "mean_token_accuracy": 0.1971517637372017,
      "num_tokens": 27501929.0,
      "step": 14760
    },
    {
      "entropy": 5.814182710647583,
      "epoch": 1.3038678912045214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004835052956572964,
      "loss": 5.3128,
      "mean_token_accuracy": 0.21751547753810882,
      "num_tokens": 27510657.0,
      "step": 14765
    },
    {
      "entropy": 5.813910007476807,
      "epoch": 1.3043094312963617,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048349345828471084,
      "loss": 5.4617,
      "mean_token_accuracy": 0.1915316805243492,
      "num_tokens": 27519726.0,
      "step": 14770
    },
    {
      "entropy": 5.758480262756348,
      "epoch": 1.3047509713882022,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048348161682785704,
      "loss": 5.3638,
      "mean_token_accuracy": 0.20772658437490463,
      "num_tokens": 27528876.0,
      "step": 14775
    },
    {
      "entropy": 5.843346357345581,
      "epoch": 1.3051925114800424,
      "grad_norm": 1.0,
      "learning_rate": 0.0004834697712869669,
      "loss": 5.4125,
      "mean_token_accuracy": 0.19880662858486176,
      "num_tokens": 27537782.0,
      "step": 14780
    },
    {
      "entropy": 5.834340810775757,
      "epoch": 1.3056340515718827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004834579216622725,
      "loss": 5.4365,
      "mean_token_accuracy": 0.19541001170873643,
      "num_tokens": 27547322.0,
      "step": 14785
    },
    {
      "entropy": 5.821652841567993,
      "epoch": 1.306075591663723,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000483446067954006,
      "loss": 5.3946,
      "mean_token_accuracy": 0.19541295021772384,
      "num_tokens": 27556070.0,
      "step": 14790
    },
    {
      "entropy": 5.753350639343262,
      "epoch": 1.3065171317555633,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048343421016239945,
      "loss": 5.2822,
      "mean_token_accuracy": 0.20969274044036865,
      "num_tokens": 27565951.0,
      "step": 14795
    },
    {
      "entropy": 5.823746204376221,
      "epoch": 1.3069586718474038,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004834223482876854,
      "loss": 5.3518,
      "mean_token_accuracy": 0.2040032207965851,
      "num_tokens": 27575102.0,
      "step": 14800
    },
    {
      "entropy": 5.863844966888427,
      "epoch": 1.307400211939244,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004834104823300961,
      "loss": 5.4472,
      "mean_token_accuracy": 0.19539563953876496,
      "num_tokens": 27584854.0,
      "step": 14805
    },
    {
      "entropy": 5.815785455703735,
      "epoch": 1.3078417520310843,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048339861228986395,
      "loss": 5.4101,
      "mean_token_accuracy": 0.2083549067378044,
      "num_tokens": 27593403.0,
      "step": 14810
    },
    {
      "entropy": 5.882075691223145,
      "epoch": 1.3082832921229248,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048338673816722145,
      "loss": 5.458,
      "mean_token_accuracy": 0.20046875178813933,
      "num_tokens": 27602162.0,
      "step": 14815
    },
    {
      "entropy": 5.817253637313843,
      "epoch": 1.308724832214765,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048337485996240127,
      "loss": 5.46,
      "mean_token_accuracy": 0.19198155850172044,
      "num_tokens": 27611433.0,
      "step": 14820
    },
    {
      "entropy": 5.785982418060303,
      "epoch": 1.3091663723066054,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004833629776756362,
      "loss": 5.2838,
      "mean_token_accuracy": 0.20872964859008789,
      "num_tokens": 27620443.0,
      "step": 14825
    },
    {
      "entropy": 5.880227565765381,
      "epoch": 1.3096079123984459,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048335109130715875,
      "loss": 5.364,
      "mean_token_accuracy": 0.20242440700531006,
      "num_tokens": 27629507.0,
      "step": 14830
    },
    {
      "entropy": 5.820908832550049,
      "epoch": 1.3100494524902861,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004833392008572021,
      "loss": 5.3751,
      "mean_token_accuracy": 0.20181168913841246,
      "num_tokens": 27638355.0,
      "step": 14835
    },
    {
      "entropy": 5.874566602706909,
      "epoch": 1.3104909925821264,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048332730632599883,
      "loss": 5.4643,
      "mean_token_accuracy": 0.19332274347543715,
      "num_tokens": 27649074.0,
      "step": 14840
    },
    {
      "entropy": 5.839673137664795,
      "epoch": 1.310932532673967,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004833154077137821,
      "loss": 5.4399,
      "mean_token_accuracy": 0.19462324380874635,
      "num_tokens": 27658564.0,
      "step": 14845
    },
    {
      "entropy": 5.8184332847595215,
      "epoch": 1.3113740727658072,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048330350502078515,
      "loss": 5.3781,
      "mean_token_accuracy": 0.20303722470998764,
      "num_tokens": 27667963.0,
      "step": 14850
    },
    {
      "entropy": 5.831605005264282,
      "epoch": 1.3118156128576475,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048329159824724087,
      "loss": 5.4442,
      "mean_token_accuracy": 0.196271875500679,
      "num_tokens": 27677528.0,
      "step": 14855
    },
    {
      "entropy": 5.852317953109742,
      "epoch": 1.3122571529494877,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048327968739338275,
      "loss": 5.3646,
      "mean_token_accuracy": 0.20680750906467438,
      "num_tokens": 27687659.0,
      "step": 14860
    },
    {
      "entropy": 5.761237287521363,
      "epoch": 1.312698693041328,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048326777245944396,
      "loss": 5.3298,
      "mean_token_accuracy": 0.2123411312699318,
      "num_tokens": 27697222.0,
      "step": 14865
    },
    {
      "entropy": 5.798789739608765,
      "epoch": 1.3131402331331685,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000483255853445658,
      "loss": 5.3964,
      "mean_token_accuracy": 0.20664604902267455,
      "num_tokens": 27705664.0,
      "step": 14870
    },
    {
      "entropy": 5.752559185028076,
      "epoch": 1.3135817732250088,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048324393035225824,
      "loss": 5.3445,
      "mean_token_accuracy": 0.20538181066513062,
      "num_tokens": 27714879.0,
      "step": 14875
    },
    {
      "entropy": 5.8446681022644045,
      "epoch": 1.314023313316849,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004832320031794783,
      "loss": 5.4747,
      "mean_token_accuracy": 0.1937073811888695,
      "num_tokens": 27724749.0,
      "step": 14880
    },
    {
      "entropy": 5.851373910903931,
      "epoch": 1.3144648534086896,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004832200719275519,
      "loss": 5.4121,
      "mean_token_accuracy": 0.19959086924791336,
      "num_tokens": 27733109.0,
      "step": 14885
    },
    {
      "entropy": 5.642890977859497,
      "epoch": 1.3149063935005298,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004832081365967128,
      "loss": 5.2137,
      "mean_token_accuracy": 0.21851505488157272,
      "num_tokens": 27742079.0,
      "step": 14890
    },
    {
      "entropy": 5.874283123016357,
      "epoch": 1.3153479335923701,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048319619718719455,
      "loss": 5.485,
      "mean_token_accuracy": 0.19008238315582277,
      "num_tokens": 27750900.0,
      "step": 14895
    },
    {
      "entropy": 5.858968925476074,
      "epoch": 1.3157894736842106,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048318425369923133,
      "loss": 5.3506,
      "mean_token_accuracy": 0.20756099969148636,
      "num_tokens": 27760392.0,
      "step": 14900
    },
    {
      "entropy": 5.851071166992187,
      "epoch": 1.316231013776051,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004831723061330569,
      "loss": 5.4097,
      "mean_token_accuracy": 0.2031494364142418,
      "num_tokens": 27768805.0,
      "step": 14905
    },
    {
      "entropy": 5.866053915023803,
      "epoch": 1.3166725538678912,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048316035448890546,
      "loss": 5.5032,
      "mean_token_accuracy": 0.19315202683210372,
      "num_tokens": 27777165.0,
      "step": 14910
    },
    {
      "entropy": 5.897108268737793,
      "epoch": 1.3171140939597317,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000483148398767011,
      "loss": 5.4208,
      "mean_token_accuracy": 0.20184513926506042,
      "num_tokens": 27786709.0,
      "step": 14915
    },
    {
      "entropy": 5.79182014465332,
      "epoch": 1.317555634051572,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00048313643896760776,
      "loss": 5.3721,
      "mean_token_accuracy": 0.19768935292959214,
      "num_tokens": 27795977.0,
      "step": 14920
    },
    {
      "entropy": 5.723812103271484,
      "epoch": 1.3179971741434122,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048312447509093017,
      "loss": 5.3736,
      "mean_token_accuracy": 0.1965114951133728,
      "num_tokens": 27804838.0,
      "step": 14925
    },
    {
      "entropy": 5.808836793899536,
      "epoch": 1.3184387142352525,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004831125071372124,
      "loss": 5.3478,
      "mean_token_accuracy": 0.20593465566635133,
      "num_tokens": 27813245.0,
      "step": 14930
    },
    {
      "entropy": 5.889650917053222,
      "epoch": 1.3188802543270928,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004831005351066889,
      "loss": 5.4948,
      "mean_token_accuracy": 0.19392691999673844,
      "num_tokens": 27822509.0,
      "step": 14935
    },
    {
      "entropy": 5.794647741317749,
      "epoch": 1.3193217944189333,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004830885589995944,
      "loss": 5.2938,
      "mean_token_accuracy": 0.21173144578933717,
      "num_tokens": 27831140.0,
      "step": 14940
    },
    {
      "entropy": 5.87294454574585,
      "epoch": 1.3197633345107735,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048307657881616333,
      "loss": 5.414,
      "mean_token_accuracy": 0.20045745968818665,
      "num_tokens": 27839537.0,
      "step": 14945
    },
    {
      "entropy": 5.830281543731689,
      "epoch": 1.3202048746026138,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048306459455663045,
      "loss": 5.3789,
      "mean_token_accuracy": 0.20270113795995712,
      "num_tokens": 27848888.0,
      "step": 14950
    },
    {
      "entropy": 5.858214712142944,
      "epoch": 1.3206464146944543,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004830526062212305,
      "loss": 5.4358,
      "mean_token_accuracy": 0.19450279623270034,
      "num_tokens": 27858910.0,
      "step": 14955
    },
    {
      "entropy": 5.749471426010132,
      "epoch": 1.3210879547862946,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048304061381019826,
      "loss": 5.336,
      "mean_token_accuracy": 0.21143770664930345,
      "num_tokens": 27867967.0,
      "step": 14960
    },
    {
      "entropy": 5.775195741653443,
      "epoch": 1.3215294948781349,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048302861732376875,
      "loss": 5.3548,
      "mean_token_accuracy": 0.2055211618542671,
      "num_tokens": 27877317.0,
      "step": 14965
    },
    {
      "entropy": 5.871386289596558,
      "epoch": 1.3219710349699754,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048301661676217693,
      "loss": 5.4188,
      "mean_token_accuracy": 0.1976461797952652,
      "num_tokens": 27886400.0,
      "step": 14970
    },
    {
      "entropy": 5.8330635070800785,
      "epoch": 1.3224125750618156,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004830046121256579,
      "loss": 5.3992,
      "mean_token_accuracy": 0.19833968579769135,
      "num_tokens": 27895816.0,
      "step": 14975
    },
    {
      "entropy": 5.876082181930542,
      "epoch": 1.322854115153656,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004829926034144468,
      "loss": 5.4319,
      "mean_token_accuracy": 0.19784627556800843,
      "num_tokens": 27905614.0,
      "step": 14980
    },
    {
      "entropy": 5.830043506622315,
      "epoch": 1.3232956552454964,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004829805906287789,
      "loss": 5.4393,
      "mean_token_accuracy": 0.19370155036449432,
      "num_tokens": 27914714.0,
      "step": 14985
    },
    {
      "entropy": 5.782928419113159,
      "epoch": 1.3237371953373367,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004829685737688895,
      "loss": 5.3692,
      "mean_token_accuracy": 0.20484673529863356,
      "num_tokens": 27924821.0,
      "step": 14990
    },
    {
      "entropy": 5.920098209381104,
      "epoch": 1.324178735429177,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048295655283501403,
      "loss": 5.4454,
      "mean_token_accuracy": 0.1988457813858986,
      "num_tokens": 27934028.0,
      "step": 14995
    },
    {
      "entropy": 5.854908895492554,
      "epoch": 1.3246202755210172,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000482944527827388,
      "loss": 5.3684,
      "mean_token_accuracy": 0.19951189011335374,
      "num_tokens": 27943550.0,
      "step": 15000
    },
    {
      "epoch": 1.3246202755210172,
      "eval_entropy": 5.613093136769054,
      "eval_loss": 5.462072372436523,
      "eval_mean_token_accuracy": 0.20655342451553044,
      "eval_num_tokens": 27943550.0,
      "eval_runtime": 20.3613,
      "eval_samples_per_second": 1734.415,
      "eval_steps_per_second": 216.833,
      "step": 15000
    },
    {
      "entropy": 5.84151930809021,
      "epoch": 1.3250618156128575,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004829324987462469,
      "loss": 5.3709,
      "mean_token_accuracy": 0.19965676069259644,
      "num_tokens": 27952752.0,
      "step": 15005
    },
    {
      "entropy": 5.850358676910401,
      "epoch": 1.325503355704698,
      "grad_norm": 1.0,
      "learning_rate": 0.0004829204655918264,
      "loss": 5.4534,
      "mean_token_accuracy": 0.19576463252305984,
      "num_tokens": 27961535.0,
      "step": 15010
    },
    {
      "entropy": 5.83775749206543,
      "epoch": 1.3259448957965383,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048290842836436226,
      "loss": 5.3779,
      "mean_token_accuracy": 0.20184564888477324,
      "num_tokens": 27969644.0,
      "step": 15015
    },
    {
      "entropy": 5.861055469512939,
      "epoch": 1.3263864358883786,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004828963870640902,
      "loss": 5.4925,
      "mean_token_accuracy": 0.19010561108589172,
      "num_tokens": 27978859.0,
      "step": 15020
    },
    {
      "entropy": 5.857515335083008,
      "epoch": 1.326827975980219,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048288434169124625,
      "loss": 5.4397,
      "mean_token_accuracy": 0.19789244681596757,
      "num_tokens": 27988693.0,
      "step": 15025
    },
    {
      "entropy": 5.8798247337341305,
      "epoch": 1.3272695160720593,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004828722922460662,
      "loss": 5.3666,
      "mean_token_accuracy": 0.20213449597358704,
      "num_tokens": 27998196.0,
      "step": 15030
    },
    {
      "entropy": 5.864420604705811,
      "epoch": 1.3277110561638996,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048286023872878614,
      "loss": 5.5026,
      "mean_token_accuracy": 0.19181415140628816,
      "num_tokens": 28008743.0,
      "step": 15035
    },
    {
      "entropy": 5.875790739059449,
      "epoch": 1.3281525962557401,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004828481811396423,
      "loss": 5.4687,
      "mean_token_accuracy": 0.18993117660284042,
      "num_tokens": 28017684.0,
      "step": 15040
    },
    {
      "entropy": 5.78806357383728,
      "epoch": 1.3285941363475804,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004828361194788708,
      "loss": 5.4342,
      "mean_token_accuracy": 0.2010320618748665,
      "num_tokens": 28027992.0,
      "step": 15045
    },
    {
      "entropy": 5.909226417541504,
      "epoch": 1.3290356764394207,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048282405374670785,
      "loss": 5.5007,
      "mean_token_accuracy": 0.19204058051109313,
      "num_tokens": 28037457.0,
      "step": 15050
    },
    {
      "entropy": 5.914190483093262,
      "epoch": 1.3294772165312612,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048281198394338986,
      "loss": 5.477,
      "mean_token_accuracy": 0.19456910640001296,
      "num_tokens": 28046881.0,
      "step": 15055
    },
    {
      "entropy": 5.846889543533325,
      "epoch": 1.3299187566231014,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048279991006915335,
      "loss": 5.3257,
      "mean_token_accuracy": 0.20549699366092683,
      "num_tokens": 28056932.0,
      "step": 15060
    },
    {
      "entropy": 5.843326091766357,
      "epoch": 1.3303602967149417,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00048278783212423484,
      "loss": 5.5046,
      "mean_token_accuracy": 0.19454202502965928,
      "num_tokens": 28066843.0,
      "step": 15065
    },
    {
      "entropy": 5.850677394866944,
      "epoch": 1.330801836806782,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004827757501088708,
      "loss": 5.413,
      "mean_token_accuracy": 0.19561834186315535,
      "num_tokens": 28076329.0,
      "step": 15070
    },
    {
      "entropy": 5.887483263015747,
      "epoch": 1.3312433768986223,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048276366402329803,
      "loss": 5.4739,
      "mean_token_accuracy": 0.19989312887191774,
      "num_tokens": 28086402.0,
      "step": 15075
    },
    {
      "entropy": 5.866004419326782,
      "epoch": 1.3316849169904628,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004827515738677532,
      "loss": 5.3964,
      "mean_token_accuracy": 0.19703940600156783,
      "num_tokens": 28095030.0,
      "step": 15080
    },
    {
      "entropy": 5.790664720535278,
      "epoch": 1.332126457082303,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004827394796424732,
      "loss": 5.3914,
      "mean_token_accuracy": 0.19742633551359176,
      "num_tokens": 28105138.0,
      "step": 15085
    },
    {
      "entropy": 5.811975955963135,
      "epoch": 1.3325679971741433,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048272738134769503,
      "loss": 5.4257,
      "mean_token_accuracy": 0.19294812083244323,
      "num_tokens": 28114359.0,
      "step": 15090
    },
    {
      "entropy": 5.839499759674072,
      "epoch": 1.3330095372659838,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048271527898365545,
      "loss": 5.3737,
      "mean_token_accuracy": 0.21175361126661302,
      "num_tokens": 28124045.0,
      "step": 15095
    },
    {
      "entropy": 5.823828506469726,
      "epoch": 1.333451077357824,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004827031725505917,
      "loss": 5.4192,
      "mean_token_accuracy": 0.2025573283433914,
      "num_tokens": 28132990.0,
      "step": 15100
    },
    {
      "entropy": 5.829469156265259,
      "epoch": 1.3338926174496644,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004826910620487411,
      "loss": 5.3975,
      "mean_token_accuracy": 0.20342407673597335,
      "num_tokens": 28142504.0,
      "step": 15105
    },
    {
      "entropy": 5.860262537002564,
      "epoch": 1.3343341575415049,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004826789474783405,
      "loss": 5.4865,
      "mean_token_accuracy": 0.19261177778244018,
      "num_tokens": 28151227.0,
      "step": 15110
    },
    {
      "entropy": 5.869546890258789,
      "epoch": 1.3347756976333451,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004826668288396276,
      "loss": 5.4061,
      "mean_token_accuracy": 0.19601559787988662,
      "num_tokens": 28160244.0,
      "step": 15115
    },
    {
      "entropy": 5.810975837707519,
      "epoch": 1.3352172377251854,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048265470613283954,
      "loss": 5.4119,
      "mean_token_accuracy": 0.19679898917675018,
      "num_tokens": 28169295.0,
      "step": 15120
    },
    {
      "entropy": 5.92709321975708,
      "epoch": 1.335658777817026,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048264257935821394,
      "loss": 5.4294,
      "mean_token_accuracy": 0.19897546470165253,
      "num_tokens": 28179199.0,
      "step": 15125
    },
    {
      "entropy": 5.867468976974488,
      "epoch": 1.3361003179088662,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004826304485159882,
      "loss": 5.4658,
      "mean_token_accuracy": 0.20500693321228028,
      "num_tokens": 28188315.0,
      "step": 15130
    },
    {
      "entropy": 5.831170654296875,
      "epoch": 1.3365418580007065,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048261831360640016,
      "loss": 5.4588,
      "mean_token_accuracy": 0.1936066895723343,
      "num_tokens": 28197776.0,
      "step": 15135
    },
    {
      "entropy": 5.849354600906372,
      "epoch": 1.3369833980925467,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00048260617462968735,
      "loss": 5.4519,
      "mean_token_accuracy": 0.1989843487739563,
      "num_tokens": 28207985.0,
      "step": 15140
    },
    {
      "entropy": 5.800958681106567,
      "epoch": 1.337424938184387,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048259403158608773,
      "loss": 5.2568,
      "mean_token_accuracy": 0.2158833384513855,
      "num_tokens": 28216321.0,
      "step": 15145
    },
    {
      "entropy": 5.82542724609375,
      "epoch": 1.3378664782762275,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004825818844758391,
      "loss": 5.4859,
      "mean_token_accuracy": 0.19664452522993087,
      "num_tokens": 28226763.0,
      "step": 15150
    },
    {
      "entropy": 5.89154281616211,
      "epoch": 1.3383080183680678,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048256973329917943,
      "loss": 5.392,
      "mean_token_accuracy": 0.20025351643562317,
      "num_tokens": 28236123.0,
      "step": 15155
    },
    {
      "entropy": 5.916205167770386,
      "epoch": 1.338749558459908,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048255757805634665,
      "loss": 5.4752,
      "mean_token_accuracy": 0.19791246205568314,
      "num_tokens": 28245848.0,
      "step": 15160
    },
    {
      "entropy": 5.7232136726379395,
      "epoch": 1.3391910985517486,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048254541874757895,
      "loss": 5.4534,
      "mean_token_accuracy": 0.20464010238647462,
      "num_tokens": 28255414.0,
      "step": 15165
    },
    {
      "entropy": 5.812811946868896,
      "epoch": 1.3396326386435888,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048253325537311463,
      "loss": 5.3255,
      "mean_token_accuracy": 0.20433811247348785,
      "num_tokens": 28264162.0,
      "step": 15170
    },
    {
      "entropy": 5.843650054931641,
      "epoch": 1.3400741787354291,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00048252108793319184,
      "loss": 5.3874,
      "mean_token_accuracy": 0.2048656165599823,
      "num_tokens": 28273146.0,
      "step": 15175
    },
    {
      "entropy": 5.854528188705444,
      "epoch": 1.3405157188272696,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004825089164280489,
      "loss": 5.3403,
      "mean_token_accuracy": 0.19782265722751619,
      "num_tokens": 28281865.0,
      "step": 15180
    },
    {
      "entropy": 5.778171062469482,
      "epoch": 1.3409572589191099,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048249674085792433,
      "loss": 5.3031,
      "mean_token_accuracy": 0.20337386280298234,
      "num_tokens": 28291197.0,
      "step": 15185
    },
    {
      "entropy": 5.8210736274719235,
      "epoch": 1.3413987990109502,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048248456122305667,
      "loss": 5.487,
      "mean_token_accuracy": 0.1921490639448166,
      "num_tokens": 28300283.0,
      "step": 15190
    },
    {
      "entropy": 5.9150454044342045,
      "epoch": 1.3418403391027907,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004824723775236844,
      "loss": 5.5433,
      "mean_token_accuracy": 0.19164900481700897,
      "num_tokens": 28309121.0,
      "step": 15195
    },
    {
      "entropy": 5.874976348876953,
      "epoch": 1.342281879194631,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004824601897600463,
      "loss": 5.3485,
      "mean_token_accuracy": 0.2040175050497055,
      "num_tokens": 28318377.0,
      "step": 15200
    },
    {
      "entropy": 5.851177406311035,
      "epoch": 1.3427234192864712,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048244799793238104,
      "loss": 5.3507,
      "mean_token_accuracy": 0.19854596853256226,
      "num_tokens": 28328341.0,
      "step": 15205
    },
    {
      "entropy": 5.853946399688721,
      "epoch": 1.3431649593783115,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004824358020409275,
      "loss": 5.5451,
      "mean_token_accuracy": 0.1914572909474373,
      "num_tokens": 28337340.0,
      "step": 15210
    },
    {
      "entropy": 5.756021308898926,
      "epoch": 1.3436064994701518,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004824236020859246,
      "loss": 5.4735,
      "mean_token_accuracy": 0.1991165727376938,
      "num_tokens": 28347653.0,
      "step": 15215
    },
    {
      "entropy": 5.92853627204895,
      "epoch": 1.3440480395619923,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048241139806761124,
      "loss": 5.5124,
      "mean_token_accuracy": 0.19249071627855302,
      "num_tokens": 28356596.0,
      "step": 15220
    },
    {
      "entropy": 5.910774898529053,
      "epoch": 1.3444895796538325,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048239918998622657,
      "loss": 5.4885,
      "mean_token_accuracy": 0.19089748859405517,
      "num_tokens": 28366253.0,
      "step": 15225
    },
    {
      "entropy": 5.760193538665772,
      "epoch": 1.3449311197456728,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004823869778420097,
      "loss": 5.3199,
      "mean_token_accuracy": 0.20183297097682953,
      "num_tokens": 28374815.0,
      "step": 15230
    },
    {
      "entropy": 5.813390016555786,
      "epoch": 1.3453726598375133,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048237476163519994,
      "loss": 5.4291,
      "mean_token_accuracy": 0.19437561482191085,
      "num_tokens": 28384221.0,
      "step": 15235
    },
    {
      "entropy": 5.823602247238159,
      "epoch": 1.3458141999293536,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004823625413660365,
      "loss": 5.4171,
      "mean_token_accuracy": 0.20221185237169265,
      "num_tokens": 28393453.0,
      "step": 15240
    },
    {
      "entropy": 5.761607027053833,
      "epoch": 1.3462557400211939,
      "grad_norm": 1.0,
      "learning_rate": 0.00048235031703475884,
      "loss": 5.3468,
      "mean_token_accuracy": 0.20481251925230026,
      "num_tokens": 28402343.0,
      "step": 15245
    },
    {
      "entropy": 5.791515302658081,
      "epoch": 1.3466972801130344,
      "grad_norm": 1.0,
      "learning_rate": 0.0004823380886416064,
      "loss": 5.3749,
      "mean_token_accuracy": 0.2084643617272377,
      "num_tokens": 28411527.0,
      "step": 15250
    },
    {
      "entropy": 5.849292278289795,
      "epoch": 1.3471388202048746,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004823258561868187,
      "loss": 5.4545,
      "mean_token_accuracy": 0.20024822801351547,
      "num_tokens": 28422518.0,
      "step": 15255
    },
    {
      "entropy": 5.889913558959961,
      "epoch": 1.347580360296715,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048231361967063534,
      "loss": 5.4154,
      "mean_token_accuracy": 0.19866528809070588,
      "num_tokens": 28431692.0,
      "step": 15260
    },
    {
      "entropy": 5.851217794418335,
      "epoch": 1.3480219003885554,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004823013790932961,
      "loss": 5.4589,
      "mean_token_accuracy": 0.19662691056728362,
      "num_tokens": 28440304.0,
      "step": 15265
    },
    {
      "entropy": 5.881387567520141,
      "epoch": 1.3484634404803957,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004822891344550408,
      "loss": 5.5074,
      "mean_token_accuracy": 0.18970640897750854,
      "num_tokens": 28450265.0,
      "step": 15270
    },
    {
      "entropy": 5.822591161727905,
      "epoch": 1.348904980572236,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048227688575610915,
      "loss": 5.3915,
      "mean_token_accuracy": 0.20024269670248032,
      "num_tokens": 28459699.0,
      "step": 15275
    },
    {
      "entropy": 5.808038568496704,
      "epoch": 1.3493465206640762,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004822646329967413,
      "loss": 5.4117,
      "mean_token_accuracy": 0.19073693603277206,
      "num_tokens": 28468523.0,
      "step": 15280
    },
    {
      "entropy": 5.9093286991119385,
      "epoch": 1.3497880607559165,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004822523761771771,
      "loss": 5.4613,
      "mean_token_accuracy": 0.1921578973531723,
      "num_tokens": 28477629.0,
      "step": 15285
    },
    {
      "entropy": 5.795480585098266,
      "epoch": 1.350229600847757,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004822401152976568,
      "loss": 5.3612,
      "mean_token_accuracy": 0.20787472426891326,
      "num_tokens": 28486517.0,
      "step": 15290
    },
    {
      "entropy": 5.8463352680206295,
      "epoch": 1.3506711409395973,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048222785035842045,
      "loss": 5.472,
      "mean_token_accuracy": 0.19570790082216263,
      "num_tokens": 28496103.0,
      "step": 15295
    },
    {
      "entropy": 5.816338348388672,
      "epoch": 1.3511126810314376,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048221558135970834,
      "loss": 5.4246,
      "mean_token_accuracy": 0.2021983817219734,
      "num_tokens": 28505538.0,
      "step": 15300
    },
    {
      "entropy": 5.730299663543701,
      "epoch": 1.351554221123278,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00048220330830176086,
      "loss": 5.2925,
      "mean_token_accuracy": 0.20554286241531372,
      "num_tokens": 28514322.0,
      "step": 15305
    },
    {
      "entropy": 5.902645683288574,
      "epoch": 1.3519957612151183,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004821910311848184,
      "loss": 5.4004,
      "mean_token_accuracy": 0.20479389876127244,
      "num_tokens": 28524298.0,
      "step": 15310
    },
    {
      "entropy": 5.882126235961914,
      "epoch": 1.3524373013069586,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004821787500091215,
      "loss": 5.4384,
      "mean_token_accuracy": 0.1970038443803787,
      "num_tokens": 28533567.0,
      "step": 15315
    },
    {
      "entropy": 5.78786997795105,
      "epoch": 1.352878841398799,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048216646477491074,
      "loss": 5.4105,
      "mean_token_accuracy": 0.19877147078514099,
      "num_tokens": 28541746.0,
      "step": 15320
    },
    {
      "entropy": 5.820613098144531,
      "epoch": 1.3533203814906394,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004821541754824267,
      "loss": 5.4475,
      "mean_token_accuracy": 0.19675872772932052,
      "num_tokens": 28551685.0,
      "step": 15325
    },
    {
      "entropy": 5.890408897399903,
      "epoch": 1.3537619215824797,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004821418821319102,
      "loss": 5.48,
      "mean_token_accuracy": 0.19745792001485823,
      "num_tokens": 28560935.0,
      "step": 15330
    },
    {
      "entropy": 5.794356346130371,
      "epoch": 1.3542034616743202,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004821295847236021,
      "loss": 5.3207,
      "mean_token_accuracy": 0.2062166601419449,
      "num_tokens": 28569506.0,
      "step": 15335
    },
    {
      "entropy": 5.800172233581543,
      "epoch": 1.3546450017661604,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004821172832577431,
      "loss": 5.3741,
      "mean_token_accuracy": 0.20243366211652755,
      "num_tokens": 28579792.0,
      "step": 15340
    },
    {
      "entropy": 5.892455387115478,
      "epoch": 1.3550865418580007,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004821049777345744,
      "loss": 5.4835,
      "mean_token_accuracy": 0.19847968816757203,
      "num_tokens": 28589756.0,
      "step": 15345
    },
    {
      "entropy": 5.8243976593017575,
      "epoch": 1.355528081949841,
      "grad_norm": 0.984375,
      "learning_rate": 0.000482092668154337,
      "loss": 5.4099,
      "mean_token_accuracy": 0.2011106237769127,
      "num_tokens": 28599853.0,
      "step": 15350
    },
    {
      "entropy": 5.785915994644165,
      "epoch": 1.3559696220416813,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048208035451727195,
      "loss": 5.4119,
      "mean_token_accuracy": 0.19574137032032013,
      "num_tokens": 28608576.0,
      "step": 15355
    },
    {
      "entropy": 5.867419624328614,
      "epoch": 1.3564111621335218,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004820680368236206,
      "loss": 5.4618,
      "mean_token_accuracy": 0.19275035262107848,
      "num_tokens": 28618820.0,
      "step": 15360
    },
    {
      "entropy": 5.886230325698852,
      "epoch": 1.356852702225362,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004820557150736241,
      "loss": 5.3857,
      "mean_token_accuracy": 0.20416640788316726,
      "num_tokens": 28628092.0,
      "step": 15365
    },
    {
      "entropy": 5.89946608543396,
      "epoch": 1.3572942423172023,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048204338926752396,
      "loss": 5.4637,
      "mean_token_accuracy": 0.1959486111998558,
      "num_tokens": 28637849.0,
      "step": 15370
    },
    {
      "entropy": 5.76736364364624,
      "epoch": 1.3577357824090428,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004820310594055616,
      "loss": 5.3645,
      "mean_token_accuracy": 0.20419100373983384,
      "num_tokens": 28646957.0,
      "step": 15375
    },
    {
      "entropy": 5.886047315597534,
      "epoch": 1.358177322500883,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004820187254879784,
      "loss": 5.459,
      "mean_token_accuracy": 0.20299917906522752,
      "num_tokens": 28655768.0,
      "step": 15380
    },
    {
      "entropy": 5.8470703125,
      "epoch": 1.3586188625927234,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048200638751501616,
      "loss": 5.4191,
      "mean_token_accuracy": 0.1953986570239067,
      "num_tokens": 28664719.0,
      "step": 15385
    },
    {
      "entropy": 5.806668853759765,
      "epoch": 1.3590604026845639,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004819940454869166,
      "loss": 5.3995,
      "mean_token_accuracy": 0.1941104605793953,
      "num_tokens": 28673568.0,
      "step": 15390
    },
    {
      "entropy": 5.817203044891357,
      "epoch": 1.3595019427764041,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004819816994039213,
      "loss": 5.3822,
      "mean_token_accuracy": 0.20431563556194304,
      "num_tokens": 28682558.0,
      "step": 15395
    },
    {
      "entropy": 5.883448076248169,
      "epoch": 1.3599434828682444,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048196934926627227,
      "loss": 5.4692,
      "mean_token_accuracy": 0.1973310723900795,
      "num_tokens": 28691965.0,
      "step": 15400
    },
    {
      "entropy": 5.796695947647095,
      "epoch": 1.360385022960085,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004819569950742114,
      "loss": 5.3362,
      "mean_token_accuracy": 0.21011397689580918,
      "num_tokens": 28700805.0,
      "step": 15405
    },
    {
      "entropy": 5.797438716888427,
      "epoch": 1.3608265630519252,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004819446368279808,
      "loss": 5.3644,
      "mean_token_accuracy": 0.20311632007360458,
      "num_tokens": 28710169.0,
      "step": 15410
    },
    {
      "entropy": 5.92458667755127,
      "epoch": 1.3612681031437655,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048193227452782237,
      "loss": 5.5127,
      "mean_token_accuracy": 0.19419682323932647,
      "num_tokens": 28718656.0,
      "step": 15415
    },
    {
      "entropy": 5.90979380607605,
      "epoch": 1.3617096432356057,
      "grad_norm": 1.0,
      "learning_rate": 0.0004819199081739783,
      "loss": 5.464,
      "mean_token_accuracy": 0.19747065156698226,
      "num_tokens": 28728546.0,
      "step": 15420
    },
    {
      "entropy": 5.87447772026062,
      "epoch": 1.362151183327446,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000481907537766691,
      "loss": 5.341,
      "mean_token_accuracy": 0.20292591601610183,
      "num_tokens": 28737267.0,
      "step": 15425
    },
    {
      "entropy": 5.776773118972779,
      "epoch": 1.3625927234192865,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004818951633062027,
      "loss": 5.4487,
      "mean_token_accuracy": 0.20051742792129518,
      "num_tokens": 28746721.0,
      "step": 15430
    },
    {
      "entropy": 5.88642725944519,
      "epoch": 1.3630342635111268,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048188278479275584,
      "loss": 5.4724,
      "mean_token_accuracy": 0.1973584532737732,
      "num_tokens": 28756099.0,
      "step": 15435
    },
    {
      "entropy": 5.8891956329345705,
      "epoch": 1.363475803602967,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004818704022265928,
      "loss": 5.4233,
      "mean_token_accuracy": 0.1957658648490906,
      "num_tokens": 28765463.0,
      "step": 15440
    },
    {
      "entropy": 5.7847511768341064,
      "epoch": 1.3639173436948075,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048185801560795627,
      "loss": 5.3661,
      "mean_token_accuracy": 0.2016372039914131,
      "num_tokens": 28775284.0,
      "step": 15445
    },
    {
      "entropy": 5.875983715057373,
      "epoch": 1.3643588837866478,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048184562493708895,
      "loss": 5.4169,
      "mean_token_accuracy": 0.20593764781951904,
      "num_tokens": 28784524.0,
      "step": 15450
    },
    {
      "entropy": 5.799777412414551,
      "epoch": 1.364800423878488,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048183323021423334,
      "loss": 5.37,
      "mean_token_accuracy": 0.2009094089269638,
      "num_tokens": 28792942.0,
      "step": 15455
    },
    {
      "entropy": 5.820668077468872,
      "epoch": 1.3652419639703286,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048182083143963247,
      "loss": 5.3778,
      "mean_token_accuracy": 0.20059644281864167,
      "num_tokens": 28801746.0,
      "step": 15460
    },
    {
      "entropy": 5.839265966415406,
      "epoch": 1.3656835040621689,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004818084286135291,
      "loss": 5.4454,
      "mean_token_accuracy": 0.19646372199058532,
      "num_tokens": 28810276.0,
      "step": 15465
    },
    {
      "entropy": 5.876328182220459,
      "epoch": 1.3661250441540091,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004817960217361663,
      "loss": 5.4282,
      "mean_token_accuracy": 0.1994615763425827,
      "num_tokens": 28819545.0,
      "step": 15470
    },
    {
      "entropy": 5.847974491119385,
      "epoch": 1.3665665842458496,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000481783610807787,
      "loss": 5.319,
      "mean_token_accuracy": 0.2033898949623108,
      "num_tokens": 28828163.0,
      "step": 15475
    },
    {
      "entropy": 5.8331139087677,
      "epoch": 1.36700812433769,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004817711958286343,
      "loss": 5.4122,
      "mean_token_accuracy": 0.20001124441623688,
      "num_tokens": 28837952.0,
      "step": 15480
    },
    {
      "entropy": 5.846285486221314,
      "epoch": 1.3674496644295302,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004817587767989516,
      "loss": 5.4752,
      "mean_token_accuracy": 0.19427438229322433,
      "num_tokens": 28847702.0,
      "step": 15485
    },
    {
      "entropy": 5.885765027999878,
      "epoch": 1.3678912045213705,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048174635371898197,
      "loss": 5.4316,
      "mean_token_accuracy": 0.1955796003341675,
      "num_tokens": 28856936.0,
      "step": 15490
    },
    {
      "entropy": 5.840321588516235,
      "epoch": 1.3683327446132107,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004817339265889689,
      "loss": 5.3616,
      "mean_token_accuracy": 0.20821680575609208,
      "num_tokens": 28867072.0,
      "step": 15495
    },
    {
      "entropy": 5.72699875831604,
      "epoch": 1.3687742847050512,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004817214954091557,
      "loss": 5.3109,
      "mean_token_accuracy": 0.20816775858402253,
      "num_tokens": 28877546.0,
      "step": 15500
    },
    {
      "entropy": 5.853149509429931,
      "epoch": 1.3692158247968915,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048170906017978605,
      "loss": 5.4027,
      "mean_token_accuracy": 0.2011694148182869,
      "num_tokens": 28886804.0,
      "step": 15505
    },
    {
      "entropy": 5.871610164642334,
      "epoch": 1.3696573648887318,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004816966209011034,
      "loss": 5.4448,
      "mean_token_accuracy": 0.19629024118185043,
      "num_tokens": 28896476.0,
      "step": 15510
    },
    {
      "entropy": 5.905157041549683,
      "epoch": 1.3700989049805723,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048168417757335155,
      "loss": 5.4373,
      "mean_token_accuracy": 0.19920352846384048,
      "num_tokens": 28906269.0,
      "step": 15515
    },
    {
      "entropy": 5.782048845291138,
      "epoch": 1.3705404450724126,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004816717301967743,
      "loss": 5.3629,
      "mean_token_accuracy": 0.20571523308753967,
      "num_tokens": 28915792.0,
      "step": 15520
    },
    {
      "entropy": 5.8342865943908695,
      "epoch": 1.3709819851642528,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004816592787716152,
      "loss": 5.4243,
      "mean_token_accuracy": 0.19673479050397874,
      "num_tokens": 28925065.0,
      "step": 15525
    },
    {
      "entropy": 5.843438053131104,
      "epoch": 1.3714235252560933,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048164682329811855,
      "loss": 5.4206,
      "mean_token_accuracy": 0.19803667068481445,
      "num_tokens": 28933879.0,
      "step": 15530
    },
    {
      "entropy": 5.76366868019104,
      "epoch": 1.3718650653479336,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048163436377652804,
      "loss": 5.2675,
      "mean_token_accuracy": 0.21203935593366624,
      "num_tokens": 28943176.0,
      "step": 15535
    },
    {
      "entropy": 5.910560131072998,
      "epoch": 1.372306605439774,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004816219002070879,
      "loss": 5.4624,
      "mean_token_accuracy": 0.1993066370487213,
      "num_tokens": 28952770.0,
      "step": 15540
    },
    {
      "entropy": 5.737296867370605,
      "epoch": 1.3727481455316144,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004816094325900422,
      "loss": 5.282,
      "mean_token_accuracy": 0.20551921278238297,
      "num_tokens": 28961186.0,
      "step": 15545
    },
    {
      "entropy": 5.821661472320557,
      "epoch": 1.3731896856234547,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048159696092563533,
      "loss": 5.414,
      "mean_token_accuracy": 0.2063349574804306,
      "num_tokens": 28971787.0,
      "step": 15550
    },
    {
      "entropy": 5.872710752487182,
      "epoch": 1.373631225715295,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004815844852141114,
      "loss": 5.5081,
      "mean_token_accuracy": 0.19428355395793914,
      "num_tokens": 28982095.0,
      "step": 15555
    },
    {
      "entropy": 5.867133617401123,
      "epoch": 1.3740727658071352,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048157200545571497,
      "loss": 5.4204,
      "mean_token_accuracy": 0.2025141552090645,
      "num_tokens": 28991084.0,
      "step": 15560
    },
    {
      "entropy": 5.830327415466309,
      "epoch": 1.3745143058989755,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004815595216506903,
      "loss": 5.4007,
      "mean_token_accuracy": 0.20169737935066223,
      "num_tokens": 29000544.0,
      "step": 15565
    },
    {
      "entropy": 5.8615076541900635,
      "epoch": 1.374955845990816,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004815470337992822,
      "loss": 5.3009,
      "mean_token_accuracy": 0.20388926416635514,
      "num_tokens": 29009366.0,
      "step": 15570
    },
    {
      "entropy": 5.818301630020142,
      "epoch": 1.3753973860826563,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048153454190173517,
      "loss": 5.4771,
      "mean_token_accuracy": 0.19607685506343842,
      "num_tokens": 29018730.0,
      "step": 15575
    },
    {
      "entropy": 5.873378038406372,
      "epoch": 1.3758389261744965,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048152204595829394,
      "loss": 5.4973,
      "mean_token_accuracy": 0.18955930918455124,
      "num_tokens": 29030255.0,
      "step": 15580
    },
    {
      "entropy": 5.902130937576294,
      "epoch": 1.376280466266337,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004815095459692032,
      "loss": 5.3443,
      "mean_token_accuracy": 0.19729165881872177,
      "num_tokens": 29039030.0,
      "step": 15585
    },
    {
      "entropy": 5.861922359466552,
      "epoch": 1.3767220063581773,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000481497041934708,
      "loss": 5.3839,
      "mean_token_accuracy": 0.20678475350141526,
      "num_tokens": 29047504.0,
      "step": 15590
    },
    {
      "entropy": 5.859125852584839,
      "epoch": 1.3771635464500176,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048148453385505317,
      "loss": 5.4769,
      "mean_token_accuracy": 0.2015798032283783,
      "num_tokens": 29056488.0,
      "step": 15595
    },
    {
      "entropy": 5.829116582870483,
      "epoch": 1.377605086541858,
      "grad_norm": 0.953125,
      "learning_rate": 0.00048147202173048366,
      "loss": 5.4136,
      "mean_token_accuracy": 0.21554109305143357,
      "num_tokens": 29066067.0,
      "step": 15600
    },
    {
      "entropy": 5.915482711791992,
      "epoch": 1.3780466266336984,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004814595055612447,
      "loss": 5.4197,
      "mean_token_accuracy": 0.19795806407928468,
      "num_tokens": 29075136.0,
      "step": 15605
    },
    {
      "entropy": 5.772874355316162,
      "epoch": 1.3784881667255386,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048144698534758153,
      "loss": 5.3638,
      "mean_token_accuracy": 0.20789763033390046,
      "num_tokens": 29084970.0,
      "step": 15610
    },
    {
      "entropy": 5.668832015991211,
      "epoch": 1.3789297068173791,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048143446108973924,
      "loss": 5.3126,
      "mean_token_accuracy": 0.20958250313997268,
      "num_tokens": 29093234.0,
      "step": 15615
    },
    {
      "entropy": 5.777319002151489,
      "epoch": 1.3793712469092194,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048142193278796335,
      "loss": 5.3653,
      "mean_token_accuracy": 0.20201972424983977,
      "num_tokens": 29102115.0,
      "step": 15620
    },
    {
      "entropy": 5.875241136550903,
      "epoch": 1.3798127870010597,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004814094004424992,
      "loss": 5.4186,
      "mean_token_accuracy": 0.19845666438341142,
      "num_tokens": 29111065.0,
      "step": 15625
    },
    {
      "entropy": 5.814388656616211,
      "epoch": 1.3802543270929,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004813968640535922,
      "loss": 5.4358,
      "mean_token_accuracy": 0.1984635293483734,
      "num_tokens": 29120238.0,
      "step": 15630
    },
    {
      "entropy": 5.819582605361939,
      "epoch": 1.3806958671847402,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000481384323621488,
      "loss": 5.3498,
      "mean_token_accuracy": 0.19713763296604156,
      "num_tokens": 29129062.0,
      "step": 15635
    },
    {
      "entropy": 5.820842027664185,
      "epoch": 1.3811374072765807,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004813717791464324,
      "loss": 5.475,
      "mean_token_accuracy": 0.19261003732681276,
      "num_tokens": 29138990.0,
      "step": 15640
    },
    {
      "entropy": 5.821529245376587,
      "epoch": 1.381578947368421,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048135923062867094,
      "loss": 5.3442,
      "mean_token_accuracy": 0.20845245122909545,
      "num_tokens": 29147507.0,
      "step": 15645
    },
    {
      "entropy": 5.80028886795044,
      "epoch": 1.3820204874602613,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004813466780684495,
      "loss": 5.4679,
      "mean_token_accuracy": 0.19808721244335176,
      "num_tokens": 29156277.0,
      "step": 15650
    },
    {
      "entropy": 5.883765363693238,
      "epoch": 1.3824620275521018,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004813341214660141,
      "loss": 5.5728,
      "mean_token_accuracy": 0.1865738958120346,
      "num_tokens": 29165925.0,
      "step": 15655
    },
    {
      "entropy": 5.96726016998291,
      "epoch": 1.382903567643942,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004813215608216105,
      "loss": 5.4953,
      "mean_token_accuracy": 0.1907041296362877,
      "num_tokens": 29175811.0,
      "step": 15660
    },
    {
      "entropy": 5.873224258422852,
      "epoch": 1.3833451077357823,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048130899613548487,
      "loss": 5.4619,
      "mean_token_accuracy": 0.2033369868993759,
      "num_tokens": 29185100.0,
      "step": 15665
    },
    {
      "entropy": 5.836682558059692,
      "epoch": 1.3837866478276228,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00048129642740788337,
      "loss": 5.4429,
      "mean_token_accuracy": 0.19301197975873946,
      "num_tokens": 29194430.0,
      "step": 15670
    },
    {
      "entropy": 5.799141454696655,
      "epoch": 1.3842281879194631,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004812838546390521,
      "loss": 5.3786,
      "mean_token_accuracy": 0.201407028734684,
      "num_tokens": 29203897.0,
      "step": 15675
    },
    {
      "entropy": 5.851935958862304,
      "epoch": 1.3846697280113034,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004812712778292375,
      "loss": 5.5135,
      "mean_token_accuracy": 0.19677763432264328,
      "num_tokens": 29213474.0,
      "step": 15680
    },
    {
      "entropy": 5.938237094879151,
      "epoch": 1.3851112681031439,
      "grad_norm": 1.03125,
      "learning_rate": 0.000481258696978686,
      "loss": 5.4701,
      "mean_token_accuracy": 0.1967015489935875,
      "num_tokens": 29222176.0,
      "step": 15685
    },
    {
      "entropy": 5.785627746582032,
      "epoch": 1.3855528081949842,
      "grad_norm": 0.9375,
      "learning_rate": 0.00048124611208764375,
      "loss": 5.3687,
      "mean_token_accuracy": 0.19589827209711075,
      "num_tokens": 29231885.0,
      "step": 15690
    },
    {
      "entropy": 5.8095043182373045,
      "epoch": 1.3859943482868244,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004812335231563576,
      "loss": 5.3928,
      "mean_token_accuracy": 0.20141471773386002,
      "num_tokens": 29241921.0,
      "step": 15695
    },
    {
      "entropy": 5.859870910644531,
      "epoch": 1.3864358883786647,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00048122093018507396,
      "loss": 5.4818,
      "mean_token_accuracy": 0.19717176556587218,
      "num_tokens": 29251308.0,
      "step": 15700
    },
    {
      "entropy": 5.847552108764648,
      "epoch": 1.386877428470505,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004812083331740396,
      "loss": 5.4057,
      "mean_token_accuracy": 0.2079196572303772,
      "num_tokens": 29260628.0,
      "step": 15705
    },
    {
      "entropy": 5.880796718597412,
      "epoch": 1.3873189685623455,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048119573212350127,
      "loss": 5.4223,
      "mean_token_accuracy": 0.20460325926542283,
      "num_tokens": 29269530.0,
      "step": 15710
    },
    {
      "entropy": 5.852338266372681,
      "epoch": 1.3877605086541858,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048118312703370576,
      "loss": 5.4423,
      "mean_token_accuracy": 0.19531672596931457,
      "num_tokens": 29279846.0,
      "step": 15715
    },
    {
      "entropy": 5.890779876708985,
      "epoch": 1.388202048746026,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048117051790490007,
      "loss": 5.4318,
      "mean_token_accuracy": 0.19866774827241898,
      "num_tokens": 29289822.0,
      "step": 15720
    },
    {
      "entropy": 5.767150974273681,
      "epoch": 1.3886435888378665,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048115790473733124,
      "loss": 5.4258,
      "mean_token_accuracy": 0.19821067303419113,
      "num_tokens": 29299454.0,
      "step": 15725
    },
    {
      "entropy": 5.836018514633179,
      "epoch": 1.3890851289297068,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048114528753124625,
      "loss": 5.4575,
      "mean_token_accuracy": 0.19333788752555847,
      "num_tokens": 29308398.0,
      "step": 15730
    },
    {
      "entropy": 5.838931560516357,
      "epoch": 1.389526669021547,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00048113266628689236,
      "loss": 5.4201,
      "mean_token_accuracy": 0.2065933644771576,
      "num_tokens": 29318968.0,
      "step": 15735
    },
    {
      "entropy": 5.8998829364776615,
      "epoch": 1.3899682091133876,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004811200410045167,
      "loss": 5.4955,
      "mean_token_accuracy": 0.18485084027051926,
      "num_tokens": 29329745.0,
      "step": 15740
    },
    {
      "entropy": 5.811136388778687,
      "epoch": 1.3904097492052279,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004811074116843667,
      "loss": 5.3131,
      "mean_token_accuracy": 0.20991339683532714,
      "num_tokens": 29338438.0,
      "step": 15745
    },
    {
      "entropy": 5.779156637191773,
      "epoch": 1.3908512892970681,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048109477832668977,
      "loss": 5.3881,
      "mean_token_accuracy": 0.19703128039836884,
      "num_tokens": 29347783.0,
      "step": 15750
    },
    {
      "entropy": 5.817676305770874,
      "epoch": 1.3912928293889086,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004810821409317334,
      "loss": 5.451,
      "mean_token_accuracy": 0.20315441936254502,
      "num_tokens": 29357062.0,
      "step": 15755
    },
    {
      "entropy": 5.853411912918091,
      "epoch": 1.391734369480749,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048106949949974494,
      "loss": 5.4146,
      "mean_token_accuracy": 0.20485300123691558,
      "num_tokens": 29366361.0,
      "step": 15760
    },
    {
      "entropy": 5.779572868347168,
      "epoch": 1.3921759095725892,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004810568540309723,
      "loss": 5.4461,
      "mean_token_accuracy": 0.19637374430894852,
      "num_tokens": 29375992.0,
      "step": 15765
    },
    {
      "entropy": 5.84677357673645,
      "epoch": 1.3926174496644295,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00048104420452566303,
      "loss": 5.4957,
      "mean_token_accuracy": 0.19725880324840545,
      "num_tokens": 29386408.0,
      "step": 15770
    },
    {
      "entropy": 5.929966163635254,
      "epoch": 1.3930589897562697,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048103155098406505,
      "loss": 5.5096,
      "mean_token_accuracy": 0.1918272316455841,
      "num_tokens": 29395881.0,
      "step": 15775
    },
    {
      "entropy": 5.824849271774292,
      "epoch": 1.3935005298481102,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048101889340642604,
      "loss": 5.3238,
      "mean_token_accuracy": 0.20882944762706757,
      "num_tokens": 29404397.0,
      "step": 15780
    },
    {
      "entropy": 5.768926668167114,
      "epoch": 1.3939420699399505,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004810062317929941,
      "loss": 5.3853,
      "mean_token_accuracy": 0.20247988849878312,
      "num_tokens": 29413219.0,
      "step": 15785
    },
    {
      "entropy": 5.858049106597901,
      "epoch": 1.3943836100317908,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004809935661440172,
      "loss": 5.4772,
      "mean_token_accuracy": 0.19423361569643022,
      "num_tokens": 29422727.0,
      "step": 15790
    },
    {
      "entropy": 5.911684560775757,
      "epoch": 1.3948251501236313,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004809808964597436,
      "loss": 5.4717,
      "mean_token_accuracy": 0.19965349435806273,
      "num_tokens": 29432183.0,
      "step": 15795
    },
    {
      "entropy": 5.823192358016968,
      "epoch": 1.3952666902154716,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048096822274042145,
      "loss": 5.3866,
      "mean_token_accuracy": 0.2023373946547508,
      "num_tokens": 29440821.0,
      "step": 15800
    },
    {
      "entropy": 5.864645719528198,
      "epoch": 1.3957082303073118,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048095554498629885,
      "loss": 5.4923,
      "mean_token_accuracy": 0.19196839183568953,
      "num_tokens": 29450128.0,
      "step": 15805
    },
    {
      "entropy": 5.82134780883789,
      "epoch": 1.3961497703991523,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048094286319762426,
      "loss": 5.3799,
      "mean_token_accuracy": 0.2021815672516823,
      "num_tokens": 29459271.0,
      "step": 15810
    },
    {
      "entropy": 5.82841420173645,
      "epoch": 1.3965913104909926,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048093017737464613,
      "loss": 5.3105,
      "mean_token_accuracy": 0.21012148708105088,
      "num_tokens": 29468517.0,
      "step": 15815
    },
    {
      "entropy": 5.800443077087403,
      "epoch": 1.3970328505828329,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048091748751761295,
      "loss": 5.5048,
      "mean_token_accuracy": 0.19221622943878175,
      "num_tokens": 29478330.0,
      "step": 15820
    },
    {
      "entropy": 5.881490993499756,
      "epoch": 1.3974743906746734,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004809047936267732,
      "loss": 5.4419,
      "mean_token_accuracy": 0.19671426713466644,
      "num_tokens": 29487190.0,
      "step": 15825
    },
    {
      "entropy": 5.894149541854858,
      "epoch": 1.3979159307665137,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048089209570237573,
      "loss": 5.5076,
      "mean_token_accuracy": 0.1914274886250496,
      "num_tokens": 29495885.0,
      "step": 15830
    },
    {
      "entropy": 5.834230709075928,
      "epoch": 1.398357470858354,
      "grad_norm": 1.0,
      "learning_rate": 0.0004808793937446692,
      "loss": 5.3744,
      "mean_token_accuracy": 0.19707161337137222,
      "num_tokens": 29505524.0,
      "step": 15835
    },
    {
      "entropy": 5.8079897403717045,
      "epoch": 1.3987990109501942,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004808666877539025,
      "loss": 5.3666,
      "mean_token_accuracy": 0.20344207137823106,
      "num_tokens": 29514003.0,
      "step": 15840
    },
    {
      "entropy": 5.819286441802978,
      "epoch": 1.3992405510420345,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048085397773032433,
      "loss": 5.3936,
      "mean_token_accuracy": 0.20126733928918839,
      "num_tokens": 29523236.0,
      "step": 15845
    },
    {
      "entropy": 5.875626707077027,
      "epoch": 1.399682091133875,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004808412636741839,
      "loss": 5.4526,
      "mean_token_accuracy": 0.20521328896284102,
      "num_tokens": 29532852.0,
      "step": 15850
    },
    {
      "entropy": 5.809407520294189,
      "epoch": 1.4001236312257153,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048082854558573017,
      "loss": 5.4183,
      "mean_token_accuracy": 0.20285482555627823,
      "num_tokens": 29542035.0,
      "step": 15855
    },
    {
      "entropy": 5.804487419128418,
      "epoch": 1.4005651713175555,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048081582346521236,
      "loss": 5.3698,
      "mean_token_accuracy": 0.20402562767267227,
      "num_tokens": 29551978.0,
      "step": 15860
    },
    {
      "entropy": 5.9571596622467045,
      "epoch": 1.401006711409396,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004808030973128795,
      "loss": 5.5206,
      "mean_token_accuracy": 0.19067658483982086,
      "num_tokens": 29562011.0,
      "step": 15865
    },
    {
      "entropy": 5.859301424026489,
      "epoch": 1.4014482515012363,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048079036712898106,
      "loss": 5.398,
      "mean_token_accuracy": 0.204546457529068,
      "num_tokens": 29571498.0,
      "step": 15870
    },
    {
      "entropy": 5.846553802490234,
      "epoch": 1.4018897915930766,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048077763291376637,
      "loss": 5.353,
      "mean_token_accuracy": 0.21147865504026414,
      "num_tokens": 29580704.0,
      "step": 15875
    },
    {
      "entropy": 5.852332878112793,
      "epoch": 1.402331331684917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004807648946674849,
      "loss": 5.3271,
      "mean_token_accuracy": 0.20979054421186447,
      "num_tokens": 29590066.0,
      "step": 15880
    },
    {
      "entropy": 5.84045147895813,
      "epoch": 1.4027728717767574,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004807521523903862,
      "loss": 5.4466,
      "mean_token_accuracy": 0.19593202024698259,
      "num_tokens": 29599473.0,
      "step": 15885
    },
    {
      "entropy": 5.80124797821045,
      "epoch": 1.4032144118685976,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048073940608271974,
      "loss": 5.4356,
      "mean_token_accuracy": 0.1963476449251175,
      "num_tokens": 29609371.0,
      "step": 15890
    },
    {
      "entropy": 5.835998010635376,
      "epoch": 1.4036559519604381,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004807266557447354,
      "loss": 5.2955,
      "mean_token_accuracy": 0.20974289923906325,
      "num_tokens": 29618821.0,
      "step": 15895
    },
    {
      "entropy": 5.832048559188843,
      "epoch": 1.4040974920522784,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004807139013766829,
      "loss": 5.3281,
      "mean_token_accuracy": 0.21409380435943604,
      "num_tokens": 29627491.0,
      "step": 15900
    },
    {
      "entropy": 5.778929615020752,
      "epoch": 1.4045390321441187,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048070114297881205,
      "loss": 5.4345,
      "mean_token_accuracy": 0.20080992728471755,
      "num_tokens": 29637699.0,
      "step": 15905
    },
    {
      "entropy": 5.86026930809021,
      "epoch": 1.404980572235959,
      "grad_norm": 1.0,
      "learning_rate": 0.0004806883805513728,
      "loss": 5.3928,
      "mean_token_accuracy": 0.20685179382562638,
      "num_tokens": 29646439.0,
      "step": 15910
    },
    {
      "entropy": 5.87608232498169,
      "epoch": 1.4054221123277992,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004806756140946151,
      "loss": 5.4643,
      "mean_token_accuracy": 0.19283811748027802,
      "num_tokens": 29655704.0,
      "step": 15915
    },
    {
      "entropy": 5.909402322769165,
      "epoch": 1.4058636524196397,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048066284360878913,
      "loss": 5.5232,
      "mean_token_accuracy": 0.19277155250310898,
      "num_tokens": 29665163.0,
      "step": 15920
    },
    {
      "entropy": 5.838028860092163,
      "epoch": 1.40630519251148,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048065006909414507,
      "loss": 5.4532,
      "mean_token_accuracy": 0.2018750011920929,
      "num_tokens": 29674803.0,
      "step": 15925
    },
    {
      "entropy": 5.85578031539917,
      "epoch": 1.4067467326033203,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004806372905509331,
      "loss": 5.4448,
      "mean_token_accuracy": 0.19554970562458038,
      "num_tokens": 29685149.0,
      "step": 15930
    },
    {
      "entropy": 5.747699022293091,
      "epoch": 1.4071882726951608,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048062450797940347,
      "loss": 5.2908,
      "mean_token_accuracy": 0.21219534277915955,
      "num_tokens": 29694224.0,
      "step": 15935
    },
    {
      "entropy": 5.819090509414673,
      "epoch": 1.407629812787001,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004806117213798068,
      "loss": 5.4124,
      "mean_token_accuracy": 0.20134179890155793,
      "num_tokens": 29704512.0,
      "step": 15940
    },
    {
      "entropy": 5.799300861358643,
      "epoch": 1.4080713528788413,
      "grad_norm": 0.984375,
      "learning_rate": 0.00048059893075239334,
      "loss": 5.3279,
      "mean_token_accuracy": 0.20564483851194382,
      "num_tokens": 29712990.0,
      "step": 15945
    },
    {
      "entropy": 5.856165552139283,
      "epoch": 1.4085128929706818,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048058613609741374,
      "loss": 5.4645,
      "mean_token_accuracy": 0.19290683269500733,
      "num_tokens": 29722844.0,
      "step": 15950
    },
    {
      "entropy": 5.8596423149108885,
      "epoch": 1.408954433062522,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004805733374151188,
      "loss": 5.3404,
      "mean_token_accuracy": 0.19993604123592376,
      "num_tokens": 29731593.0,
      "step": 15955
    },
    {
      "entropy": 5.820978832244873,
      "epoch": 1.4093959731543624,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048056053470575894,
      "loss": 5.4446,
      "mean_token_accuracy": 0.19267890006303787,
      "num_tokens": 29741052.0,
      "step": 15960
    },
    {
      "entropy": 5.846028423309326,
      "epoch": 1.4098375132462029,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048054772796958517,
      "loss": 5.429,
      "mean_token_accuracy": 0.197341625392437,
      "num_tokens": 29750187.0,
      "step": 15965
    },
    {
      "entropy": 5.770731258392334,
      "epoch": 1.4102790533380432,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00048053491720684836,
      "loss": 5.3562,
      "mean_token_accuracy": 0.19872645139694214,
      "num_tokens": 29759574.0,
      "step": 15970
    },
    {
      "entropy": 5.780177783966065,
      "epoch": 1.4107205934298834,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004805221024177994,
      "loss": 5.4024,
      "mean_token_accuracy": 0.20029570162296295,
      "num_tokens": 29769549.0,
      "step": 15975
    },
    {
      "entropy": 5.880591773986817,
      "epoch": 1.4111621335217237,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004805092836026893,
      "loss": 5.4477,
      "mean_token_accuracy": 0.1947784662246704,
      "num_tokens": 29778643.0,
      "step": 15980
    },
    {
      "entropy": 5.864433097839355,
      "epoch": 1.411603673613564,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004804964607617693,
      "loss": 5.391,
      "mean_token_accuracy": 0.20077617317438126,
      "num_tokens": 29788685.0,
      "step": 15985
    },
    {
      "entropy": 5.836127805709839,
      "epoch": 1.4120452137054045,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00048048363389529045,
      "loss": 5.3648,
      "mean_token_accuracy": 0.2051248475909233,
      "num_tokens": 29798648.0,
      "step": 15990
    },
    {
      "entropy": 5.820440435409546,
      "epoch": 1.4124867537972448,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000480470803003504,
      "loss": 5.4195,
      "mean_token_accuracy": 0.1922850042581558,
      "num_tokens": 29807582.0,
      "step": 15995
    },
    {
      "entropy": 5.840159368515015,
      "epoch": 1.412928293889085,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004804579680866614,
      "loss": 5.4258,
      "mean_token_accuracy": 0.1995154470205307,
      "num_tokens": 29817443.0,
      "step": 16000
    },
    {
      "entropy": 5.875101613998413,
      "epoch": 1.4133698339809255,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004804451291450142,
      "loss": 5.4363,
      "mean_token_accuracy": 0.20084422528743745,
      "num_tokens": 29825796.0,
      "step": 16005
    },
    {
      "entropy": 5.837054777145386,
      "epoch": 1.4138113740727658,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048043228617881365,
      "loss": 5.3585,
      "mean_token_accuracy": 0.20034700334072114,
      "num_tokens": 29834874.0,
      "step": 16010
    },
    {
      "entropy": 5.848668813705444,
      "epoch": 1.414252914164606,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004804194391883114,
      "loss": 5.4117,
      "mean_token_accuracy": 0.19753310233354568,
      "num_tokens": 29843884.0,
      "step": 16015
    },
    {
      "entropy": 5.919211912155151,
      "epoch": 1.4146944542564466,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048040658817375927,
      "loss": 5.5199,
      "mean_token_accuracy": 0.1868203714489937,
      "num_tokens": 29853180.0,
      "step": 16020
    },
    {
      "entropy": 5.853323316574096,
      "epoch": 1.4151359943482869,
      "grad_norm": 0.96875,
      "learning_rate": 0.00048039373313540885,
      "loss": 5.384,
      "mean_token_accuracy": 0.19485014379024507,
      "num_tokens": 29862926.0,
      "step": 16025
    },
    {
      "entropy": 5.773474931716919,
      "epoch": 1.4155775344401271,
      "grad_norm": 1.109375,
      "learning_rate": 0.000480380874073512,
      "loss": 5.3709,
      "mean_token_accuracy": 0.20578792691230774,
      "num_tokens": 29871450.0,
      "step": 16030
    },
    {
      "entropy": 5.770298099517822,
      "epoch": 1.4160190745319676,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00048036801098832067,
      "loss": 5.3539,
      "mean_token_accuracy": 0.20704010576009751,
      "num_tokens": 29880142.0,
      "step": 16035
    },
    {
      "entropy": 5.808156871795655,
      "epoch": 1.416460614623808,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004803551438800868,
      "loss": 5.4037,
      "mean_token_accuracy": 0.20299105495214462,
      "num_tokens": 29888668.0,
      "step": 16040
    },
    {
      "entropy": 5.906285238265991,
      "epoch": 1.4169021547156482,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048034227274906235,
      "loss": 5.5293,
      "mean_token_accuracy": 0.19299840480089187,
      "num_tokens": 29898727.0,
      "step": 16045
    },
    {
      "entropy": 5.859495830535889,
      "epoch": 1.4173436948074885,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048032939759549964,
      "loss": 5.441,
      "mean_token_accuracy": 0.18627241849899293,
      "num_tokens": 29909089.0,
      "step": 16050
    },
    {
      "entropy": 5.883098411560058,
      "epoch": 1.4177852348993287,
      "grad_norm": 1.0,
      "learning_rate": 0.00048031651841965085,
      "loss": 5.3967,
      "mean_token_accuracy": 0.2024922713637352,
      "num_tokens": 29918169.0,
      "step": 16055
    },
    {
      "entropy": 5.800869464874268,
      "epoch": 1.4182267749911692,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004803036352217682,
      "loss": 5.4241,
      "mean_token_accuracy": 0.2045391783118248,
      "num_tokens": 29928160.0,
      "step": 16060
    },
    {
      "entropy": 5.808574485778808,
      "epoch": 1.4186683150830095,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048029074800210413,
      "loss": 5.4323,
      "mean_token_accuracy": 0.1945517599582672,
      "num_tokens": 29937695.0,
      "step": 16065
    },
    {
      "entropy": 5.9447166442871096,
      "epoch": 1.4191098551748498,
      "grad_norm": 0.953125,
      "learning_rate": 0.000480277856760911,
      "loss": 5.4366,
      "mean_token_accuracy": 0.19815507382154465,
      "num_tokens": 29947380.0,
      "step": 16070
    },
    {
      "entropy": 5.945921468734741,
      "epoch": 1.4195513952666903,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048026496149844146,
      "loss": 5.4953,
      "mean_token_accuracy": 0.1980571538209915,
      "num_tokens": 29956313.0,
      "step": 16075
    },
    {
      "entropy": 5.783976030349732,
      "epoch": 1.4199929353585306,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00048025206221494806,
      "loss": 5.3398,
      "mean_token_accuracy": 0.20875429213047028,
      "num_tokens": 29965502.0,
      "step": 16080
    },
    {
      "entropy": 5.830239534378052,
      "epoch": 1.4204344754503708,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004802391589106835,
      "loss": 5.3564,
      "mean_token_accuracy": 0.20533740669488906,
      "num_tokens": 29974243.0,
      "step": 16085
    },
    {
      "entropy": 5.792404699325561,
      "epoch": 1.4208760155422113,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004802262515859005,
      "loss": 5.3209,
      "mean_token_accuracy": 0.21151241958141326,
      "num_tokens": 29983966.0,
      "step": 16090
    },
    {
      "entropy": 5.828254079818725,
      "epoch": 1.4213175556340516,
      "grad_norm": 1.0,
      "learning_rate": 0.0004802133402408519,
      "loss": 5.4248,
      "mean_token_accuracy": 0.20044248104095458,
      "num_tokens": 29993365.0,
      "step": 16095
    },
    {
      "entropy": 5.9269355773925785,
      "epoch": 1.4217590957258919,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048020042487579076,
      "loss": 5.4326,
      "mean_token_accuracy": 0.2006969228386879,
      "num_tokens": 30001919.0,
      "step": 16100
    },
    {
      "entropy": 5.782835054397583,
      "epoch": 1.4222006358177324,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048018750549097,
      "loss": 5.3833,
      "mean_token_accuracy": 0.20162815004587173,
      "num_tokens": 30011353.0,
      "step": 16105
    },
    {
      "entropy": 5.840275001525879,
      "epoch": 1.4226421759095726,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048017458208664265,
      "loss": 5.4483,
      "mean_token_accuracy": 0.19493011832237245,
      "num_tokens": 30019994.0,
      "step": 16110
    },
    {
      "entropy": 5.760211086273193,
      "epoch": 1.423083716001413,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000480161654663062,
      "loss": 5.3076,
      "mean_token_accuracy": 0.20146130621433259,
      "num_tokens": 30028934.0,
      "step": 16115
    },
    {
      "entropy": 5.8535082817077635,
      "epoch": 1.4235252560932532,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004801487232204811,
      "loss": 5.3229,
      "mean_token_accuracy": 0.1995198518037796,
      "num_tokens": 30038069.0,
      "step": 16120
    },
    {
      "entropy": 5.909199523925781,
      "epoch": 1.4239667961850935,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004801357877591535,
      "loss": 5.4848,
      "mean_token_accuracy": 0.19611310362815856,
      "num_tokens": 30047307.0,
      "step": 16125
    },
    {
      "entropy": 5.845629167556763,
      "epoch": 1.424408336276934,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048012284827933245,
      "loss": 5.4188,
      "mean_token_accuracy": 0.20130494832992554,
      "num_tokens": 30057429.0,
      "step": 16130
    },
    {
      "entropy": 5.752436351776123,
      "epoch": 1.4248498763687742,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004801099047812715,
      "loss": 5.3537,
      "mean_token_accuracy": 0.20477347522974015,
      "num_tokens": 30067327.0,
      "step": 16135
    },
    {
      "entropy": 5.880493450164795,
      "epoch": 1.4252914164606145,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004800969572652241,
      "loss": 5.4739,
      "mean_token_accuracy": 0.19458369463682174,
      "num_tokens": 30077103.0,
      "step": 16140
    },
    {
      "entropy": 5.926301002502441,
      "epoch": 1.425732956552455,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048008400573144394,
      "loss": 5.5277,
      "mean_token_accuracy": 0.1887054920196533,
      "num_tokens": 30087539.0,
      "step": 16145
    },
    {
      "entropy": 5.919013166427613,
      "epoch": 1.4261744966442953,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00048007105018018487,
      "loss": 5.3959,
      "mean_token_accuracy": 0.20030292570590974,
      "num_tokens": 30097153.0,
      "step": 16150
    },
    {
      "entropy": 5.850428581237793,
      "epoch": 1.4266160367361356,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048005809061170044,
      "loss": 5.4753,
      "mean_token_accuracy": 0.1989583507180214,
      "num_tokens": 30106291.0,
      "step": 16155
    },
    {
      "entropy": 5.855349159240722,
      "epoch": 1.427057576827976,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004800451270262447,
      "loss": 5.4798,
      "mean_token_accuracy": 0.19883015006780624,
      "num_tokens": 30116441.0,
      "step": 16160
    },
    {
      "entropy": 5.824636077880859,
      "epoch": 1.4274991169198163,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004800321594240714,
      "loss": 5.3588,
      "mean_token_accuracy": 0.21407222151756286,
      "num_tokens": 30125040.0,
      "step": 16165
    },
    {
      "entropy": 5.8391622543334964,
      "epoch": 1.4279406570116566,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004800191878054349,
      "loss": 5.3444,
      "mean_token_accuracy": 0.20534966886043549,
      "num_tokens": 30134284.0,
      "step": 16170
    },
    {
      "entropy": 5.86709623336792,
      "epoch": 1.4283821971034971,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048000621217058904,
      "loss": 5.3328,
      "mean_token_accuracy": 0.20978496223688126,
      "num_tokens": 30143679.0,
      "step": 16175
    },
    {
      "entropy": 5.7082648277282715,
      "epoch": 1.4288237371953374,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047999323251978805,
      "loss": 5.2711,
      "mean_token_accuracy": 0.21223915368318558,
      "num_tokens": 30152011.0,
      "step": 16180
    },
    {
      "entropy": 5.734315299987793,
      "epoch": 1.4292652772871777,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047998024885328625,
      "loss": 5.3649,
      "mean_token_accuracy": 0.1998768299818039,
      "num_tokens": 30161949.0,
      "step": 16185
    },
    {
      "entropy": 5.853041696548462,
      "epoch": 1.429706817379018,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047996726117133795,
      "loss": 5.3856,
      "mean_token_accuracy": 0.2079288735985756,
      "num_tokens": 30172003.0,
      "step": 16190
    },
    {
      "entropy": 5.744950151443481,
      "epoch": 1.4301483574708582,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004799542694741975,
      "loss": 5.3959,
      "mean_token_accuracy": 0.20896569937467574,
      "num_tokens": 30181918.0,
      "step": 16195
    },
    {
      "entropy": 5.811241388320923,
      "epoch": 1.4305898975626987,
      "grad_norm": 1.0,
      "learning_rate": 0.00047994127376211953,
      "loss": 5.3457,
      "mean_token_accuracy": 0.20625263452529907,
      "num_tokens": 30191032.0,
      "step": 16200
    },
    {
      "entropy": 5.8363714694976805,
      "epoch": 1.431031437654539,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004799282740353586,
      "loss": 5.3662,
      "mean_token_accuracy": 0.20566608756780624,
      "num_tokens": 30200276.0,
      "step": 16205
    },
    {
      "entropy": 5.909934043884277,
      "epoch": 1.4314729777463793,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047991527029416924,
      "loss": 5.4271,
      "mean_token_accuracy": 0.20007840394973755,
      "num_tokens": 30208574.0,
      "step": 16210
    },
    {
      "entropy": 5.809680318832397,
      "epoch": 1.4319145178382198,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004799022625388064,
      "loss": 5.4345,
      "mean_token_accuracy": 0.20435420274734498,
      "num_tokens": 30218912.0,
      "step": 16215
    },
    {
      "entropy": 5.885587930679321,
      "epoch": 1.43235605793006,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004798892507695247,
      "loss": 5.5081,
      "mean_token_accuracy": 0.19482046067714692,
      "num_tokens": 30228325.0,
      "step": 16220
    },
    {
      "entropy": 5.860631465911865,
      "epoch": 1.4327975980219003,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004798762349865791,
      "loss": 5.3984,
      "mean_token_accuracy": 0.19849643260240554,
      "num_tokens": 30237169.0,
      "step": 16225
    },
    {
      "entropy": 5.825541543960571,
      "epoch": 1.4332391381137408,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004798632151902246,
      "loss": 5.3692,
      "mean_token_accuracy": 0.20470220744609832,
      "num_tokens": 30246452.0,
      "step": 16230
    },
    {
      "entropy": 5.887117719650268,
      "epoch": 1.433680678205581,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047985019138071616,
      "loss": 5.4256,
      "mean_token_accuracy": 0.20170413553714753,
      "num_tokens": 30256055.0,
      "step": 16235
    },
    {
      "entropy": 5.784398937225342,
      "epoch": 1.4341222182974214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047983716355830903,
      "loss": 5.2876,
      "mean_token_accuracy": 0.20524161607027053,
      "num_tokens": 30265764.0,
      "step": 16240
    },
    {
      "entropy": 5.830113935470581,
      "epoch": 1.4345637583892619,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047982413172325845,
      "loss": 5.4367,
      "mean_token_accuracy": 0.20573489218950272,
      "num_tokens": 30274490.0,
      "step": 16245
    },
    {
      "entropy": 5.811930418014526,
      "epoch": 1.4350052984811021,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047981109587581945,
      "loss": 5.3746,
      "mean_token_accuracy": 0.20781485587358475,
      "num_tokens": 30283330.0,
      "step": 16250
    },
    {
      "entropy": 5.882568454742431,
      "epoch": 1.4354468385729424,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004797980560162477,
      "loss": 5.5341,
      "mean_token_accuracy": 0.1854379415512085,
      "num_tokens": 30293542.0,
      "step": 16255
    },
    {
      "entropy": 5.853444290161133,
      "epoch": 1.4358883786647827,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004797850121447984,
      "loss": 5.4736,
      "mean_token_accuracy": 0.2044121265411377,
      "num_tokens": 30302122.0,
      "step": 16260
    },
    {
      "entropy": 5.838605976104736,
      "epoch": 1.436329918756623,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047977196426172725,
      "loss": 5.5062,
      "mean_token_accuracy": 0.2023172616958618,
      "num_tokens": 30311490.0,
      "step": 16265
    },
    {
      "entropy": 5.813236713409424,
      "epoch": 1.4367714588484635,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047975891236728973,
      "loss": 5.3439,
      "mean_token_accuracy": 0.2104988604784012,
      "num_tokens": 30321199.0,
      "step": 16270
    },
    {
      "entropy": 5.876508617401123,
      "epoch": 1.4372129989403037,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047974585646174165,
      "loss": 5.4044,
      "mean_token_accuracy": 0.1991390511393547,
      "num_tokens": 30330114.0,
      "step": 16275
    },
    {
      "entropy": 5.832000017166138,
      "epoch": 1.437654539032144,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047973279654533866,
      "loss": 5.4317,
      "mean_token_accuracy": 0.20468027740716935,
      "num_tokens": 30339290.0,
      "step": 16280
    },
    {
      "entropy": 5.803961992263794,
      "epoch": 1.4380960791239845,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004797197326183365,
      "loss": 5.3772,
      "mean_token_accuracy": 0.20148180723190307,
      "num_tokens": 30349595.0,
      "step": 16285
    },
    {
      "entropy": 5.9002259254455565,
      "epoch": 1.4385376192158248,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004797066646809914,
      "loss": 5.4182,
      "mean_token_accuracy": 0.1969104915857315,
      "num_tokens": 30358010.0,
      "step": 16290
    },
    {
      "entropy": 5.88068585395813,
      "epoch": 1.438979159307665,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047969359273355907,
      "loss": 5.4968,
      "mean_token_accuracy": 0.19931793510913848,
      "num_tokens": 30367408.0,
      "step": 16295
    },
    {
      "entropy": 5.829802513122559,
      "epoch": 1.4394206993995056,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047968051677629565,
      "loss": 5.4762,
      "mean_token_accuracy": 0.20065601766109467,
      "num_tokens": 30376475.0,
      "step": 16300
    },
    {
      "entropy": 5.870932626724243,
      "epoch": 1.4398622394913458,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047966743680945734,
      "loss": 5.3926,
      "mean_token_accuracy": 0.19485222548246384,
      "num_tokens": 30385361.0,
      "step": 16305
    },
    {
      "entropy": 5.937409687042236,
      "epoch": 1.4403037795831861,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004796543528333004,
      "loss": 5.4216,
      "mean_token_accuracy": 0.200754514336586,
      "num_tokens": 30393594.0,
      "step": 16310
    },
    {
      "entropy": 5.845773601531983,
      "epoch": 1.4407453196750266,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000479641264848081,
      "loss": 5.406,
      "mean_token_accuracy": 0.20125416219234465,
      "num_tokens": 30403444.0,
      "step": 16315
    },
    {
      "entropy": 5.743726015090942,
      "epoch": 1.441186859766867,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004796281728540556,
      "loss": 5.2932,
      "mean_token_accuracy": 0.21228694766759873,
      "num_tokens": 30412485.0,
      "step": 16320
    },
    {
      "entropy": 5.808514165878296,
      "epoch": 1.4416283998587072,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047961507685148077,
      "loss": 5.3553,
      "mean_token_accuracy": 0.20754000097513198,
      "num_tokens": 30421104.0,
      "step": 16325
    },
    {
      "entropy": 5.804881477355957,
      "epoch": 1.4420699399505474,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047960197684061287,
      "loss": 5.3779,
      "mean_token_accuracy": 0.20005034208297728,
      "num_tokens": 30430061.0,
      "step": 16330
    },
    {
      "entropy": 5.856679153442383,
      "epoch": 1.4425114800423877,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004795888728217086,
      "loss": 5.3118,
      "mean_token_accuracy": 0.20085889101028442,
      "num_tokens": 30437645.0,
      "step": 16335
    },
    {
      "entropy": 5.7644130229949955,
      "epoch": 1.4429530201342282,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047957576479502466,
      "loss": 5.2905,
      "mean_token_accuracy": 0.21740503013134002,
      "num_tokens": 30447805.0,
      "step": 16340
    },
    {
      "entropy": 5.870812320709229,
      "epoch": 1.4433945602260685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004795626527608179,
      "loss": 5.4356,
      "mean_token_accuracy": 0.19969286918640136,
      "num_tokens": 30457524.0,
      "step": 16345
    },
    {
      "entropy": 5.907554388046265,
      "epoch": 1.4438361003179088,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004795495367193451,
      "loss": 5.4057,
      "mean_token_accuracy": 0.2024523138999939,
      "num_tokens": 30467082.0,
      "step": 16350
    },
    {
      "entropy": 5.876122283935547,
      "epoch": 1.4442776404097493,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047953641667086315,
      "loss": 5.456,
      "mean_token_accuracy": 0.20295267701148986,
      "num_tokens": 30476251.0,
      "step": 16355
    },
    {
      "entropy": 5.787760448455811,
      "epoch": 1.4447191805015895,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047952329261562914,
      "loss": 5.3478,
      "mean_token_accuracy": 0.20849206298589706,
      "num_tokens": 30485997.0,
      "step": 16360
    },
    {
      "entropy": 5.886525630950928,
      "epoch": 1.4451607205934298,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047951016455390023,
      "loss": 5.4898,
      "mean_token_accuracy": 0.1977571427822113,
      "num_tokens": 30494886.0,
      "step": 16365
    },
    {
      "entropy": 5.880224132537842,
      "epoch": 1.4456022606852703,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004794970324859334,
      "loss": 5.4754,
      "mean_token_accuracy": 0.19239621758460998,
      "num_tokens": 30504194.0,
      "step": 16370
    },
    {
      "entropy": 5.791299486160279,
      "epoch": 1.4460438007771106,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047948389641198615,
      "loss": 5.343,
      "mean_token_accuracy": 0.20522153228521348,
      "num_tokens": 30513077.0,
      "step": 16375
    },
    {
      "entropy": 5.824699258804321,
      "epoch": 1.4464853408689509,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047947075633231553,
      "loss": 5.4307,
      "mean_token_accuracy": 0.19653559178113938,
      "num_tokens": 30521999.0,
      "step": 16380
    },
    {
      "entropy": 5.914393568038941,
      "epoch": 1.4469268809607914,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004794576122471791,
      "loss": 5.502,
      "mean_token_accuracy": 0.19254199117422105,
      "num_tokens": 30530138.0,
      "step": 16385
    },
    {
      "entropy": 5.813262462615967,
      "epoch": 1.4473684210526316,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004794444641568344,
      "loss": 5.3256,
      "mean_token_accuracy": 0.2041580557823181,
      "num_tokens": 30538850.0,
      "step": 16390
    },
    {
      "entropy": 5.8017425537109375,
      "epoch": 1.447809961144472,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004794313120615388,
      "loss": 5.4109,
      "mean_token_accuracy": 0.19747596085071564,
      "num_tokens": 30547593.0,
      "step": 16395
    },
    {
      "entropy": 5.81569275856018,
      "epoch": 1.4482515012363122,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004794181559615502,
      "loss": 5.375,
      "mean_token_accuracy": 0.20397378653287887,
      "num_tokens": 30557154.0,
      "step": 16400
    },
    {
      "entropy": 5.920955467224121,
      "epoch": 1.4486930413281525,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047940499585712615,
      "loss": 5.3749,
      "mean_token_accuracy": 0.20880068987607955,
      "num_tokens": 30566379.0,
      "step": 16405
    },
    {
      "entropy": 5.820366382598877,
      "epoch": 1.449134581419993,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047939183174852446,
      "loss": 5.3284,
      "mean_token_accuracy": 0.21193971037864684,
      "num_tokens": 30575146.0,
      "step": 16410
    },
    {
      "entropy": 5.720854997634888,
      "epoch": 1.4495761215118332,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047937866363600295,
      "loss": 5.4111,
      "mean_token_accuracy": 0.20401481986045839,
      "num_tokens": 30584444.0,
      "step": 16415
    },
    {
      "entropy": 5.8042151927948,
      "epoch": 1.4500176616036735,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004793654915198198,
      "loss": 5.3861,
      "mean_token_accuracy": 0.21130629181861876,
      "num_tokens": 30593779.0,
      "step": 16420
    },
    {
      "entropy": 5.881678485870362,
      "epoch": 1.450459201695514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004793523154002328,
      "loss": 5.424,
      "mean_token_accuracy": 0.2025240734219551,
      "num_tokens": 30603762.0,
      "step": 16425
    },
    {
      "entropy": 5.867666101455688,
      "epoch": 1.4509007417873543,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047933913527750015,
      "loss": 5.3971,
      "mean_token_accuracy": 0.19658233374357223,
      "num_tokens": 30612907.0,
      "step": 16430
    },
    {
      "entropy": 5.781331539154053,
      "epoch": 1.4513422818791946,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004793259511518801,
      "loss": 5.3337,
      "mean_token_accuracy": 0.21251970529556274,
      "num_tokens": 30621537.0,
      "step": 16435
    },
    {
      "entropy": 5.820788431167602,
      "epoch": 1.451783821971035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004793127630236309,
      "loss": 5.4593,
      "mean_token_accuracy": 0.19122121036052703,
      "num_tokens": 30632143.0,
      "step": 16440
    },
    {
      "entropy": 5.90724024772644,
      "epoch": 1.4522253620628753,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047929957089301085,
      "loss": 5.4049,
      "mean_token_accuracy": 0.19678455889225005,
      "num_tokens": 30641145.0,
      "step": 16445
    },
    {
      "entropy": 5.860531234741211,
      "epoch": 1.4526669021547156,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047928637476027836,
      "loss": 5.415,
      "mean_token_accuracy": 0.20139241963624954,
      "num_tokens": 30650036.0,
      "step": 16450
    },
    {
      "entropy": 5.873429489135742,
      "epoch": 1.453108442246556,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047927317462569206,
      "loss": 5.3963,
      "mean_token_accuracy": 0.21078886091709137,
      "num_tokens": 30659552.0,
      "step": 16455
    },
    {
      "entropy": 5.8764019966125485,
      "epoch": 1.4535499823383964,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004792599704895103,
      "loss": 5.3763,
      "mean_token_accuracy": 0.20104623436927796,
      "num_tokens": 30668652.0,
      "step": 16460
    },
    {
      "entropy": 5.844933605194091,
      "epoch": 1.4539915224302367,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000479246762351992,
      "loss": 5.4471,
      "mean_token_accuracy": 0.19569310247898103,
      "num_tokens": 30678613.0,
      "step": 16465
    },
    {
      "entropy": 5.840695190429687,
      "epoch": 1.454433062522077,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047923355021339576,
      "loss": 5.4022,
      "mean_token_accuracy": 0.20952788591384888,
      "num_tokens": 30688259.0,
      "step": 16470
    },
    {
      "entropy": 5.840805149078369,
      "epoch": 1.4548746026139172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004792203340739805,
      "loss": 5.4196,
      "mean_token_accuracy": 0.19956467747688295,
      "num_tokens": 30697309.0,
      "step": 16475
    },
    {
      "entropy": 5.902931213378906,
      "epoch": 1.4553161427057577,
      "grad_norm": 1.0,
      "learning_rate": 0.0004792071139340049,
      "loss": 5.4687,
      "mean_token_accuracy": 0.19744110852479935,
      "num_tokens": 30707278.0,
      "step": 16480
    },
    {
      "entropy": 5.849676942825317,
      "epoch": 1.455757682797598,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004791938897937281,
      "loss": 5.3574,
      "mean_token_accuracy": 0.20679961442947387,
      "num_tokens": 30716840.0,
      "step": 16485
    },
    {
      "entropy": 5.816055059432983,
      "epoch": 1.4561992228894383,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047918066165340926,
      "loss": 5.449,
      "mean_token_accuracy": 0.19277482032775878,
      "num_tokens": 30727120.0,
      "step": 16490
    },
    {
      "entropy": 5.786322355270386,
      "epoch": 1.4566407629812788,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004791674295133072,
      "loss": 5.344,
      "mean_token_accuracy": 0.20551401525735855,
      "num_tokens": 30735543.0,
      "step": 16495
    },
    {
      "entropy": 5.892681026458741,
      "epoch": 1.457082303073119,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047915419337368146,
      "loss": 5.4388,
      "mean_token_accuracy": 0.19324137419462203,
      "num_tokens": 30744971.0,
      "step": 16500
    },
    {
      "entropy": 5.79856390953064,
      "epoch": 1.4575238431649593,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047914095323479115,
      "loss": 5.2939,
      "mean_token_accuracy": 0.2104108974337578,
      "num_tokens": 30753444.0,
      "step": 16505
    },
    {
      "entropy": 5.828524255752564,
      "epoch": 1.4579653832567998,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004791277090968956,
      "loss": 5.3969,
      "mean_token_accuracy": 0.20633172988891602,
      "num_tokens": 30761469.0,
      "step": 16510
    },
    {
      "entropy": 5.801257085800171,
      "epoch": 1.45840692334864,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004791144609602544,
      "loss": 5.333,
      "mean_token_accuracy": 0.20391583293676377,
      "num_tokens": 30770773.0,
      "step": 16515
    },
    {
      "entropy": 5.832178211212158,
      "epoch": 1.4588484634404804,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047910120882512696,
      "loss": 5.4515,
      "mean_token_accuracy": 0.19713431894779204,
      "num_tokens": 30779624.0,
      "step": 16520
    },
    {
      "entropy": 5.818113851547241,
      "epoch": 1.4592900035323209,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004790879526917729,
      "loss": 5.348,
      "mean_token_accuracy": 0.19848013520240784,
      "num_tokens": 30788671.0,
      "step": 16525
    },
    {
      "entropy": 5.8809247493743895,
      "epoch": 1.4597315436241611,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004790746925604519,
      "loss": 5.4282,
      "mean_token_accuracy": 0.20235230326652526,
      "num_tokens": 30797470.0,
      "step": 16530
    },
    {
      "entropy": 5.871769428253174,
      "epoch": 1.4601730837160014,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047906142843142383,
      "loss": 5.4105,
      "mean_token_accuracy": 0.20621410459280015,
      "num_tokens": 30807209.0,
      "step": 16535
    },
    {
      "entropy": 5.897931671142578,
      "epoch": 1.4606146238078417,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047904816030494834,
      "loss": 5.4344,
      "mean_token_accuracy": 0.1966389685869217,
      "num_tokens": 30816156.0,
      "step": 16540
    },
    {
      "entropy": 5.78818564414978,
      "epoch": 1.461056163899682,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004790348881812855,
      "loss": 5.3509,
      "mean_token_accuracy": 0.21043713688850402,
      "num_tokens": 30825439.0,
      "step": 16545
    },
    {
      "entropy": 5.749357604980469,
      "epoch": 1.4614977039915225,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047902161206069514,
      "loss": 5.3694,
      "mean_token_accuracy": 0.20392918437719346,
      "num_tokens": 30834953.0,
      "step": 16550
    },
    {
      "entropy": 5.8226083278656,
      "epoch": 1.4619392440833627,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047900833194343764,
      "loss": 5.4254,
      "mean_token_accuracy": 0.1996046185493469,
      "num_tokens": 30843534.0,
      "step": 16555
    },
    {
      "entropy": 5.79872236251831,
      "epoch": 1.462380784175203,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004789950478297727,
      "loss": 5.3252,
      "mean_token_accuracy": 0.21067256182432176,
      "num_tokens": 30851869.0,
      "step": 16560
    },
    {
      "entropy": 5.835751295089722,
      "epoch": 1.4628223242670435,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047898175971996095,
      "loss": 5.4571,
      "mean_token_accuracy": 0.19869925528764726,
      "num_tokens": 30862262.0,
      "step": 16565
    },
    {
      "entropy": 5.849185752868652,
      "epoch": 1.4632638643588838,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047896846761426255,
      "loss": 5.4567,
      "mean_token_accuracy": 0.19498667269945144,
      "num_tokens": 30872223.0,
      "step": 16570
    },
    {
      "entropy": 5.840133619308472,
      "epoch": 1.463705404450724,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004789551715129378,
      "loss": 5.3781,
      "mean_token_accuracy": 0.19859335124492644,
      "num_tokens": 30881834.0,
      "step": 16575
    },
    {
      "entropy": 5.875336408615112,
      "epoch": 1.4641469445425646,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004789418714162472,
      "loss": 5.4652,
      "mean_token_accuracy": 0.19814420938491822,
      "num_tokens": 30893414.0,
      "step": 16580
    },
    {
      "entropy": 5.881783103942871,
      "epoch": 1.4645884846344048,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004789285673244514,
      "loss": 5.4899,
      "mean_token_accuracy": 0.19113173335790634,
      "num_tokens": 30902345.0,
      "step": 16585
    },
    {
      "entropy": 5.699627304077149,
      "epoch": 1.465030024726245,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004789152592378109,
      "loss": 5.2066,
      "mean_token_accuracy": 0.22669571340084077,
      "num_tokens": 30911619.0,
      "step": 16590
    },
    {
      "entropy": 5.832030534744263,
      "epoch": 1.4654715648180856,
      "grad_norm": 0.9375,
      "learning_rate": 0.00047890194715658645,
      "loss": 5.4191,
      "mean_token_accuracy": 0.1936587542295456,
      "num_tokens": 30921515.0,
      "step": 16595
    },
    {
      "entropy": 5.884548616409302,
      "epoch": 1.4659131049099259,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004788886310810389,
      "loss": 5.3857,
      "mean_token_accuracy": 0.20652967393398286,
      "num_tokens": 30930807.0,
      "step": 16600
    },
    {
      "entropy": 5.923854541778565,
      "epoch": 1.4663546450017662,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047887531101142893,
      "loss": 5.4646,
      "mean_token_accuracy": 0.19024052768945693,
      "num_tokens": 30939266.0,
      "step": 16605
    },
    {
      "entropy": 5.85533881187439,
      "epoch": 1.4667961850936064,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047886198694801764,
      "loss": 5.363,
      "mean_token_accuracy": 0.19712022840976715,
      "num_tokens": 30948119.0,
      "step": 16610
    },
    {
      "entropy": 5.741694116592408,
      "epoch": 1.4672377251854467,
      "grad_norm": 1.0,
      "learning_rate": 0.00047884865889106593,
      "loss": 5.2957,
      "mean_token_accuracy": 0.21649323254823685,
      "num_tokens": 30957005.0,
      "step": 16615
    },
    {
      "entropy": 5.720598888397217,
      "epoch": 1.4676792652772872,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047883532684083493,
      "loss": 5.3568,
      "mean_token_accuracy": 0.2143753722310066,
      "num_tokens": 30967144.0,
      "step": 16620
    },
    {
      "entropy": 5.863882112503052,
      "epoch": 1.4681208053691275,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004788219907975858,
      "loss": 5.3128,
      "mean_token_accuracy": 0.2041093736886978,
      "num_tokens": 30975315.0,
      "step": 16625
    },
    {
      "entropy": 5.909586763381958,
      "epoch": 1.4685623454609678,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004788086507615798,
      "loss": 5.4839,
      "mean_token_accuracy": 0.19469848722219468,
      "num_tokens": 30984705.0,
      "step": 16630
    },
    {
      "entropy": 5.834614658355713,
      "epoch": 1.4690038855528083,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004787953067330782,
      "loss": 5.4708,
      "mean_token_accuracy": 0.19439245462417604,
      "num_tokens": 30994936.0,
      "step": 16635
    },
    {
      "entropy": 5.760412549972534,
      "epoch": 1.4694454256446485,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004787819587123425,
      "loss": 5.3862,
      "mean_token_accuracy": 0.19505118429660798,
      "num_tokens": 31004011.0,
      "step": 16640
    },
    {
      "entropy": 5.916839361190796,
      "epoch": 1.4698869657364888,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004787686066996341,
      "loss": 5.4285,
      "mean_token_accuracy": 0.19813136905431747,
      "num_tokens": 31013852.0,
      "step": 16645
    },
    {
      "entropy": 5.831754207611084,
      "epoch": 1.4703285058283293,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047875525069521457,
      "loss": 5.4023,
      "mean_token_accuracy": 0.20186276137828826,
      "num_tokens": 31023456.0,
      "step": 16650
    },
    {
      "entropy": 5.887347459793091,
      "epoch": 1.4707700459201696,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004787418906993456,
      "loss": 5.4488,
      "mean_token_accuracy": 0.1886138692498207,
      "num_tokens": 31032828.0,
      "step": 16655
    },
    {
      "entropy": 5.810969734191895,
      "epoch": 1.4712115860120099,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047872852671228884,
      "loss": 5.3751,
      "mean_token_accuracy": 0.21169993728399278,
      "num_tokens": 31043017.0,
      "step": 16660
    },
    {
      "entropy": 5.74207067489624,
      "epoch": 1.4716531261038504,
      "grad_norm": 0.921875,
      "learning_rate": 0.00047871515873430614,
      "loss": 5.2896,
      "mean_token_accuracy": 0.21016721278429032,
      "num_tokens": 31051491.0,
      "step": 16665
    },
    {
      "entropy": 5.761693859100342,
      "epoch": 1.4720946661956906,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004787017867656593,
      "loss": 5.2881,
      "mean_token_accuracy": 0.2228364810347557,
      "num_tokens": 31060375.0,
      "step": 16670
    },
    {
      "entropy": 5.9136261463165285,
      "epoch": 1.472536206287531,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004786884108066103,
      "loss": 5.5058,
      "mean_token_accuracy": 0.18943930566310882,
      "num_tokens": 31071410.0,
      "step": 16675
    },
    {
      "entropy": 5.833731985092163,
      "epoch": 1.4729777463793712,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004786750308574212,
      "loss": 5.2514,
      "mean_token_accuracy": 0.22078309804201127,
      "num_tokens": 31080123.0,
      "step": 16680
    },
    {
      "entropy": 5.823306083679199,
      "epoch": 1.4734192864712115,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004786616469183541,
      "loss": 5.4124,
      "mean_token_accuracy": 0.20207696706056594,
      "num_tokens": 31089870.0,
      "step": 16685
    },
    {
      "entropy": 5.803851175308227,
      "epoch": 1.473860826563052,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004786482589896711,
      "loss": 5.4401,
      "mean_token_accuracy": 0.19819240123033524,
      "num_tokens": 31098682.0,
      "step": 16690
    },
    {
      "entropy": 5.885953378677368,
      "epoch": 1.4743023666548922,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004786348670716346,
      "loss": 5.4537,
      "mean_token_accuracy": 0.1934850126504898,
      "num_tokens": 31107785.0,
      "step": 16695
    },
    {
      "entropy": 5.81049337387085,
      "epoch": 1.4747439067467325,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004786214711645068,
      "loss": 5.2774,
      "mean_token_accuracy": 0.21126044541597366,
      "num_tokens": 31116853.0,
      "step": 16700
    },
    {
      "entropy": 5.820914459228516,
      "epoch": 1.475185446838573,
      "grad_norm": 1.0,
      "learning_rate": 0.0004786080712685503,
      "loss": 5.3776,
      "mean_token_accuracy": 0.20030050873756408,
      "num_tokens": 31126260.0,
      "step": 16705
    },
    {
      "entropy": 5.85260272026062,
      "epoch": 1.4756269869304133,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047859466738402743,
      "loss": 5.4482,
      "mean_token_accuracy": 0.19973967522382735,
      "num_tokens": 31135047.0,
      "step": 16710
    },
    {
      "entropy": 5.844044208526611,
      "epoch": 1.4760685270222536,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047858125951120073,
      "loss": 5.376,
      "mean_token_accuracy": 0.2031434953212738,
      "num_tokens": 31144177.0,
      "step": 16715
    },
    {
      "entropy": 5.85214319229126,
      "epoch": 1.476510067114094,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000478567847650333,
      "loss": 5.464,
      "mean_token_accuracy": 0.19586363881826402,
      "num_tokens": 31153586.0,
      "step": 16720
    },
    {
      "entropy": 5.929026174545288,
      "epoch": 1.4769516072059343,
      "grad_norm": 1.046875,
      "learning_rate": 0.000478554431801687,
      "loss": 5.5425,
      "mean_token_accuracy": 0.19936016350984573,
      "num_tokens": 31163252.0,
      "step": 16725
    },
    {
      "entropy": 5.7872968196868895,
      "epoch": 1.4773931472977746,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004785410119655253,
      "loss": 5.3365,
      "mean_token_accuracy": 0.20308424681425094,
      "num_tokens": 31171947.0,
      "step": 16730
    },
    {
      "entropy": 5.809985160827637,
      "epoch": 1.477834687389615,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047852758814211104,
      "loss": 5.3693,
      "mean_token_accuracy": 0.20936822295188903,
      "num_tokens": 31180502.0,
      "step": 16735
    },
    {
      "entropy": 5.8745352268219,
      "epoch": 1.4782762274814554,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047851416033170704,
      "loss": 5.4681,
      "mean_token_accuracy": 0.19506845772266387,
      "num_tokens": 31190744.0,
      "step": 16740
    },
    {
      "entropy": 5.9168616771698,
      "epoch": 1.4787177675732956,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004785007285345764,
      "loss": 5.4904,
      "mean_token_accuracy": 0.19723411202430724,
      "num_tokens": 31200869.0,
      "step": 16745
    },
    {
      "entropy": 5.8352379322052,
      "epoch": 1.479159307665136,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004784872927509822,
      "loss": 5.3846,
      "mean_token_accuracy": 0.20750546008348464,
      "num_tokens": 31210046.0,
      "step": 16750
    },
    {
      "entropy": 5.818807458877563,
      "epoch": 1.4796008477569762,
      "grad_norm": 1.125,
      "learning_rate": 0.0004784738529811877,
      "loss": 5.3184,
      "mean_token_accuracy": 0.20589302629232406,
      "num_tokens": 31219140.0,
      "step": 16755
    },
    {
      "entropy": 5.808164453506469,
      "epoch": 1.4800423878488167,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047846040922545604,
      "loss": 5.3339,
      "mean_token_accuracy": 0.20964421331882477,
      "num_tokens": 31227993.0,
      "step": 16760
    },
    {
      "entropy": 5.819470500946045,
      "epoch": 1.480483927940657,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047844696148405074,
      "loss": 5.3803,
      "mean_token_accuracy": 0.20366198271512986,
      "num_tokens": 31236959.0,
      "step": 16765
    },
    {
      "entropy": 5.843044471740723,
      "epoch": 1.4809254680324972,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047843350975723523,
      "loss": 5.291,
      "mean_token_accuracy": 0.21459991186857225,
      "num_tokens": 31246373.0,
      "step": 16770
    },
    {
      "entropy": 5.809376668930054,
      "epoch": 1.4813670081243377,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004784200540452729,
      "loss": 5.4085,
      "mean_token_accuracy": 0.20238426178693772,
      "num_tokens": 31255556.0,
      "step": 16775
    },
    {
      "entropy": 5.785092353820801,
      "epoch": 1.481808548216178,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004784065943484274,
      "loss": 5.3466,
      "mean_token_accuracy": 0.20924828052520753,
      "num_tokens": 31264824.0,
      "step": 16780
    },
    {
      "entropy": 5.776709032058716,
      "epoch": 1.4822500883080183,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047839313066696246,
      "loss": 5.3098,
      "mean_token_accuracy": 0.2110050603747368,
      "num_tokens": 31273946.0,
      "step": 16785
    },
    {
      "entropy": 5.8087320804595945,
      "epoch": 1.4826916283998588,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047837966300114164,
      "loss": 5.3888,
      "mean_token_accuracy": 0.1979956343770027,
      "num_tokens": 31283492.0,
      "step": 16790
    },
    {
      "entropy": 5.894581937789917,
      "epoch": 1.483133168491699,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047836619135122906,
      "loss": 5.4219,
      "mean_token_accuracy": 0.20325595438480376,
      "num_tokens": 31292250.0,
      "step": 16795
    },
    {
      "entropy": 5.889112043380737,
      "epoch": 1.4835747085835393,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047835271571748834,
      "loss": 5.5386,
      "mean_token_accuracy": 0.18938992619514466,
      "num_tokens": 31302868.0,
      "step": 16800
    },
    {
      "entropy": 5.839573812484741,
      "epoch": 1.4840162486753798,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004783392361001836,
      "loss": 5.3218,
      "mean_token_accuracy": 0.20803227424621581,
      "num_tokens": 31311918.0,
      "step": 16805
    },
    {
      "entropy": 5.7997101783752445,
      "epoch": 1.4844577887672201,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004783257524995788,
      "loss": 5.3219,
      "mean_token_accuracy": 0.20955585688352585,
      "num_tokens": 31321339.0,
      "step": 16810
    },
    {
      "entropy": 5.81311264038086,
      "epoch": 1.4848993288590604,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004783122649159383,
      "loss": 5.3268,
      "mean_token_accuracy": 0.20057355910539626,
      "num_tokens": 31330498.0,
      "step": 16815
    },
    {
      "entropy": 5.801687622070313,
      "epoch": 1.4853408689509007,
      "grad_norm": 1.0625,
      "learning_rate": 0.000478298773349526,
      "loss": 5.3472,
      "mean_token_accuracy": 0.2093326896429062,
      "num_tokens": 31339676.0,
      "step": 16820
    },
    {
      "entropy": 5.912374877929688,
      "epoch": 1.485782409042741,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004782852778006064,
      "loss": 5.4837,
      "mean_token_accuracy": 0.19155890196561814,
      "num_tokens": 31350072.0,
      "step": 16825
    },
    {
      "entropy": 5.8931622982025145,
      "epoch": 1.4862239491345814,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047827177826944396,
      "loss": 5.3733,
      "mean_token_accuracy": 0.20686384290456772,
      "num_tokens": 31359992.0,
      "step": 16830
    },
    {
      "entropy": 5.859904432296753,
      "epoch": 1.4866654892264217,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047825827475630285,
      "loss": 5.3776,
      "mean_token_accuracy": 0.2051248475909233,
      "num_tokens": 31369434.0,
      "step": 16835
    },
    {
      "entropy": 5.804741477966308,
      "epoch": 1.487107029318262,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047824476726144775,
      "loss": 5.3757,
      "mean_token_accuracy": 0.20290971994400026,
      "num_tokens": 31378773.0,
      "step": 16840
    },
    {
      "entropy": 5.8703711986541744,
      "epoch": 1.4875485694101025,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047823125578514325,
      "loss": 5.4194,
      "mean_token_accuracy": 0.2034597411751747,
      "num_tokens": 31388219.0,
      "step": 16845
    },
    {
      "entropy": 5.788036298751831,
      "epoch": 1.4879901095019428,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047821774032765407,
      "loss": 5.3301,
      "mean_token_accuracy": 0.20831093341112136,
      "num_tokens": 31397216.0,
      "step": 16850
    },
    {
      "entropy": 5.775175380706787,
      "epoch": 1.488431649593783,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004782042208892449,
      "loss": 5.3165,
      "mean_token_accuracy": 0.21075984090566635,
      "num_tokens": 31406861.0,
      "step": 16855
    },
    {
      "entropy": 5.803642892837525,
      "epoch": 1.4888731896856235,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047819069747018065,
      "loss": 5.3967,
      "mean_token_accuracy": 0.20211237967014312,
      "num_tokens": 31416387.0,
      "step": 16860
    },
    {
      "entropy": 5.804005718231201,
      "epoch": 1.4893147297774638,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004781771700707262,
      "loss": 5.299,
      "mean_token_accuracy": 0.20730532258749007,
      "num_tokens": 31425056.0,
      "step": 16865
    },
    {
      "entropy": 5.893209075927734,
      "epoch": 1.489756269869304,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004781636386911465,
      "loss": 5.5122,
      "mean_token_accuracy": 0.19564283192157744,
      "num_tokens": 31435746.0,
      "step": 16870
    },
    {
      "entropy": 5.914508438110351,
      "epoch": 1.4901978099611446,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004781501033317067,
      "loss": 5.5373,
      "mean_token_accuracy": 0.1911771312355995,
      "num_tokens": 31444808.0,
      "step": 16875
    },
    {
      "entropy": 5.8947224617004395,
      "epoch": 1.4906393500529849,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004781365639926718,
      "loss": 5.4274,
      "mean_token_accuracy": 0.20052673369646073,
      "num_tokens": 31454205.0,
      "step": 16880
    },
    {
      "entropy": 5.874072456359864,
      "epoch": 1.4910808901448251,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004781230206743073,
      "loss": 5.4409,
      "mean_token_accuracy": 0.19775310009717942,
      "num_tokens": 31463446.0,
      "step": 16885
    },
    {
      "entropy": 5.805663919448852,
      "epoch": 1.4915224302366654,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004781094733768783,
      "loss": 5.3927,
      "mean_token_accuracy": 0.20490449219942092,
      "num_tokens": 31473168.0,
      "step": 16890
    },
    {
      "entropy": 5.82881269454956,
      "epoch": 1.4919639703285057,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047809592210065017,
      "loss": 5.3025,
      "mean_token_accuracy": 0.20378587692975997,
      "num_tokens": 31482899.0,
      "step": 16895
    },
    {
      "entropy": 5.861395931243896,
      "epoch": 1.4924055104203462,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047808236684588847,
      "loss": 5.2961,
      "mean_token_accuracy": 0.20091051012277603,
      "num_tokens": 31492503.0,
      "step": 16900
    },
    {
      "entropy": 5.72953085899353,
      "epoch": 1.4928470505121865,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004780688076128587,
      "loss": 5.2802,
      "mean_token_accuracy": 0.21831628531217576,
      "num_tokens": 31501520.0,
      "step": 16905
    },
    {
      "entropy": 5.8094563484191895,
      "epoch": 1.4932885906040267,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047805524440182654,
      "loss": 5.433,
      "mean_token_accuracy": 0.20691922903060914,
      "num_tokens": 31510821.0,
      "step": 16910
    },
    {
      "entropy": 5.851106786727906,
      "epoch": 1.4937301306958672,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047804167721305756,
      "loss": 5.3937,
      "mean_token_accuracy": 0.20294977724552155,
      "num_tokens": 31520472.0,
      "step": 16915
    },
    {
      "entropy": 5.836772918701172,
      "epoch": 1.4941716707877075,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047802810604681757,
      "loss": 5.4105,
      "mean_token_accuracy": 0.20710429102182387,
      "num_tokens": 31529129.0,
      "step": 16920
    },
    {
      "entropy": 5.7587299823760985,
      "epoch": 1.4946132108795478,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047801453090337256,
      "loss": 5.3587,
      "mean_token_accuracy": 0.20782049894332885,
      "num_tokens": 31537897.0,
      "step": 16925
    },
    {
      "entropy": 5.860333633422852,
      "epoch": 1.4950547509713883,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004780009517829882,
      "loss": 5.3624,
      "mean_token_accuracy": 0.2069771721959114,
      "num_tokens": 31547656.0,
      "step": 16930
    },
    {
      "entropy": 5.825451326370239,
      "epoch": 1.4954962910632286,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004779873686859308,
      "loss": 5.3736,
      "mean_token_accuracy": 0.20781843662261962,
      "num_tokens": 31557303.0,
      "step": 16935
    },
    {
      "entropy": 5.906974411010742,
      "epoch": 1.4959378311550688,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047797378161246613,
      "loss": 5.3926,
      "mean_token_accuracy": 0.1985720783472061,
      "num_tokens": 31565749.0,
      "step": 16940
    },
    {
      "entropy": 5.854325914382935,
      "epoch": 1.4963793712469093,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047796019056286067,
      "loss": 5.4327,
      "mean_token_accuracy": 0.20241083800792695,
      "num_tokens": 31575286.0,
      "step": 16945
    },
    {
      "entropy": 5.887064981460571,
      "epoch": 1.4968209113387496,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004779465955373804,
      "loss": 5.375,
      "mean_token_accuracy": 0.20670202523469924,
      "num_tokens": 31583723.0,
      "step": 16950
    },
    {
      "entropy": 5.742811679840088,
      "epoch": 1.49726245143059,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004779329965362918,
      "loss": 5.2181,
      "mean_token_accuracy": 0.22568588852882385,
      "num_tokens": 31593171.0,
      "step": 16955
    },
    {
      "entropy": 5.792586898803711,
      "epoch": 1.4977039915224302,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047791939355986123,
      "loss": 5.4014,
      "mean_token_accuracy": 0.19927307814359665,
      "num_tokens": 31603958.0,
      "step": 16960
    },
    {
      "entropy": 5.823423433303833,
      "epoch": 1.4981455316142704,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004779057866083551,
      "loss": 5.3901,
      "mean_token_accuracy": 0.19900340884923934,
      "num_tokens": 31612460.0,
      "step": 16965
    },
    {
      "entropy": 5.917697954177856,
      "epoch": 1.498587071706111,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047789217568204004,
      "loss": 5.446,
      "mean_token_accuracy": 0.19984558671712876,
      "num_tokens": 31622114.0,
      "step": 16970
    },
    {
      "entropy": 5.877597141265869,
      "epoch": 1.4990286117979512,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047787856078118266,
      "loss": 5.3875,
      "mean_token_accuracy": 0.20543920993804932,
      "num_tokens": 31632272.0,
      "step": 16975
    },
    {
      "entropy": 5.819973516464233,
      "epoch": 1.4994701518897915,
      "grad_norm": 1.0,
      "learning_rate": 0.0004778649419060496,
      "loss": 5.352,
      "mean_token_accuracy": 0.20504435002803803,
      "num_tokens": 31642235.0,
      "step": 16980
    },
    {
      "entropy": 5.876267766952514,
      "epoch": 1.499911691981632,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004778513190569077,
      "loss": 5.4591,
      "mean_token_accuracy": 0.1948750615119934,
      "num_tokens": 31652773.0,
      "step": 16985
    },
    {
      "entropy": 5.9093646049499515,
      "epoch": 1.5003532320734723,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004778376922340239,
      "loss": 5.4106,
      "mean_token_accuracy": 0.2044292151927948,
      "num_tokens": 31662681.0,
      "step": 16990
    },
    {
      "entropy": 5.7939286708831785,
      "epoch": 1.5007947721653125,
      "grad_norm": 1.0625,
      "learning_rate": 0.000477824061437665,
      "loss": 5.3156,
      "mean_token_accuracy": 0.20705605000257493,
      "num_tokens": 31671576.0,
      "step": 16995
    },
    {
      "entropy": 5.883949708938599,
      "epoch": 1.501236312257153,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047781042666809816,
      "loss": 5.554,
      "mean_token_accuracy": 0.18697230219841005,
      "num_tokens": 31681149.0,
      "step": 17000
    },
    {
      "entropy": 5.8254352569580075,
      "epoch": 1.5016778523489933,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004777967879255904,
      "loss": 5.3622,
      "mean_token_accuracy": 0.20532894134521484,
      "num_tokens": 31690007.0,
      "step": 17005
    },
    {
      "entropy": 5.958875370025635,
      "epoch": 1.5021193924408336,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004777831452104088,
      "loss": 5.4937,
      "mean_token_accuracy": 0.19368649274110794,
      "num_tokens": 31700055.0,
      "step": 17010
    },
    {
      "entropy": 5.853254556655884,
      "epoch": 1.502560932532674,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004777694985228208,
      "loss": 5.3876,
      "mean_token_accuracy": 0.20470905303955078,
      "num_tokens": 31708255.0,
      "step": 17015
    },
    {
      "entropy": 5.821598052978516,
      "epoch": 1.5030024726245141,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047775584786309365,
      "loss": 5.4398,
      "mean_token_accuracy": 0.19863502830266952,
      "num_tokens": 31716855.0,
      "step": 17020
    },
    {
      "entropy": 5.816057014465332,
      "epoch": 1.5034440127163546,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047774219323149474,
      "loss": 5.3706,
      "mean_token_accuracy": 0.2109145477414131,
      "num_tokens": 31726153.0,
      "step": 17025
    },
    {
      "entropy": 5.812814950942993,
      "epoch": 1.5038855528081951,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004777285346282916,
      "loss": 5.3558,
      "mean_token_accuracy": 0.2040833353996277,
      "num_tokens": 31736935.0,
      "step": 17030
    },
    {
      "entropy": 5.881922674179077,
      "epoch": 1.5043270929000352,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004777148720537517,
      "loss": 5.4544,
      "mean_token_accuracy": 0.1992572620511055,
      "num_tokens": 31746402.0,
      "step": 17035
    },
    {
      "entropy": 5.831804084777832,
      "epoch": 1.5047686329918757,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047770120550814276,
      "loss": 5.2661,
      "mean_token_accuracy": 0.2085268184542656,
      "num_tokens": 31755615.0,
      "step": 17040
    },
    {
      "entropy": 5.788525199890136,
      "epoch": 1.505210173083716,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004776875349917325,
      "loss": 5.4364,
      "mean_token_accuracy": 0.2040776476264,
      "num_tokens": 31765701.0,
      "step": 17045
    },
    {
      "entropy": 5.850245380401612,
      "epoch": 1.5056517131755562,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004776738605047887,
      "loss": 5.3964,
      "mean_token_accuracy": 0.20048863142728807,
      "num_tokens": 31774674.0,
      "step": 17050
    },
    {
      "entropy": 5.85404896736145,
      "epoch": 1.5060932532673967,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004776601820475793,
      "loss": 5.3733,
      "mean_token_accuracy": 0.20120461881160737,
      "num_tokens": 31784871.0,
      "step": 17055
    },
    {
      "entropy": 5.8014312267303465,
      "epoch": 1.506534793359237,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004776464996203721,
      "loss": 5.3261,
      "mean_token_accuracy": 0.21462402045726775,
      "num_tokens": 31793274.0,
      "step": 17060
    },
    {
      "entropy": 5.838164901733398,
      "epoch": 1.5069763334510773,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004776328132234353,
      "loss": 5.5001,
      "mean_token_accuracy": 0.19429777264595033,
      "num_tokens": 31803543.0,
      "step": 17065
    },
    {
      "entropy": 5.971688938140869,
      "epoch": 1.5074178735429178,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004776191228570368,
      "loss": 5.4657,
      "mean_token_accuracy": 0.19177936017513275,
      "num_tokens": 31814529.0,
      "step": 17070
    },
    {
      "entropy": 5.94747486114502,
      "epoch": 1.507859413634758,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004776054285214451,
      "loss": 5.4404,
      "mean_token_accuracy": 0.20195518881082536,
      "num_tokens": 31824064.0,
      "step": 17075
    },
    {
      "entropy": 5.798696613311767,
      "epoch": 1.5083009537265983,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047759173021692813,
      "loss": 5.4117,
      "mean_token_accuracy": 0.1961733281612396,
      "num_tokens": 31833881.0,
      "step": 17080
    },
    {
      "entropy": 5.822088432312012,
      "epoch": 1.5087424938184388,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004775780279437545,
      "loss": 5.3575,
      "mean_token_accuracy": 0.21328300088644028,
      "num_tokens": 31842987.0,
      "step": 17085
    },
    {
      "entropy": 5.8510472774505615,
      "epoch": 1.5091840339102789,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047756432170219243,
      "loss": 5.3193,
      "mean_token_accuracy": 0.20450857579708098,
      "num_tokens": 31851806.0,
      "step": 17090
    },
    {
      "entropy": 5.8785346984863285,
      "epoch": 1.5096255740021194,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004775506114925106,
      "loss": 5.4773,
      "mean_token_accuracy": 0.19472805857658387,
      "num_tokens": 31861857.0,
      "step": 17095
    },
    {
      "entropy": 5.862979936599731,
      "epoch": 1.5100671140939599,
      "grad_norm": 1.0,
      "learning_rate": 0.00047753689731497734,
      "loss": 5.3781,
      "mean_token_accuracy": 0.1985655054450035,
      "num_tokens": 31871252.0,
      "step": 17100
    },
    {
      "entropy": 5.912623882293701,
      "epoch": 1.5105086541858,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004775231791698616,
      "loss": 5.4435,
      "mean_token_accuracy": 0.18940001875162124,
      "num_tokens": 31880734.0,
      "step": 17105
    },
    {
      "entropy": 5.855919408798218,
      "epoch": 1.5109501942776404,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047750945705743187,
      "loss": 5.3634,
      "mean_token_accuracy": 0.19766008257865905,
      "num_tokens": 31889304.0,
      "step": 17110
    },
    {
      "entropy": 5.807578039169312,
      "epoch": 1.5113917343694807,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004774957309779571,
      "loss": 5.4,
      "mean_token_accuracy": 0.19846741408109664,
      "num_tokens": 31898586.0,
      "step": 17115
    },
    {
      "entropy": 5.90633111000061,
      "epoch": 1.511833274461321,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004774820009317061,
      "loss": 5.4523,
      "mean_token_accuracy": 0.19630279690027236,
      "num_tokens": 31908153.0,
      "step": 17120
    },
    {
      "entropy": 5.8275305271148685,
      "epoch": 1.5122748145531615,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047746826691894787,
      "loss": 5.2505,
      "mean_token_accuracy": 0.21836305111646653,
      "num_tokens": 31916467.0,
      "step": 17125
    },
    {
      "entropy": 5.788586235046386,
      "epoch": 1.5127163546450018,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047745452893995145,
      "loss": 5.3643,
      "mean_token_accuracy": 0.19612461030483247,
      "num_tokens": 31925469.0,
      "step": 17130
    },
    {
      "entropy": 5.793550443649292,
      "epoch": 1.513157894736842,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047744078699498587,
      "loss": 5.4631,
      "mean_token_accuracy": 0.1943300947546959,
      "num_tokens": 31934747.0,
      "step": 17135
    },
    {
      "entropy": 5.818152809143067,
      "epoch": 1.5135994348286825,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047742704108432046,
      "loss": 5.3534,
      "mean_token_accuracy": 0.1993481770157814,
      "num_tokens": 31943773.0,
      "step": 17140
    },
    {
      "entropy": 5.8493987083435055,
      "epoch": 1.5140409749205228,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047741329120822446,
      "loss": 5.4224,
      "mean_token_accuracy": 0.20505952835083008,
      "num_tokens": 31952699.0,
      "step": 17145
    },
    {
      "entropy": 5.840079307556152,
      "epoch": 1.514482515012363,
      "grad_norm": 1.125,
      "learning_rate": 0.0004773995373669672,
      "loss": 5.3944,
      "mean_token_accuracy": 0.19767861813306808,
      "num_tokens": 31962845.0,
      "step": 17150
    },
    {
      "entropy": 5.8120654106140135,
      "epoch": 1.5149240551042036,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004773857795608181,
      "loss": 5.2894,
      "mean_token_accuracy": 0.21276037842035295,
      "num_tokens": 31972437.0,
      "step": 17155
    },
    {
      "entropy": 5.875831747055054,
      "epoch": 1.5153655951960436,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047737201779004667,
      "loss": 5.4996,
      "mean_token_accuracy": 0.1989943951368332,
      "num_tokens": 31981652.0,
      "step": 17160
    },
    {
      "entropy": 5.855569887161255,
      "epoch": 1.5158071352878841,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004773582520549226,
      "loss": 5.3873,
      "mean_token_accuracy": 0.20434178709983825,
      "num_tokens": 31990088.0,
      "step": 17165
    },
    {
      "entropy": 5.870453596115112,
      "epoch": 1.5162486753797246,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004773444823557153,
      "loss": 5.4056,
      "mean_token_accuracy": 0.20224764496088027,
      "num_tokens": 31999386.0,
      "step": 17170
    },
    {
      "entropy": 5.831242275238037,
      "epoch": 1.5166902154715647,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004773307086926948,
      "loss": 5.3313,
      "mean_token_accuracy": 0.20481727570295333,
      "num_tokens": 32007695.0,
      "step": 17175
    },
    {
      "entropy": 5.8826985359191895,
      "epoch": 1.5171317555634052,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047731693106613075,
      "loss": 5.4619,
      "mean_token_accuracy": 0.2010740280151367,
      "num_tokens": 32018300.0,
      "step": 17180
    },
    {
      "entropy": 5.893084335327148,
      "epoch": 1.5175732956552455,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004773031494762931,
      "loss": 5.448,
      "mean_token_accuracy": 0.1982845038175583,
      "num_tokens": 32028386.0,
      "step": 17185
    },
    {
      "entropy": 5.907029438018799,
      "epoch": 1.5180148357470857,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004772893639234518,
      "loss": 5.4529,
      "mean_token_accuracy": 0.1987977385520935,
      "num_tokens": 32037260.0,
      "step": 17190
    },
    {
      "entropy": 5.8555513381958,
      "epoch": 1.5184563758389262,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004772755744078769,
      "loss": 5.4685,
      "mean_token_accuracy": 0.19956057518720627,
      "num_tokens": 32047654.0,
      "step": 17195
    },
    {
      "entropy": 5.840859317779541,
      "epoch": 1.5188979159307665,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004772617809298385,
      "loss": 5.3707,
      "mean_token_accuracy": 0.20469409376382827,
      "num_tokens": 32056791.0,
      "step": 17200
    },
    {
      "entropy": 5.886715650558472,
      "epoch": 1.5193394560226068,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004772479834896069,
      "loss": 5.3741,
      "mean_token_accuracy": 0.20050286501646042,
      "num_tokens": 32066561.0,
      "step": 17205
    },
    {
      "entropy": 5.923398065567016,
      "epoch": 1.5197809961144473,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047723418208745226,
      "loss": 5.422,
      "mean_token_accuracy": 0.1947730541229248,
      "num_tokens": 32075801.0,
      "step": 17210
    },
    {
      "entropy": 5.81295075416565,
      "epoch": 1.5202225362062876,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004772203767236451,
      "loss": 5.3556,
      "mean_token_accuracy": 0.20861087292432784,
      "num_tokens": 32085274.0,
      "step": 17215
    },
    {
      "entropy": 5.9020044803619385,
      "epoch": 1.5206640762981278,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047720656739845573,
      "loss": 5.3902,
      "mean_token_accuracy": 0.2032547190785408,
      "num_tokens": 32094481.0,
      "step": 17220
    },
    {
      "entropy": 5.80965485572815,
      "epoch": 1.5211056163899683,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004771927541121547,
      "loss": 5.3669,
      "mean_token_accuracy": 0.2061656892299652,
      "num_tokens": 32103094.0,
      "step": 17225
    },
    {
      "entropy": 5.789358472824096,
      "epoch": 1.5215471564818084,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004771789368650126,
      "loss": 5.356,
      "mean_token_accuracy": 0.20408538430929185,
      "num_tokens": 32112230.0,
      "step": 17230
    },
    {
      "entropy": 5.845158672332763,
      "epoch": 1.5219886965736489,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047716511565730013,
      "loss": 5.3794,
      "mean_token_accuracy": 0.2095283031463623,
      "num_tokens": 32121472.0,
      "step": 17235
    },
    {
      "entropy": 5.850609827041626,
      "epoch": 1.5224302366654894,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000477151290489288,
      "loss": 5.4185,
      "mean_token_accuracy": 0.200775046646595,
      "num_tokens": 32130849.0,
      "step": 17240
    },
    {
      "entropy": 5.839935541152954,
      "epoch": 1.5228717767573294,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004771374613612471,
      "loss": 5.3251,
      "mean_token_accuracy": 0.20684054791927337,
      "num_tokens": 32140874.0,
      "step": 17245
    },
    {
      "entropy": 5.850006771087647,
      "epoch": 1.52331331684917,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004771236282734482,
      "loss": 5.3445,
      "mean_token_accuracy": 0.20047370046377183,
      "num_tokens": 32150774.0,
      "step": 17250
    },
    {
      "entropy": 5.805947971343994,
      "epoch": 1.5237548569410102,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047710979122616247,
      "loss": 5.3327,
      "mean_token_accuracy": 0.20334020704030992,
      "num_tokens": 32159825.0,
      "step": 17255
    },
    {
      "entropy": 5.812037086486816,
      "epoch": 1.5241963970328505,
      "grad_norm": 1.125,
      "learning_rate": 0.0004770959502196609,
      "loss": 5.3886,
      "mean_token_accuracy": 0.19752940088510512,
      "num_tokens": 32168443.0,
      "step": 17260
    },
    {
      "entropy": 5.835806941986084,
      "epoch": 1.524637937124691,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047708210525421445,
      "loss": 5.3436,
      "mean_token_accuracy": 0.20664990395307542,
      "num_tokens": 32178065.0,
      "step": 17265
    },
    {
      "entropy": 5.792875337600708,
      "epoch": 1.5250794772165313,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047706825633009466,
      "loss": 5.2512,
      "mean_token_accuracy": 0.211979740858078,
      "num_tokens": 32187274.0,
      "step": 17270
    },
    {
      "entropy": 5.822767353057861,
      "epoch": 1.5255210173083715,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047705440344757264,
      "loss": 5.2235,
      "mean_token_accuracy": 0.2181275174021721,
      "num_tokens": 32195811.0,
      "step": 17275
    },
    {
      "entropy": 5.754204750061035,
      "epoch": 1.525962557400212,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047704054660691964,
      "loss": 5.3172,
      "mean_token_accuracy": 0.20618013441562652,
      "num_tokens": 32204958.0,
      "step": 17280
    },
    {
      "entropy": 5.796719312667847,
      "epoch": 1.5264040974920523,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004770266858084074,
      "loss": 5.351,
      "mean_token_accuracy": 0.20292788743972778,
      "num_tokens": 32214299.0,
      "step": 17285
    },
    {
      "entropy": 5.878060579299927,
      "epoch": 1.5268456375838926,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047701282105230716,
      "loss": 5.4198,
      "mean_token_accuracy": 0.2011943593621254,
      "num_tokens": 32223353.0,
      "step": 17290
    },
    {
      "entropy": 5.852122783660889,
      "epoch": 1.527287177675733,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004769989523388906,
      "loss": 5.4203,
      "mean_token_accuracy": 0.18991311341524125,
      "num_tokens": 32232641.0,
      "step": 17295
    },
    {
      "entropy": 5.8938398361206055,
      "epoch": 1.5277287177675731,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004769850796684296,
      "loss": 5.4508,
      "mean_token_accuracy": 0.199055977165699,
      "num_tokens": 32243028.0,
      "step": 17300
    },
    {
      "entropy": 5.875230169296264,
      "epoch": 1.5281702578594136,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004769712030411957,
      "loss": 5.3962,
      "mean_token_accuracy": 0.20871222019195557,
      "num_tokens": 32253660.0,
      "step": 17305
    },
    {
      "entropy": 5.822922468185425,
      "epoch": 1.5286117979512541,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047695732245746076,
      "loss": 5.3859,
      "mean_token_accuracy": 0.2028472825884819,
      "num_tokens": 32263267.0,
      "step": 17310
    },
    {
      "entropy": 5.848821353912354,
      "epoch": 1.5290533380430942,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004769434379174967,
      "loss": 5.3469,
      "mean_token_accuracy": 0.20837787985801698,
      "num_tokens": 32273929.0,
      "step": 17315
    },
    {
      "entropy": 5.809392261505127,
      "epoch": 1.5294948781349347,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004769295494215757,
      "loss": 5.366,
      "mean_token_accuracy": 0.20906019657850267,
      "num_tokens": 32282717.0,
      "step": 17320
    },
    {
      "entropy": 5.8649170875549315,
      "epoch": 1.529936418226775,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047691565696996954,
      "loss": 5.3914,
      "mean_token_accuracy": 0.20684265494346618,
      "num_tokens": 32293225.0,
      "step": 17325
    },
    {
      "entropy": 5.890305137634277,
      "epoch": 1.5303779583186152,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004769017605629505,
      "loss": 5.3779,
      "mean_token_accuracy": 0.20241788625717164,
      "num_tokens": 32301582.0,
      "step": 17330
    },
    {
      "entropy": 5.8151891231536865,
      "epoch": 1.5308194984104557,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004768878602007909,
      "loss": 5.3187,
      "mean_token_accuracy": 0.20678892135620117,
      "num_tokens": 32310131.0,
      "step": 17335
    },
    {
      "entropy": 5.791430807113647,
      "epoch": 1.531261038502296,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004768739558837628,
      "loss": 5.3816,
      "mean_token_accuracy": 0.20074218064546584,
      "num_tokens": 32320038.0,
      "step": 17340
    },
    {
      "entropy": 5.834036922454834,
      "epoch": 1.5317025785941363,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004768600476121388,
      "loss": 5.4501,
      "mean_token_accuracy": 0.19252036958932878,
      "num_tokens": 32329568.0,
      "step": 17345
    },
    {
      "entropy": 5.9029418468475345,
      "epoch": 1.5321441186859768,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047684613538619125,
      "loss": 5.3819,
      "mean_token_accuracy": 0.2056812733411789,
      "num_tokens": 32338940.0,
      "step": 17350
    },
    {
      "entropy": 5.857225322723389,
      "epoch": 1.532585658777817,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004768322192061927,
      "loss": 5.3345,
      "mean_token_accuracy": 0.20741656720638274,
      "num_tokens": 32347757.0,
      "step": 17355
    },
    {
      "entropy": 5.781398344039917,
      "epoch": 1.5330271988696573,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004768182990724158,
      "loss": 5.2749,
      "mean_token_accuracy": 0.2075476586818695,
      "num_tokens": 32357099.0,
      "step": 17360
    },
    {
      "entropy": 5.823176717758178,
      "epoch": 1.5334687389614978,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047680437498513317,
      "loss": 5.3983,
      "mean_token_accuracy": 0.1995120480656624,
      "num_tokens": 32366548.0,
      "step": 17365
    },
    {
      "entropy": 5.9196630954742435,
      "epoch": 1.5339102790533379,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047679044694461765,
      "loss": 5.5132,
      "mean_token_accuracy": 0.19590681344270705,
      "num_tokens": 32377435.0,
      "step": 17370
    },
    {
      "entropy": 5.92530927658081,
      "epoch": 1.5343518191451784,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000476776514951142,
      "loss": 5.406,
      "mean_token_accuracy": 0.19295999258756638,
      "num_tokens": 32386922.0,
      "step": 17375
    },
    {
      "entropy": 5.840206813812256,
      "epoch": 1.5347933592370189,
      "grad_norm": 1.0,
      "learning_rate": 0.00047676257900497915,
      "loss": 5.4434,
      "mean_token_accuracy": 0.19346087723970412,
      "num_tokens": 32396350.0,
      "step": 17380
    },
    {
      "entropy": 5.754315328598023,
      "epoch": 1.535234899328859,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047674863910640217,
      "loss": 5.2431,
      "mean_token_accuracy": 0.2165575474500656,
      "num_tokens": 32405678.0,
      "step": 17385
    },
    {
      "entropy": 5.845928001403808,
      "epoch": 1.5356764394206994,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047673469525568406,
      "loss": 5.3411,
      "mean_token_accuracy": 0.20299955904483796,
      "num_tokens": 32415793.0,
      "step": 17390
    },
    {
      "entropy": 5.806760215759278,
      "epoch": 1.5361179795125397,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047672074745309804,
      "loss": 5.2636,
      "mean_token_accuracy": 0.2122264862060547,
      "num_tokens": 32424735.0,
      "step": 17395
    },
    {
      "entropy": 5.790989017486572,
      "epoch": 1.53655951960438,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004767067956989172,
      "loss": 5.405,
      "mean_token_accuracy": 0.20501989275217056,
      "num_tokens": 32435213.0,
      "step": 17400
    },
    {
      "entropy": 5.788079929351807,
      "epoch": 1.5370010596962205,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004766928399934151,
      "loss": 5.2675,
      "mean_token_accuracy": 0.20899792164564132,
      "num_tokens": 32443932.0,
      "step": 17405
    },
    {
      "entropy": 5.792341327667236,
      "epoch": 1.5374425997880607,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004766788803368649,
      "loss": 5.2905,
      "mean_token_accuracy": 0.20316837579011918,
      "num_tokens": 32453851.0,
      "step": 17410
    },
    {
      "entropy": 5.819828987121582,
      "epoch": 1.537884139879901,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004766649167295401,
      "loss": 5.3785,
      "mean_token_accuracy": 0.20670698285102845,
      "num_tokens": 32462906.0,
      "step": 17415
    },
    {
      "entropy": 5.788566398620605,
      "epoch": 1.5383256799717415,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004766509491717144,
      "loss": 5.3314,
      "mean_token_accuracy": 0.2036026418209076,
      "num_tokens": 32471840.0,
      "step": 17420
    },
    {
      "entropy": 5.799122381210327,
      "epoch": 1.5387672200635818,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047663697766366114,
      "loss": 5.2938,
      "mean_token_accuracy": 0.20713522881269456,
      "num_tokens": 32481111.0,
      "step": 17425
    },
    {
      "entropy": 5.855897712707519,
      "epoch": 1.539208760155422,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004766230022056542,
      "loss": 5.4008,
      "mean_token_accuracy": 0.19948508590459824,
      "num_tokens": 32489827.0,
      "step": 17430
    },
    {
      "entropy": 5.82600245475769,
      "epoch": 1.5396503002472626,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047660902279796736,
      "loss": 5.3174,
      "mean_token_accuracy": 0.21005891263484955,
      "num_tokens": 32499106.0,
      "step": 17435
    },
    {
      "entropy": 5.971836709976197,
      "epoch": 1.5400918403391026,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004765950394408744,
      "loss": 5.5817,
      "mean_token_accuracy": 0.18005635142326354,
      "num_tokens": 32507858.0,
      "step": 17440
    },
    {
      "entropy": 5.840009021759033,
      "epoch": 1.5405333804309431,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004765810521346493,
      "loss": 5.3555,
      "mean_token_accuracy": 0.21056333929300308,
      "num_tokens": 32516761.0,
      "step": 17445
    },
    {
      "entropy": 5.819193315505982,
      "epoch": 1.5409749205227836,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047656706087956604,
      "loss": 5.3778,
      "mean_token_accuracy": 0.20764204114675522,
      "num_tokens": 32526585.0,
      "step": 17450
    },
    {
      "entropy": 5.870976734161377,
      "epoch": 1.5414164606146237,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047655306567589864,
      "loss": 5.3791,
      "mean_token_accuracy": 0.20405740588903426,
      "num_tokens": 32535607.0,
      "step": 17455
    },
    {
      "entropy": 5.774896669387817,
      "epoch": 1.5418580007064642,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047653906652392135,
      "loss": 5.2853,
      "mean_token_accuracy": 0.21170148849487305,
      "num_tokens": 32545401.0,
      "step": 17460
    },
    {
      "entropy": 5.836571121215821,
      "epoch": 1.5422995407983044,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004765250634239083,
      "loss": 5.3645,
      "mean_token_accuracy": 0.20433194935321808,
      "num_tokens": 32554421.0,
      "step": 17465
    },
    {
      "entropy": 5.869767808914185,
      "epoch": 1.5427410808901447,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047651105637613395,
      "loss": 5.4561,
      "mean_token_accuracy": 0.19872623682022095,
      "num_tokens": 32563652.0,
      "step": 17470
    },
    {
      "entropy": 5.901002073287964,
      "epoch": 1.5431826209819852,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004764970453808726,
      "loss": 5.4422,
      "mean_token_accuracy": 0.20603567808866502,
      "num_tokens": 32572812.0,
      "step": 17475
    },
    {
      "entropy": 5.7957408905029295,
      "epoch": 1.5436241610738255,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047648303043839875,
      "loss": 5.3615,
      "mean_token_accuracy": 0.20554557889699937,
      "num_tokens": 32581746.0,
      "step": 17480
    },
    {
      "entropy": 5.906637954711914,
      "epoch": 1.5440657011656658,
      "grad_norm": 1.0,
      "learning_rate": 0.0004764690115489869,
      "loss": 5.3766,
      "mean_token_accuracy": 0.2001782089471817,
      "num_tokens": 32591596.0,
      "step": 17485
    },
    {
      "entropy": 5.860601902008057,
      "epoch": 1.5445072412575063,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047645498871291163,
      "loss": 5.3757,
      "mean_token_accuracy": 0.203918893635273,
      "num_tokens": 32600610.0,
      "step": 17490
    },
    {
      "entropy": 5.841221666336059,
      "epoch": 1.5449487813493465,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004764409619304478,
      "loss": 5.3928,
      "mean_token_accuracy": 0.20123014003038406,
      "num_tokens": 32609827.0,
      "step": 17495
    },
    {
      "entropy": 5.781210422515869,
      "epoch": 1.5453903214411868,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047642693120187007,
      "loss": 5.4205,
      "mean_token_accuracy": 0.2035561054944992,
      "num_tokens": 32618258.0,
      "step": 17500
    },
    {
      "entropy": 5.886530780792237,
      "epoch": 1.5458318615330273,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004764128965274533,
      "loss": 5.3906,
      "mean_token_accuracy": 0.19994398057460785,
      "num_tokens": 32628155.0,
      "step": 17505
    },
    {
      "entropy": 5.939802551269532,
      "epoch": 1.5462734016248674,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047639885790747244,
      "loss": 5.4503,
      "mean_token_accuracy": 0.2002129688858986,
      "num_tokens": 32638347.0,
      "step": 17510
    },
    {
      "entropy": 5.838560485839844,
      "epoch": 1.5467149417167079,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047638481534220244,
      "loss": 5.3168,
      "mean_token_accuracy": 0.20974133610725404,
      "num_tokens": 32648636.0,
      "step": 17515
    },
    {
      "entropy": 5.791447496414184,
      "epoch": 1.5471564818085484,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004763707688319185,
      "loss": 5.3922,
      "mean_token_accuracy": 0.20038572549819947,
      "num_tokens": 32657762.0,
      "step": 17520
    },
    {
      "entropy": 5.830028676986695,
      "epoch": 1.5475980219003884,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004763567183768958,
      "loss": 5.3519,
      "mean_token_accuracy": 0.2025613531470299,
      "num_tokens": 32668265.0,
      "step": 17525
    },
    {
      "entropy": 5.824889755249023,
      "epoch": 1.548039561992229,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047634266397740934,
      "loss": 5.2836,
      "mean_token_accuracy": 0.20865013748407363,
      "num_tokens": 32678051.0,
      "step": 17530
    },
    {
      "entropy": 5.940329456329346,
      "epoch": 1.5484811020840692,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047632860563373466,
      "loss": 5.4788,
      "mean_token_accuracy": 0.20264360904693604,
      "num_tokens": 32687186.0,
      "step": 17535
    },
    {
      "entropy": 5.863598728179932,
      "epoch": 1.5489226421759095,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004763145433461472,
      "loss": 5.3882,
      "mean_token_accuracy": 0.21008908301591872,
      "num_tokens": 32697082.0,
      "step": 17540
    },
    {
      "entropy": 5.857601547241211,
      "epoch": 1.54936418226775,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004763004771149222,
      "loss": 5.3782,
      "mean_token_accuracy": 0.1941447600722313,
      "num_tokens": 32706502.0,
      "step": 17545
    },
    {
      "entropy": 5.860434532165527,
      "epoch": 1.5498057223595902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004762864069403354,
      "loss": 5.3662,
      "mean_token_accuracy": 0.20193805396556855,
      "num_tokens": 32715035.0,
      "step": 17550
    },
    {
      "entropy": 5.836504745483398,
      "epoch": 1.5502472624514305,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047627233282266236,
      "loss": 5.3839,
      "mean_token_accuracy": 0.20119669586420058,
      "num_tokens": 32724713.0,
      "step": 17555
    },
    {
      "entropy": 5.850826597213745,
      "epoch": 1.550688802543271,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004762582547621788,
      "loss": 5.3756,
      "mean_token_accuracy": 0.20523784756660463,
      "num_tokens": 32734584.0,
      "step": 17560
    },
    {
      "entropy": 5.87843599319458,
      "epoch": 1.5511303426351113,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047624417275916045,
      "loss": 5.3417,
      "mean_token_accuracy": 0.2070808857679367,
      "num_tokens": 32742744.0,
      "step": 17565
    },
    {
      "entropy": 5.776598167419434,
      "epoch": 1.5515718827269516,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047623008681388324,
      "loss": 5.2969,
      "mean_token_accuracy": 0.21753205358982086,
      "num_tokens": 32752051.0,
      "step": 17570
    },
    {
      "entropy": 5.817727661132812,
      "epoch": 1.552013422818792,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047621599692662306,
      "loss": 5.3734,
      "mean_token_accuracy": 0.2033172905445099,
      "num_tokens": 32761740.0,
      "step": 17575
    },
    {
      "entropy": 5.770384740829468,
      "epoch": 1.5524549629106321,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047620190309765595,
      "loss": 5.3229,
      "mean_token_accuracy": 0.2133523106575012,
      "num_tokens": 32770809.0,
      "step": 17580
    },
    {
      "entropy": 5.879725694656372,
      "epoch": 1.5528965030024726,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047618780532725805,
      "loss": 5.3917,
      "mean_token_accuracy": 0.20065007656812667,
      "num_tokens": 32780614.0,
      "step": 17585
    },
    {
      "entropy": 5.905575323104858,
      "epoch": 1.5533380430943131,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047617370361570534,
      "loss": 5.3893,
      "mean_token_accuracy": 0.19946929812431335,
      "num_tokens": 32790336.0,
      "step": 17590
    },
    {
      "entropy": 5.77630763053894,
      "epoch": 1.5537795831861532,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004761595979632743,
      "loss": 5.3561,
      "mean_token_accuracy": 0.20518322438001632,
      "num_tokens": 32798991.0,
      "step": 17595
    },
    {
      "entropy": 5.906645965576172,
      "epoch": 1.5542211232779937,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004761454883702411,
      "loss": 5.4053,
      "mean_token_accuracy": 0.20123518705368043,
      "num_tokens": 32807616.0,
      "step": 17600
    },
    {
      "entropy": 5.8880298137664795,
      "epoch": 1.554662663369834,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004761313748368821,
      "loss": 5.4445,
      "mean_token_accuracy": 0.19556560218334199,
      "num_tokens": 32817107.0,
      "step": 17605
    },
    {
      "entropy": 5.789643144607544,
      "epoch": 1.5551042034616742,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004761172573634739,
      "loss": 5.3752,
      "mean_token_accuracy": 0.20864083021879196,
      "num_tokens": 32827205.0,
      "step": 17610
    },
    {
      "entropy": 5.868758249282837,
      "epoch": 1.5555457435535147,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004761031359502931,
      "loss": 5.3161,
      "mean_token_accuracy": 0.21065206825733185,
      "num_tokens": 32836102.0,
      "step": 17615
    },
    {
      "entropy": 5.892943382263184,
      "epoch": 1.555987283645355,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047608901059761617,
      "loss": 5.4211,
      "mean_token_accuracy": 0.19518836587667465,
      "num_tokens": 32845389.0,
      "step": 17620
    },
    {
      "entropy": 5.777494430541992,
      "epoch": 1.5564288237371953,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047607488130571997,
      "loss": 5.3258,
      "mean_token_accuracy": 0.20757322162389755,
      "num_tokens": 32854186.0,
      "step": 17625
    },
    {
      "entropy": 5.78282356262207,
      "epoch": 1.5568703638290358,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004760607480748811,
      "loss": 5.3271,
      "mean_token_accuracy": 0.20399378687143327,
      "num_tokens": 32863498.0,
      "step": 17630
    },
    {
      "entropy": 5.853580284118652,
      "epoch": 1.557311903920876,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047604661090537657,
      "loss": 5.4329,
      "mean_token_accuracy": 0.19530525505542756,
      "num_tokens": 32872483.0,
      "step": 17635
    },
    {
      "entropy": 5.897101497650146,
      "epoch": 1.5577534440127163,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004760324697974832,
      "loss": 5.4635,
      "mean_token_accuracy": 0.19322464615106583,
      "num_tokens": 32882739.0,
      "step": 17640
    },
    {
      "entropy": 5.912864351272583,
      "epoch": 1.5581949841045568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004760183247514782,
      "loss": 5.513,
      "mean_token_accuracy": 0.193892902135849,
      "num_tokens": 32892707.0,
      "step": 17645
    },
    {
      "entropy": 5.890587854385376,
      "epoch": 1.5586365241963969,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047600417576763843,
      "loss": 5.4084,
      "mean_token_accuracy": 0.20191147923469543,
      "num_tokens": 32901916.0,
      "step": 17650
    },
    {
      "entropy": 5.823947334289551,
      "epoch": 1.5590780642882374,
      "grad_norm": 1.0,
      "learning_rate": 0.0004759900228462412,
      "loss": 5.3467,
      "mean_token_accuracy": 0.2064092352986336,
      "num_tokens": 32910656.0,
      "step": 17655
    },
    {
      "entropy": 5.923448896408081,
      "epoch": 1.5595196043800779,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004759758659875638,
      "loss": 5.4651,
      "mean_token_accuracy": 0.20218765288591384,
      "num_tokens": 32919579.0,
      "step": 17660
    },
    {
      "entropy": 5.895467233657837,
      "epoch": 1.559961144471918,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047596170519188335,
      "loss": 5.4146,
      "mean_token_accuracy": 0.19703866839408873,
      "num_tokens": 32928271.0,
      "step": 17665
    },
    {
      "entropy": 5.855054187774658,
      "epoch": 1.5604026845637584,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004759475404594776,
      "loss": 5.3159,
      "mean_token_accuracy": 0.2074134573340416,
      "num_tokens": 32937671.0,
      "step": 17670
    },
    {
      "entropy": 5.869637680053711,
      "epoch": 1.5608442246555987,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004759333717906236,
      "loss": 5.4886,
      "mean_token_accuracy": 0.19632164835929872,
      "num_tokens": 32946896.0,
      "step": 17675
    },
    {
      "entropy": 5.911066055297852,
      "epoch": 1.561285764747439,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047591919918559927,
      "loss": 5.4229,
      "mean_token_accuracy": 0.2013778120279312,
      "num_tokens": 32956706.0,
      "step": 17680
    },
    {
      "entropy": 5.8082053661346436,
      "epoch": 1.5617273048392795,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004759050226446821,
      "loss": 5.2845,
      "mean_token_accuracy": 0.2147594213485718,
      "num_tokens": 32965990.0,
      "step": 17685
    },
    {
      "entropy": 5.7986860275268555,
      "epoch": 1.5621688449311197,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004758908421681497,
      "loss": 5.2748,
      "mean_token_accuracy": 0.21321889013051987,
      "num_tokens": 32975311.0,
      "step": 17690
    },
    {
      "entropy": 5.865177249908447,
      "epoch": 1.56261038502296,
      "grad_norm": 1.0,
      "learning_rate": 0.00047587665775627997,
      "loss": 5.374,
      "mean_token_accuracy": 0.2026787981390953,
      "num_tokens": 32984641.0,
      "step": 17695
    },
    {
      "entropy": 5.866060256958008,
      "epoch": 1.5630519251148005,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00047586246940935087,
      "loss": 5.4386,
      "mean_token_accuracy": 0.1965424045920372,
      "num_tokens": 32994696.0,
      "step": 17700
    },
    {
      "entropy": 5.878091669082641,
      "epoch": 1.5634934652066408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047584827712764013,
      "loss": 5.33,
      "mean_token_accuracy": 0.20413600504398347,
      "num_tokens": 33003392.0,
      "step": 17705
    },
    {
      "entropy": 5.839124488830566,
      "epoch": 1.563935005298481,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004758340809114259,
      "loss": 5.3866,
      "mean_token_accuracy": 0.19434957653284074,
      "num_tokens": 33012896.0,
      "step": 17710
    },
    {
      "entropy": 5.8867676734924315,
      "epoch": 1.5643765453903216,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004758198807609862,
      "loss": 5.4388,
      "mean_token_accuracy": 0.19910742938518525,
      "num_tokens": 33022710.0,
      "step": 17715
    },
    {
      "entropy": 5.917848587036133,
      "epoch": 1.5648180854821618,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004758056766765993,
      "loss": 5.364,
      "mean_token_accuracy": 0.20309918969869614,
      "num_tokens": 33031982.0,
      "step": 17720
    },
    {
      "entropy": 5.893701696395874,
      "epoch": 1.5652596255740021,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047579146865854333,
      "loss": 5.4731,
      "mean_token_accuracy": 0.19704148918390274,
      "num_tokens": 33041496.0,
      "step": 17725
    },
    {
      "entropy": 5.842383813858032,
      "epoch": 1.5657011656658426,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047577725670709675,
      "loss": 5.3805,
      "mean_token_accuracy": 0.20649636387825013,
      "num_tokens": 33050764.0,
      "step": 17730
    },
    {
      "entropy": 5.847108793258667,
      "epoch": 1.5661427057576827,
      "grad_norm": 0.96875,
      "learning_rate": 0.000475763040822538,
      "loss": 5.4108,
      "mean_token_accuracy": 0.20037575215101242,
      "num_tokens": 33060809.0,
      "step": 17735
    },
    {
      "entropy": 5.770617151260376,
      "epoch": 1.5665842458495232,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004757488210051453,
      "loss": 5.2755,
      "mean_token_accuracy": 0.20983420461416244,
      "num_tokens": 33070155.0,
      "step": 17740
    },
    {
      "entropy": 5.8552123546600345,
      "epoch": 1.5670257859413634,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004757345972551974,
      "loss": 5.4023,
      "mean_token_accuracy": 0.20433754473924637,
      "num_tokens": 33079354.0,
      "step": 17745
    },
    {
      "entropy": 5.899097728729248,
      "epoch": 1.5674673260332037,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004757203695729729,
      "loss": 5.331,
      "mean_token_accuracy": 0.2078401491045952,
      "num_tokens": 33087877.0,
      "step": 17750
    },
    {
      "entropy": 5.837959575653076,
      "epoch": 1.5679088661250442,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047570613795875055,
      "loss": 5.4258,
      "mean_token_accuracy": 0.19665904343128204,
      "num_tokens": 33097375.0,
      "step": 17755
    },
    {
      "entropy": 5.874375104904175,
      "epoch": 1.5683504062168845,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004756919024128091,
      "loss": 5.3893,
      "mean_token_accuracy": 0.20355249643325807,
      "num_tokens": 33106336.0,
      "step": 17760
    },
    {
      "entropy": 5.779546785354614,
      "epoch": 1.5687919463087248,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004756776629354274,
      "loss": 5.2528,
      "mean_token_accuracy": 0.21987343728542327,
      "num_tokens": 33115753.0,
      "step": 17765
    },
    {
      "entropy": 5.876402950286865,
      "epoch": 1.5692334864005653,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004756634195268844,
      "loss": 5.4202,
      "mean_token_accuracy": 0.19458318799734114,
      "num_tokens": 33125205.0,
      "step": 17770
    },
    {
      "entropy": 5.792306280136108,
      "epoch": 1.5696750264924055,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004756491721874592,
      "loss": 5.3319,
      "mean_token_accuracy": 0.21207024306058883,
      "num_tokens": 33134470.0,
      "step": 17775
    },
    {
      "entropy": 5.799787712097168,
      "epoch": 1.5701165665842458,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047563492091743075,
      "loss": 5.3122,
      "mean_token_accuracy": 0.215229594707489,
      "num_tokens": 33143346.0,
      "step": 17780
    },
    {
      "entropy": 5.792463970184326,
      "epoch": 1.5705581066760863,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047562066571707836,
      "loss": 5.3202,
      "mean_token_accuracy": 0.20465569496154784,
      "num_tokens": 33153291.0,
      "step": 17785
    },
    {
      "entropy": 5.948005247116089,
      "epoch": 1.5709996467679266,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047560640658668126,
      "loss": 5.5063,
      "mean_token_accuracy": 0.19285668730735778,
      "num_tokens": 33163642.0,
      "step": 17790
    },
    {
      "entropy": 5.909404325485229,
      "epoch": 1.5714411868597669,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047559214352651863,
      "loss": 5.3406,
      "mean_token_accuracy": 0.20744375586509706,
      "num_tokens": 33173529.0,
      "step": 17795
    },
    {
      "entropy": 5.874653053283692,
      "epoch": 1.5718827269516074,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047557787653687014,
      "loss": 5.3306,
      "mean_token_accuracy": 0.20597842633724212,
      "num_tokens": 33180877.0,
      "step": 17800
    },
    {
      "entropy": 5.81123719215393,
      "epoch": 1.5723242670434474,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000475563605618015,
      "loss": 5.4059,
      "mean_token_accuracy": 0.20450567007064818,
      "num_tokens": 33190831.0,
      "step": 17805
    },
    {
      "entropy": 5.8893884181976315,
      "epoch": 1.572765807135288,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000475549330770233,
      "loss": 5.4117,
      "mean_token_accuracy": 0.19570378214120865,
      "num_tokens": 33200313.0,
      "step": 17810
    },
    {
      "entropy": 5.896144580841065,
      "epoch": 1.5732073472271282,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004755350519938036,
      "loss": 5.3941,
      "mean_token_accuracy": 0.20024604946374894,
      "num_tokens": 33209272.0,
      "step": 17815
    },
    {
      "entropy": 5.822745084762573,
      "epoch": 1.5736488873189685,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047552076928900663,
      "loss": 5.3276,
      "mean_token_accuracy": 0.21161558032035827,
      "num_tokens": 33218712.0,
      "step": 17820
    },
    {
      "entropy": 5.8529187679290775,
      "epoch": 1.574090427410809,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004755064826561217,
      "loss": 5.4518,
      "mean_token_accuracy": 0.20207060724496842,
      "num_tokens": 33227480.0,
      "step": 17825
    },
    {
      "entropy": 5.821786403656006,
      "epoch": 1.5745319675026492,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004754921920954289,
      "loss": 5.3239,
      "mean_token_accuracy": 0.20574539750814438,
      "num_tokens": 33236044.0,
      "step": 17830
    },
    {
      "entropy": 5.816183853149414,
      "epoch": 1.5749735075944895,
      "grad_norm": 1.125,
      "learning_rate": 0.0004754778976072081,
      "loss": 5.3981,
      "mean_token_accuracy": 0.20449062287807465,
      "num_tokens": 33244770.0,
      "step": 17835
    },
    {
      "entropy": 5.8485435962677,
      "epoch": 1.57541504768633,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047546359919173933,
      "loss": 5.305,
      "mean_token_accuracy": 0.20976659506559373,
      "num_tokens": 33254178.0,
      "step": 17840
    },
    {
      "entropy": 5.775556182861328,
      "epoch": 1.5758565877781703,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004754492968493026,
      "loss": 5.3698,
      "mean_token_accuracy": 0.2095904141664505,
      "num_tokens": 33264210.0,
      "step": 17845
    },
    {
      "entropy": 5.73330717086792,
      "epoch": 1.5762981278700106,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004754349905801782,
      "loss": 5.2925,
      "mean_token_accuracy": 0.21290307343006135,
      "num_tokens": 33273348.0,
      "step": 17850
    },
    {
      "entropy": 5.810392284393311,
      "epoch": 1.576739667961851,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047542068038464625,
      "loss": 5.3244,
      "mean_token_accuracy": 0.2157539740204811,
      "num_tokens": 33281906.0,
      "step": 17855
    },
    {
      "entropy": 5.941177701950073,
      "epoch": 1.5771812080536913,
      "grad_norm": 1.125,
      "learning_rate": 0.0004754063662629872,
      "loss": 5.4449,
      "mean_token_accuracy": 0.19736992567777634,
      "num_tokens": 33291363.0,
      "step": 17860
    },
    {
      "entropy": 5.865746259689331,
      "epoch": 1.5776227481455316,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047539204821548146,
      "loss": 5.3827,
      "mean_token_accuracy": 0.20610157400369644,
      "num_tokens": 33299785.0,
      "step": 17865
    },
    {
      "entropy": 5.907922983169556,
      "epoch": 1.578064288237372,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004753777262424094,
      "loss": 5.4795,
      "mean_token_accuracy": 0.19846486300230026,
      "num_tokens": 33309655.0,
      "step": 17870
    },
    {
      "entropy": 5.830729198455811,
      "epoch": 1.5785058283292122,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047536340034405166,
      "loss": 5.4502,
      "mean_token_accuracy": 0.2007785901427269,
      "num_tokens": 33320158.0,
      "step": 17875
    },
    {
      "entropy": 5.783361387252808,
      "epoch": 1.5789473684210527,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047534907052068886,
      "loss": 5.3196,
      "mean_token_accuracy": 0.2107610061764717,
      "num_tokens": 33328959.0,
      "step": 17880
    },
    {
      "entropy": 5.907349157333374,
      "epoch": 1.579388908512893,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004753347367726017,
      "loss": 5.4308,
      "mean_token_accuracy": 0.2007957726716995,
      "num_tokens": 33338624.0,
      "step": 17885
    },
    {
      "entropy": 5.863383054733276,
      "epoch": 1.5798304486047332,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047532039910007097,
      "loss": 5.3984,
      "mean_token_accuracy": 0.20311308205127715,
      "num_tokens": 33348083.0,
      "step": 17890
    },
    {
      "entropy": 5.839151430130005,
      "epoch": 1.5802719886965737,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047530605750337765,
      "loss": 5.328,
      "mean_token_accuracy": 0.2022930294275284,
      "num_tokens": 33358226.0,
      "step": 17895
    },
    {
      "entropy": 5.888965082168579,
      "epoch": 1.580713528788414,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047529171198280244,
      "loss": 5.4721,
      "mean_token_accuracy": 0.19955771863460542,
      "num_tokens": 33367997.0,
      "step": 17900
    },
    {
      "entropy": 5.886206912994385,
      "epoch": 1.5811550688802543,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047527736253862664,
      "loss": 5.4247,
      "mean_token_accuracy": 0.19584109187126159,
      "num_tokens": 33377574.0,
      "step": 17905
    },
    {
      "entropy": 5.845288419723511,
      "epoch": 1.5815966089720948,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047526300917113116,
      "loss": 5.4399,
      "mean_token_accuracy": 0.20210800617933272,
      "num_tokens": 33387016.0,
      "step": 17910
    },
    {
      "entropy": 5.871938276290893,
      "epoch": 1.582038149063935,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004752486518805972,
      "loss": 5.3612,
      "mean_token_accuracy": 0.20363385379314422,
      "num_tokens": 33396239.0,
      "step": 17915
    },
    {
      "entropy": 5.851468086242676,
      "epoch": 1.5824796891557753,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047523429066730603,
      "loss": 5.2824,
      "mean_token_accuracy": 0.21232890486717224,
      "num_tokens": 33404738.0,
      "step": 17920
    },
    {
      "entropy": 5.79391655921936,
      "epoch": 1.5829212292476158,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047521992553153903,
      "loss": 5.2687,
      "mean_token_accuracy": 0.21605887711048127,
      "num_tokens": 33413670.0,
      "step": 17925
    },
    {
      "entropy": 5.802855110168457,
      "epoch": 1.583362769339456,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004752055564735775,
      "loss": 5.3973,
      "mean_token_accuracy": 0.19847149699926375,
      "num_tokens": 33422841.0,
      "step": 17930
    },
    {
      "entropy": 5.92512845993042,
      "epoch": 1.5838043094312964,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047519118349370306,
      "loss": 5.4156,
      "mean_token_accuracy": 0.2049744799733162,
      "num_tokens": 33432023.0,
      "step": 17935
    },
    {
      "entropy": 5.865643548965454,
      "epoch": 1.5842458495231369,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004751768065921971,
      "loss": 5.3967,
      "mean_token_accuracy": 0.19575154781341553,
      "num_tokens": 33442181.0,
      "step": 17940
    },
    {
      "entropy": 5.928330183029175,
      "epoch": 1.584687389614977,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047516242576934144,
      "loss": 5.454,
      "mean_token_accuracy": 0.19018019288778304,
      "num_tokens": 33451847.0,
      "step": 17945
    },
    {
      "entropy": 5.887420177459717,
      "epoch": 1.5851289297068174,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047514804102541765,
      "loss": 5.4281,
      "mean_token_accuracy": 0.20378214567899705,
      "num_tokens": 33461587.0,
      "step": 17950
    },
    {
      "entropy": 5.919276428222656,
      "epoch": 1.5855704697986577,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047513365236070753,
      "loss": 5.4536,
      "mean_token_accuracy": 0.19712756872177123,
      "num_tokens": 33471553.0,
      "step": 17955
    },
    {
      "entropy": 5.7478841781616214,
      "epoch": 1.586012009890498,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047511925977549303,
      "loss": 5.2852,
      "mean_token_accuracy": 0.21106737554073335,
      "num_tokens": 33480984.0,
      "step": 17960
    },
    {
      "entropy": 5.8340246200561525,
      "epoch": 1.5864535499823385,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047510486327005604,
      "loss": 5.4365,
      "mean_token_accuracy": 0.2043757900595665,
      "num_tokens": 33491196.0,
      "step": 17965
    },
    {
      "entropy": 5.801984882354736,
      "epoch": 1.5868950900741787,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047509046284467855,
      "loss": 5.2733,
      "mean_token_accuracy": 0.20755961239337922,
      "num_tokens": 33500404.0,
      "step": 17970
    },
    {
      "entropy": 5.937586784362793,
      "epoch": 1.587336630166019,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047507605849964274,
      "loss": 5.4694,
      "mean_token_accuracy": 0.19570547789335252,
      "num_tokens": 33509281.0,
      "step": 17975
    },
    {
      "entropy": 5.924002933502197,
      "epoch": 1.5877781702578595,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004750616502352307,
      "loss": 5.3458,
      "mean_token_accuracy": 0.20356533229351043,
      "num_tokens": 33518556.0,
      "step": 17980
    },
    {
      "entropy": 5.91500301361084,
      "epoch": 1.5882197103496998,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047504723805172474,
      "loss": 5.5362,
      "mean_token_accuracy": 0.1891017362475395,
      "num_tokens": 33529266.0,
      "step": 17985
    },
    {
      "entropy": 5.849728584289551,
      "epoch": 1.58866125044154,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047503282194940715,
      "loss": 5.4006,
      "mean_token_accuracy": 0.20334022492170334,
      "num_tokens": 33538972.0,
      "step": 17990
    },
    {
      "entropy": 5.850444889068603,
      "epoch": 1.5891027905333805,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004750184019285603,
      "loss": 5.3433,
      "mean_token_accuracy": 0.2019957810640335,
      "num_tokens": 33549528.0,
      "step": 17995
    },
    {
      "entropy": 5.809695053100586,
      "epoch": 1.5895443306252208,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004750039779894668,
      "loss": 5.2668,
      "mean_token_accuracy": 0.20987723767757416,
      "num_tokens": 33557828.0,
      "step": 18000
    },
    {
      "epoch": 1.5895443306252208,
      "eval_entropy": 5.575396425510722,
      "eval_loss": 5.42067813873291,
      "eval_mean_token_accuracy": 0.21030604640277165,
      "eval_num_tokens": 33557828.0,
      "eval_runtime": 20.1472,
      "eval_samples_per_second": 1752.847,
      "eval_steps_per_second": 219.137,
      "step": 18000
    },
    {
      "entropy": 5.881388092041016,
      "epoch": 1.589985870717061,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047498955013240907,
      "loss": 5.4308,
      "mean_token_accuracy": 0.197257362306118,
      "num_tokens": 33567087.0,
      "step": 18005
    },
    {
      "entropy": 5.878854417800904,
      "epoch": 1.5904274108089016,
      "grad_norm": 1.0,
      "learning_rate": 0.00047497511835766984,
      "loss": 5.4012,
      "mean_token_accuracy": 0.19553140550851822,
      "num_tokens": 33577442.0,
      "step": 18010
    },
    {
      "entropy": 5.893789529800415,
      "epoch": 1.5908689509007417,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047496068266553165,
      "loss": 5.4126,
      "mean_token_accuracy": 0.2019266277551651,
      "num_tokens": 33585655.0,
      "step": 18015
    },
    {
      "entropy": 5.873789072036743,
      "epoch": 1.5913104909925821,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047494624305627754,
      "loss": 5.414,
      "mean_token_accuracy": 0.20742429941892623,
      "num_tokens": 33594808.0,
      "step": 18020
    },
    {
      "entropy": 5.933353424072266,
      "epoch": 1.5917520310844224,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004749317995301902,
      "loss": 5.4607,
      "mean_token_accuracy": 0.19954823702573776,
      "num_tokens": 33603924.0,
      "step": 18025
    },
    {
      "entropy": 5.84439697265625,
      "epoch": 1.5921935711762627,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004749173520875525,
      "loss": 5.3446,
      "mean_token_accuracy": 0.20497927963733673,
      "num_tokens": 33613870.0,
      "step": 18030
    },
    {
      "entropy": 5.817981767654419,
      "epoch": 1.5926351112681032,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047490290072864773,
      "loss": 5.382,
      "mean_token_accuracy": 0.20596250593662263,
      "num_tokens": 33622938.0,
      "step": 18035
    },
    {
      "entropy": 5.865955924987793,
      "epoch": 1.5930766513599435,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047488844545375886,
      "loss": 5.4436,
      "mean_token_accuracy": 0.20010052770376205,
      "num_tokens": 33631520.0,
      "step": 18040
    },
    {
      "entropy": 5.925880241394043,
      "epoch": 1.5935181914517837,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004748739862631689,
      "loss": 5.4307,
      "mean_token_accuracy": 0.20103561133146286,
      "num_tokens": 33641365.0,
      "step": 18045
    },
    {
      "entropy": 5.831374645233154,
      "epoch": 1.5939597315436242,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047485952315716126,
      "loss": 5.2374,
      "mean_token_accuracy": 0.2163265123963356,
      "num_tokens": 33650425.0,
      "step": 18050
    },
    {
      "entropy": 5.857035827636719,
      "epoch": 1.5944012716354645,
      "grad_norm": 1.125,
      "learning_rate": 0.00047484505613601917,
      "loss": 5.446,
      "mean_token_accuracy": 0.19701125025749205,
      "num_tokens": 33659928.0,
      "step": 18055
    },
    {
      "entropy": 5.838379144668579,
      "epoch": 1.5948428117273048,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004748305852000262,
      "loss": 5.4421,
      "mean_token_accuracy": 0.20323508083820344,
      "num_tokens": 33668860.0,
      "step": 18060
    },
    {
      "entropy": 5.901949787139893,
      "epoch": 1.5952843518191453,
      "grad_norm": 1.0,
      "learning_rate": 0.00047481611034946565,
      "loss": 5.3716,
      "mean_token_accuracy": 0.19717168658971787,
      "num_tokens": 33679343.0,
      "step": 18065
    },
    {
      "entropy": 5.805941200256347,
      "epoch": 1.5957258919109856,
      "grad_norm": 1.0,
      "learning_rate": 0.0004748016315846212,
      "loss": 5.3419,
      "mean_token_accuracy": 0.2030827969312668,
      "num_tokens": 33688749.0,
      "step": 18070
    },
    {
      "entropy": 5.857469367980957,
      "epoch": 1.5961674320028258,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047478714890577633,
      "loss": 5.4197,
      "mean_token_accuracy": 0.19789186418056487,
      "num_tokens": 33698359.0,
      "step": 18075
    },
    {
      "entropy": 5.925813961029053,
      "epoch": 1.5966089720946663,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047477266231321484,
      "loss": 5.3982,
      "mean_token_accuracy": 0.19987578094005584,
      "num_tokens": 33707402.0,
      "step": 18080
    },
    {
      "entropy": 5.869510984420776,
      "epoch": 1.5970505121865064,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047475817180722056,
      "loss": 5.3421,
      "mean_token_accuracy": 0.2070999413728714,
      "num_tokens": 33717043.0,
      "step": 18085
    },
    {
      "entropy": 5.750817203521729,
      "epoch": 1.597492052278347,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00047474367738807734,
      "loss": 5.3132,
      "mean_token_accuracy": 0.2143101140856743,
      "num_tokens": 33726527.0,
      "step": 18090
    },
    {
      "entropy": 5.792326211929321,
      "epoch": 1.5979335923701872,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047472917905606903,
      "loss": 5.3179,
      "mean_token_accuracy": 0.20932081192731858,
      "num_tokens": 33735357.0,
      "step": 18095
    },
    {
      "entropy": 5.7805774211883545,
      "epoch": 1.5983751324620274,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047471467681147974,
      "loss": 5.209,
      "mean_token_accuracy": 0.22694893032312394,
      "num_tokens": 33743660.0,
      "step": 18100
    },
    {
      "entropy": 5.848110008239746,
      "epoch": 1.598816672553868,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004747001706545935,
      "loss": 5.3991,
      "mean_token_accuracy": 0.201560440659523,
      "num_tokens": 33752899.0,
      "step": 18105
    },
    {
      "entropy": 5.867289304733276,
      "epoch": 1.5992582126457082,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004746856605856944,
      "loss": 5.3977,
      "mean_token_accuracy": 0.2011600747704506,
      "num_tokens": 33761333.0,
      "step": 18110
    },
    {
      "entropy": 5.876346731185913,
      "epoch": 1.5996997527375485,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047467114660506687,
      "loss": 5.4301,
      "mean_token_accuracy": 0.19527532607316972,
      "num_tokens": 33770507.0,
      "step": 18115
    },
    {
      "entropy": 5.765473651885986,
      "epoch": 1.600141292829389,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004746566287129951,
      "loss": 5.3256,
      "mean_token_accuracy": 0.20503395348787307,
      "num_tokens": 33781658.0,
      "step": 18120
    },
    {
      "entropy": 5.878257131576538,
      "epoch": 1.6005828329212293,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004746421069097636,
      "loss": 5.4533,
      "mean_token_accuracy": 0.1909184128046036,
      "num_tokens": 33792229.0,
      "step": 18125
    },
    {
      "entropy": 5.928519010543823,
      "epoch": 1.6010243730130695,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047462758119565663,
      "loss": 5.4091,
      "mean_token_accuracy": 0.20041222125291824,
      "num_tokens": 33801674.0,
      "step": 18130
    },
    {
      "entropy": 5.8398048877716064,
      "epoch": 1.60146591310491,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000474613051570959,
      "loss": 5.345,
      "mean_token_accuracy": 0.21145925521850586,
      "num_tokens": 33812066.0,
      "step": 18135
    },
    {
      "entropy": 5.8932593822479244,
      "epoch": 1.6019074531967503,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047459851803595524,
      "loss": 5.468,
      "mean_token_accuracy": 0.19464504420757295,
      "num_tokens": 33822206.0,
      "step": 18140
    },
    {
      "entropy": 5.913393831253051,
      "epoch": 1.6023489932885906,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047458398059092995,
      "loss": 5.454,
      "mean_token_accuracy": 0.19806140959262847,
      "num_tokens": 33832075.0,
      "step": 18145
    },
    {
      "entropy": 5.817515468597412,
      "epoch": 1.602790533380431,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047456943923616793,
      "loss": 5.2806,
      "mean_token_accuracy": 0.21167910546064378,
      "num_tokens": 33841523.0,
      "step": 18150
    },
    {
      "entropy": 5.870386791229248,
      "epoch": 1.6032320734722711,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047455489397195416,
      "loss": 5.3393,
      "mean_token_accuracy": 0.207174251973629,
      "num_tokens": 33850756.0,
      "step": 18155
    },
    {
      "entropy": 5.868167924880981,
      "epoch": 1.6036736135641116,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047454034479857356,
      "loss": 5.4247,
      "mean_token_accuracy": 0.20147025883197783,
      "num_tokens": 33859611.0,
      "step": 18160
    },
    {
      "entropy": 5.852504253387451,
      "epoch": 1.604115153655952,
      "grad_norm": 1.09375,
      "learning_rate": 0.000474525791716311,
      "loss": 5.4122,
      "mean_token_accuracy": 0.20078572481870652,
      "num_tokens": 33867993.0,
      "step": 18165
    },
    {
      "entropy": 5.923367261886597,
      "epoch": 1.6045566937477922,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047451123472545175,
      "loss": 5.4812,
      "mean_token_accuracy": 0.19497620314359665,
      "num_tokens": 33876785.0,
      "step": 18170
    },
    {
      "entropy": 5.877623701095581,
      "epoch": 1.6049982338396327,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047449667382628083,
      "loss": 5.3153,
      "mean_token_accuracy": 0.2080849826335907,
      "num_tokens": 33885729.0,
      "step": 18175
    },
    {
      "entropy": 5.8686470031738285,
      "epoch": 1.605439773931473,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004744821090190835,
      "loss": 5.3477,
      "mean_token_accuracy": 0.20913055837154387,
      "num_tokens": 33894622.0,
      "step": 18180
    },
    {
      "entropy": 5.867928886413575,
      "epoch": 1.6058813140233132,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047446754030414513,
      "loss": 5.4419,
      "mean_token_accuracy": 0.19912338107824326,
      "num_tokens": 33904187.0,
      "step": 18185
    },
    {
      "entropy": 5.8461769104003904,
      "epoch": 1.6063228541151537,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004744529676817511,
      "loss": 5.4355,
      "mean_token_accuracy": 0.2053565114736557,
      "num_tokens": 33914327.0,
      "step": 18190
    },
    {
      "entropy": 5.895674133300782,
      "epoch": 1.606764394206994,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047443839115218685,
      "loss": 5.2759,
      "mean_token_accuracy": 0.20811100602149962,
      "num_tokens": 33923424.0,
      "step": 18195
    },
    {
      "entropy": 5.852606964111328,
      "epoch": 1.6072059342988343,
      "grad_norm": 1.09375,
      "learning_rate": 0.000474423810715738,
      "loss": 5.321,
      "mean_token_accuracy": 0.20784832090139388,
      "num_tokens": 33932145.0,
      "step": 18200
    },
    {
      "entropy": 5.789508914947509,
      "epoch": 1.6076474743906748,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004744092263726901,
      "loss": 5.3339,
      "mean_token_accuracy": 0.2098607525229454,
      "num_tokens": 33942505.0,
      "step": 18205
    },
    {
      "entropy": 5.848215389251709,
      "epoch": 1.608089014482515,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004743946381233288,
      "loss": 5.3973,
      "mean_token_accuracy": 0.20695634037256241,
      "num_tokens": 33952156.0,
      "step": 18210
    },
    {
      "entropy": 5.91343207359314,
      "epoch": 1.6085305545743553,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047438004596794,
      "loss": 5.4195,
      "mean_token_accuracy": 0.19986581057310104,
      "num_tokens": 33962478.0,
      "step": 18215
    },
    {
      "entropy": 5.866097068786621,
      "epoch": 1.6089720946661958,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004743654499068095,
      "loss": 5.3509,
      "mean_token_accuracy": 0.19943293333053588,
      "num_tokens": 33971581.0,
      "step": 18220
    },
    {
      "entropy": 5.816497421264648,
      "epoch": 1.609413634758036,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004743508499402232,
      "loss": 5.3722,
      "mean_token_accuracy": 0.20656681656837464,
      "num_tokens": 33981262.0,
      "step": 18225
    },
    {
      "entropy": 5.899464416503906,
      "epoch": 1.6098551748498764,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004743362460684671,
      "loss": 5.4086,
      "mean_token_accuracy": 0.203243488073349,
      "num_tokens": 33989865.0,
      "step": 18230
    },
    {
      "entropy": 5.893782281875611,
      "epoch": 1.6102967149417167,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047432163829182737,
      "loss": 5.3748,
      "mean_token_accuracy": 0.20195204466581346,
      "num_tokens": 33999261.0,
      "step": 18235
    },
    {
      "entropy": 5.925965452194214,
      "epoch": 1.610738255033557,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047430702661059007,
      "loss": 5.385,
      "mean_token_accuracy": 0.19887196868658066,
      "num_tokens": 34008471.0,
      "step": 18240
    },
    {
      "entropy": 5.780384492874146,
      "epoch": 1.6111797951253974,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047429241102504144,
      "loss": 5.2668,
      "mean_token_accuracy": 0.21342891454696655,
      "num_tokens": 34018305.0,
      "step": 18245
    },
    {
      "entropy": 5.877861070632934,
      "epoch": 1.6116213352172377,
      "grad_norm": 1.0,
      "learning_rate": 0.00047427779153546787,
      "loss": 5.4222,
      "mean_token_accuracy": 0.18988059014081954,
      "num_tokens": 34027393.0,
      "step": 18250
    },
    {
      "entropy": 5.898454761505127,
      "epoch": 1.612062875309078,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047426316814215563,
      "loss": 5.3608,
      "mean_token_accuracy": 0.20616988688707352,
      "num_tokens": 34035835.0,
      "step": 18255
    },
    {
      "entropy": 5.811767196655273,
      "epoch": 1.6125044154009185,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047424854084539134,
      "loss": 5.4164,
      "mean_token_accuracy": 0.20323833525180818,
      "num_tokens": 34045946.0,
      "step": 18260
    },
    {
      "entropy": 5.925228691101074,
      "epoch": 1.6129459554927588,
      "grad_norm": 1.0,
      "learning_rate": 0.00047423390964546145,
      "loss": 5.4678,
      "mean_token_accuracy": 0.19535911977291107,
      "num_tokens": 34055356.0,
      "step": 18265
    },
    {
      "entropy": 5.977757596969605,
      "epoch": 1.613387495584599,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004742192745426525,
      "loss": 5.5347,
      "mean_token_accuracy": 0.18955558091402053,
      "num_tokens": 34065862.0,
      "step": 18270
    },
    {
      "entropy": 5.963193941116333,
      "epoch": 1.6138290356764395,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004742046355372513,
      "loss": 5.4159,
      "mean_token_accuracy": 0.20128547996282578,
      "num_tokens": 34074980.0,
      "step": 18275
    },
    {
      "entropy": 5.8671472549438475,
      "epoch": 1.6142705757682798,
      "grad_norm": 1.0,
      "learning_rate": 0.00047418999262954466,
      "loss": 5.3847,
      "mean_token_accuracy": 0.19881257563829421,
      "num_tokens": 34085501.0,
      "step": 18280
    },
    {
      "entropy": 5.845548009872436,
      "epoch": 1.61471211586012,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004741753458198192,
      "loss": 5.3587,
      "mean_token_accuracy": 0.20400122851133345,
      "num_tokens": 34095564.0,
      "step": 18285
    },
    {
      "entropy": 5.896277904510498,
      "epoch": 1.6151536559519606,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047416069510836215,
      "loss": 5.4195,
      "mean_token_accuracy": 0.19548761695623398,
      "num_tokens": 34104571.0,
      "step": 18290
    },
    {
      "entropy": 5.878162956237793,
      "epoch": 1.6155951960438006,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004741460404954602,
      "loss": 5.4421,
      "mean_token_accuracy": 0.19408143907785416,
      "num_tokens": 34114272.0,
      "step": 18295
    },
    {
      "entropy": 5.76869306564331,
      "epoch": 1.6160367361356411,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004741313819814007,
      "loss": 5.352,
      "mean_token_accuracy": 0.20639338493347167,
      "num_tokens": 34125521.0,
      "step": 18300
    },
    {
      "entropy": 5.844787216186523,
      "epoch": 1.6164782762274814,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004741167195664706,
      "loss": 5.4392,
      "mean_token_accuracy": 0.19454080611467361,
      "num_tokens": 34134447.0,
      "step": 18305
    },
    {
      "entropy": 5.8771710872650145,
      "epoch": 1.6169198163193217,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047410205325095725,
      "loss": 5.3178,
      "mean_token_accuracy": 0.20895783901214598,
      "num_tokens": 34143207.0,
      "step": 18310
    },
    {
      "entropy": 5.873680162429809,
      "epoch": 1.6173613564111622,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004740873830351479,
      "loss": 5.3493,
      "mean_token_accuracy": 0.20246138870716096,
      "num_tokens": 34153025.0,
      "step": 18315
    },
    {
      "entropy": 5.865786027908325,
      "epoch": 1.6178028965030025,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047407270891932995,
      "loss": 5.3558,
      "mean_token_accuracy": 0.20315389782190324,
      "num_tokens": 34162793.0,
      "step": 18320
    },
    {
      "entropy": 5.862230348587036,
      "epoch": 1.6182444365948427,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047405803090379083,
      "loss": 5.4092,
      "mean_token_accuracy": 0.1985282376408577,
      "num_tokens": 34171941.0,
      "step": 18325
    },
    {
      "entropy": 5.887270593643189,
      "epoch": 1.6186859766866832,
      "grad_norm": 1.0,
      "learning_rate": 0.0004740433489888182,
      "loss": 5.382,
      "mean_token_accuracy": 0.20298713445663452,
      "num_tokens": 34180993.0,
      "step": 18330
    },
    {
      "entropy": 5.89012336730957,
      "epoch": 1.6191275167785235,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047402866317469944,
      "loss": 5.3029,
      "mean_token_accuracy": 0.21341717392206191,
      "num_tokens": 34190078.0,
      "step": 18335
    },
    {
      "entropy": 5.774645376205444,
      "epoch": 1.6195690568703638,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004740139734617224,
      "loss": 5.3122,
      "mean_token_accuracy": 0.21125834286212922,
      "num_tokens": 34199091.0,
      "step": 18340
    },
    {
      "entropy": 5.848790168762207,
      "epoch": 1.6200105969622043,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004739992798501748,
      "loss": 5.4204,
      "mean_token_accuracy": 0.20708162635564803,
      "num_tokens": 34209042.0,
      "step": 18345
    },
    {
      "entropy": 5.9101039409637455,
      "epoch": 1.6204521370540446,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047398458234034445,
      "loss": 5.3888,
      "mean_token_accuracy": 0.20516490340232849,
      "num_tokens": 34217311.0,
      "step": 18350
    },
    {
      "entropy": 5.782332801818848,
      "epoch": 1.6208936771458848,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047396988093251934,
      "loss": 5.3518,
      "mean_token_accuracy": 0.2104898452758789,
      "num_tokens": 34227001.0,
      "step": 18355
    },
    {
      "entropy": 5.827291965484619,
      "epoch": 1.6213352172377253,
      "grad_norm": 1.0,
      "learning_rate": 0.0004739551756269875,
      "loss": 5.3817,
      "mean_token_accuracy": 0.20240366756916045,
      "num_tokens": 34236214.0,
      "step": 18360
    },
    {
      "entropy": 5.875600099563599,
      "epoch": 1.6217767573295654,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004739404664240368,
      "loss": 5.3561,
      "mean_token_accuracy": 0.2042583554983139,
      "num_tokens": 34244808.0,
      "step": 18365
    },
    {
      "entropy": 5.841067981719971,
      "epoch": 1.6222182974214059,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047392575332395556,
      "loss": 5.3669,
      "mean_token_accuracy": 0.21230745315551758,
      "num_tokens": 34253355.0,
      "step": 18370
    },
    {
      "entropy": 5.9156270027160645,
      "epoch": 1.6226598375132462,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004739110363270319,
      "loss": 5.4646,
      "mean_token_accuracy": 0.19377294778823853,
      "num_tokens": 34262235.0,
      "step": 18375
    },
    {
      "entropy": 5.871703672409057,
      "epoch": 1.6231013776050864,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047389631543355424,
      "loss": 5.3273,
      "mean_token_accuracy": 0.2104901224374771,
      "num_tokens": 34272174.0,
      "step": 18380
    },
    {
      "entropy": 5.890290784835815,
      "epoch": 1.623542917696927,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004738815906438108,
      "loss": 5.3552,
      "mean_token_accuracy": 0.20510149002075195,
      "num_tokens": 34281153.0,
      "step": 18385
    },
    {
      "entropy": 5.851149082183838,
      "epoch": 1.6239844577887672,
      "grad_norm": 1.0,
      "learning_rate": 0.00047386686195809017,
      "loss": 5.3791,
      "mean_token_accuracy": 0.1983202025294304,
      "num_tokens": 34291045.0,
      "step": 18390
    },
    {
      "entropy": 5.82644214630127,
      "epoch": 1.6244259978806075,
      "grad_norm": 1.125,
      "learning_rate": 0.00047385212937668084,
      "loss": 5.3332,
      "mean_token_accuracy": 0.20592467188835145,
      "num_tokens": 34299234.0,
      "step": 18395
    },
    {
      "entropy": 5.822642421722412,
      "epoch": 1.624867537972448,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047383739289987134,
      "loss": 5.4037,
      "mean_token_accuracy": 0.2028586596250534,
      "num_tokens": 34309084.0,
      "step": 18400
    },
    {
      "entropy": 5.932138252258301,
      "epoch": 1.6253090780642883,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004738226525279504,
      "loss": 5.47,
      "mean_token_accuracy": 0.19281615167856217,
      "num_tokens": 34320039.0,
      "step": 18405
    },
    {
      "entropy": 5.83445372581482,
      "epoch": 1.6257506181561285,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004738079082612068,
      "loss": 5.2778,
      "mean_token_accuracy": 0.2126028507947922,
      "num_tokens": 34329020.0,
      "step": 18410
    },
    {
      "entropy": 5.821759986877441,
      "epoch": 1.626192158247969,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047379316009992927,
      "loss": 5.3986,
      "mean_token_accuracy": 0.19910838752985,
      "num_tokens": 34337953.0,
      "step": 18415
    },
    {
      "entropy": 5.7514075756073,
      "epoch": 1.6266336983398093,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047377840804440686,
      "loss": 5.2854,
      "mean_token_accuracy": 0.21218639463186265,
      "num_tokens": 34346564.0,
      "step": 18420
    },
    {
      "entropy": 5.89910740852356,
      "epoch": 1.6270752384316496,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004737636520949285,
      "loss": 5.5159,
      "mean_token_accuracy": 0.1931287094950676,
      "num_tokens": 34357163.0,
      "step": 18425
    },
    {
      "entropy": 5.906308269500732,
      "epoch": 1.62751677852349,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004737488922517832,
      "loss": 5.4759,
      "mean_token_accuracy": 0.19571054875850677,
      "num_tokens": 34366731.0,
      "step": 18430
    },
    {
      "entropy": 5.814606523513794,
      "epoch": 1.6279583186153301,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00047373412851526013,
      "loss": 5.3377,
      "mean_token_accuracy": 0.20792262107133866,
      "num_tokens": 34375845.0,
      "step": 18435
    },
    {
      "entropy": 5.864726448059082,
      "epoch": 1.6283998587071706,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047371936088564856,
      "loss": 5.3398,
      "mean_token_accuracy": 0.21043552607297897,
      "num_tokens": 34385335.0,
      "step": 18440
    },
    {
      "entropy": 5.8956443786621096,
      "epoch": 1.628841398799011,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004737045893632377,
      "loss": 5.4335,
      "mean_token_accuracy": 0.20258648693561554,
      "num_tokens": 34394356.0,
      "step": 18445
    },
    {
      "entropy": 5.829281616210937,
      "epoch": 1.6292829388908512,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047368981394831696,
      "loss": 5.3708,
      "mean_token_accuracy": 0.2046119302511215,
      "num_tokens": 34403727.0,
      "step": 18450
    },
    {
      "entropy": 5.924932670593262,
      "epoch": 1.6297244789826917,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047367503464117573,
      "loss": 5.4336,
      "mean_token_accuracy": 0.19864198565483093,
      "num_tokens": 34413253.0,
      "step": 18455
    },
    {
      "entropy": 5.813119983673095,
      "epoch": 1.630166019074532,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047366025144210356,
      "loss": 5.339,
      "mean_token_accuracy": 0.20886782556772232,
      "num_tokens": 34422430.0,
      "step": 18460
    },
    {
      "entropy": 5.824563121795654,
      "epoch": 1.6306075591663722,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047364546435139004,
      "loss": 5.3796,
      "mean_token_accuracy": 0.2088511675596237,
      "num_tokens": 34431111.0,
      "step": 18465
    },
    {
      "entropy": 5.8530223846435545,
      "epoch": 1.6310490992582127,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047363067336932493,
      "loss": 5.3678,
      "mean_token_accuracy": 0.20952707827091216,
      "num_tokens": 34440044.0,
      "step": 18470
    },
    {
      "entropy": 5.826060247421265,
      "epoch": 1.631490639350053,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004736158784961978,
      "loss": 5.3683,
      "mean_token_accuracy": 0.21185309141874314,
      "num_tokens": 34450315.0,
      "step": 18475
    },
    {
      "entropy": 5.820285654067993,
      "epoch": 1.6319321794418933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047360107973229865,
      "loss": 5.2941,
      "mean_token_accuracy": 0.21950102746486663,
      "num_tokens": 34459380.0,
      "step": 18480
    },
    {
      "entropy": 5.89492883682251,
      "epoch": 1.6323737195337338,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004735862770779172,
      "loss": 5.4187,
      "mean_token_accuracy": 0.20642026364803315,
      "num_tokens": 34468224.0,
      "step": 18485
    },
    {
      "entropy": 5.780845785140992,
      "epoch": 1.632815259625574,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004735714705333436,
      "loss": 5.2894,
      "mean_token_accuracy": 0.20784697830677032,
      "num_tokens": 34476917.0,
      "step": 18490
    },
    {
      "entropy": 5.784996795654297,
      "epoch": 1.6332567997174143,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047355666009886785,
      "loss": 5.3613,
      "mean_token_accuracy": 0.21024878323078156,
      "num_tokens": 34487192.0,
      "step": 18495
    },
    {
      "entropy": 5.862994718551636,
      "epoch": 1.6336983398092548,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00047354184577478,
      "loss": 5.3609,
      "mean_token_accuracy": 0.20948976129293442,
      "num_tokens": 34496658.0,
      "step": 18500
    },
    {
      "entropy": 5.878349256515503,
      "epoch": 1.6341398799010949,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047352702756137035,
      "loss": 5.2955,
      "mean_token_accuracy": 0.21331375241279601,
      "num_tokens": 34505126.0,
      "step": 18505
    },
    {
      "entropy": 5.845956468582154,
      "epoch": 1.6345814199929354,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004735122054589291,
      "loss": 5.3476,
      "mean_token_accuracy": 0.2015761613845825,
      "num_tokens": 34514236.0,
      "step": 18510
    },
    {
      "entropy": 5.735983371734619,
      "epoch": 1.6350229600847757,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004734973794677467,
      "loss": 5.273,
      "mean_token_accuracy": 0.21282364428043365,
      "num_tokens": 34522609.0,
      "step": 18515
    },
    {
      "entropy": 5.772179412841797,
      "epoch": 1.635464500176616,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047348254958811347,
      "loss": 5.3932,
      "mean_token_accuracy": 0.20632438063621522,
      "num_tokens": 34531742.0,
      "step": 18520
    },
    {
      "entropy": 5.805711889266968,
      "epoch": 1.6359060402684564,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047346771582031994,
      "loss": 5.3198,
      "mean_token_accuracy": 0.20902577936649322,
      "num_tokens": 34540821.0,
      "step": 18525
    },
    {
      "entropy": 5.914322185516357,
      "epoch": 1.6363475803602967,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004734528781646568,
      "loss": 5.3986,
      "mean_token_accuracy": 0.20035960674285888,
      "num_tokens": 34549795.0,
      "step": 18530
    },
    {
      "entropy": 5.883692359924316,
      "epoch": 1.636789120452137,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004734380366214146,
      "loss": 5.4239,
      "mean_token_accuracy": 0.19963254034519196,
      "num_tokens": 34559455.0,
      "step": 18535
    },
    {
      "entropy": 5.880136299133301,
      "epoch": 1.6372306605439775,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004734231911908842,
      "loss": 5.4154,
      "mean_token_accuracy": 0.2007180407643318,
      "num_tokens": 34568832.0,
      "step": 18540
    },
    {
      "entropy": 5.80340805053711,
      "epoch": 1.6376722006358178,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00047340834187335626,
      "loss": 5.3684,
      "mean_token_accuracy": 0.207799032330513,
      "num_tokens": 34578630.0,
      "step": 18545
    },
    {
      "entropy": 5.888047599792481,
      "epoch": 1.638113740727658,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047339348866912173,
      "loss": 5.4145,
      "mean_token_accuracy": 0.19899772703647614,
      "num_tokens": 34587782.0,
      "step": 18550
    },
    {
      "entropy": 5.838915061950684,
      "epoch": 1.6385552808194985,
      "grad_norm": 1.125,
      "learning_rate": 0.0004733786315784716,
      "loss": 5.343,
      "mean_token_accuracy": 0.20412253439426423,
      "num_tokens": 34597133.0,
      "step": 18555
    },
    {
      "entropy": 5.820776557922363,
      "epoch": 1.6389968209113388,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047336377060169696,
      "loss": 5.4203,
      "mean_token_accuracy": 0.19890081137418747,
      "num_tokens": 34606697.0,
      "step": 18560
    },
    {
      "entropy": 5.866944885253906,
      "epoch": 1.639438361003179,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004733489057390888,
      "loss": 5.373,
      "mean_token_accuracy": 0.20852273404598237,
      "num_tokens": 34615806.0,
      "step": 18565
    },
    {
      "entropy": 5.891376066207886,
      "epoch": 1.6398799010950196,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004733340369909384,
      "loss": 5.5243,
      "mean_token_accuracy": 0.19144340604543686,
      "num_tokens": 34626563.0,
      "step": 18570
    },
    {
      "entropy": 5.879660797119141,
      "epoch": 1.6403214411868596,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047331916435753706,
      "loss": 5.4042,
      "mean_token_accuracy": 0.20562248826026916,
      "num_tokens": 34636077.0,
      "step": 18575
    },
    {
      "entropy": 5.878395032882691,
      "epoch": 1.6407629812787001,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047330428783917607,
      "loss": 5.4508,
      "mean_token_accuracy": 0.1952090084552765,
      "num_tokens": 34646164.0,
      "step": 18580
    },
    {
      "entropy": 5.882407474517822,
      "epoch": 1.6412045213705404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047328940743614683,
      "loss": 5.3852,
      "mean_token_accuracy": 0.19838855415582657,
      "num_tokens": 34655577.0,
      "step": 18585
    },
    {
      "entropy": 5.875201749801636,
      "epoch": 1.6416460614623807,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047327452314874094,
      "loss": 5.4252,
      "mean_token_accuracy": 0.192774598300457,
      "num_tokens": 34664520.0,
      "step": 18590
    },
    {
      "entropy": 5.820896291732788,
      "epoch": 1.6420876015542212,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047325963497724983,
      "loss": 5.3218,
      "mean_token_accuracy": 0.205902536213398,
      "num_tokens": 34673528.0,
      "step": 18595
    },
    {
      "entropy": 5.865429306030274,
      "epoch": 1.6425291416460615,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004732447429219652,
      "loss": 5.3969,
      "mean_token_accuracy": 0.20284605920314788,
      "num_tokens": 34682563.0,
      "step": 18600
    },
    {
      "entropy": 5.89749960899353,
      "epoch": 1.6429706817379017,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004732298469831789,
      "loss": 5.3048,
      "mean_token_accuracy": 0.20845038294792176,
      "num_tokens": 34691932.0,
      "step": 18605
    },
    {
      "entropy": 5.887390661239624,
      "epoch": 1.6434122218297422,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047321494716118256,
      "loss": 5.3446,
      "mean_token_accuracy": 0.2051656261086464,
      "num_tokens": 34701376.0,
      "step": 18610
    },
    {
      "entropy": 5.795174598693848,
      "epoch": 1.6438537619215825,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00047320004345626816,
      "loss": 5.3459,
      "mean_token_accuracy": 0.20236467868089675,
      "num_tokens": 34711112.0,
      "step": 18615
    },
    {
      "entropy": 5.837432384490967,
      "epoch": 1.6442953020134228,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047318513586872766,
      "loss": 5.351,
      "mean_token_accuracy": 0.21455023884773256,
      "num_tokens": 34720429.0,
      "step": 18620
    },
    {
      "entropy": 5.854586839675903,
      "epoch": 1.6447368421052633,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000473170224398853,
      "loss": 5.3693,
      "mean_token_accuracy": 0.20948509573936464,
      "num_tokens": 34729422.0,
      "step": 18625
    },
    {
      "entropy": 5.915212249755859,
      "epoch": 1.6451783821971036,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047315530904693643,
      "loss": 5.3917,
      "mean_token_accuracy": 0.20037249922752381,
      "num_tokens": 34738231.0,
      "step": 18630
    },
    {
      "entropy": 5.9505791664123535,
      "epoch": 1.6456199222889438,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047314038981327,
      "loss": 5.4269,
      "mean_token_accuracy": 0.20173603296279907,
      "num_tokens": 34747002.0,
      "step": 18635
    },
    {
      "entropy": 5.7558191299438475,
      "epoch": 1.6460614623807843,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000473125466698146,
      "loss": 5.3056,
      "mean_token_accuracy": 0.215034019947052,
      "num_tokens": 34756093.0,
      "step": 18640
    },
    {
      "entropy": 5.770248460769653,
      "epoch": 1.6465030024726244,
      "grad_norm": 1.0,
      "learning_rate": 0.0004731105397018569,
      "loss": 5.3776,
      "mean_token_accuracy": 0.20310365110635759,
      "num_tokens": 34766031.0,
      "step": 18645
    },
    {
      "entropy": 5.918416357040405,
      "epoch": 1.6469445425644649,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004730956088246949,
      "loss": 5.4459,
      "mean_token_accuracy": 0.20414311438798904,
      "num_tokens": 34776487.0,
      "step": 18650
    },
    {
      "entropy": 5.903948783874512,
      "epoch": 1.6473860826563052,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047308067406695255,
      "loss": 5.3058,
      "mean_token_accuracy": 0.2076905757188797,
      "num_tokens": 34785372.0,
      "step": 18655
    },
    {
      "entropy": 5.789320850372315,
      "epoch": 1.6478276227481454,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047306573542892257,
      "loss": 5.3299,
      "mean_token_accuracy": 0.20968879461288453,
      "num_tokens": 34795453.0,
      "step": 18660
    },
    {
      "entropy": 5.8344183444976805,
      "epoch": 1.648269162839986,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047305079291089735,
      "loss": 5.4095,
      "mean_token_accuracy": 0.20453164875507354,
      "num_tokens": 34804818.0,
      "step": 18665
    },
    {
      "entropy": 5.847234725952148,
      "epoch": 1.6487107029318262,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047303584651316985,
      "loss": 5.4392,
      "mean_token_accuracy": 0.19708086997270585,
      "num_tokens": 34814031.0,
      "step": 18670
    },
    {
      "entropy": 5.816784191131592,
      "epoch": 1.6491522430236665,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004730208962360327,
      "loss": 5.3254,
      "mean_token_accuracy": 0.20470742732286454,
      "num_tokens": 34823560.0,
      "step": 18675
    },
    {
      "entropy": 5.883140230178833,
      "epoch": 1.649593783115507,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004730059420797788,
      "loss": 5.3244,
      "mean_token_accuracy": 0.20900787711143493,
      "num_tokens": 34833662.0,
      "step": 18680
    },
    {
      "entropy": 5.897165203094483,
      "epoch": 1.6500353232073472,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004729909840447011,
      "loss": 5.4622,
      "mean_token_accuracy": 0.1925043895840645,
      "num_tokens": 34844264.0,
      "step": 18685
    },
    {
      "entropy": 5.902440738677979,
      "epoch": 1.6504768632991875,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047297602213109265,
      "loss": 5.3679,
      "mean_token_accuracy": 0.20841593146324158,
      "num_tokens": 34853708.0,
      "step": 18690
    },
    {
      "entropy": 5.937035799026489,
      "epoch": 1.650918403391028,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004729610563392465,
      "loss": 5.4398,
      "mean_token_accuracy": 0.19457600265741348,
      "num_tokens": 34861926.0,
      "step": 18695
    },
    {
      "entropy": 5.8199704647064205,
      "epoch": 1.6513599434828683,
      "grad_norm": 1.125,
      "learning_rate": 0.0004729460866694558,
      "loss": 5.3512,
      "mean_token_accuracy": 0.2098531037569046,
      "num_tokens": 34870919.0,
      "step": 18700
    },
    {
      "entropy": 5.851771879196167,
      "epoch": 1.6518014835747086,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000472931113122014,
      "loss": 5.3979,
      "mean_token_accuracy": 0.2007477343082428,
      "num_tokens": 34879801.0,
      "step": 18705
    },
    {
      "entropy": 5.8653754711151125,
      "epoch": 1.652243023666549,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004729161356972141,
      "loss": 5.3702,
      "mean_token_accuracy": 0.20218620598316192,
      "num_tokens": 34888605.0,
      "step": 18710
    },
    {
      "entropy": 5.882697534561157,
      "epoch": 1.6526845637583891,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047290115439534977,
      "loss": 5.4375,
      "mean_token_accuracy": 0.19343627393245696,
      "num_tokens": 34899294.0,
      "step": 18715
    },
    {
      "entropy": 5.9275675296783445,
      "epoch": 1.6531261038502296,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004728861692167143,
      "loss": 5.437,
      "mean_token_accuracy": 0.19946447908878326,
      "num_tokens": 34908624.0,
      "step": 18720
    },
    {
      "entropy": 5.847243595123291,
      "epoch": 1.65356764394207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004728711801616013,
      "loss": 5.3221,
      "mean_token_accuracy": 0.20005699694156648,
      "num_tokens": 34917390.0,
      "step": 18725
    },
    {
      "entropy": 5.853270483016968,
      "epoch": 1.6540091840339102,
      "grad_norm": 1.0,
      "learning_rate": 0.0004728561872303045,
      "loss": 5.3924,
      "mean_token_accuracy": 0.2000747174024582,
      "num_tokens": 34926652.0,
      "step": 18730
    },
    {
      "entropy": 5.840923500061035,
      "epoch": 1.6544507241257507,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004728411904231175,
      "loss": 5.4278,
      "mean_token_accuracy": 0.19802002906799315,
      "num_tokens": 34936900.0,
      "step": 18735
    },
    {
      "entropy": 5.9109937191009525,
      "epoch": 1.654892264217591,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004728261897403341,
      "loss": 5.4059,
      "mean_token_accuracy": 0.2028878390789032,
      "num_tokens": 34946827.0,
      "step": 18740
    },
    {
      "entropy": 5.887080526351928,
      "epoch": 1.6553338043094312,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047281118518224807,
      "loss": 5.4529,
      "mean_token_accuracy": 0.1989116445183754,
      "num_tokens": 34955987.0,
      "step": 18745
    },
    {
      "entropy": 5.810631990432739,
      "epoch": 1.6557753444012717,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004727961767491535,
      "loss": 5.3405,
      "mean_token_accuracy": 0.20601450204849242,
      "num_tokens": 34965001.0,
      "step": 18750
    },
    {
      "entropy": 5.88202395439148,
      "epoch": 1.656216884493112,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004727811644413443,
      "loss": 5.4249,
      "mean_token_accuracy": 0.19914984703063965,
      "num_tokens": 34974113.0,
      "step": 18755
    },
    {
      "entropy": 5.803858470916748,
      "epoch": 1.6566584245849523,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004727661482591146,
      "loss": 5.2662,
      "mean_token_accuracy": 0.21207729279994963,
      "num_tokens": 34982535.0,
      "step": 18760
    },
    {
      "entropy": 5.839359855651855,
      "epoch": 1.6570999646767928,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004727511282027585,
      "loss": 5.4013,
      "mean_token_accuracy": 0.20235300362110137,
      "num_tokens": 34991828.0,
      "step": 18765
    },
    {
      "entropy": 5.86194076538086,
      "epoch": 1.657541504768633,
      "grad_norm": 1.0,
      "learning_rate": 0.0004727361042725703,
      "loss": 5.3522,
      "mean_token_accuracy": 0.20406592339277269,
      "num_tokens": 35001670.0,
      "step": 18770
    },
    {
      "entropy": 5.82854733467102,
      "epoch": 1.6579830448604733,
      "grad_norm": 1.0,
      "learning_rate": 0.0004727210764688442,
      "loss": 5.3053,
      "mean_token_accuracy": 0.21206227391958238,
      "num_tokens": 35011407.0,
      "step": 18775
    },
    {
      "entropy": 5.759233093261718,
      "epoch": 1.6584245849523138,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004727060447918748,
      "loss": 5.2626,
      "mean_token_accuracy": 0.21727235168218612,
      "num_tokens": 35020753.0,
      "step": 18780
    },
    {
      "entropy": 5.832109642028809,
      "epoch": 1.6588661250441539,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00047269100924195623,
      "loss": 5.3895,
      "mean_token_accuracy": 0.20545273274183273,
      "num_tokens": 35030599.0,
      "step": 18785
    },
    {
      "entropy": 5.880756568908692,
      "epoch": 1.6593076651359944,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047267596981938337,
      "loss": 5.4193,
      "mean_token_accuracy": 0.20258154422044755,
      "num_tokens": 35039098.0,
      "step": 18790
    },
    {
      "entropy": 5.892031860351563,
      "epoch": 1.6597492052278346,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047266092652445066,
      "loss": 5.3551,
      "mean_token_accuracy": 0.20884549170732497,
      "num_tokens": 35048286.0,
      "step": 18795
    },
    {
      "entropy": 5.917185974121094,
      "epoch": 1.660190745319675,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004726458793574528,
      "loss": 5.4458,
      "mean_token_accuracy": 0.19773625135421752,
      "num_tokens": 35057331.0,
      "step": 18800
    },
    {
      "entropy": 5.88164587020874,
      "epoch": 1.6606322854115154,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004726308283186846,
      "loss": 5.399,
      "mean_token_accuracy": 0.1980038672685623,
      "num_tokens": 35066321.0,
      "step": 18805
    },
    {
      "entropy": 5.773690843582154,
      "epoch": 1.6610738255033557,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004726157734084409,
      "loss": 5.339,
      "mean_token_accuracy": 0.20270509272813797,
      "num_tokens": 35075218.0,
      "step": 18810
    },
    {
      "entropy": 5.808038377761841,
      "epoch": 1.661515365595196,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004726007146270164,
      "loss": 5.3983,
      "mean_token_accuracy": 0.20630132853984834,
      "num_tokens": 35084563.0,
      "step": 18815
    },
    {
      "entropy": 5.79873571395874,
      "epoch": 1.6619569056870365,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004725856519747065,
      "loss": 5.2779,
      "mean_token_accuracy": 0.21271708309650422,
      "num_tokens": 35093346.0,
      "step": 18820
    },
    {
      "entropy": 5.857068824768066,
      "epoch": 1.6623984457788767,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004725705854518059,
      "loss": 5.3805,
      "mean_token_accuracy": 0.19910885095596315,
      "num_tokens": 35102501.0,
      "step": 18825
    },
    {
      "entropy": 5.855516576766968,
      "epoch": 1.662839985870717,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047255551505861007,
      "loss": 5.3788,
      "mean_token_accuracy": 0.2015061244368553,
      "num_tokens": 35112083.0,
      "step": 18830
    },
    {
      "entropy": 5.809852600097656,
      "epoch": 1.6632815259625575,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047254044079541387,
      "loss": 5.2996,
      "mean_token_accuracy": 0.2174018293619156,
      "num_tokens": 35121380.0,
      "step": 18835
    },
    {
      "entropy": 5.855141305923462,
      "epoch": 1.6637230660543978,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004725253626625129,
      "loss": 5.4386,
      "mean_token_accuracy": 0.19308264851570128,
      "num_tokens": 35130083.0,
      "step": 18840
    },
    {
      "entropy": 5.848509502410889,
      "epoch": 1.664164606146238,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047251028066020237,
      "loss": 5.3052,
      "mean_token_accuracy": 0.21321660578250884,
      "num_tokens": 35139001.0,
      "step": 18845
    },
    {
      "entropy": 5.81809778213501,
      "epoch": 1.6646061462380786,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047249519478877774,
      "loss": 5.2762,
      "mean_token_accuracy": 0.2110721617937088,
      "num_tokens": 35147701.0,
      "step": 18850
    },
    {
      "entropy": 5.805716037750244,
      "epoch": 1.6650476863299186,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004724801050485346,
      "loss": 5.285,
      "mean_token_accuracy": 0.21251884400844573,
      "num_tokens": 35157334.0,
      "step": 18855
    },
    {
      "entropy": 5.817666625976562,
      "epoch": 1.6654892264217591,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004724650114397685,
      "loss": 5.3325,
      "mean_token_accuracy": 0.2110571950674057,
      "num_tokens": 35166647.0,
      "step": 18860
    },
    {
      "entropy": 5.8946551322937015,
      "epoch": 1.6659307665135994,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047244991396277515,
      "loss": 5.3328,
      "mean_token_accuracy": 0.2022341623902321,
      "num_tokens": 35176345.0,
      "step": 18865
    },
    {
      "entropy": 5.8745075225830075,
      "epoch": 1.6663723066054397,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047243481261785027,
      "loss": 5.3335,
      "mean_token_accuracy": 0.20662084966897964,
      "num_tokens": 35185793.0,
      "step": 18870
    },
    {
      "entropy": 5.837036418914795,
      "epoch": 1.6668138466972802,
      "grad_norm": 1.0,
      "learning_rate": 0.00047241970740528975,
      "loss": 5.4159,
      "mean_token_accuracy": 0.20366771072149276,
      "num_tokens": 35195905.0,
      "step": 18875
    },
    {
      "entropy": 5.760806751251221,
      "epoch": 1.6672553867891204,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047240459832538934,
      "loss": 5.2823,
      "mean_token_accuracy": 0.2141048491001129,
      "num_tokens": 35204340.0,
      "step": 18880
    },
    {
      "entropy": 5.842314195632935,
      "epoch": 1.6676969268809607,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004723894853784452,
      "loss": 5.3789,
      "mean_token_accuracy": 0.20157068520784377,
      "num_tokens": 35213270.0,
      "step": 18885
    },
    {
      "entropy": 5.786710500717163,
      "epoch": 1.6681384669728012,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004723743685647533,
      "loss": 5.3618,
      "mean_token_accuracy": 0.2016805350780487,
      "num_tokens": 35222652.0,
      "step": 18890
    },
    {
      "entropy": 5.872417354583741,
      "epoch": 1.6685800070646415,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004723592478846097,
      "loss": 5.3571,
      "mean_token_accuracy": 0.2074940726161003,
      "num_tokens": 35231684.0,
      "step": 18895
    },
    {
      "entropy": 5.881430053710938,
      "epoch": 1.6690215471564818,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047234412333831073,
      "loss": 5.4161,
      "mean_token_accuracy": 0.20022842437028884,
      "num_tokens": 35241589.0,
      "step": 18900
    },
    {
      "entropy": 5.9737931251525875,
      "epoch": 1.6694630872483223,
      "grad_norm": 1.0,
      "learning_rate": 0.0004723289949261527,
      "loss": 5.5155,
      "mean_token_accuracy": 0.18734343498945236,
      "num_tokens": 35250903.0,
      "step": 18905
    },
    {
      "entropy": 5.850294208526611,
      "epoch": 1.6699046273401625,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004723138626484318,
      "loss": 5.375,
      "mean_token_accuracy": 0.2025665521621704,
      "num_tokens": 35260556.0,
      "step": 18910
    },
    {
      "entropy": 5.778967475891113,
      "epoch": 1.6703461674320028,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047229872650544446,
      "loss": 5.2654,
      "mean_token_accuracy": 0.2170637235045433,
      "num_tokens": 35270237.0,
      "step": 18915
    },
    {
      "entropy": 5.769305229187012,
      "epoch": 1.6707877075238433,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047228358649748745,
      "loss": 5.3051,
      "mean_token_accuracy": 0.2084587976336479,
      "num_tokens": 35279017.0,
      "step": 18920
    },
    {
      "entropy": 5.787815856933594,
      "epoch": 1.6712292476156834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004722684426248571,
      "loss": 5.3269,
      "mean_token_accuracy": 0.20815094709396362,
      "num_tokens": 35288150.0,
      "step": 18925
    },
    {
      "entropy": 5.880165719985962,
      "epoch": 1.6716707877075239,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047225329488785015,
      "loss": 5.4056,
      "mean_token_accuracy": 0.20196864753961563,
      "num_tokens": 35297463.0,
      "step": 18930
    },
    {
      "entropy": 5.85289740562439,
      "epoch": 1.6721123277993641,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047223814328676336,
      "loss": 5.3692,
      "mean_token_accuracy": 0.20511680692434311,
      "num_tokens": 35307409.0,
      "step": 18935
    },
    {
      "entropy": 5.79543023109436,
      "epoch": 1.6725538678912044,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004722229878218935,
      "loss": 5.3172,
      "mean_token_accuracy": 0.20564032942056656,
      "num_tokens": 35315613.0,
      "step": 18940
    },
    {
      "entropy": 5.8183387279510494,
      "epoch": 1.672995407983045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047220782849353743,
      "loss": 5.3838,
      "mean_token_accuracy": 0.2033582389354706,
      "num_tokens": 35325297.0,
      "step": 18945
    },
    {
      "entropy": 5.8459968090057375,
      "epoch": 1.6734369480748852,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004721926653019922,
      "loss": 5.3654,
      "mean_token_accuracy": 0.19865651726722716,
      "num_tokens": 35334014.0,
      "step": 18950
    },
    {
      "entropy": 5.82857460975647,
      "epoch": 1.6738784881667255,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004721774982475548,
      "loss": 5.3758,
      "mean_token_accuracy": 0.20930878072977066,
      "num_tokens": 35342784.0,
      "step": 18955
    },
    {
      "entropy": 5.930045795440674,
      "epoch": 1.674320028258566,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047216232733052237,
      "loss": 5.4783,
      "mean_token_accuracy": 0.1991959422826767,
      "num_tokens": 35351985.0,
      "step": 18960
    },
    {
      "entropy": 5.867730951309204,
      "epoch": 1.6747615683504062,
      "grad_norm": 1.125,
      "learning_rate": 0.0004721471525511921,
      "loss": 5.3389,
      "mean_token_accuracy": 0.1984658047556877,
      "num_tokens": 35360857.0,
      "step": 18965
    },
    {
      "entropy": 5.805293941497803,
      "epoch": 1.6752031084422465,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004721319739098611,
      "loss": 5.3304,
      "mean_token_accuracy": 0.2052135795354843,
      "num_tokens": 35370200.0,
      "step": 18970
    },
    {
      "entropy": 5.837771272659301,
      "epoch": 1.675644648534087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000472116791406827,
      "loss": 5.3998,
      "mean_token_accuracy": 0.20853707939386368,
      "num_tokens": 35379481.0,
      "step": 18975
    },
    {
      "entropy": 5.810681009292603,
      "epoch": 1.6760861886259273,
      "grad_norm": 1.03125,
      "learning_rate": 0.000472101605042387,
      "loss": 5.3705,
      "mean_token_accuracy": 0.211078479886055,
      "num_tokens": 35387951.0,
      "step": 18980
    },
    {
      "entropy": 5.936203479766846,
      "epoch": 1.6765277287177676,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004720864148168387,
      "loss": 5.397,
      "mean_token_accuracy": 0.19801660627126694,
      "num_tokens": 35396200.0,
      "step": 18985
    },
    {
      "entropy": 5.892510652542114,
      "epoch": 1.676969268809608,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004720712207304796,
      "loss": 5.3794,
      "mean_token_accuracy": 0.19763886779546738,
      "num_tokens": 35404474.0,
      "step": 18990
    },
    {
      "entropy": 5.919994831085205,
      "epoch": 1.6774108089014481,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004720560227836073,
      "loss": 5.4726,
      "mean_token_accuracy": 0.19684788435697556,
      "num_tokens": 35414419.0,
      "step": 18995
    },
    {
      "entropy": 5.822895193099976,
      "epoch": 1.6778523489932886,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004720408209765197,
      "loss": 5.3023,
      "mean_token_accuracy": 0.21042004972696304,
      "num_tokens": 35423727.0,
      "step": 19000
    },
    {
      "entropy": 5.760235261917114,
      "epoch": 1.6782938890851289,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00047202561530951443,
      "loss": 5.2702,
      "mean_token_accuracy": 0.21885372698307037,
      "num_tokens": 35433364.0,
      "step": 19005
    },
    {
      "entropy": 5.919628286361695,
      "epoch": 1.6787354291769692,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004720104057828894,
      "loss": 5.4346,
      "mean_token_accuracy": 0.19773211181163788,
      "num_tokens": 35442665.0,
      "step": 19010
    },
    {
      "entropy": 5.8532500743865965,
      "epoch": 1.6791769692688097,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047199519239694257,
      "loss": 5.2658,
      "mean_token_accuracy": 0.21819208264350892,
      "num_tokens": 35451443.0,
      "step": 19015
    },
    {
      "entropy": 5.762861824035644,
      "epoch": 1.67961850936065,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047197997515197195,
      "loss": 5.2665,
      "mean_token_accuracy": 0.22124966681003572,
      "num_tokens": 35460622.0,
      "step": 19020
    },
    {
      "entropy": 5.757262182235718,
      "epoch": 1.6800600494524902,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047196475404827565,
      "loss": 5.409,
      "mean_token_accuracy": 0.20361010432243348,
      "num_tokens": 35470345.0,
      "step": 19025
    },
    {
      "entropy": 5.923474836349487,
      "epoch": 1.6805015895443307,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047194952908615183,
      "loss": 5.4904,
      "mean_token_accuracy": 0.19669502973556519,
      "num_tokens": 35480460.0,
      "step": 19030
    },
    {
      "entropy": 5.9418315410614015,
      "epoch": 1.680943129636171,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004719343002658987,
      "loss": 5.4182,
      "mean_token_accuracy": 0.1977526530623436,
      "num_tokens": 35489983.0,
      "step": 19035
    },
    {
      "entropy": 5.886590909957886,
      "epoch": 1.6813846697280113,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047191906758781475,
      "loss": 5.3489,
      "mean_token_accuracy": 0.20787034481763839,
      "num_tokens": 35498963.0,
      "step": 19040
    },
    {
      "entropy": 5.824855089187622,
      "epoch": 1.6818262098198518,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004719038310521982,
      "loss": 5.3827,
      "mean_token_accuracy": 0.2082638829946518,
      "num_tokens": 35509054.0,
      "step": 19045
    },
    {
      "entropy": 5.876221179962158,
      "epoch": 1.682267749911692,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004718885906593476,
      "loss": 5.5138,
      "mean_token_accuracy": 0.19023756682872772,
      "num_tokens": 35518093.0,
      "step": 19050
    },
    {
      "entropy": 5.888548374176025,
      "epoch": 1.6827092900035323,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004718733464095614,
      "loss": 5.3959,
      "mean_token_accuracy": 0.20310216248035431,
      "num_tokens": 35527330.0,
      "step": 19055
    },
    {
      "entropy": 5.882849168777466,
      "epoch": 1.6831508300953728,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047185809830313844,
      "loss": 5.3672,
      "mean_token_accuracy": 0.20466699302196503,
      "num_tokens": 35538147.0,
      "step": 19060
    },
    {
      "entropy": 5.962263298034668,
      "epoch": 1.6835923701872129,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047184284634037727,
      "loss": 5.4588,
      "mean_token_accuracy": 0.19409112334251405,
      "num_tokens": 35547930.0,
      "step": 19065
    },
    {
      "entropy": 5.858351755142212,
      "epoch": 1.6840339102790534,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047182759052157665,
      "loss": 5.2959,
      "mean_token_accuracy": 0.2073899418115616,
      "num_tokens": 35557111.0,
      "step": 19070
    },
    {
      "entropy": 5.8369776725769045,
      "epoch": 1.6844754503708936,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004718123308470355,
      "loss": 5.4761,
      "mean_token_accuracy": 0.19234734028577805,
      "num_tokens": 35565739.0,
      "step": 19075
    },
    {
      "entropy": 5.918671321868897,
      "epoch": 1.684916990462734,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047179706731705263,
      "loss": 5.3051,
      "mean_token_accuracy": 0.20207255184650422,
      "num_tokens": 35574164.0,
      "step": 19080
    },
    {
      "entropy": 5.855103445053101,
      "epoch": 1.6853585305545744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047178179993192723,
      "loss": 5.3752,
      "mean_token_accuracy": 0.19642841666936875,
      "num_tokens": 35582813.0,
      "step": 19085
    },
    {
      "entropy": 5.870343112945557,
      "epoch": 1.6858000706464147,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047176652869195825,
      "loss": 5.3455,
      "mean_token_accuracy": 0.1986942782998085,
      "num_tokens": 35591884.0,
      "step": 19090
    },
    {
      "entropy": 5.872808456420898,
      "epoch": 1.686241610738255,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047175125359744487,
      "loss": 5.3648,
      "mean_token_accuracy": 0.201996086537838,
      "num_tokens": 35601824.0,
      "step": 19095
    },
    {
      "entropy": 5.8208756923675535,
      "epoch": 1.6866831508300955,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047173597464868634,
      "loss": 5.3101,
      "mean_token_accuracy": 0.2113776221871376,
      "num_tokens": 35610756.0,
      "step": 19100
    },
    {
      "entropy": 5.795255994796753,
      "epoch": 1.6871246909219357,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000471720691845982,
      "loss": 5.3149,
      "mean_token_accuracy": 0.20665600448846816,
      "num_tokens": 35620160.0,
      "step": 19105
    },
    {
      "entropy": 5.824592638015747,
      "epoch": 1.687566231013776,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047170540518963114,
      "loss": 5.4617,
      "mean_token_accuracy": 0.19557576924562453,
      "num_tokens": 35630487.0,
      "step": 19110
    },
    {
      "entropy": 5.980372428894043,
      "epoch": 1.6880077711056165,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004716901146799333,
      "loss": 5.5166,
      "mean_token_accuracy": 0.1869860053062439,
      "num_tokens": 35641924.0,
      "step": 19115
    },
    {
      "entropy": 5.903148031234741,
      "epoch": 1.6884493111974568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004716748203171879,
      "loss": 5.3355,
      "mean_token_accuracy": 0.2022925317287445,
      "num_tokens": 35651575.0,
      "step": 19120
    },
    {
      "entropy": 5.825525188446045,
      "epoch": 1.688890851289297,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047165952210169475,
      "loss": 5.3211,
      "mean_token_accuracy": 0.20450648814439773,
      "num_tokens": 35661481.0,
      "step": 19125
    },
    {
      "entropy": 5.7702141284942625,
      "epoch": 1.6893323913811376,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004716442200337533,
      "loss": 5.2953,
      "mean_token_accuracy": 0.20387887954711914,
      "num_tokens": 35672049.0,
      "step": 19130
    },
    {
      "entropy": 5.893923759460449,
      "epoch": 1.6897739314729776,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047162891411366343,
      "loss": 5.3663,
      "mean_token_accuracy": 0.2059311419725418,
      "num_tokens": 35681653.0,
      "step": 19135
    },
    {
      "entropy": 5.908714342117309,
      "epoch": 1.690215471564818,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00047161360434172506,
      "loss": 5.429,
      "mean_token_accuracy": 0.21284251362085344,
      "num_tokens": 35690868.0,
      "step": 19140
    },
    {
      "entropy": 5.822940111160278,
      "epoch": 1.6906570116566584,
      "grad_norm": 1.125,
      "learning_rate": 0.0004715982907182379,
      "loss": 5.3511,
      "mean_token_accuracy": 0.2021711513400078,
      "num_tokens": 35699062.0,
      "step": 19145
    },
    {
      "entropy": 5.790659379959107,
      "epoch": 1.6910985517484987,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004715829732435021,
      "loss": 5.3988,
      "mean_token_accuracy": 0.20533424168825148,
      "num_tokens": 35708411.0,
      "step": 19150
    },
    {
      "entropy": 5.82406325340271,
      "epoch": 1.6915400918403392,
      "grad_norm": 0.96875,
      "learning_rate": 0.00047156765191781763,
      "loss": 5.3191,
      "mean_token_accuracy": 0.2091885730624199,
      "num_tokens": 35717937.0,
      "step": 19155
    },
    {
      "entropy": 5.8888551712036135,
      "epoch": 1.6919816319321794,
      "grad_norm": 1.0,
      "learning_rate": 0.00047155232674148465,
      "loss": 5.3308,
      "mean_token_accuracy": 0.20362054705619811,
      "num_tokens": 35727712.0,
      "step": 19160
    },
    {
      "entropy": 5.902427530288696,
      "epoch": 1.6924231720240197,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047153699771480335,
      "loss": 5.4092,
      "mean_token_accuracy": 0.2020666107535362,
      "num_tokens": 35737076.0,
      "step": 19165
    },
    {
      "entropy": 5.824070167541504,
      "epoch": 1.6928647121158602,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000471521664838074,
      "loss": 5.2314,
      "mean_token_accuracy": 0.21509755849838258,
      "num_tokens": 35746436.0,
      "step": 19170
    },
    {
      "entropy": 5.830689811706543,
      "epoch": 1.6933062522077005,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004715063281115971,
      "loss": 5.4214,
      "mean_token_accuracy": 0.20655142813920974,
      "num_tokens": 35755786.0,
      "step": 19175
    },
    {
      "entropy": 5.871500778198242,
      "epoch": 1.6937477922995408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000471490987535673,
      "loss": 5.398,
      "mean_token_accuracy": 0.20298216193914415,
      "num_tokens": 35764424.0,
      "step": 19180
    },
    {
      "entropy": 5.887312316894532,
      "epoch": 1.6941893323913813,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047147564311060216,
      "loss": 5.3626,
      "mean_token_accuracy": 0.20557391196489333,
      "num_tokens": 35773792.0,
      "step": 19185
    },
    {
      "entropy": 5.825941038131714,
      "epoch": 1.6946308724832215,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047146029483668517,
      "loss": 5.3065,
      "mean_token_accuracy": 0.20931385159492494,
      "num_tokens": 35783285.0,
      "step": 19190
    },
    {
      "entropy": 5.8302113056182865,
      "epoch": 1.6950724125750618,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004714449427142228,
      "loss": 5.4321,
      "mean_token_accuracy": 0.19831449091434478,
      "num_tokens": 35792476.0,
      "step": 19195
    },
    {
      "entropy": 5.861258459091187,
      "epoch": 1.6955139526669023,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047142958674351563,
      "loss": 5.3686,
      "mean_token_accuracy": 0.20720222294330598,
      "num_tokens": 35801640.0,
      "step": 19200
    },
    {
      "entropy": 5.8794529914855955,
      "epoch": 1.6959554927587424,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047141422692486466,
      "loss": 5.3574,
      "mean_token_accuracy": 0.21005713939666748,
      "num_tokens": 35811073.0,
      "step": 19205
    },
    {
      "entropy": 5.860421895980835,
      "epoch": 1.6963970328505829,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047139886325857074,
      "loss": 5.3364,
      "mean_token_accuracy": 0.21036551147699356,
      "num_tokens": 35819814.0,
      "step": 19210
    },
    {
      "entropy": 5.8425548553466795,
      "epoch": 1.6968385729424231,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004713834957449347,
      "loss": 5.3207,
      "mean_token_accuracy": 0.20893237739801407,
      "num_tokens": 35828363.0,
      "step": 19215
    },
    {
      "entropy": 5.867175626754761,
      "epoch": 1.6972801130342634,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004713681243842577,
      "loss": 5.2849,
      "mean_token_accuracy": 0.20875922739505767,
      "num_tokens": 35836868.0,
      "step": 19220
    },
    {
      "entropy": 5.859682321548462,
      "epoch": 1.697721653126104,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047135274917684087,
      "loss": 5.3439,
      "mean_token_accuracy": 0.2039727121591568,
      "num_tokens": 35846030.0,
      "step": 19225
    },
    {
      "entropy": 5.782294034957886,
      "epoch": 1.6981631932179442,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004713373701229852,
      "loss": 5.2554,
      "mean_token_accuracy": 0.2128818228840828,
      "num_tokens": 35854862.0,
      "step": 19230
    },
    {
      "entropy": 5.820282459259033,
      "epoch": 1.6986047333097845,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004713219872229923,
      "loss": 5.3034,
      "mean_token_accuracy": 0.21242024898529052,
      "num_tokens": 35864238.0,
      "step": 19235
    },
    {
      "entropy": 5.845459508895874,
      "epoch": 1.699046273401625,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004713066004771633,
      "loss": 5.3233,
      "mean_token_accuracy": 0.20124825537204744,
      "num_tokens": 35872542.0,
      "step": 19240
    },
    {
      "entropy": 5.882920169830323,
      "epoch": 1.6994878134934652,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047129120988579953,
      "loss": 5.4001,
      "mean_token_accuracy": 0.20187266319990158,
      "num_tokens": 35881899.0,
      "step": 19245
    },
    {
      "entropy": 5.870782089233399,
      "epoch": 1.6999293535853055,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047127581544920263,
      "loss": 5.3855,
      "mean_token_accuracy": 0.19976069480180741,
      "num_tokens": 35891970.0,
      "step": 19250
    },
    {
      "entropy": 5.813725709915161,
      "epoch": 1.700370893677146,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004712604171676742,
      "loss": 5.2545,
      "mean_token_accuracy": 0.20687177926301956,
      "num_tokens": 35901286.0,
      "step": 19255
    },
    {
      "entropy": 5.792936611175537,
      "epoch": 1.7008124337689863,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00047124501504151574,
      "loss": 5.478,
      "mean_token_accuracy": 0.19672661423683166,
      "num_tokens": 35912301.0,
      "step": 19260
    },
    {
      "entropy": 5.8657081604003904,
      "epoch": 1.7012539738608266,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047122960907102906,
      "loss": 5.3591,
      "mean_token_accuracy": 0.2061811089515686,
      "num_tokens": 35920767.0,
      "step": 19265
    },
    {
      "entropy": 5.921802234649658,
      "epoch": 1.701695513952667,
      "grad_norm": 1.03125,
      "learning_rate": 0.000471214199256516,
      "loss": 5.4451,
      "mean_token_accuracy": 0.20307040959596634,
      "num_tokens": 35930235.0,
      "step": 19270
    },
    {
      "entropy": 5.912538194656372,
      "epoch": 1.702137054044507,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047119878559827825,
      "loss": 5.3888,
      "mean_token_accuracy": 0.1973385512828827,
      "num_tokens": 35939213.0,
      "step": 19275
    },
    {
      "entropy": 5.912663698196411,
      "epoch": 1.7025785941363476,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047118336809661793,
      "loss": 5.4631,
      "mean_token_accuracy": 0.1904297724366188,
      "num_tokens": 35948221.0,
      "step": 19280
    },
    {
      "entropy": 5.872070264816284,
      "epoch": 1.7030201342281879,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004711679467518371,
      "loss": 5.3659,
      "mean_token_accuracy": 0.1964646816253662,
      "num_tokens": 35957961.0,
      "step": 19285
    },
    {
      "entropy": 5.861671447753906,
      "epoch": 1.7034616743200282,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047115252156423767,
      "loss": 5.4075,
      "mean_token_accuracy": 0.19784276336431503,
      "num_tokens": 35966742.0,
      "step": 19290
    },
    {
      "entropy": 5.852165079116821,
      "epoch": 1.7039032144118686,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004711370925341219,
      "loss": 5.3424,
      "mean_token_accuracy": 0.19969536662101744,
      "num_tokens": 35976312.0,
      "step": 19295
    },
    {
      "entropy": 5.88031039237976,
      "epoch": 1.704344754503709,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047112165966179206,
      "loss": 5.4497,
      "mean_token_accuracy": 0.19996579140424728,
      "num_tokens": 35985537.0,
      "step": 19300
    },
    {
      "entropy": 5.875399875640869,
      "epoch": 1.7047862945955492,
      "grad_norm": 1.0,
      "learning_rate": 0.00047110622294755045,
      "loss": 5.352,
      "mean_token_accuracy": 0.20590488314628602,
      "num_tokens": 35995004.0,
      "step": 19305
    },
    {
      "entropy": 5.860301446914673,
      "epoch": 1.7052278346873897,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047109078239169946,
      "loss": 5.3819,
      "mean_token_accuracy": 0.19609652310609818,
      "num_tokens": 36004786.0,
      "step": 19310
    },
    {
      "entropy": 5.814460706710816,
      "epoch": 1.70566937477923,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004710753379945415,
      "loss": 5.3075,
      "mean_token_accuracy": 0.2170813649892807,
      "num_tokens": 36014047.0,
      "step": 19315
    },
    {
      "entropy": 5.8558917999267575,
      "epoch": 1.7061109148710702,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047105988975637927,
      "loss": 5.3342,
      "mean_token_accuracy": 0.21030336767435073,
      "num_tokens": 36022921.0,
      "step": 19320
    },
    {
      "entropy": 5.8279869079589846,
      "epoch": 1.7065524549629107,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004710444376775153,
      "loss": 5.3297,
      "mean_token_accuracy": 0.20479645133018493,
      "num_tokens": 36032713.0,
      "step": 19325
    },
    {
      "entropy": 5.837489318847656,
      "epoch": 1.706993995054751,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047102898175825224,
      "loss": 5.3658,
      "mean_token_accuracy": 0.20159162431955338,
      "num_tokens": 36041595.0,
      "step": 19330
    },
    {
      "entropy": 5.861054944992065,
      "epoch": 1.7074355351465913,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047101352199889295,
      "loss": 5.3502,
      "mean_token_accuracy": 0.19800026714801788,
      "num_tokens": 36051993.0,
      "step": 19335
    },
    {
      "entropy": 5.784552431106567,
      "epoch": 1.7078770752384318,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004709980583997402,
      "loss": 5.3483,
      "mean_token_accuracy": 0.20705602020025254,
      "num_tokens": 36061806.0,
      "step": 19340
    },
    {
      "entropy": 5.8445803165435795,
      "epoch": 1.7083186153302719,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000470982590961097,
      "loss": 5.3988,
      "mean_token_accuracy": 0.20039329826831817,
      "num_tokens": 36071390.0,
      "step": 19345
    },
    {
      "entropy": 5.84760890007019,
      "epoch": 1.7087601554221123,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047096711968326626,
      "loss": 5.3843,
      "mean_token_accuracy": 0.20336378067731858,
      "num_tokens": 36080561.0,
      "step": 19350
    },
    {
      "entropy": 5.865473937988281,
      "epoch": 1.7092016955139526,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004709516445665512,
      "loss": 5.3233,
      "mean_token_accuracy": 0.20303022712469102,
      "num_tokens": 36089568.0,
      "step": 19355
    },
    {
      "entropy": 5.825430583953858,
      "epoch": 1.709643235605793,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004709361656112547,
      "loss": 5.2948,
      "mean_token_accuracy": 0.213546322286129,
      "num_tokens": 36098881.0,
      "step": 19360
    },
    {
      "entropy": 5.835883045196534,
      "epoch": 1.7100847756976334,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004709206828176804,
      "loss": 5.3588,
      "mean_token_accuracy": 0.21046178936958312,
      "num_tokens": 36108209.0,
      "step": 19365
    },
    {
      "entropy": 5.8493904113769535,
      "epoch": 1.7105263157894737,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004709051961861311,
      "loss": 5.3865,
      "mean_token_accuracy": 0.20557094365358353,
      "num_tokens": 36117573.0,
      "step": 19370
    },
    {
      "entropy": 5.861238622665406,
      "epoch": 1.710967855881314,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004708897057169105,
      "loss": 5.3747,
      "mean_token_accuracy": 0.20894356369972228,
      "num_tokens": 36127405.0,
      "step": 19375
    },
    {
      "entropy": 5.935452270507812,
      "epoch": 1.7114093959731544,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004708742114103221,
      "loss": 5.3765,
      "mean_token_accuracy": 0.2082043021917343,
      "num_tokens": 36136148.0,
      "step": 19380
    },
    {
      "entropy": 5.837888240814209,
      "epoch": 1.7118509360649947,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004708587132666692,
      "loss": 5.375,
      "mean_token_accuracy": 0.20259551852941513,
      "num_tokens": 36146164.0,
      "step": 19385
    },
    {
      "entropy": 5.863293027877807,
      "epoch": 1.712292476156835,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047084321128625545,
      "loss": 5.3689,
      "mean_token_accuracy": 0.1976170539855957,
      "num_tokens": 36155313.0,
      "step": 19390
    },
    {
      "entropy": 5.822325849533081,
      "epoch": 1.7127340162486755,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047082770546938457,
      "loss": 5.3739,
      "mean_token_accuracy": 0.2107451856136322,
      "num_tokens": 36164902.0,
      "step": 19395
    },
    {
      "entropy": 5.869547176361084,
      "epoch": 1.7131755563405158,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004708121958163604,
      "loss": 5.3986,
      "mean_token_accuracy": 0.21294075548648833,
      "num_tokens": 36174128.0,
      "step": 19400
    },
    {
      "entropy": 5.898516464233398,
      "epoch": 1.713617096432356,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004707966823274866,
      "loss": 5.4469,
      "mean_token_accuracy": 0.19630146622657776,
      "num_tokens": 36183084.0,
      "step": 19405
    },
    {
      "entropy": 5.932181167602539,
      "epoch": 1.7140586365241965,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047078116500306723,
      "loss": 5.4008,
      "mean_token_accuracy": 0.20068621933460234,
      "num_tokens": 36191650.0,
      "step": 19410
    },
    {
      "entropy": 5.870257472991943,
      "epoch": 1.7145001766160366,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047076564384340616,
      "loss": 5.3869,
      "mean_token_accuracy": 0.2046503245830536,
      "num_tokens": 36200712.0,
      "step": 19415
    },
    {
      "entropy": 5.818039178848267,
      "epoch": 1.714941716707877,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004707501188488074,
      "loss": 5.3581,
      "mean_token_accuracy": 0.2052019789814949,
      "num_tokens": 36209347.0,
      "step": 19420
    },
    {
      "entropy": 5.869766092300415,
      "epoch": 1.7153832567997174,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047073459001957524,
      "loss": 5.3546,
      "mean_token_accuracy": 0.20100276470184325,
      "num_tokens": 36218233.0,
      "step": 19425
    },
    {
      "entropy": 5.930176305770874,
      "epoch": 1.7158247968915576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004707190573560137,
      "loss": 5.4033,
      "mean_token_accuracy": 0.2007612481713295,
      "num_tokens": 36227627.0,
      "step": 19430
    },
    {
      "entropy": 5.843554735183716,
      "epoch": 1.7162663369833981,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004707035208584272,
      "loss": 5.3669,
      "mean_token_accuracy": 0.2091110333800316,
      "num_tokens": 36236298.0,
      "step": 19435
    },
    {
      "entropy": 5.8809840202331545,
      "epoch": 1.7167078770752384,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047068798052711996,
      "loss": 5.4268,
      "mean_token_accuracy": 0.20471113473176955,
      "num_tokens": 36245623.0,
      "step": 19440
    },
    {
      "entropy": 5.8725379467010494,
      "epoch": 1.7171494171670787,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004706724363623965,
      "loss": 5.2879,
      "mean_token_accuracy": 0.21204689890146255,
      "num_tokens": 36255117.0,
      "step": 19445
    },
    {
      "entropy": 5.809520244598389,
      "epoch": 1.7175909572589192,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047065688836456134,
      "loss": 5.3386,
      "mean_token_accuracy": 0.21131264716386794,
      "num_tokens": 36264535.0,
      "step": 19450
    },
    {
      "entropy": 5.877058744430542,
      "epoch": 1.7180324973507595,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047064133653391895,
      "loss": 5.3843,
      "mean_token_accuracy": 0.20071642249822616,
      "num_tokens": 36274432.0,
      "step": 19455
    },
    {
      "entropy": 5.84521894454956,
      "epoch": 1.7184740374425997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004706257808707741,
      "loss": 5.2955,
      "mean_token_accuracy": 0.21847237795591354,
      "num_tokens": 36282644.0,
      "step": 19460
    },
    {
      "entropy": 5.839371538162231,
      "epoch": 1.7189155775344402,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047061022137543144,
      "loss": 5.3538,
      "mean_token_accuracy": 0.20643906444311141,
      "num_tokens": 36290885.0,
      "step": 19465
    },
    {
      "entropy": 5.913707494735718,
      "epoch": 1.7193571176262805,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047059465804819576,
      "loss": 5.4586,
      "mean_token_accuracy": 0.20248591303825378,
      "num_tokens": 36299477.0,
      "step": 19470
    },
    {
      "entropy": 5.84574408531189,
      "epoch": 1.7197986577181208,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004705790908893721,
      "loss": 5.4272,
      "mean_token_accuracy": 0.20245751589536667,
      "num_tokens": 36309638.0,
      "step": 19475
    },
    {
      "entropy": 5.9297737121582035,
      "epoch": 1.7202401978099613,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004705635198992651,
      "loss": 5.4782,
      "mean_token_accuracy": 0.19718853384256363,
      "num_tokens": 36319027.0,
      "step": 19480
    },
    {
      "entropy": 5.856172370910644,
      "epoch": 1.7206817379018013,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047054794507818,
      "loss": 5.3223,
      "mean_token_accuracy": 0.19970888942480086,
      "num_tokens": 36327936.0,
      "step": 19485
    },
    {
      "entropy": 5.900039005279541,
      "epoch": 1.7211232779936418,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047053236642642193,
      "loss": 5.5262,
      "mean_token_accuracy": 0.19295619279146195,
      "num_tokens": 36337162.0,
      "step": 19490
    },
    {
      "entropy": 5.8792760372161865,
      "epoch": 1.7215648180854821,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000470516783944296,
      "loss": 5.3747,
      "mean_token_accuracy": 0.2089792400598526,
      "num_tokens": 36345587.0,
      "step": 19495
    },
    {
      "entropy": 5.868822526931763,
      "epoch": 1.7220063581773224,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047050119763210747,
      "loss": 5.4098,
      "mean_token_accuracy": 0.2047387808561325,
      "num_tokens": 36354536.0,
      "step": 19500
    },
    {
      "entropy": 5.872997760772705,
      "epoch": 1.722447898269163,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047048560749016167,
      "loss": 5.4632,
      "mean_token_accuracy": 0.19283899962902068,
      "num_tokens": 36363691.0,
      "step": 19505
    },
    {
      "entropy": 5.9692363262176515,
      "epoch": 1.7228894383610032,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047047001351876403,
      "loss": 5.4985,
      "mean_token_accuracy": 0.19898141473531722,
      "num_tokens": 36373898.0,
      "step": 19510
    },
    {
      "entropy": 5.9161478042602536,
      "epoch": 1.7233309784528434,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004704544157182199,
      "loss": 5.4788,
      "mean_token_accuracy": 0.19759699404239656,
      "num_tokens": 36384345.0,
      "step": 19515
    },
    {
      "entropy": 5.849001693725586,
      "epoch": 1.723772518544684,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047043881408883497,
      "loss": 5.361,
      "mean_token_accuracy": 0.2048543244600296,
      "num_tokens": 36393021.0,
      "step": 19520
    },
    {
      "entropy": 5.940569639205933,
      "epoch": 1.7242140586365242,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004704232086309148,
      "loss": 5.357,
      "mean_token_accuracy": 0.20328376740217208,
      "num_tokens": 36402729.0,
      "step": 19525
    },
    {
      "entropy": 5.796016836166382,
      "epoch": 1.7246555987283645,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047040759934476517,
      "loss": 5.2884,
      "mean_token_accuracy": 0.20681367069482803,
      "num_tokens": 36412360.0,
      "step": 19530
    },
    {
      "entropy": 5.779270648956299,
      "epoch": 1.725097138820205,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004703919862306918,
      "loss": 5.3021,
      "mean_token_accuracy": 0.20544631779193878,
      "num_tokens": 36420245.0,
      "step": 19535
    },
    {
      "entropy": 5.821827411651611,
      "epoch": 1.7255386789120453,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004703763692890006,
      "loss": 5.2665,
      "mean_token_accuracy": 0.21545811444520951,
      "num_tokens": 36428758.0,
      "step": 19540
    },
    {
      "entropy": 5.86518120765686,
      "epoch": 1.7259802190038855,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004703607485199973,
      "loss": 5.3507,
      "mean_token_accuracy": 0.20525264889001846,
      "num_tokens": 36437883.0,
      "step": 19545
    },
    {
      "entropy": 5.83111162185669,
      "epoch": 1.726421759095726,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047034512392398815,
      "loss": 5.3521,
      "mean_token_accuracy": 0.20012767761945724,
      "num_tokens": 36447950.0,
      "step": 19550
    },
    {
      "entropy": 5.891711902618408,
      "epoch": 1.726863299187566,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004703294955012791,
      "loss": 5.3968,
      "mean_token_accuracy": 0.20180790722370148,
      "num_tokens": 36456993.0,
      "step": 19555
    },
    {
      "entropy": 5.908048009872436,
      "epoch": 1.7273048392794066,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004703138632521762,
      "loss": 5.4438,
      "mean_token_accuracy": 0.1943171054124832,
      "num_tokens": 36465942.0,
      "step": 19560
    },
    {
      "entropy": 5.8576130867004395,
      "epoch": 1.7277463793712469,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000470298227176986,
      "loss": 5.4374,
      "mean_token_accuracy": 0.19999451488256453,
      "num_tokens": 36474906.0,
      "step": 19565
    },
    {
      "entropy": 5.904548978805542,
      "epoch": 1.7281879194630871,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004702825872760145,
      "loss": 5.3558,
      "mean_token_accuracy": 0.20032448470592498,
      "num_tokens": 36483579.0,
      "step": 19570
    },
    {
      "entropy": 5.862172698974609,
      "epoch": 1.7286294595549276,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004702669435495681,
      "loss": 5.3593,
      "mean_token_accuracy": 0.20729408860206605,
      "num_tokens": 36492541.0,
      "step": 19575
    },
    {
      "entropy": 5.898200178146363,
      "epoch": 1.729070999646768,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004702512959979535,
      "loss": 5.4942,
      "mean_token_accuracy": 0.19613170325756074,
      "num_tokens": 36502833.0,
      "step": 19580
    },
    {
      "entropy": 5.887970066070556,
      "epoch": 1.7295125397386082,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00047023564462147697,
      "loss": 5.3655,
      "mean_token_accuracy": 0.2045273721218109,
      "num_tokens": 36513091.0,
      "step": 19585
    },
    {
      "entropy": 5.830621767044067,
      "epoch": 1.7299540798304487,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047021998942044517,
      "loss": 5.3695,
      "mean_token_accuracy": 0.20839569270610808,
      "num_tokens": 36522911.0,
      "step": 19590
    },
    {
      "entropy": 5.887117099761963,
      "epoch": 1.730395619922289,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047020433039516484,
      "loss": 5.4355,
      "mean_token_accuracy": 0.20266944468021392,
      "num_tokens": 36531934.0,
      "step": 19595
    },
    {
      "entropy": 5.859900617599488,
      "epoch": 1.7308371600141292,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004701886675459427,
      "loss": 5.3646,
      "mean_token_accuracy": 0.20867136120796204,
      "num_tokens": 36541823.0,
      "step": 19600
    },
    {
      "entropy": 5.910891008377075,
      "epoch": 1.7312787001059697,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047017300087308557,
      "loss": 5.4687,
      "mean_token_accuracy": 0.1963314786553383,
      "num_tokens": 36550954.0,
      "step": 19605
    },
    {
      "entropy": 5.871105909347534,
      "epoch": 1.73172024019781,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004701573303769004,
      "loss": 5.3638,
      "mean_token_accuracy": 0.20587648451328278,
      "num_tokens": 36560689.0,
      "step": 19610
    },
    {
      "entropy": 5.8392587184906,
      "epoch": 1.7321617802896503,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000470141656057694,
      "loss": 5.3521,
      "mean_token_accuracy": 0.2004813715815544,
      "num_tokens": 36569587.0,
      "step": 19615
    },
    {
      "entropy": 5.83994312286377,
      "epoch": 1.7326033203814908,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00047012597791577364,
      "loss": 5.3763,
      "mean_token_accuracy": 0.2069721922278404,
      "num_tokens": 36579609.0,
      "step": 19620
    },
    {
      "entropy": 5.857770109176636,
      "epoch": 1.7330448604733308,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004701102959514463,
      "loss": 5.3849,
      "mean_token_accuracy": 0.20199847519397734,
      "num_tokens": 36589685.0,
      "step": 19625
    },
    {
      "entropy": 5.89018497467041,
      "epoch": 1.7334864005651713,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047009461016501926,
      "loss": 5.339,
      "mean_token_accuracy": 0.19970999509096146,
      "num_tokens": 36598225.0,
      "step": 19630
    },
    {
      "entropy": 5.759757232666016,
      "epoch": 1.7339279406570116,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047007892055679976,
      "loss": 5.2741,
      "mean_token_accuracy": 0.2128346160054207,
      "num_tokens": 36607481.0,
      "step": 19635
    },
    {
      "entropy": 5.864145040512085,
      "epoch": 1.7343694807488519,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047006322712709513,
      "loss": 5.3642,
      "mean_token_accuracy": 0.2028595194220543,
      "num_tokens": 36616707.0,
      "step": 19640
    },
    {
      "entropy": 5.926380443572998,
      "epoch": 1.7348110208406924,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047004752987621286,
      "loss": 5.4997,
      "mean_token_accuracy": 0.19786589741706848,
      "num_tokens": 36626290.0,
      "step": 19645
    },
    {
      "entropy": 5.85288290977478,
      "epoch": 1.7352525609325327,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004700318288044603,
      "loss": 5.3845,
      "mean_token_accuracy": 0.20945847332477568,
      "num_tokens": 36635155.0,
      "step": 19650
    },
    {
      "entropy": 5.878585720062256,
      "epoch": 1.735694101024373,
      "grad_norm": 0.984375,
      "learning_rate": 0.00047001612391214527,
      "loss": 5.3789,
      "mean_token_accuracy": 0.20974762439727784,
      "num_tokens": 36644632.0,
      "step": 19655
    },
    {
      "entropy": 5.898300361633301,
      "epoch": 1.7361356411162134,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004700004151995753,
      "loss": 5.4212,
      "mean_token_accuracy": 0.19723953157663346,
      "num_tokens": 36654818.0,
      "step": 19660
    },
    {
      "entropy": 5.8098838329315186,
      "epoch": 1.7365771812080537,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046998470266705795,
      "loss": 5.2586,
      "mean_token_accuracy": 0.2128022387623787,
      "num_tokens": 36664149.0,
      "step": 19665
    },
    {
      "entropy": 5.85724892616272,
      "epoch": 1.737018721299894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046996898631490125,
      "loss": 5.3207,
      "mean_token_accuracy": 0.21209775507450104,
      "num_tokens": 36673447.0,
      "step": 19670
    },
    {
      "entropy": 5.989524602890015,
      "epoch": 1.7374602613917345,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000469953266143413,
      "loss": 5.5064,
      "mean_token_accuracy": 0.1909137487411499,
      "num_tokens": 36683426.0,
      "step": 19675
    },
    {
      "entropy": 5.9506079196929935,
      "epoch": 1.7379018014835748,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046993754215290116,
      "loss": 5.3937,
      "mean_token_accuracy": 0.2005843475461006,
      "num_tokens": 36692708.0,
      "step": 19680
    },
    {
      "entropy": 5.7777101516723635,
      "epoch": 1.738343341575415,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004699218143436737,
      "loss": 5.2959,
      "mean_token_accuracy": 0.20974681824445723,
      "num_tokens": 36701527.0,
      "step": 19685
    },
    {
      "entropy": 5.869098091125489,
      "epoch": 1.7387848816672555,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046990608271603874,
      "loss": 5.4436,
      "mean_token_accuracy": 0.201623073220253,
      "num_tokens": 36711458.0,
      "step": 19690
    },
    {
      "entropy": 5.8512145519256595,
      "epoch": 1.7392264217590956,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004698903472703045,
      "loss": 5.3629,
      "mean_token_accuracy": 0.20296697765588761,
      "num_tokens": 36721260.0,
      "step": 19695
    },
    {
      "entropy": 5.909906721115112,
      "epoch": 1.739667961850936,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046987460800677915,
      "loss": 5.3755,
      "mean_token_accuracy": 0.2055405154824257,
      "num_tokens": 36730496.0,
      "step": 19700
    },
    {
      "entropy": 5.840519809722901,
      "epoch": 1.7401095019427764,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046985886492577106,
      "loss": 5.337,
      "mean_token_accuracy": 0.20332951694726945,
      "num_tokens": 36739016.0,
      "step": 19705
    },
    {
      "entropy": 5.922931909561157,
      "epoch": 1.7405510420346166,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004698431180275887,
      "loss": 5.4611,
      "mean_token_accuracy": 0.19510802030563354,
      "num_tokens": 36748259.0,
      "step": 19710
    },
    {
      "entropy": 6.0068436622619625,
      "epoch": 1.7409925821264571,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046982736731254044,
      "loss": 5.4818,
      "mean_token_accuracy": 0.19545356631278993,
      "num_tokens": 36758174.0,
      "step": 19715
    },
    {
      "entropy": 5.9150248050689695,
      "epoch": 1.7414341222182974,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046981161278093487,
      "loss": 5.4696,
      "mean_token_accuracy": 0.19996329993009568,
      "num_tokens": 36767286.0,
      "step": 19720
    },
    {
      "entropy": 5.8429395198822025,
      "epoch": 1.7418756623101377,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046979585443308056,
      "loss": 5.3046,
      "mean_token_accuracy": 0.2088826209306717,
      "num_tokens": 36775818.0,
      "step": 19725
    },
    {
      "entropy": 5.806203556060791,
      "epoch": 1.7423172024019782,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046978009226928633,
      "loss": 5.3612,
      "mean_token_accuracy": 0.20912386924028398,
      "num_tokens": 36785489.0,
      "step": 19730
    },
    {
      "entropy": 5.8916933059692385,
      "epoch": 1.7427587424938185,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004697643262898608,
      "loss": 5.379,
      "mean_token_accuracy": 0.20245468020439147,
      "num_tokens": 36794241.0,
      "step": 19735
    },
    {
      "entropy": 5.917622709274292,
      "epoch": 1.7432002825856587,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004697485564951128,
      "loss": 5.3809,
      "mean_token_accuracy": 0.2006335362792015,
      "num_tokens": 36803477.0,
      "step": 19740
    },
    {
      "entropy": 5.871290874481201,
      "epoch": 1.7436418226774992,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004697327828853515,
      "loss": 5.3756,
      "mean_token_accuracy": 0.20136721283197404,
      "num_tokens": 36811974.0,
      "step": 19745
    },
    {
      "entropy": 5.881715536117554,
      "epoch": 1.7440833627693395,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004697170054608857,
      "loss": 5.4335,
      "mean_token_accuracy": 0.20168333500623703,
      "num_tokens": 36822049.0,
      "step": 19750
    },
    {
      "entropy": 5.893443298339844,
      "epoch": 1.7445249028611798,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004697012242220245,
      "loss": 5.374,
      "mean_token_accuracy": 0.20690135806798934,
      "num_tokens": 36831989.0,
      "step": 19755
    },
    {
      "entropy": 5.902186012268066,
      "epoch": 1.7449664429530203,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046968543916907704,
      "loss": 5.409,
      "mean_token_accuracy": 0.1928035080432892,
      "num_tokens": 36841141.0,
      "step": 19760
    },
    {
      "entropy": 5.828170442581177,
      "epoch": 1.7454079830448603,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046966965030235253,
      "loss": 5.2616,
      "mean_token_accuracy": 0.2130754292011261,
      "num_tokens": 36850101.0,
      "step": 19765
    },
    {
      "entropy": 5.850824928283691,
      "epoch": 1.7458495231367008,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046965385762216024,
      "loss": 5.3773,
      "mean_token_accuracy": 0.2056165024638176,
      "num_tokens": 36859234.0,
      "step": 19770
    },
    {
      "entropy": 5.952357816696167,
      "epoch": 1.746291063228541,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004696380611288096,
      "loss": 5.4591,
      "mean_token_accuracy": 0.20363277941942215,
      "num_tokens": 36868937.0,
      "step": 19775
    },
    {
      "entropy": 5.823459577560425,
      "epoch": 1.7467326033203814,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046962226082261014,
      "loss": 5.2983,
      "mean_token_accuracy": 0.20970505475997925,
      "num_tokens": 36878380.0,
      "step": 19780
    },
    {
      "entropy": 5.855659341812133,
      "epoch": 1.7471741434122219,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004696064567038712,
      "loss": 5.4302,
      "mean_token_accuracy": 0.2020963564515114,
      "num_tokens": 36887702.0,
      "step": 19785
    },
    {
      "entropy": 5.877581596374512,
      "epoch": 1.7476156835040622,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046959064877290246,
      "loss": 5.2773,
      "mean_token_accuracy": 0.2190902128815651,
      "num_tokens": 36896777.0,
      "step": 19790
    },
    {
      "entropy": 5.896066093444825,
      "epoch": 1.7480572235959024,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046957483703001366,
      "loss": 5.3274,
      "mean_token_accuracy": 0.20655572563409805,
      "num_tokens": 36905256.0,
      "step": 19795
    },
    {
      "entropy": 5.838788747787476,
      "epoch": 1.748498763687743,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004695590214755143,
      "loss": 5.3537,
      "mean_token_accuracy": 0.20737936347723007,
      "num_tokens": 36913767.0,
      "step": 19800
    },
    {
      "entropy": 5.82216157913208,
      "epoch": 1.7489403037795832,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046954320210971444,
      "loss": 5.3378,
      "mean_token_accuracy": 0.21220385134220124,
      "num_tokens": 36922838.0,
      "step": 19805
    },
    {
      "entropy": 5.869567346572876,
      "epoch": 1.7493818438714235,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046952737893292396,
      "loss": 5.2205,
      "mean_token_accuracy": 0.21445235013961791,
      "num_tokens": 36931955.0,
      "step": 19810
    },
    {
      "entropy": 5.818287324905396,
      "epoch": 1.749823383963264,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046951155194545265,
      "loss": 5.3198,
      "mean_token_accuracy": 0.20727813988924026,
      "num_tokens": 36941505.0,
      "step": 19815
    },
    {
      "entropy": 5.858857727050781,
      "epoch": 1.7502649240551043,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046949572114761076,
      "loss": 5.4319,
      "mean_token_accuracy": 0.19525412917137147,
      "num_tokens": 36951007.0,
      "step": 19820
    },
    {
      "entropy": 5.891967630386352,
      "epoch": 1.7507064641469445,
      "grad_norm": 0.9375,
      "learning_rate": 0.00046947988653970826,
      "loss": 5.3919,
      "mean_token_accuracy": 0.21053603440523147,
      "num_tokens": 36961111.0,
      "step": 19825
    },
    {
      "entropy": 5.883198499679565,
      "epoch": 1.751148004238785,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004694640481220554,
      "loss": 5.4018,
      "mean_token_accuracy": 0.20545209646224977,
      "num_tokens": 36969605.0,
      "step": 19830
    },
    {
      "entropy": 5.902576494216919,
      "epoch": 1.751589544330625,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004694482058949624,
      "loss": 5.3924,
      "mean_token_accuracy": 0.1992309421300888,
      "num_tokens": 36979051.0,
      "step": 19835
    },
    {
      "entropy": 5.822027969360351,
      "epoch": 1.7520310844224656,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046943235985873975,
      "loss": 5.3414,
      "mean_token_accuracy": 0.20949007272720338,
      "num_tokens": 36988807.0,
      "step": 19840
    },
    {
      "entropy": 5.864878177642822,
      "epoch": 1.7524726245143059,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004694165100136976,
      "loss": 5.2909,
      "mean_token_accuracy": 0.21140781044960022,
      "num_tokens": 36997813.0,
      "step": 19845
    },
    {
      "entropy": 5.89201397895813,
      "epoch": 1.7529141646061461,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046940065636014665,
      "loss": 5.3421,
      "mean_token_accuracy": 0.20457070618867873,
      "num_tokens": 37007849.0,
      "step": 19850
    },
    {
      "entropy": 5.815494537353516,
      "epoch": 1.7533557046979866,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004693847988983974,
      "loss": 5.2468,
      "mean_token_accuracy": 0.21503526866436004,
      "num_tokens": 37016109.0,
      "step": 19855
    },
    {
      "entropy": 5.831973314285278,
      "epoch": 1.753797244789827,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004693689376287605,
      "loss": 5.3692,
      "mean_token_accuracy": 0.19793687313795089,
      "num_tokens": 37025156.0,
      "step": 19860
    },
    {
      "entropy": 5.887754487991333,
      "epoch": 1.7542387848816672,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046935307255154663,
      "loss": 5.4666,
      "mean_token_accuracy": 0.19373612403869628,
      "num_tokens": 37035361.0,
      "step": 19865
    },
    {
      "entropy": 5.93493971824646,
      "epoch": 1.7546803249735077,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004693372036670667,
      "loss": 5.3973,
      "mean_token_accuracy": 0.1965683400630951,
      "num_tokens": 37045249.0,
      "step": 19870
    },
    {
      "entropy": 5.965833044052124,
      "epoch": 1.755121865065348,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046932133097563137,
      "loss": 5.4105,
      "mean_token_accuracy": 0.20748149901628493,
      "num_tokens": 37054637.0,
      "step": 19875
    },
    {
      "entropy": 5.883833026885986,
      "epoch": 1.7555634051571882,
      "grad_norm": 1.0,
      "learning_rate": 0.00046930545447755167,
      "loss": 5.4061,
      "mean_token_accuracy": 0.20418444871902466,
      "num_tokens": 37064557.0,
      "step": 19880
    },
    {
      "entropy": 5.870358991622925,
      "epoch": 1.7560049452490287,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004692895741731387,
      "loss": 5.4176,
      "mean_token_accuracy": 0.1922706186771393,
      "num_tokens": 37073915.0,
      "step": 19885
    },
    {
      "entropy": 5.920988941192627,
      "epoch": 1.756446485340869,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004692736900627034,
      "loss": 5.3856,
      "mean_token_accuracy": 0.19981521517038345,
      "num_tokens": 37083008.0,
      "step": 19890
    },
    {
      "entropy": 5.937457704544068,
      "epoch": 1.7568880254327093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046925780214655713,
      "loss": 5.4136,
      "mean_token_accuracy": 0.20175595134496688,
      "num_tokens": 37092742.0,
      "step": 19895
    },
    {
      "entropy": 5.934474658966065,
      "epoch": 1.7573295655245498,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004692419104250108,
      "loss": 5.4582,
      "mean_token_accuracy": 0.19153337329626083,
      "num_tokens": 37102242.0,
      "step": 19900
    },
    {
      "entropy": 5.846590995788574,
      "epoch": 1.7577711056163898,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004692260148983761,
      "loss": 5.3376,
      "mean_token_accuracy": 0.21341789662837982,
      "num_tokens": 37111304.0,
      "step": 19905
    },
    {
      "entropy": 5.873947095870972,
      "epoch": 1.7582126457082303,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004692101155669641,
      "loss": 5.3434,
      "mean_token_accuracy": 0.20788217186927796,
      "num_tokens": 37121420.0,
      "step": 19910
    },
    {
      "entropy": 5.9507365226745605,
      "epoch": 1.7586541858000706,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004691942124310865,
      "loss": 5.4119,
      "mean_token_accuracy": 0.20299457162618637,
      "num_tokens": 37131265.0,
      "step": 19915
    },
    {
      "entropy": 5.830027723312378,
      "epoch": 1.7590957258919109,
      "grad_norm": 0.953125,
      "learning_rate": 0.00046917830549105465,
      "loss": 5.3081,
      "mean_token_accuracy": 0.21015240848064423,
      "num_tokens": 37141097.0,
      "step": 19920
    },
    {
      "entropy": 5.721772575378418,
      "epoch": 1.7595372659837514,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004691623947471803,
      "loss": 5.2226,
      "mean_token_accuracy": 0.21801266819238663,
      "num_tokens": 37149687.0,
      "step": 19925
    },
    {
      "entropy": 5.851511716842651,
      "epoch": 1.7599788060755917,
      "grad_norm": 1.0,
      "learning_rate": 0.00046914648019977495,
      "loss": 5.3536,
      "mean_token_accuracy": 0.2111783504486084,
      "num_tokens": 37159219.0,
      "step": 19930
    },
    {
      "entropy": 5.841753673553467,
      "epoch": 1.760420346167432,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046913056184915054,
      "loss": 5.3601,
      "mean_token_accuracy": 0.20732667595148085,
      "num_tokens": 37168681.0,
      "step": 19935
    },
    {
      "entropy": 5.7493514060974125,
      "epoch": 1.7608618862592724,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004691146396956188,
      "loss": 5.2552,
      "mean_token_accuracy": 0.21570795625448227,
      "num_tokens": 37177932.0,
      "step": 19940
    },
    {
      "entropy": 5.844285154342652,
      "epoch": 1.7613034263511127,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004690987137394917,
      "loss": 5.2732,
      "mean_token_accuracy": 0.20587903410196304,
      "num_tokens": 37186783.0,
      "step": 19945
    },
    {
      "entropy": 5.902088165283203,
      "epoch": 1.761744966442953,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004690827839810812,
      "loss": 5.4538,
      "mean_token_accuracy": 0.19667417407035828,
      "num_tokens": 37196290.0,
      "step": 19950
    },
    {
      "entropy": 5.8575023174285885,
      "epoch": 1.7621865065347935,
      "grad_norm": 1.125,
      "learning_rate": 0.00046906685042069927,
      "loss": 5.3296,
      "mean_token_accuracy": 0.20638182163238525,
      "num_tokens": 37205057.0,
      "step": 19955
    },
    {
      "entropy": 5.8945387840271,
      "epoch": 1.7626280466266337,
      "grad_norm": 1.0,
      "learning_rate": 0.0004690509130586582,
      "loss": 5.3757,
      "mean_token_accuracy": 0.20165475010871886,
      "num_tokens": 37214078.0,
      "step": 19960
    },
    {
      "entropy": 5.828520059585571,
      "epoch": 1.763069586718474,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046903497189527006,
      "loss": 5.3298,
      "mean_token_accuracy": 0.2077646389603615,
      "num_tokens": 37223232.0,
      "step": 19965
    },
    {
      "entropy": 5.844703960418701,
      "epoch": 1.7635111268103145,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004690190269308472,
      "loss": 5.3322,
      "mean_token_accuracy": 0.20792658030986785,
      "num_tokens": 37232054.0,
      "step": 19970
    },
    {
      "entropy": 5.786932897567749,
      "epoch": 1.7639526669021546,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004690030781657019,
      "loss": 5.293,
      "mean_token_accuracy": 0.21082770228385925,
      "num_tokens": 37241451.0,
      "step": 19975
    },
    {
      "entropy": 5.85056791305542,
      "epoch": 1.764394206993995,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004689871256001467,
      "loss": 5.3471,
      "mean_token_accuracy": 0.20343182533979415,
      "num_tokens": 37251029.0,
      "step": 19980
    },
    {
      "entropy": 5.8258668899536135,
      "epoch": 1.7648357470858353,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046897116923449403,
      "loss": 5.3484,
      "mean_token_accuracy": 0.2111079141497612,
      "num_tokens": 37259594.0,
      "step": 19985
    },
    {
      "entropy": 5.808517980575561,
      "epoch": 1.7652772871776756,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046895520906905635,
      "loss": 5.2942,
      "mean_token_accuracy": 0.2086445465683937,
      "num_tokens": 37268824.0,
      "step": 19990
    },
    {
      "entropy": 5.873378944396973,
      "epoch": 1.7657188272695161,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046893924510414654,
      "loss": 5.3428,
      "mean_token_accuracy": 0.20504022538661956,
      "num_tokens": 37278494.0,
      "step": 19995
    },
    {
      "entropy": 5.713855457305908,
      "epoch": 1.7661603673613564,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004689232773400772,
      "loss": 5.164,
      "mean_token_accuracy": 0.23042462021112442,
      "num_tokens": 37287648.0,
      "step": 20000
    },
    {
      "entropy": 5.843976974487305,
      "epoch": 1.7666019074531967,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004689073057771611,
      "loss": 5.3266,
      "mean_token_accuracy": 0.2016871228814125,
      "num_tokens": 37295938.0,
      "step": 20005
    },
    {
      "entropy": 5.829558897018432,
      "epoch": 1.7670434475450372,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046889133041571126,
      "loss": 5.3263,
      "mean_token_accuracy": 0.2090665102005005,
      "num_tokens": 37305097.0,
      "step": 20010
    },
    {
      "entropy": 5.809263563156128,
      "epoch": 1.7674849876368774,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004688753512560405,
      "loss": 5.3465,
      "mean_token_accuracy": 0.20816302597522734,
      "num_tokens": 37313672.0,
      "step": 20015
    },
    {
      "entropy": 5.83147964477539,
      "epoch": 1.7679265277287177,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004688593682984619,
      "loss": 5.4079,
      "mean_token_accuracy": 0.20093484371900558,
      "num_tokens": 37323342.0,
      "step": 20020
    },
    {
      "entropy": 5.878684139251709,
      "epoch": 1.7683680678205582,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004688433815432884,
      "loss": 5.3675,
      "mean_token_accuracy": 0.20378257781267167,
      "num_tokens": 37332578.0,
      "step": 20025
    },
    {
      "entropy": 5.928590631484985,
      "epoch": 1.7688096079123985,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046882739099083335,
      "loss": 5.3809,
      "mean_token_accuracy": 0.1996745079755783,
      "num_tokens": 37341769.0,
      "step": 20030
    },
    {
      "entropy": 5.929939413070679,
      "epoch": 1.7692511480042388,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046881139664141006,
      "loss": 5.4369,
      "mean_token_accuracy": 0.19845349639654158,
      "num_tokens": 37351158.0,
      "step": 20035
    },
    {
      "entropy": 5.881866216659546,
      "epoch": 1.7696926880960793,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004687953984953316,
      "loss": 5.3704,
      "mean_token_accuracy": 0.20912910103797913,
      "num_tokens": 37360924.0,
      "step": 20040
    },
    {
      "entropy": 5.798215103149414,
      "epoch": 1.7701342281879193,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004687793965529115,
      "loss": 5.3142,
      "mean_token_accuracy": 0.21615274101495743,
      "num_tokens": 37369295.0,
      "step": 20045
    },
    {
      "entropy": 5.812409973144531,
      "epoch": 1.7705757682797598,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004687633908144633,
      "loss": 5.3617,
      "mean_token_accuracy": 0.20486588031053543,
      "num_tokens": 37379135.0,
      "step": 20050
    },
    {
      "entropy": 5.833855152130127,
      "epoch": 1.7710173083716,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046874738128030045,
      "loss": 5.2546,
      "mean_token_accuracy": 0.21422645896673204,
      "num_tokens": 37387927.0,
      "step": 20055
    },
    {
      "entropy": 5.776106119155884,
      "epoch": 1.7714588484634404,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004687313679507366,
      "loss": 5.2944,
      "mean_token_accuracy": 0.21309812664985656,
      "num_tokens": 37397629.0,
      "step": 20060
    },
    {
      "entropy": 5.78966007232666,
      "epoch": 1.7719003885552809,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046871535082608544,
      "loss": 5.3182,
      "mean_token_accuracy": 0.20881196856498718,
      "num_tokens": 37407688.0,
      "step": 20065
    },
    {
      "entropy": 5.893298101425171,
      "epoch": 1.7723419286471211,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004686993299066607,
      "loss": 5.2992,
      "mean_token_accuracy": 0.2105863317847252,
      "num_tokens": 37417019.0,
      "step": 20070
    },
    {
      "entropy": 5.867874240875244,
      "epoch": 1.7727834687389614,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004686833051927763,
      "loss": 5.3629,
      "mean_token_accuracy": 0.20231543034315108,
      "num_tokens": 37427685.0,
      "step": 20075
    },
    {
      "entropy": 5.797391891479492,
      "epoch": 1.773225008830802,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046866727668474605,
      "loss": 5.2991,
      "mean_token_accuracy": 0.2138638973236084,
      "num_tokens": 37436844.0,
      "step": 20080
    },
    {
      "entropy": 5.782184791564942,
      "epoch": 1.7736665489226422,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000468651244382884,
      "loss": 5.2698,
      "mean_token_accuracy": 0.21559372395277024,
      "num_tokens": 37445755.0,
      "step": 20085
    },
    {
      "entropy": 5.855338668823242,
      "epoch": 1.7741080890144825,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004686352082875043,
      "loss": 5.4357,
      "mean_token_accuracy": 0.1951846346259117,
      "num_tokens": 37454176.0,
      "step": 20090
    },
    {
      "entropy": 5.871099424362183,
      "epoch": 1.774549629106323,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004686191683989209,
      "loss": 5.2938,
      "mean_token_accuracy": 0.21474916040897368,
      "num_tokens": 37463696.0,
      "step": 20095
    },
    {
      "entropy": 5.90635724067688,
      "epoch": 1.7749911691981632,
      "grad_norm": 1.0,
      "learning_rate": 0.0004686031247174482,
      "loss": 5.3519,
      "mean_token_accuracy": 0.2026623472571373,
      "num_tokens": 37473651.0,
      "step": 20100
    },
    {
      "entropy": 5.868699455261231,
      "epoch": 1.7754327092900035,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004685870772434002,
      "loss": 5.4312,
      "mean_token_accuracy": 0.19730533808469772,
      "num_tokens": 37483137.0,
      "step": 20105
    },
    {
      "entropy": 5.912846851348877,
      "epoch": 1.775874249381844,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004685710259770916,
      "loss": 5.4006,
      "mean_token_accuracy": 0.19890353232622146,
      "num_tokens": 37492029.0,
      "step": 20110
    },
    {
      "entropy": 5.893759250640869,
      "epoch": 1.776315789473684,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046855497091883664,
      "loss": 5.3572,
      "mean_token_accuracy": 0.20875534415245056,
      "num_tokens": 37500421.0,
      "step": 20115
    },
    {
      "entropy": 5.820963430404663,
      "epoch": 1.7767573295655246,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004685389120689499,
      "loss": 5.2993,
      "mean_token_accuracy": 0.20525917112827302,
      "num_tokens": 37510332.0,
      "step": 20120
    },
    {
      "entropy": 5.798265886306763,
      "epoch": 1.7771988696573648,
      "grad_norm": 1.09375,
      "learning_rate": 0.000468522849427746,
      "loss": 5.343,
      "mean_token_accuracy": 0.2129892036318779,
      "num_tokens": 37519802.0,
      "step": 20125
    },
    {
      "entropy": 5.951801538467407,
      "epoch": 1.7776404097492051,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046850678299553943,
      "loss": 5.4744,
      "mean_token_accuracy": 0.1945207715034485,
      "num_tokens": 37529622.0,
      "step": 20130
    },
    {
      "entropy": 5.783419799804688,
      "epoch": 1.7780819498410456,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004684907127726451,
      "loss": 5.2962,
      "mean_token_accuracy": 0.21682336628437043,
      "num_tokens": 37539446.0,
      "step": 20135
    },
    {
      "entropy": 5.807795906066895,
      "epoch": 1.778523489932886,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004684746387593778,
      "loss": 5.2873,
      "mean_token_accuracy": 0.2105346664786339,
      "num_tokens": 37547877.0,
      "step": 20140
    },
    {
      "entropy": 5.8958127975463865,
      "epoch": 1.7789650300247262,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004684585609560523,
      "loss": 5.5018,
      "mean_token_accuracy": 0.1957994669675827,
      "num_tokens": 37557841.0,
      "step": 20145
    },
    {
      "entropy": 5.947286176681518,
      "epoch": 1.7794065701165667,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004684424793629836,
      "loss": 5.4075,
      "mean_token_accuracy": 0.19818983525037764,
      "num_tokens": 37566876.0,
      "step": 20150
    },
    {
      "entropy": 5.906966066360473,
      "epoch": 1.779848110208407,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046842639398048683,
      "loss": 5.4073,
      "mean_token_accuracy": 0.20548618882894515,
      "num_tokens": 37575571.0,
      "step": 20155
    },
    {
      "entropy": 5.896219158172608,
      "epoch": 1.7802896503002472,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004684103048088769,
      "loss": 5.4797,
      "mean_token_accuracy": 0.20168900340795518,
      "num_tokens": 37585972.0,
      "step": 20160
    },
    {
      "entropy": 5.9010529041290285,
      "epoch": 1.7807311903920877,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004683942118484692,
      "loss": 5.4147,
      "mean_token_accuracy": 0.20082128047943115,
      "num_tokens": 37594849.0,
      "step": 20165
    },
    {
      "entropy": 5.890091323852539,
      "epoch": 1.781172730483928,
      "grad_norm": 1.0,
      "learning_rate": 0.00046837811509957883,
      "loss": 5.3469,
      "mean_token_accuracy": 0.2057483211159706,
      "num_tokens": 37603634.0,
      "step": 20170
    },
    {
      "entropy": 5.869364023208618,
      "epoch": 1.7816142705757683,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046836201456252116,
      "loss": 5.2968,
      "mean_token_accuracy": 0.21155764311552047,
      "num_tokens": 37612518.0,
      "step": 20175
    },
    {
      "entropy": 5.8324672222137455,
      "epoch": 1.7820558106676088,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004683459102376116,
      "loss": 5.3033,
      "mean_token_accuracy": 0.20891628116369249,
      "num_tokens": 37621553.0,
      "step": 20180
    },
    {
      "entropy": 5.829755783081055,
      "epoch": 1.7824973507594488,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046832980212516557,
      "loss": 5.4122,
      "mean_token_accuracy": 0.20100240558385848,
      "num_tokens": 37631263.0,
      "step": 20185
    },
    {
      "entropy": 5.879919385910034,
      "epoch": 1.7829388908512893,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046831369022549876,
      "loss": 5.2989,
      "mean_token_accuracy": 0.2082589492201805,
      "num_tokens": 37640810.0,
      "step": 20190
    },
    {
      "entropy": 5.89871563911438,
      "epoch": 1.7833804309431296,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004682975745389267,
      "loss": 5.4003,
      "mean_token_accuracy": 0.2027232214808464,
      "num_tokens": 37650848.0,
      "step": 20195
    },
    {
      "entropy": 5.897505903244019,
      "epoch": 1.7838219710349699,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046828145506576504,
      "loss": 5.3668,
      "mean_token_accuracy": 0.20478703081607819,
      "num_tokens": 37660370.0,
      "step": 20200
    },
    {
      "entropy": 5.817348146438599,
      "epoch": 1.7842635111268104,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004682653318063296,
      "loss": 5.3343,
      "mean_token_accuracy": 0.2099165752530098,
      "num_tokens": 37669570.0,
      "step": 20205
    },
    {
      "entropy": 5.802600765228272,
      "epoch": 1.7847050512186506,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046824920476093624,
      "loss": 5.4042,
      "mean_token_accuracy": 0.20283826887607576,
      "num_tokens": 37679585.0,
      "step": 20210
    },
    {
      "entropy": 5.915673637390137,
      "epoch": 1.785146591310491,
      "grad_norm": 1.125,
      "learning_rate": 0.00046823307392990086,
      "loss": 5.3891,
      "mean_token_accuracy": 0.2046528786420822,
      "num_tokens": 37688736.0,
      "step": 20215
    },
    {
      "entropy": 5.977085590362549,
      "epoch": 1.7855881314023314,
      "grad_norm": 1.609375,
      "learning_rate": 0.00046821693931353947,
      "loss": 5.4315,
      "mean_token_accuracy": 0.19578227698802947,
      "num_tokens": 37699104.0,
      "step": 20220
    },
    {
      "entropy": 5.911003971099854,
      "epoch": 1.7860296714941717,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004682008009121681,
      "loss": 5.4324,
      "mean_token_accuracy": 0.20133746415376663,
      "num_tokens": 37708260.0,
      "step": 20225
    },
    {
      "entropy": 5.814093494415284,
      "epoch": 1.786471211586012,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000468184658726103,
      "loss": 5.3673,
      "mean_token_accuracy": 0.20417275428771972,
      "num_tokens": 37717442.0,
      "step": 20230
    },
    {
      "entropy": 5.884583139419556,
      "epoch": 1.7869127516778525,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004681685127556602,
      "loss": 5.4064,
      "mean_token_accuracy": 0.19773323088884354,
      "num_tokens": 37727063.0,
      "step": 20235
    },
    {
      "entropy": 5.875490999221801,
      "epoch": 1.7873542917696927,
      "grad_norm": 1.0,
      "learning_rate": 0.00046815236300115615,
      "loss": 5.3248,
      "mean_token_accuracy": 0.20776914954185485,
      "num_tokens": 37736441.0,
      "step": 20240
    },
    {
      "entropy": 5.76911768913269,
      "epoch": 1.787795831861533,
      "grad_norm": 1.125,
      "learning_rate": 0.00046813620946290716,
      "loss": 5.2723,
      "mean_token_accuracy": 0.2133665129542351,
      "num_tokens": 37745290.0,
      "step": 20245
    },
    {
      "entropy": 5.900927352905273,
      "epoch": 1.7882373719533735,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004681200521412296,
      "loss": 5.3875,
      "mean_token_accuracy": 0.20325565040111543,
      "num_tokens": 37754786.0,
      "step": 20250
    },
    {
      "entropy": 5.908439207077026,
      "epoch": 1.7886789120452136,
      "grad_norm": 1.0,
      "learning_rate": 0.00046810389103644015,
      "loss": 5.4023,
      "mean_token_accuracy": 0.20222677737474443,
      "num_tokens": 37763513.0,
      "step": 20255
    },
    {
      "entropy": 5.862041330337524,
      "epoch": 1.789120452137054,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004680877261488552,
      "loss": 5.3817,
      "mean_token_accuracy": 0.1998486340045929,
      "num_tokens": 37771752.0,
      "step": 20260
    },
    {
      "entropy": 5.865772724151611,
      "epoch": 1.7895619922288943,
      "grad_norm": 1.0,
      "learning_rate": 0.0004680715574787916,
      "loss": 5.3866,
      "mean_token_accuracy": 0.19422840029001237,
      "num_tokens": 37781531.0,
      "step": 20265
    },
    {
      "entropy": 5.838648986816406,
      "epoch": 1.7900035323207346,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046805538502656596,
      "loss": 5.3202,
      "mean_token_accuracy": 0.20190480947494507,
      "num_tokens": 37791469.0,
      "step": 20270
    },
    {
      "entropy": 5.900450754165649,
      "epoch": 1.7904450724125751,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004680392087924952,
      "loss": 5.3937,
      "mean_token_accuracy": 0.20292436927556992,
      "num_tokens": 37800301.0,
      "step": 20275
    },
    {
      "entropy": 5.8141848087310795,
      "epoch": 1.7908866125044154,
      "grad_norm": 1.078125,
      "learning_rate": 0.000468023028776896,
      "loss": 5.2347,
      "mean_token_accuracy": 0.21848550885915757,
      "num_tokens": 37808849.0,
      "step": 20280
    },
    {
      "entropy": 5.88062219619751,
      "epoch": 1.7913281525962557,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004680068449800855,
      "loss": 5.4048,
      "mean_token_accuracy": 0.2034169316291809,
      "num_tokens": 37818632.0,
      "step": 20285
    },
    {
      "entropy": 5.898212194442749,
      "epoch": 1.7917696926880962,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004679906574023807,
      "loss": 5.3949,
      "mean_token_accuracy": 0.19932047724723817,
      "num_tokens": 37828158.0,
      "step": 20290
    },
    {
      "entropy": 5.8748620510101315,
      "epoch": 1.7922112327799364,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004679744660440987,
      "loss": 5.3289,
      "mean_token_accuracy": 0.20268318355083464,
      "num_tokens": 37836242.0,
      "step": 20295
    },
    {
      "entropy": 5.866003608703613,
      "epoch": 1.7926527728717767,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046795827090555654,
      "loss": 5.3535,
      "mean_token_accuracy": 0.202984955906868,
      "num_tokens": 37844286.0,
      "step": 20300
    },
    {
      "entropy": 5.847696304321289,
      "epoch": 1.7930943129636172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004679420719870717,
      "loss": 5.3068,
      "mean_token_accuracy": 0.20818886011838914,
      "num_tokens": 37853440.0,
      "step": 20305
    },
    {
      "entropy": 5.872677278518677,
      "epoch": 1.7935358530554575,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004679258692889613,
      "loss": 5.2455,
      "mean_token_accuracy": 0.21110901981592178,
      "num_tokens": 37862511.0,
      "step": 20310
    },
    {
      "entropy": 5.942249822616577,
      "epoch": 1.7939773931472978,
      "grad_norm": 1.046875,
      "learning_rate": 0.000467909662811543,
      "loss": 5.4542,
      "mean_token_accuracy": 0.19627978652715683,
      "num_tokens": 37872797.0,
      "step": 20315
    },
    {
      "entropy": 5.844317817687989,
      "epoch": 1.7944189332391383,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000467893452555134,
      "loss": 5.3134,
      "mean_token_accuracy": 0.21182480156421662,
      "num_tokens": 37881859.0,
      "step": 20320
    },
    {
      "entropy": 5.825922060012817,
      "epoch": 1.7948604733309783,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046787723852005197,
      "loss": 5.3365,
      "mean_token_accuracy": 0.2158793330192566,
      "num_tokens": 37891381.0,
      "step": 20325
    },
    {
      "entropy": 5.741138362884522,
      "epoch": 1.7953020134228188,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004678610207066146,
      "loss": 5.2496,
      "mean_token_accuracy": 0.21238402128219605,
      "num_tokens": 37900554.0,
      "step": 20330
    },
    {
      "entropy": 5.89889121055603,
      "epoch": 1.795743553514659,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004678447991151395,
      "loss": 5.3023,
      "mean_token_accuracy": 0.2121421828866005,
      "num_tokens": 37908965.0,
      "step": 20335
    },
    {
      "entropy": 5.792270231246948,
      "epoch": 1.7961850936064994,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046782857374594447,
      "loss": 5.296,
      "mean_token_accuracy": 0.21209638863801955,
      "num_tokens": 37917555.0,
      "step": 20340
    },
    {
      "entropy": 5.810970449447632,
      "epoch": 1.7966266336983399,
      "grad_norm": 1.125,
      "learning_rate": 0.0004678123445993473,
      "loss": 5.3585,
      "mean_token_accuracy": 0.21037175506353378,
      "num_tokens": 37926334.0,
      "step": 20345
    },
    {
      "entropy": 5.803108406066895,
      "epoch": 1.7970681737901801,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000467796111675666,
      "loss": 5.3096,
      "mean_token_accuracy": 0.2096547305583954,
      "num_tokens": 37935912.0,
      "step": 20350
    },
    {
      "entropy": 5.768332433700562,
      "epoch": 1.7975097138820204,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046777987497521845,
      "loss": 5.3272,
      "mean_token_accuracy": 0.21359478533267975,
      "num_tokens": 37945238.0,
      "step": 20355
    },
    {
      "entropy": 5.833979177474975,
      "epoch": 1.797951253973861,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046776363449832286,
      "loss": 5.3248,
      "mean_token_accuracy": 0.21153734475374222,
      "num_tokens": 37954945.0,
      "step": 20360
    },
    {
      "entropy": 5.851474046707153,
      "epoch": 1.7983927940657012,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046774739024529736,
      "loss": 5.2451,
      "mean_token_accuracy": 0.21519911140203477,
      "num_tokens": 37964114.0,
      "step": 20365
    },
    {
      "entropy": 5.908031940460205,
      "epoch": 1.7988343341575415,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046773114221645995,
      "loss": 5.3995,
      "mean_token_accuracy": 0.1979236423969269,
      "num_tokens": 37973774.0,
      "step": 20370
    },
    {
      "entropy": 5.910662221908569,
      "epoch": 1.799275874249382,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046771489041212916,
      "loss": 5.3562,
      "mean_token_accuracy": 0.20842334032058715,
      "num_tokens": 37982763.0,
      "step": 20375
    },
    {
      "entropy": 5.931088447570801,
      "epoch": 1.7997174143412222,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004676986348326233,
      "loss": 5.4658,
      "mean_token_accuracy": 0.20048643052577972,
      "num_tokens": 37992302.0,
      "step": 20380
    },
    {
      "entropy": 5.952190208435058,
      "epoch": 1.8001589544330625,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004676823754782607,
      "loss": 5.4894,
      "mean_token_accuracy": 0.19188101589679718,
      "num_tokens": 38001785.0,
      "step": 20385
    },
    {
      "entropy": 5.8138501167297365,
      "epoch": 1.800600494524903,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046766611234936,
      "loss": 5.2994,
      "mean_token_accuracy": 0.20842182636260986,
      "num_tokens": 38010823.0,
      "step": 20390
    },
    {
      "entropy": 5.829443883895874,
      "epoch": 1.801042034616743,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004676498454462398,
      "loss": 5.3183,
      "mean_token_accuracy": 0.214272178709507,
      "num_tokens": 38019718.0,
      "step": 20395
    },
    {
      "entropy": 5.8638537406921385,
      "epoch": 1.8014835747085836,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046763357476921856,
      "loss": 5.3213,
      "mean_token_accuracy": 0.20627929270267487,
      "num_tokens": 38028962.0,
      "step": 20400
    },
    {
      "entropy": 5.863303899765015,
      "epoch": 1.8019251148004238,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004676173003186152,
      "loss": 5.2979,
      "mean_token_accuracy": 0.21785591393709183,
      "num_tokens": 38037824.0,
      "step": 20405
    },
    {
      "entropy": 5.882257127761841,
      "epoch": 1.802366654892264,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046760102209474855,
      "loss": 5.4427,
      "mean_token_accuracy": 0.19779300838708877,
      "num_tokens": 38046764.0,
      "step": 20410
    },
    {
      "entropy": 5.906739568710327,
      "epoch": 1.8028081949841046,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004675847400979373,
      "loss": 5.3682,
      "mean_token_accuracy": 0.2031184047460556,
      "num_tokens": 38055363.0,
      "step": 20415
    },
    {
      "entropy": 5.864423513412476,
      "epoch": 1.8032497350759449,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046756845432850057,
      "loss": 5.409,
      "mean_token_accuracy": 0.20096191316843032,
      "num_tokens": 38065471.0,
      "step": 20420
    },
    {
      "entropy": 5.9324547290802006,
      "epoch": 1.8036912751677852,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004675521647867573,
      "loss": 5.4975,
      "mean_token_accuracy": 0.19525431841611862,
      "num_tokens": 38075448.0,
      "step": 20425
    },
    {
      "entropy": 5.859852504730225,
      "epoch": 1.8041328152596257,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00046753587147302665,
      "loss": 5.4327,
      "mean_token_accuracy": 0.2063135489821434,
      "num_tokens": 38085657.0,
      "step": 20430
    },
    {
      "entropy": 5.919007635116577,
      "epoch": 1.804574355351466,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004675195743876278,
      "loss": 5.3549,
      "mean_token_accuracy": 0.2081198126077652,
      "num_tokens": 38095580.0,
      "step": 20435
    },
    {
      "entropy": 5.965777969360351,
      "epoch": 1.8050158954433062,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046750327353088,
      "loss": 5.4592,
      "mean_token_accuracy": 0.19610353410243989,
      "num_tokens": 38104969.0,
      "step": 20440
    },
    {
      "entropy": 5.905309343338013,
      "epoch": 1.8054574355351467,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004674869689031026,
      "loss": 5.4032,
      "mean_token_accuracy": 0.2008821427822113,
      "num_tokens": 38114536.0,
      "step": 20445
    },
    {
      "entropy": 5.842322158813476,
      "epoch": 1.805898975626987,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046747066050461486,
      "loss": 5.3056,
      "mean_token_accuracy": 0.2088845983147621,
      "num_tokens": 38122969.0,
      "step": 20450
    },
    {
      "entropy": 5.8650373935699465,
      "epoch": 1.8063405157188273,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046745434833573635,
      "loss": 5.3631,
      "mean_token_accuracy": 0.20031063109636307,
      "num_tokens": 38132892.0,
      "step": 20455
    },
    {
      "entropy": 5.780540418624878,
      "epoch": 1.8067820558106678,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004674380323967867,
      "loss": 5.3324,
      "mean_token_accuracy": 0.2102600157260895,
      "num_tokens": 38141950.0,
      "step": 20460
    },
    {
      "entropy": 5.939284229278565,
      "epoch": 1.8072235959025078,
      "grad_norm": 1.125,
      "learning_rate": 0.0004674217126880853,
      "loss": 5.4464,
      "mean_token_accuracy": 0.19473828077316285,
      "num_tokens": 38151314.0,
      "step": 20465
    },
    {
      "entropy": 5.97164945602417,
      "epoch": 1.8076651359943483,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004674053892099521,
      "loss": 5.4392,
      "mean_token_accuracy": 0.20110773295164108,
      "num_tokens": 38159992.0,
      "step": 20470
    },
    {
      "entropy": 5.8675071716308596,
      "epoch": 1.8081066760861886,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046738906196270663,
      "loss": 5.3446,
      "mean_token_accuracy": 0.1976835235953331,
      "num_tokens": 38168436.0,
      "step": 20475
    },
    {
      "entropy": 5.8300371170043945,
      "epoch": 1.8085482161780289,
      "grad_norm": 1.015625,
      "learning_rate": 0.000467372730946669,
      "loss": 5.3449,
      "mean_token_accuracy": 0.20450993329286576,
      "num_tokens": 38177932.0,
      "step": 20480
    },
    {
      "entropy": 5.902409601211548,
      "epoch": 1.8089897562698694,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046735639616215884,
      "loss": 5.3557,
      "mean_token_accuracy": 0.21158699095249175,
      "num_tokens": 38187727.0,
      "step": 20485
    },
    {
      "entropy": 5.843787002563476,
      "epoch": 1.8094312963617096,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046734005760949634,
      "loss": 5.4154,
      "mean_token_accuracy": 0.19625237584114075,
      "num_tokens": 38198768.0,
      "step": 20490
    },
    {
      "entropy": 5.8692399024963375,
      "epoch": 1.80987283645355,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046732371528900154,
      "loss": 5.4124,
      "mean_token_accuracy": 0.2083530083298683,
      "num_tokens": 38207538.0,
      "step": 20495
    },
    {
      "entropy": 5.932307720184326,
      "epoch": 1.8103143765453904,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004673073692009945,
      "loss": 5.449,
      "mean_token_accuracy": 0.19892775416374206,
      "num_tokens": 38216254.0,
      "step": 20500
    },
    {
      "entropy": 5.947169828414917,
      "epoch": 1.8107559166372307,
      "grad_norm": 1.0,
      "learning_rate": 0.0004672910193457955,
      "loss": 5.4194,
      "mean_token_accuracy": 0.19809736758470536,
      "num_tokens": 38225356.0,
      "step": 20505
    },
    {
      "entropy": 5.851454734802246,
      "epoch": 1.811197456729071,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004672746657237248,
      "loss": 5.2808,
      "mean_token_accuracy": 0.21226958483457564,
      "num_tokens": 38234561.0,
      "step": 20510
    },
    {
      "entropy": 5.815788793563843,
      "epoch": 1.8116389968209115,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004672583083351027,
      "loss": 5.3572,
      "mean_token_accuracy": 0.20180919766426086,
      "num_tokens": 38244418.0,
      "step": 20515
    },
    {
      "entropy": 5.874659538269043,
      "epoch": 1.8120805369127517,
      "grad_norm": 1.125,
      "learning_rate": 0.0004672419471802497,
      "loss": 5.3386,
      "mean_token_accuracy": 0.21120245158672332,
      "num_tokens": 38253639.0,
      "step": 20520
    },
    {
      "entropy": 5.896960735321045,
      "epoch": 1.812522077004592,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046722558225948627,
      "loss": 5.2595,
      "mean_token_accuracy": 0.21947896033525466,
      "num_tokens": 38262401.0,
      "step": 20525
    },
    {
      "entropy": 5.8342921257019045,
      "epoch": 1.8129636170964325,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046720921357313303,
      "loss": 5.3504,
      "mean_token_accuracy": 0.21129164844751358,
      "num_tokens": 38271070.0,
      "step": 20530
    },
    {
      "entropy": 5.88353271484375,
      "epoch": 1.8134051571882726,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004671928411215106,
      "loss": 5.3405,
      "mean_token_accuracy": 0.21325495690107346,
      "num_tokens": 38280675.0,
      "step": 20535
    },
    {
      "entropy": 5.879548454284668,
      "epoch": 1.813846697280113,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046717646490493977,
      "loss": 5.4133,
      "mean_token_accuracy": 0.1977847248315811,
      "num_tokens": 38290116.0,
      "step": 20540
    },
    {
      "entropy": 5.905095386505127,
      "epoch": 1.8142882373719533,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004671600849237412,
      "loss": 5.3818,
      "mean_token_accuracy": 0.20388080626726152,
      "num_tokens": 38299415.0,
      "step": 20545
    },
    {
      "entropy": 5.7524528980255125,
      "epoch": 1.8147297774637936,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004671437011782359,
      "loss": 5.2546,
      "mean_token_accuracy": 0.22169750183820724,
      "num_tokens": 38308761.0,
      "step": 20550
    },
    {
      "entropy": 5.912684583663941,
      "epoch": 1.815171317555634,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004671273136687448,
      "loss": 5.4034,
      "mean_token_accuracy": 0.2000546410679817,
      "num_tokens": 38318212.0,
      "step": 20555
    },
    {
      "entropy": 5.923445415496826,
      "epoch": 1.8156128576474744,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004671109223955889,
      "loss": 5.3775,
      "mean_token_accuracy": 0.20213733166456221,
      "num_tokens": 38327233.0,
      "step": 20560
    },
    {
      "entropy": 5.8516336441040036,
      "epoch": 1.8160543977393147,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004670945273590892,
      "loss": 5.2799,
      "mean_token_accuracy": 0.21335710287094117,
      "num_tokens": 38336473.0,
      "step": 20565
    },
    {
      "entropy": 5.808154964447022,
      "epoch": 1.8164959378311551,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004670781285595671,
      "loss": 5.3954,
      "mean_token_accuracy": 0.20852681398391723,
      "num_tokens": 38345286.0,
      "step": 20570
    },
    {
      "entropy": 5.894282007217408,
      "epoch": 1.8169374779229954,
      "grad_norm": 1.0,
      "learning_rate": 0.0004670617259973435,
      "loss": 5.3161,
      "mean_token_accuracy": 0.2049289181828499,
      "num_tokens": 38355191.0,
      "step": 20575
    },
    {
      "entropy": 5.878620290756226,
      "epoch": 1.8173790180148357,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046704531967274007,
      "loss": 5.309,
      "mean_token_accuracy": 0.20789871513843536,
      "num_tokens": 38363902.0,
      "step": 20580
    },
    {
      "entropy": 5.8712554454803465,
      "epoch": 1.8178205581066762,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00046702890958607804,
      "loss": 5.4714,
      "mean_token_accuracy": 0.19961811155080794,
      "num_tokens": 38374828.0,
      "step": 20585
    },
    {
      "entropy": 5.814764404296875,
      "epoch": 1.8182620981985165,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004670124957376788,
      "loss": 5.3378,
      "mean_token_accuracy": 0.2025281995534897,
      "num_tokens": 38384918.0,
      "step": 20590
    },
    {
      "entropy": 5.942892074584961,
      "epoch": 1.8187036382903567,
      "grad_norm": 1.0,
      "learning_rate": 0.0004669960781278641,
      "loss": 5.4034,
      "mean_token_accuracy": 0.20002204030752183,
      "num_tokens": 38394005.0,
      "step": 20595
    },
    {
      "entropy": 5.973907041549682,
      "epoch": 1.8191451783821972,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004669796567569553,
      "loss": 5.4208,
      "mean_token_accuracy": 0.20189059376716614,
      "num_tokens": 38402722.0,
      "step": 20600
    },
    {
      "entropy": 5.843851327896118,
      "epoch": 1.8195867184740373,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004669632316252743,
      "loss": 5.3348,
      "mean_token_accuracy": 0.21077610552310944,
      "num_tokens": 38411726.0,
      "step": 20605
    },
    {
      "entropy": 5.899236488342285,
      "epoch": 1.8200282585658778,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046694680273314275,
      "loss": 5.396,
      "mean_token_accuracy": 0.2030399203300476,
      "num_tokens": 38421187.0,
      "step": 20610
    },
    {
      "entropy": 5.877706670761109,
      "epoch": 1.820469798657718,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004669303700808824,
      "loss": 5.4301,
      "mean_token_accuracy": 0.1965902864933014,
      "num_tokens": 38430722.0,
      "step": 20615
    },
    {
      "entropy": 5.840076351165772,
      "epoch": 1.8209113387495584,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004669139336688154,
      "loss": 5.3489,
      "mean_token_accuracy": 0.2044423222541809,
      "num_tokens": 38440763.0,
      "step": 20620
    },
    {
      "entropy": 5.823544597625732,
      "epoch": 1.8213528788413988,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046689749349726353,
      "loss": 5.3848,
      "mean_token_accuracy": 0.2033069536089897,
      "num_tokens": 38449298.0,
      "step": 20625
    },
    {
      "entropy": 5.859510135650635,
      "epoch": 1.8217944189332391,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004668810495665489,
      "loss": 5.3558,
      "mean_token_accuracy": 0.20442309677600862,
      "num_tokens": 38458941.0,
      "step": 20630
    },
    {
      "entropy": 5.876007127761841,
      "epoch": 1.8222359590250794,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004668646018769936,
      "loss": 5.3266,
      "mean_token_accuracy": 0.21604098826646806,
      "num_tokens": 38468381.0,
      "step": 20635
    },
    {
      "entropy": 5.9373369216918945,
      "epoch": 1.82267749911692,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046684815042891993,
      "loss": 5.3783,
      "mean_token_accuracy": 0.202556112408638,
      "num_tokens": 38477767.0,
      "step": 20640
    },
    {
      "entropy": 5.837852525711059,
      "epoch": 1.8231190392087602,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004668316952226501,
      "loss": 5.3002,
      "mean_token_accuracy": 0.2069453090429306,
      "num_tokens": 38486884.0,
      "step": 20645
    },
    {
      "entropy": 5.843544769287109,
      "epoch": 1.8235605793006004,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004668152362585064,
      "loss": 5.4226,
      "mean_token_accuracy": 0.1992058902978897,
      "num_tokens": 38496243.0,
      "step": 20650
    },
    {
      "entropy": 5.778637886047363,
      "epoch": 1.824002119392441,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004667987735368113,
      "loss": 5.294,
      "mean_token_accuracy": 0.2042941391468048,
      "num_tokens": 38504447.0,
      "step": 20655
    },
    {
      "entropy": 5.882816457748413,
      "epoch": 1.8244436594842812,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004667823070578874,
      "loss": 5.4054,
      "mean_token_accuracy": 0.20013791173696518,
      "num_tokens": 38514423.0,
      "step": 20660
    },
    {
      "entropy": 5.884525346755981,
      "epoch": 1.8248851995761215,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046676583682205713,
      "loss": 5.3297,
      "mean_token_accuracy": 0.21381803154945372,
      "num_tokens": 38522765.0,
      "step": 20665
    },
    {
      "entropy": 5.917822170257568,
      "epoch": 1.825326739667962,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046674936282964326,
      "loss": 5.4478,
      "mean_token_accuracy": 0.19618436247110366,
      "num_tokens": 38532074.0,
      "step": 20670
    },
    {
      "entropy": 5.845418119430542,
      "epoch": 1.825768279759802,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004667328850809683,
      "loss": 5.3291,
      "mean_token_accuracy": 0.2020283207297325,
      "num_tokens": 38541423.0,
      "step": 20675
    },
    {
      "entropy": 5.869617366790772,
      "epoch": 1.8262098198516425,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004667164035763552,
      "loss": 5.4233,
      "mean_token_accuracy": 0.20130159556865693,
      "num_tokens": 38552369.0,
      "step": 20680
    },
    {
      "entropy": 5.909629583358765,
      "epoch": 1.8266513599434828,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004666999183161268,
      "loss": 5.3648,
      "mean_token_accuracy": 0.20436142832040788,
      "num_tokens": 38561310.0,
      "step": 20685
    },
    {
      "entropy": 5.919795942306519,
      "epoch": 1.827092900035323,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046668342930060604,
      "loss": 5.4257,
      "mean_token_accuracy": 0.20203170478343963,
      "num_tokens": 38571157.0,
      "step": 20690
    },
    {
      "entropy": 5.862828111648559,
      "epoch": 1.8275344401271636,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046666693653011596,
      "loss": 5.3492,
      "mean_token_accuracy": 0.2039842963218689,
      "num_tokens": 38579862.0,
      "step": 20695
    },
    {
      "entropy": 5.870121860504151,
      "epoch": 1.8279759802190039,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046665044000497954,
      "loss": 5.4234,
      "mean_token_accuracy": 0.20539480149745942,
      "num_tokens": 38589784.0,
      "step": 20700
    },
    {
      "entropy": 5.910328960418701,
      "epoch": 1.8284175203108441,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046663393972551994,
      "loss": 5.3317,
      "mean_token_accuracy": 0.20784323811531066,
      "num_tokens": 38598925.0,
      "step": 20705
    },
    {
      "entropy": 5.888635158538818,
      "epoch": 1.8288590604026846,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046661743569206056,
      "loss": 5.392,
      "mean_token_accuracy": 0.1957668572664261,
      "num_tokens": 38608245.0,
      "step": 20710
    },
    {
      "entropy": 5.883733415603638,
      "epoch": 1.829300600494525,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004666009279049245,
      "loss": 5.4586,
      "mean_token_accuracy": 0.1983842819929123,
      "num_tokens": 38617586.0,
      "step": 20715
    },
    {
      "entropy": 5.905021858215332,
      "epoch": 1.8297421405863652,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046658441636443526,
      "loss": 5.4061,
      "mean_token_accuracy": 0.2049261063337326,
      "num_tokens": 38626420.0,
      "step": 20720
    },
    {
      "entropy": 5.881510162353516,
      "epoch": 1.8301836806782057,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046656790107091627,
      "loss": 5.3007,
      "mean_token_accuracy": 0.21221004873514177,
      "num_tokens": 38635585.0,
      "step": 20725
    },
    {
      "entropy": 5.830260181427002,
      "epoch": 1.830625220770046,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004665513820246911,
      "loss": 5.3552,
      "mean_token_accuracy": 0.20323756635189055,
      "num_tokens": 38643998.0,
      "step": 20730
    },
    {
      "entropy": 5.841861295700073,
      "epoch": 1.8310667608618862,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004665348592260833,
      "loss": 5.3721,
      "mean_token_accuracy": 0.20905961394309996,
      "num_tokens": 38653018.0,
      "step": 20735
    },
    {
      "entropy": 5.84530348777771,
      "epoch": 1.8315083009537267,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004665183326754165,
      "loss": 5.3195,
      "mean_token_accuracy": 0.2096097707748413,
      "num_tokens": 38662079.0,
      "step": 20740
    },
    {
      "entropy": 5.885212516784668,
      "epoch": 1.8319498410455668,
      "grad_norm": 1.125,
      "learning_rate": 0.0004665018023730144,
      "loss": 5.3803,
      "mean_token_accuracy": 0.20793042033910752,
      "num_tokens": 38669928.0,
      "step": 20745
    },
    {
      "entropy": 5.855934047698975,
      "epoch": 1.8323913811374073,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000466485268319201,
      "loss": 5.3482,
      "mean_token_accuracy": 0.2029734492301941,
      "num_tokens": 38679475.0,
      "step": 20750
    },
    {
      "entropy": 5.873014354705811,
      "epoch": 1.8328329212292476,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004664687305143001,
      "loss": 5.4035,
      "mean_token_accuracy": 0.20414873957633972,
      "num_tokens": 38689057.0,
      "step": 20755
    },
    {
      "entropy": 5.84939866065979,
      "epoch": 1.8332744613210878,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046645218895863564,
      "loss": 5.258,
      "mean_token_accuracy": 0.21808112561702728,
      "num_tokens": 38698362.0,
      "step": 20760
    },
    {
      "entropy": 5.82479043006897,
      "epoch": 1.8337160014129283,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046643564365253173,
      "loss": 5.2754,
      "mean_token_accuracy": 0.2201460063457489,
      "num_tokens": 38708319.0,
      "step": 20765
    },
    {
      "entropy": 5.890381574630737,
      "epoch": 1.8341575415047686,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004664190945963124,
      "loss": 5.399,
      "mean_token_accuracy": 0.19942618310451507,
      "num_tokens": 38717441.0,
      "step": 20770
    },
    {
      "entropy": 5.956099128723144,
      "epoch": 1.834599081596609,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046640254179030184,
      "loss": 5.3988,
      "mean_token_accuracy": 0.1964491829276085,
      "num_tokens": 38726903.0,
      "step": 20775
    },
    {
      "entropy": 5.838819360733032,
      "epoch": 1.8350406216884494,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004663859852348244,
      "loss": 5.3292,
      "mean_token_accuracy": 0.2094813823699951,
      "num_tokens": 38735983.0,
      "step": 20780
    },
    {
      "entropy": 5.838028144836426,
      "epoch": 1.8354821617802897,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046636942493020437,
      "loss": 5.443,
      "mean_token_accuracy": 0.20704626590013503,
      "num_tokens": 38745017.0,
      "step": 20785
    },
    {
      "entropy": 5.783022260665893,
      "epoch": 1.83592370187213,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004663528608767661,
      "loss": 5.2822,
      "mean_token_accuracy": 0.20643096119165422,
      "num_tokens": 38754756.0,
      "step": 20790
    },
    {
      "entropy": 5.893237066268921,
      "epoch": 1.8363652419639704,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004663362930748341,
      "loss": 5.3616,
      "mean_token_accuracy": 0.2074210524559021,
      "num_tokens": 38763695.0,
      "step": 20795
    },
    {
      "entropy": 5.913727617263794,
      "epoch": 1.8368067820558107,
      "grad_norm": 1.0,
      "learning_rate": 0.00046631972152473295,
      "loss": 5.4402,
      "mean_token_accuracy": 0.20066530108451844,
      "num_tokens": 38773637.0,
      "step": 20800
    },
    {
      "entropy": 5.744051313400268,
      "epoch": 1.837248322147651,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046630314622678724,
      "loss": 5.2374,
      "mean_token_accuracy": 0.21938198655843735,
      "num_tokens": 38783004.0,
      "step": 20805
    },
    {
      "entropy": 5.781368541717529,
      "epoch": 1.8376898622394915,
      "grad_norm": 1.125,
      "learning_rate": 0.0004662865671813217,
      "loss": 5.3663,
      "mean_token_accuracy": 0.20539602339267732,
      "num_tokens": 38791342.0,
      "step": 20810
    },
    {
      "entropy": 5.856006097793579,
      "epoch": 1.8381314023313315,
      "grad_norm": 0.921875,
      "learning_rate": 0.00046626998438866114,
      "loss": 5.31,
      "mean_token_accuracy": 0.20879921317100525,
      "num_tokens": 38801046.0,
      "step": 20815
    },
    {
      "entropy": 5.882837390899658,
      "epoch": 1.838572942423172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046625339784913035,
      "loss": 5.3695,
      "mean_token_accuracy": 0.2043130800127983,
      "num_tokens": 38810806.0,
      "step": 20820
    },
    {
      "entropy": 5.84970908164978,
      "epoch": 1.8390144825150123,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046623680756305427,
      "loss": 5.4013,
      "mean_token_accuracy": 0.2009470283985138,
      "num_tokens": 38819362.0,
      "step": 20825
    },
    {
      "entropy": 5.897683763504029,
      "epoch": 1.8394560226068526,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004662202135307579,
      "loss": 5.3996,
      "mean_token_accuracy": 0.20360717922449112,
      "num_tokens": 38829337.0,
      "step": 20830
    },
    {
      "entropy": 5.874234962463379,
      "epoch": 1.839897562698693,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004662036157525663,
      "loss": 5.36,
      "mean_token_accuracy": 0.20298291891813278,
      "num_tokens": 38838351.0,
      "step": 20835
    },
    {
      "entropy": 5.895681667327881,
      "epoch": 1.8403391027905334,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046618701422880456,
      "loss": 5.4089,
      "mean_token_accuracy": 0.1998438224196434,
      "num_tokens": 38847830.0,
      "step": 20840
    },
    {
      "entropy": 5.9258873462677,
      "epoch": 1.8407806428823736,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000466170408959798,
      "loss": 5.4135,
      "mean_token_accuracy": 0.1994987055659294,
      "num_tokens": 38856945.0,
      "step": 20845
    },
    {
      "entropy": 5.83818883895874,
      "epoch": 1.8412221829742141,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004661537999458718,
      "loss": 5.3314,
      "mean_token_accuracy": 0.20225901603698732,
      "num_tokens": 38866493.0,
      "step": 20850
    },
    {
      "entropy": 5.7961172580719,
      "epoch": 1.8416637230660544,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046613718718735143,
      "loss": 5.2967,
      "mean_token_accuracy": 0.21373035609722138,
      "num_tokens": 38875653.0,
      "step": 20855
    },
    {
      "entropy": 5.8129431247711185,
      "epoch": 1.8421052631578947,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046612057068456226,
      "loss": 5.229,
      "mean_token_accuracy": 0.21884686946868898,
      "num_tokens": 38884740.0,
      "step": 20860
    },
    {
      "entropy": 5.852499675750733,
      "epoch": 1.8425468032497352,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046610395043782983,
      "loss": 5.3558,
      "mean_token_accuracy": 0.20886240303516387,
      "num_tokens": 38893801.0,
      "step": 20865
    },
    {
      "entropy": 5.85790696144104,
      "epoch": 1.8429883433415755,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046608732644747963,
      "loss": 5.357,
      "mean_token_accuracy": 0.2100034236907959,
      "num_tokens": 38902298.0,
      "step": 20870
    },
    {
      "entropy": 5.854526329040527,
      "epoch": 1.8434298834334157,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046607069871383745,
      "loss": 5.2604,
      "mean_token_accuracy": 0.2157875820994377,
      "num_tokens": 38910904.0,
      "step": 20875
    },
    {
      "entropy": 5.789853858947754,
      "epoch": 1.8438714235252562,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004660540672372289,
      "loss": 5.2504,
      "mean_token_accuracy": 0.2158147305250168,
      "num_tokens": 38920405.0,
      "step": 20880
    },
    {
      "entropy": 5.800595092773437,
      "epoch": 1.8443129636170963,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004660374320179798,
      "loss": 5.2519,
      "mean_token_accuracy": 0.2045004189014435,
      "num_tokens": 38930118.0,
      "step": 20885
    },
    {
      "entropy": 5.828299236297608,
      "epoch": 1.8447545037089368,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046602079305641615,
      "loss": 5.3589,
      "mean_token_accuracy": 0.20638312846422197,
      "num_tokens": 38939388.0,
      "step": 20890
    },
    {
      "entropy": 5.839706134796143,
      "epoch": 1.845196043800777,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046600415035286374,
      "loss": 5.3956,
      "mean_token_accuracy": 0.20893628001213074,
      "num_tokens": 38948632.0,
      "step": 20895
    },
    {
      "entropy": 5.926899003982544,
      "epoch": 1.8456375838926173,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004659875039076487,
      "loss": 5.3824,
      "mean_token_accuracy": 0.20359280109405517,
      "num_tokens": 38957504.0,
      "step": 20900
    },
    {
      "entropy": 5.85803861618042,
      "epoch": 1.8460791239844578,
      "grad_norm": 0.953125,
      "learning_rate": 0.000465970853721097,
      "loss": 5.3212,
      "mean_token_accuracy": 0.20657461285591125,
      "num_tokens": 38967713.0,
      "step": 20905
    },
    {
      "entropy": 5.840960073471069,
      "epoch": 1.8465206640762981,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046595419979353494,
      "loss": 5.3741,
      "mean_token_accuracy": 0.20615524798631668,
      "num_tokens": 38976982.0,
      "step": 20910
    },
    {
      "entropy": 5.816210126876831,
      "epoch": 1.8469622041681384,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004659375421252887,
      "loss": 5.3323,
      "mean_token_accuracy": 0.20973749160766603,
      "num_tokens": 38985479.0,
      "step": 20915
    },
    {
      "entropy": 5.9450376510620115,
      "epoch": 1.8474037442599789,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004659208807166846,
      "loss": 5.4157,
      "mean_token_accuracy": 0.1954493373632431,
      "num_tokens": 38995418.0,
      "step": 20920
    },
    {
      "entropy": 5.858893203735351,
      "epoch": 1.8478452843518192,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004659042155680491,
      "loss": 5.339,
      "mean_token_accuracy": 0.20293375253677368,
      "num_tokens": 39004961.0,
      "step": 20925
    },
    {
      "entropy": 5.779239082336426,
      "epoch": 1.8482868244436594,
      "grad_norm": 1.0,
      "learning_rate": 0.0004658875466797085,
      "loss": 5.3175,
      "mean_token_accuracy": 0.2112161859869957,
      "num_tokens": 39014282.0,
      "step": 20930
    },
    {
      "entropy": 5.800938892364502,
      "epoch": 1.8487283645355,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046587087405198943,
      "loss": 5.3293,
      "mean_token_accuracy": 0.2036822706460953,
      "num_tokens": 39024480.0,
      "step": 20935
    },
    {
      "entropy": 5.891244697570801,
      "epoch": 1.8491699046273402,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00046585419768521854,
      "loss": 5.3606,
      "mean_token_accuracy": 0.20762478113174437,
      "num_tokens": 39034363.0,
      "step": 20940
    },
    {
      "entropy": 5.90705361366272,
      "epoch": 1.8496114447191805,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046583751757972244,
      "loss": 5.3102,
      "mean_token_accuracy": 0.210002301633358,
      "num_tokens": 39043823.0,
      "step": 20945
    },
    {
      "entropy": 5.8491464138031,
      "epoch": 1.850052984811021,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004658208337358279,
      "loss": 5.3048,
      "mean_token_accuracy": 0.20848426967859268,
      "num_tokens": 39052219.0,
      "step": 20950
    },
    {
      "entropy": 5.882801914215088,
      "epoch": 1.850494524902861,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046580414615386175,
      "loss": 5.4307,
      "mean_token_accuracy": 0.20012145340442658,
      "num_tokens": 39062974.0,
      "step": 20955
    },
    {
      "entropy": 5.717336177825928,
      "epoch": 1.8509360649947015,
      "grad_norm": 1.125,
      "learning_rate": 0.00046578745483415087,
      "loss": 5.1892,
      "mean_token_accuracy": 0.22310186475515364,
      "num_tokens": 39072178.0,
      "step": 20960
    },
    {
      "entropy": 5.8452311038970945,
      "epoch": 1.8513776050865418,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004657707597770223,
      "loss": 5.3498,
      "mean_token_accuracy": 0.2042410597205162,
      "num_tokens": 39081494.0,
      "step": 20965
    },
    {
      "entropy": 5.826963758468628,
      "epoch": 1.851819145178382,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000465754060982803,
      "loss": 5.3756,
      "mean_token_accuracy": 0.20149335712194444,
      "num_tokens": 39090783.0,
      "step": 20970
    },
    {
      "entropy": 5.85706934928894,
      "epoch": 1.8522606852702226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004657373584518201,
      "loss": 5.3728,
      "mean_token_accuracy": 0.20822679102420807,
      "num_tokens": 39100150.0,
      "step": 20975
    },
    {
      "entropy": 5.936109209060669,
      "epoch": 1.8527022253620629,
      "grad_norm": 1.03125,
      "learning_rate": 0.000465720652184401,
      "loss": 5.4202,
      "mean_token_accuracy": 0.20123695582151413,
      "num_tokens": 39109419.0,
      "step": 20980
    },
    {
      "entropy": 5.829092693328858,
      "epoch": 1.8531437654539031,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004657039421808726,
      "loss": 5.2763,
      "mean_token_accuracy": 0.21640390008687974,
      "num_tokens": 39118312.0,
      "step": 20985
    },
    {
      "entropy": 5.874718379974365,
      "epoch": 1.8535853055457436,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046568722844156254,
      "loss": 5.4306,
      "mean_token_accuracy": 0.19490169882774352,
      "num_tokens": 39128169.0,
      "step": 20990
    },
    {
      "entropy": 5.848998928070069,
      "epoch": 1.854026845637584,
      "grad_norm": 1.0,
      "learning_rate": 0.00046567051096679807,
      "loss": 5.3597,
      "mean_token_accuracy": 0.2045652076601982,
      "num_tokens": 39136790.0,
      "step": 20995
    },
    {
      "entropy": 5.858442258834839,
      "epoch": 1.8544683857294242,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004656537897569068,
      "loss": 5.2841,
      "mean_token_accuracy": 0.21445168554782867,
      "num_tokens": 39146974.0,
      "step": 21000
    },
    {
      "epoch": 1.8544683857294242,
      "eval_entropy": 5.64301583370023,
      "eval_loss": 5.393889904022217,
      "eval_mean_token_accuracy": 0.2127020798530746,
      "eval_num_tokens": 39146974.0,
      "eval_runtime": 20.2421,
      "eval_samples_per_second": 1744.631,
      "eval_steps_per_second": 218.11,
      "step": 21000
    },
    {
      "entropy": 5.9857076644897464,
      "epoch": 1.8549099258212647,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004656370648122161,
      "loss": 5.4548,
      "mean_token_accuracy": 0.2057218387722969,
      "num_tokens": 39156910.0,
      "step": 21005
    },
    {
      "entropy": 5.912996578216553,
      "epoch": 1.855351465913105,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046562033613305377,
      "loss": 5.4353,
      "mean_token_accuracy": 0.20232681185007095,
      "num_tokens": 39165790.0,
      "step": 21010
    },
    {
      "entropy": 5.761199951171875,
      "epoch": 1.8557930060049452,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004656036037197475,
      "loss": 5.2633,
      "mean_token_accuracy": 0.22008705139160156,
      "num_tokens": 39174811.0,
      "step": 21015
    },
    {
      "entropy": 5.825755834579468,
      "epoch": 1.8562345460967857,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004655868675726251,
      "loss": 5.385,
      "mean_token_accuracy": 0.20006946474313736,
      "num_tokens": 39184325.0,
      "step": 21020
    },
    {
      "entropy": 5.925671625137329,
      "epoch": 1.8566760861886258,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046557012769201425,
      "loss": 5.3268,
      "mean_token_accuracy": 0.21048913300037383,
      "num_tokens": 39194879.0,
      "step": 21025
    },
    {
      "entropy": 5.843914127349853,
      "epoch": 1.8571176262804663,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000465553384078243,
      "loss": 5.3566,
      "mean_token_accuracy": 0.2091958314180374,
      "num_tokens": 39204702.0,
      "step": 21030
    },
    {
      "entropy": 5.831756210327148,
      "epoch": 1.8575591663723066,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004655366367316394,
      "loss": 5.3749,
      "mean_token_accuracy": 0.21127330064773558,
      "num_tokens": 39214561.0,
      "step": 21035
    },
    {
      "entropy": 5.90690426826477,
      "epoch": 1.8580007064641468,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046551988565253136,
      "loss": 5.4598,
      "mean_token_accuracy": 0.19376136362552643,
      "num_tokens": 39224411.0,
      "step": 21040
    },
    {
      "entropy": 5.877468824386597,
      "epoch": 1.8584422465559873,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004655031308412471,
      "loss": 5.3732,
      "mean_token_accuracy": 0.2032475307583809,
      "num_tokens": 39233422.0,
      "step": 21045
    },
    {
      "entropy": 5.9029398441314695,
      "epoch": 1.8588837866478276,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046548637229811495,
      "loss": 5.3929,
      "mean_token_accuracy": 0.2050471395254135,
      "num_tokens": 39243919.0,
      "step": 21050
    },
    {
      "entropy": 5.936050033569336,
      "epoch": 1.8593253267396679,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046546961002346304,
      "loss": 5.4114,
      "mean_token_accuracy": 0.20312367081642152,
      "num_tokens": 39253682.0,
      "step": 21055
    },
    {
      "entropy": 5.895182180404663,
      "epoch": 1.8597668668315084,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004654528440176198,
      "loss": 5.4499,
      "mean_token_accuracy": 0.20048205107450484,
      "num_tokens": 39263855.0,
      "step": 21060
    },
    {
      "entropy": 5.912688207626343,
      "epoch": 1.8602084069233487,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004654360742809137,
      "loss": 5.2943,
      "mean_token_accuracy": 0.21288205832242965,
      "num_tokens": 39272292.0,
      "step": 21065
    },
    {
      "entropy": 5.888539266586304,
      "epoch": 1.860649947015189,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004654193008136732,
      "loss": 5.3536,
      "mean_token_accuracy": 0.2024380460381508,
      "num_tokens": 39281321.0,
      "step": 21070
    },
    {
      "entropy": 5.936291694641113,
      "epoch": 1.8610914871070294,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046540252361622683,
      "loss": 5.45,
      "mean_token_accuracy": 0.19980594664812087,
      "num_tokens": 39292255.0,
      "step": 21075
    },
    {
      "entropy": 5.905332326889038,
      "epoch": 1.8615330271988697,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046538574268890334,
      "loss": 5.381,
      "mean_token_accuracy": 0.2068076640367508,
      "num_tokens": 39301841.0,
      "step": 21080
    },
    {
      "entropy": 5.877217149734497,
      "epoch": 1.86197456729071,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046536895803203144,
      "loss": 5.3715,
      "mean_token_accuracy": 0.20257276594638823,
      "num_tokens": 39311238.0,
      "step": 21085
    },
    {
      "entropy": 5.896858835220337,
      "epoch": 1.8624161073825505,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046535216964593997,
      "loss": 5.3787,
      "mean_token_accuracy": 0.20414014011621476,
      "num_tokens": 39319945.0,
      "step": 21090
    },
    {
      "entropy": 5.88555679321289,
      "epoch": 1.8628576474743905,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004653353775309576,
      "loss": 5.3437,
      "mean_token_accuracy": 0.20710037499666215,
      "num_tokens": 39329436.0,
      "step": 21095
    },
    {
      "entropy": 5.925942277908325,
      "epoch": 1.863299187566231,
      "grad_norm": 1.125,
      "learning_rate": 0.0004653185816874136,
      "loss": 5.3743,
      "mean_token_accuracy": 0.2059625506401062,
      "num_tokens": 39338447.0,
      "step": 21100
    },
    {
      "entropy": 5.82644624710083,
      "epoch": 1.8637407276580713,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004653017821156368,
      "loss": 5.2958,
      "mean_token_accuracy": 0.21237327456474303,
      "num_tokens": 39346744.0,
      "step": 21105
    },
    {
      "entropy": 5.853572511672974,
      "epoch": 1.8641822677499116,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004652849788159563,
      "loss": 5.406,
      "mean_token_accuracy": 0.19667177200317382,
      "num_tokens": 39357402.0,
      "step": 21110
    },
    {
      "entropy": 5.9240320205688475,
      "epoch": 1.864623807841752,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046526817178870123,
      "loss": 5.3913,
      "mean_token_accuracy": 0.1992943748831749,
      "num_tokens": 39366902.0,
      "step": 21115
    },
    {
      "entropy": 5.9621518611907955,
      "epoch": 1.8650653479335924,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046525136103420093,
      "loss": 5.3677,
      "mean_token_accuracy": 0.20447319000959396,
      "num_tokens": 39376330.0,
      "step": 21120
    },
    {
      "entropy": 5.8133303165435795,
      "epoch": 1.8655068880254326,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046523454655278466,
      "loss": 5.3279,
      "mean_token_accuracy": 0.21593234986066817,
      "num_tokens": 39385926.0,
      "step": 21125
    },
    {
      "entropy": 5.854966020584106,
      "epoch": 1.8659484281172731,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004652177283447818,
      "loss": 5.2904,
      "mean_token_accuracy": 0.20635225772857665,
      "num_tokens": 39395558.0,
      "step": 21130
    },
    {
      "entropy": 5.88421483039856,
      "epoch": 1.8663899682091134,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046520090641052187,
      "loss": 5.3576,
      "mean_token_accuracy": 0.2085113123059273,
      "num_tokens": 39405072.0,
      "step": 21135
    },
    {
      "entropy": 5.926157474517822,
      "epoch": 1.8668315083009537,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046518408075033437,
      "loss": 5.4929,
      "mean_token_accuracy": 0.19987283945083617,
      "num_tokens": 39414714.0,
      "step": 21140
    },
    {
      "entropy": 5.870391607284546,
      "epoch": 1.8672730483927942,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004651672513645488,
      "loss": 5.2767,
      "mean_token_accuracy": 0.21129160225391388,
      "num_tokens": 39423189.0,
      "step": 21145
    },
    {
      "entropy": 5.865139818191528,
      "epoch": 1.8677145884846345,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046515041825349505,
      "loss": 5.4277,
      "mean_token_accuracy": 0.19804446995258332,
      "num_tokens": 39432791.0,
      "step": 21150
    },
    {
      "entropy": 5.8997735500335695,
      "epoch": 1.8681561285764747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004651335814175027,
      "loss": 5.3748,
      "mean_token_accuracy": 0.20265783816576005,
      "num_tokens": 39442849.0,
      "step": 21155
    },
    {
      "entropy": 5.897530889511108,
      "epoch": 1.8685976686683152,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046511674085690155,
      "loss": 5.3692,
      "mean_token_accuracy": 0.2050998792052269,
      "num_tokens": 39452171.0,
      "step": 21160
    },
    {
      "entropy": 5.926449632644653,
      "epoch": 1.8690392087601553,
      "grad_norm": 1.0,
      "learning_rate": 0.0004650998965720216,
      "loss": 5.3764,
      "mean_token_accuracy": 0.20553335398435593,
      "num_tokens": 39460351.0,
      "step": 21165
    },
    {
      "entropy": 5.940393590927124,
      "epoch": 1.8694807488519958,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046508304856319283,
      "loss": 5.441,
      "mean_token_accuracy": 0.2016812741756439,
      "num_tokens": 39470050.0,
      "step": 21170
    },
    {
      "entropy": 5.8886185646057125,
      "epoch": 1.869922288943836,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046506619683074526,
      "loss": 5.3666,
      "mean_token_accuracy": 0.2059075877070427,
      "num_tokens": 39480192.0,
      "step": 21175
    },
    {
      "entropy": 5.90000433921814,
      "epoch": 1.8703638290356763,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046504934137500903,
      "loss": 5.4128,
      "mean_token_accuracy": 0.20533685088157655,
      "num_tokens": 39490118.0,
      "step": 21180
    },
    {
      "entropy": 5.928255319595337,
      "epoch": 1.8708053691275168,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046503248219631417,
      "loss": 5.4288,
      "mean_token_accuracy": 0.1986628070473671,
      "num_tokens": 39499173.0,
      "step": 21185
    },
    {
      "entropy": 5.79513783454895,
      "epoch": 1.871246909219357,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046501561929499115,
      "loss": 5.2627,
      "mean_token_accuracy": 0.2171308770775795,
      "num_tokens": 39508595.0,
      "step": 21190
    },
    {
      "entropy": 5.8287464618682865,
      "epoch": 1.8716884493111974,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046499875267137016,
      "loss": 5.3347,
      "mean_token_accuracy": 0.2113501623272896,
      "num_tokens": 39519300.0,
      "step": 21195
    },
    {
      "entropy": 5.902194929122925,
      "epoch": 1.8721299894030379,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004649818823257817,
      "loss": 5.3783,
      "mean_token_accuracy": 0.20862430334091187,
      "num_tokens": 39527722.0,
      "step": 21200
    },
    {
      "entropy": 5.890845537185669,
      "epoch": 1.8725715294948782,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004649650082585562,
      "loss": 5.3559,
      "mean_token_accuracy": 0.20628983825445174,
      "num_tokens": 39537063.0,
      "step": 21205
    },
    {
      "entropy": 5.8790748596191404,
      "epoch": 1.8730130695867184,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046494813047002424,
      "loss": 5.3915,
      "mean_token_accuracy": 0.2070586144924164,
      "num_tokens": 39546402.0,
      "step": 21210
    },
    {
      "entropy": 5.90912446975708,
      "epoch": 1.873454609678559,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046493124896051634,
      "loss": 5.38,
      "mean_token_accuracy": 0.20982841402292252,
      "num_tokens": 39555877.0,
      "step": 21215
    },
    {
      "entropy": 5.839294862747193,
      "epoch": 1.8738961497703992,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046491436373036346,
      "loss": 5.3273,
      "mean_token_accuracy": 0.20793818384408952,
      "num_tokens": 39565153.0,
      "step": 21220
    },
    {
      "entropy": 5.8650837421417235,
      "epoch": 1.8743376898622395,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004648974747798961,
      "loss": 5.3464,
      "mean_token_accuracy": 0.21023863106966018,
      "num_tokens": 39573775.0,
      "step": 21225
    },
    {
      "entropy": 5.9614763259887695,
      "epoch": 1.87477922995408,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004648805821094452,
      "loss": 5.4367,
      "mean_token_accuracy": 0.20038348138332368,
      "num_tokens": 39583159.0,
      "step": 21230
    },
    {
      "entropy": 6.000168514251709,
      "epoch": 1.87522077004592,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046486368571934175,
      "loss": 5.4559,
      "mean_token_accuracy": 0.19545531868934632,
      "num_tokens": 39592545.0,
      "step": 21235
    },
    {
      "entropy": 5.721289873123169,
      "epoch": 1.8756623101377605,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004648467856099167,
      "loss": 5.1806,
      "mean_token_accuracy": 0.22161275744438172,
      "num_tokens": 39602059.0,
      "step": 21240
    },
    {
      "entropy": 5.847084808349609,
      "epoch": 1.8761038502296008,
      "grad_norm": 1.015625,
      "learning_rate": 0.000464829881781501,
      "loss": 5.3379,
      "mean_token_accuracy": 0.20980760604143142,
      "num_tokens": 39610777.0,
      "step": 21245
    },
    {
      "entropy": 5.907437419891357,
      "epoch": 1.876545390321441,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046481297423442593,
      "loss": 5.4176,
      "mean_token_accuracy": 0.20618286430835725,
      "num_tokens": 39620782.0,
      "step": 21250
    },
    {
      "entropy": 6.001585054397583,
      "epoch": 1.8769869304132816,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004647960629690226,
      "loss": 5.4581,
      "mean_token_accuracy": 0.19751667380332946,
      "num_tokens": 39631340.0,
      "step": 21255
    },
    {
      "entropy": 5.867327404022217,
      "epoch": 1.8774284705051218,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004647791479856223,
      "loss": 5.2793,
      "mean_token_accuracy": 0.21446926891803741,
      "num_tokens": 39640361.0,
      "step": 21260
    },
    {
      "entropy": 5.942128372192383,
      "epoch": 1.8778700105969621,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004647622292845566,
      "loss": 5.3487,
      "mean_token_accuracy": 0.19976659417152404,
      "num_tokens": 39649726.0,
      "step": 21265
    },
    {
      "entropy": 5.884883880615234,
      "epoch": 1.8783115506888026,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004647453068661566,
      "loss": 5.3979,
      "mean_token_accuracy": 0.20370790958404542,
      "num_tokens": 39660226.0,
      "step": 21270
    },
    {
      "entropy": 5.824184226989746,
      "epoch": 1.878753090780643,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046472838073075397,
      "loss": 5.2184,
      "mean_token_accuracy": 0.21244992017745973,
      "num_tokens": 39668731.0,
      "step": 21275
    },
    {
      "entropy": 5.834587097167969,
      "epoch": 1.8791946308724832,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046471145087868034,
      "loss": 5.3708,
      "mean_token_accuracy": 0.20916611552238465,
      "num_tokens": 39678465.0,
      "step": 21280
    },
    {
      "entropy": 5.821716356277466,
      "epoch": 1.8796361709643237,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046469451731026725,
      "loss": 5.2415,
      "mean_token_accuracy": 0.2200051501393318,
      "num_tokens": 39687057.0,
      "step": 21285
    },
    {
      "entropy": 5.867373943328857,
      "epoch": 1.880077711056164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004646775800258464,
      "loss": 5.3661,
      "mean_token_accuracy": 0.20868303030729293,
      "num_tokens": 39696387.0,
      "step": 21290
    },
    {
      "entropy": 5.875208377838135,
      "epoch": 1.8805192511480042,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046466063902574973,
      "loss": 5.3156,
      "mean_token_accuracy": 0.21101562827825546,
      "num_tokens": 39705756.0,
      "step": 21295
    },
    {
      "entropy": 5.863362646102905,
      "epoch": 1.8809607912398447,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046464369431030894,
      "loss": 5.3434,
      "mean_token_accuracy": 0.2099301591515541,
      "num_tokens": 39713916.0,
      "step": 21300
    },
    {
      "entropy": 5.8952765464782715,
      "epoch": 1.8814023313316848,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000464626745879856,
      "loss": 5.4328,
      "mean_token_accuracy": 0.197401724755764,
      "num_tokens": 39722920.0,
      "step": 21305
    },
    {
      "entropy": 5.950076246261597,
      "epoch": 1.8818438714235253,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000464609793734723,
      "loss": 5.4163,
      "mean_token_accuracy": 0.199638032913208,
      "num_tokens": 39732217.0,
      "step": 21310
    },
    {
      "entropy": 5.821593332290649,
      "epoch": 1.8822854115153655,
      "grad_norm": 1.0,
      "learning_rate": 0.000464592837875242,
      "loss": 5.2585,
      "mean_token_accuracy": 0.21253640949726105,
      "num_tokens": 39741636.0,
      "step": 21315
    },
    {
      "entropy": 5.897186279296875,
      "epoch": 1.8827269516072058,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004645758783017451,
      "loss": 5.3943,
      "mean_token_accuracy": 0.20481854379177095,
      "num_tokens": 39751457.0,
      "step": 21320
    },
    {
      "entropy": 5.972907638549804,
      "epoch": 1.8831684916990463,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046455891501456456,
      "loss": 5.3531,
      "mean_token_accuracy": 0.20630601793527603,
      "num_tokens": 39760308.0,
      "step": 21325
    },
    {
      "entropy": 5.8535772323608395,
      "epoch": 1.8836100317908866,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046454194801403274,
      "loss": 5.3037,
      "mean_token_accuracy": 0.21301157027482986,
      "num_tokens": 39770002.0,
      "step": 21330
    },
    {
      "entropy": 5.815937948226929,
      "epoch": 1.8840515718827269,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004645249773004819,
      "loss": 5.3369,
      "mean_token_accuracy": 0.20491984188556672,
      "num_tokens": 39779043.0,
      "step": 21335
    },
    {
      "entropy": 5.884713268280029,
      "epoch": 1.8844931119745674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046450800287424455,
      "loss": 5.3563,
      "mean_token_accuracy": 0.20212042480707168,
      "num_tokens": 39788545.0,
      "step": 21340
    },
    {
      "entropy": 5.865634202957153,
      "epoch": 1.8849346520664076,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004644910247356532,
      "loss": 5.273,
      "mean_token_accuracy": 0.21859100908041001,
      "num_tokens": 39798063.0,
      "step": 21345
    },
    {
      "entropy": 5.872667264938355,
      "epoch": 1.885376192158248,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004644740428850404,
      "loss": 5.2906,
      "mean_token_accuracy": 0.21180068403482438,
      "num_tokens": 39808622.0,
      "step": 21350
    },
    {
      "entropy": 5.9357787609100345,
      "epoch": 1.8858177322500884,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004644570573227389,
      "loss": 5.4309,
      "mean_token_accuracy": 0.19891026765108108,
      "num_tokens": 39817608.0,
      "step": 21355
    },
    {
      "entropy": 5.848351001739502,
      "epoch": 1.8862592723419287,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046444006804908144,
      "loss": 5.2401,
      "mean_token_accuracy": 0.22318484485149384,
      "num_tokens": 39826055.0,
      "step": 21360
    },
    {
      "entropy": 5.762154769897461,
      "epoch": 1.886700812433769,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004644230750644007,
      "loss": 5.2688,
      "mean_token_accuracy": 0.21999183297157288,
      "num_tokens": 39835490.0,
      "step": 21365
    },
    {
      "entropy": 5.827249479293823,
      "epoch": 1.8871423525256095,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046440607836902974,
      "loss": 5.3533,
      "mean_token_accuracy": 0.21116445511579512,
      "num_tokens": 39845613.0,
      "step": 21370
    },
    {
      "entropy": 5.91176815032959,
      "epoch": 1.8875838926174495,
      "grad_norm": 1.125,
      "learning_rate": 0.00046438907796330145,
      "loss": 5.3532,
      "mean_token_accuracy": 0.20562867373228072,
      "num_tokens": 39854495.0,
      "step": 21375
    },
    {
      "entropy": 5.876791667938233,
      "epoch": 1.88802543270929,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00046437207384754874,
      "loss": 5.3752,
      "mean_token_accuracy": 0.20743363797664643,
      "num_tokens": 39864528.0,
      "step": 21380
    },
    {
      "entropy": 5.925540733337402,
      "epoch": 1.8884669728011303,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004643550660221048,
      "loss": 5.4246,
      "mean_token_accuracy": 0.20055753737688065,
      "num_tokens": 39873780.0,
      "step": 21385
    },
    {
      "entropy": 5.8628504276275635,
      "epoch": 1.8889085128929706,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046433805448730287,
      "loss": 5.2423,
      "mean_token_accuracy": 0.21738229244947432,
      "num_tokens": 39882529.0,
      "step": 21390
    },
    {
      "entropy": 5.947251176834106,
      "epoch": 1.889350052984811,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004643210392434761,
      "loss": 5.3834,
      "mean_token_accuracy": 0.2069010093808174,
      "num_tokens": 39890688.0,
      "step": 21395
    },
    {
      "entropy": 5.857395315170288,
      "epoch": 1.8897915930766513,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046430402029095784,
      "loss": 5.4122,
      "mean_token_accuracy": 0.19547487795352936,
      "num_tokens": 39900849.0,
      "step": 21400
    },
    {
      "entropy": 5.800002336502075,
      "epoch": 1.8902331331684916,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004642869976300815,
      "loss": 5.3773,
      "mean_token_accuracy": 0.20372939556837083,
      "num_tokens": 39911297.0,
      "step": 21405
    },
    {
      "entropy": 5.818081521987915,
      "epoch": 1.8906746732603321,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004642699712611806,
      "loss": 5.3169,
      "mean_token_accuracy": 0.20706899762153624,
      "num_tokens": 39920312.0,
      "step": 21410
    },
    {
      "entropy": 5.865331077575684,
      "epoch": 1.8911162133521724,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004642529411845885,
      "loss": 5.3976,
      "mean_token_accuracy": 0.20539042204618455,
      "num_tokens": 39930128.0,
      "step": 21415
    },
    {
      "entropy": 5.919701719284058,
      "epoch": 1.8915577534440127,
      "grad_norm": 1.0,
      "learning_rate": 0.0004642359074006389,
      "loss": 5.3389,
      "mean_token_accuracy": 0.21122706234455108,
      "num_tokens": 39939177.0,
      "step": 21420
    },
    {
      "entropy": 5.946272134780884,
      "epoch": 1.8919992935358532,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004642188699096656,
      "loss": 5.431,
      "mean_token_accuracy": 0.19848523736000062,
      "num_tokens": 39948537.0,
      "step": 21425
    },
    {
      "entropy": 5.893461275100708,
      "epoch": 1.8924408336276934,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004642018287120022,
      "loss": 5.3503,
      "mean_token_accuracy": 0.20792061239480972,
      "num_tokens": 39958623.0,
      "step": 21430
    },
    {
      "entropy": 5.926222562789917,
      "epoch": 1.8928823737195337,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046418478380798263,
      "loss": 5.4411,
      "mean_token_accuracy": 0.2013219639658928,
      "num_tokens": 39968736.0,
      "step": 21435
    },
    {
      "entropy": 5.769104671478272,
      "epoch": 1.8933239138113742,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004641677351979407,
      "loss": 5.2944,
      "mean_token_accuracy": 0.20479418337345123,
      "num_tokens": 39977600.0,
      "step": 21440
    },
    {
      "entropy": 5.847313785552979,
      "epoch": 1.8937654539032143,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004641506828822104,
      "loss": 5.402,
      "mean_token_accuracy": 0.1978820577263832,
      "num_tokens": 39987808.0,
      "step": 21445
    },
    {
      "entropy": 5.967379426956176,
      "epoch": 1.8942069939950548,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004641336268611258,
      "loss": 5.5056,
      "mean_token_accuracy": 0.20149057656526564,
      "num_tokens": 39998553.0,
      "step": 21450
    },
    {
      "entropy": 6.031639814376831,
      "epoch": 1.894648534086895,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046411656713502106,
      "loss": 5.4261,
      "mean_token_accuracy": 0.19868351370096207,
      "num_tokens": 40007529.0,
      "step": 21455
    },
    {
      "entropy": 5.936395454406738,
      "epoch": 1.8950900741787353,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004640995037042303,
      "loss": 5.4313,
      "mean_token_accuracy": 0.20063990950584412,
      "num_tokens": 40017116.0,
      "step": 21460
    },
    {
      "entropy": 5.888848066329956,
      "epoch": 1.8955316142705758,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004640824365690878,
      "loss": 5.4385,
      "mean_token_accuracy": 0.20070538073778152,
      "num_tokens": 40026601.0,
      "step": 21465
    },
    {
      "entropy": 5.894359588623047,
      "epoch": 1.895973154362416,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046406536572992795,
      "loss": 5.3302,
      "mean_token_accuracy": 0.21494613885879515,
      "num_tokens": 40036029.0,
      "step": 21470
    },
    {
      "entropy": 5.902152967453003,
      "epoch": 1.8964146944542564,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000464048291187085,
      "loss": 5.3687,
      "mean_token_accuracy": 0.20566665828227998,
      "num_tokens": 40045892.0,
      "step": 21475
    },
    {
      "entropy": 5.9180056095123295,
      "epoch": 1.8968562345460969,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004640312129408936,
      "loss": 5.3119,
      "mean_token_accuracy": 0.21512603461742402,
      "num_tokens": 40055411.0,
      "step": 21480
    },
    {
      "entropy": 5.936558055877685,
      "epoch": 1.8972977746379371,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004640141309916883,
      "loss": 5.405,
      "mean_token_accuracy": 0.2002963423728943,
      "num_tokens": 40063988.0,
      "step": 21485
    },
    {
      "entropy": 5.892999792098999,
      "epoch": 1.8977393147297774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004639970453398037,
      "loss": 5.354,
      "mean_token_accuracy": 0.2138413041830063,
      "num_tokens": 40072671.0,
      "step": 21490
    },
    {
      "entropy": 5.915765333175659,
      "epoch": 1.898180854821618,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046397995598557434,
      "loss": 5.4451,
      "mean_token_accuracy": 0.20091729909181594,
      "num_tokens": 40082191.0,
      "step": 21495
    },
    {
      "entropy": 5.877042198181153,
      "epoch": 1.8986223949134582,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004639628629293352,
      "loss": 5.4021,
      "mean_token_accuracy": 0.19798944145441055,
      "num_tokens": 40090823.0,
      "step": 21500
    },
    {
      "entropy": 5.910460090637207,
      "epoch": 1.8990639350052985,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000463945766171421,
      "loss": 5.3423,
      "mean_token_accuracy": 0.2063678190112114,
      "num_tokens": 40099825.0,
      "step": 21505
    },
    {
      "entropy": 5.934996223449707,
      "epoch": 1.899505475097139,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004639286657121668,
      "loss": 5.3691,
      "mean_token_accuracy": 0.20978682190179826,
      "num_tokens": 40109589.0,
      "step": 21510
    },
    {
      "entropy": 5.848345184326172,
      "epoch": 1.899947015188979,
      "grad_norm": 1.0,
      "learning_rate": 0.0004639115615519074,
      "loss": 5.3202,
      "mean_token_accuracy": 0.2055730938911438,
      "num_tokens": 40119194.0,
      "step": 21515
    },
    {
      "entropy": 5.886260843276977,
      "epoch": 1.9003885552808195,
      "grad_norm": 1.0625,
      "learning_rate": 0.000463894453690978,
      "loss": 5.4047,
      "mean_token_accuracy": 0.20396253168582917,
      "num_tokens": 40128184.0,
      "step": 21520
    },
    {
      "entropy": 5.89281997680664,
      "epoch": 1.9008300953726598,
      "grad_norm": 0.96875,
      "learning_rate": 0.00046387734212971363,
      "loss": 5.419,
      "mean_token_accuracy": 0.19890098571777343,
      "num_tokens": 40137842.0,
      "step": 21525
    },
    {
      "entropy": 5.902789640426636,
      "epoch": 1.9012716354645,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004638602268684496,
      "loss": 5.3791,
      "mean_token_accuracy": 0.2025577872991562,
      "num_tokens": 40147151.0,
      "step": 21530
    },
    {
      "entropy": 5.902479410171509,
      "epoch": 1.9017131755563406,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00046384310790752105,
      "loss": 5.3458,
      "mean_token_accuracy": 0.20398520231246947,
      "num_tokens": 40157170.0,
      "step": 21535
    },
    {
      "entropy": 5.898153638839721,
      "epoch": 1.9021547156481808,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004638259852472635,
      "loss": 5.3335,
      "mean_token_accuracy": 0.20552589893341064,
      "num_tokens": 40166787.0,
      "step": 21540
    },
    {
      "entropy": 5.887764644622803,
      "epoch": 1.9025962557400211,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004638088588880122,
      "loss": 5.2987,
      "mean_token_accuracy": 0.21487546265125274,
      "num_tokens": 40176640.0,
      "step": 21545
    },
    {
      "entropy": 5.88833384513855,
      "epoch": 1.9030377958318616,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004637917288301029,
      "loss": 5.3655,
      "mean_token_accuracy": 0.21044817119836806,
      "num_tokens": 40185732.0,
      "step": 21550
    },
    {
      "entropy": 5.9018439769744875,
      "epoch": 1.9034793359237019,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004637745950738709,
      "loss": 5.2949,
      "mean_token_accuracy": 0.21157171428203583,
      "num_tokens": 40194798.0,
      "step": 21555
    },
    {
      "entropy": 5.92890567779541,
      "epoch": 1.9039208760155422,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004637574576196519,
      "loss": 5.4597,
      "mean_token_accuracy": 0.20000536739826202,
      "num_tokens": 40204507.0,
      "step": 21560
    },
    {
      "entropy": 5.93493914604187,
      "epoch": 1.9043624161073827,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004637403164677817,
      "loss": 5.3832,
      "mean_token_accuracy": 0.1958839178085327,
      "num_tokens": 40213546.0,
      "step": 21565
    },
    {
      "entropy": 5.892339468002319,
      "epoch": 1.904803956199223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046372317161859605,
      "loss": 5.3211,
      "mean_token_accuracy": 0.20888868272304534,
      "num_tokens": 40222571.0,
      "step": 21570
    },
    {
      "entropy": 5.944922876358032,
      "epoch": 1.9052454962910632,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046370602307243083,
      "loss": 5.4081,
      "mean_token_accuracy": 0.20033878684043885,
      "num_tokens": 40231803.0,
      "step": 21575
    },
    {
      "entropy": 5.9365276336669925,
      "epoch": 1.9056870363829037,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046368887082962187,
      "loss": 5.4632,
      "mean_token_accuracy": 0.20124018490314483,
      "num_tokens": 40241755.0,
      "step": 21580
    },
    {
      "entropy": 5.888432312011719,
      "epoch": 1.9061285764747438,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004636717148905053,
      "loss": 5.3514,
      "mean_token_accuracy": 0.2052123323082924,
      "num_tokens": 40250963.0,
      "step": 21585
    },
    {
      "entropy": 5.93085503578186,
      "epoch": 1.9065701165665843,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004636545552554171,
      "loss": 5.347,
      "mean_token_accuracy": 0.20675989240407944,
      "num_tokens": 40260626.0,
      "step": 21590
    },
    {
      "entropy": 5.904647779464722,
      "epoch": 1.9070116566584245,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004636373919246935,
      "loss": 5.3249,
      "mean_token_accuracy": 0.20984554588794707,
      "num_tokens": 40269852.0,
      "step": 21595
    },
    {
      "entropy": 5.87044882774353,
      "epoch": 1.9074531967502648,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004636202248986707,
      "loss": 5.3358,
      "mean_token_accuracy": 0.2062729373574257,
      "num_tokens": 40279499.0,
      "step": 21600
    },
    {
      "entropy": 5.965744924545288,
      "epoch": 1.9078947368421053,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046360305417768484,
      "loss": 5.4459,
      "mean_token_accuracy": 0.19920772910118104,
      "num_tokens": 40288131.0,
      "step": 21605
    },
    {
      "entropy": 5.970402812957763,
      "epoch": 1.9083362769339456,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004635858797620725,
      "loss": 5.387,
      "mean_token_accuracy": 0.20197357684373857,
      "num_tokens": 40296951.0,
      "step": 21610
    },
    {
      "entropy": 5.882398843765259,
      "epoch": 1.9087778170257859,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046356870165217004,
      "loss": 5.3202,
      "mean_token_accuracy": 0.21134337037801743,
      "num_tokens": 40306250.0,
      "step": 21615
    },
    {
      "entropy": 5.811308526992798,
      "epoch": 1.9092193571176264,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046355151984831395,
      "loss": 5.2959,
      "mean_token_accuracy": 0.21645228415727616,
      "num_tokens": 40314561.0,
      "step": 21620
    },
    {
      "entropy": 5.788237428665161,
      "epoch": 1.9096608972094666,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046353433435084083,
      "loss": 5.2602,
      "mean_token_accuracy": 0.20963333547115326,
      "num_tokens": 40323580.0,
      "step": 21625
    },
    {
      "entropy": 5.838340187072754,
      "epoch": 1.910102437301307,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004635171451600872,
      "loss": 5.3438,
      "mean_token_accuracy": 0.20755885988473893,
      "num_tokens": 40333747.0,
      "step": 21630
    },
    {
      "entropy": 5.965235948562622,
      "epoch": 1.9105439773931474,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046349995227639015,
      "loss": 5.4485,
      "mean_token_accuracy": 0.19505509734153748,
      "num_tokens": 40343435.0,
      "step": 21635
    },
    {
      "entropy": 5.988936710357666,
      "epoch": 1.9109855174849877,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004634827557000861,
      "loss": 5.4062,
      "mean_token_accuracy": 0.20402436703443527,
      "num_tokens": 40353282.0,
      "step": 21640
    },
    {
      "entropy": 5.904768800735473,
      "epoch": 1.911427057576828,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000463465555431512,
      "loss": 5.4503,
      "mean_token_accuracy": 0.1991128444671631,
      "num_tokens": 40363781.0,
      "step": 21645
    },
    {
      "entropy": 5.879390859603882,
      "epoch": 1.9118685976686685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046344835147100495,
      "loss": 5.3562,
      "mean_token_accuracy": 0.20000342428684234,
      "num_tokens": 40373240.0,
      "step": 21650
    },
    {
      "entropy": 5.894801187515259,
      "epoch": 1.9123101377605085,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046343114381890187,
      "loss": 5.4388,
      "mean_token_accuracy": 0.20514141023159027,
      "num_tokens": 40383088.0,
      "step": 21655
    },
    {
      "entropy": 5.984843826293945,
      "epoch": 1.912751677852349,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004634139324755398,
      "loss": 5.3974,
      "mean_token_accuracy": 0.21017206609249114,
      "num_tokens": 40391654.0,
      "step": 21660
    },
    {
      "entropy": 5.879476881027221,
      "epoch": 1.9131932179441893,
      "grad_norm": 1.125,
      "learning_rate": 0.000463396717441256,
      "loss": 5.2974,
      "mean_token_accuracy": 0.2165357306599617,
      "num_tokens": 40400616.0,
      "step": 21665
    },
    {
      "entropy": 5.811811876296997,
      "epoch": 1.9136347580360296,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004633794987163876,
      "loss": 5.3681,
      "mean_token_accuracy": 0.19875801354646683,
      "num_tokens": 40411271.0,
      "step": 21670
    },
    {
      "entropy": 5.89321117401123,
      "epoch": 1.91407629812787,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046336227630127205,
      "loss": 5.2857,
      "mean_token_accuracy": 0.20427657961845397,
      "num_tokens": 40420019.0,
      "step": 21675
    },
    {
      "entropy": 5.9374107837677,
      "epoch": 1.9145178382197103,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046334505019624656,
      "loss": 5.3641,
      "mean_token_accuracy": 0.20939703285694122,
      "num_tokens": 40429858.0,
      "step": 21680
    },
    {
      "entropy": 5.84266414642334,
      "epoch": 1.9149593783115506,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046332782040164867,
      "loss": 5.3403,
      "mean_token_accuracy": 0.21174548715353012,
      "num_tokens": 40439566.0,
      "step": 21685
    },
    {
      "entropy": 5.845394515991211,
      "epoch": 1.915400918403391,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004633105869178159,
      "loss": 5.2912,
      "mean_token_accuracy": 0.21488560289144515,
      "num_tokens": 40448527.0,
      "step": 21690
    },
    {
      "entropy": 5.875892210006714,
      "epoch": 1.9158424584952314,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004632933497450858,
      "loss": 5.3107,
      "mean_token_accuracy": 0.21092088520526886,
      "num_tokens": 40458076.0,
      "step": 21695
    },
    {
      "entropy": 5.843246698379517,
      "epoch": 1.9162839985870717,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004632761088837961,
      "loss": 5.2335,
      "mean_token_accuracy": 0.22885193526744843,
      "num_tokens": 40467639.0,
      "step": 21700
    },
    {
      "entropy": 5.881030607223511,
      "epoch": 1.9167255386789122,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046325886433428446,
      "loss": 5.3242,
      "mean_token_accuracy": 0.20737940818071365,
      "num_tokens": 40476243.0,
      "step": 21705
    },
    {
      "entropy": 5.891556739807129,
      "epoch": 1.9171670787707524,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046324161609688876,
      "loss": 5.3737,
      "mean_token_accuracy": 0.2093565955758095,
      "num_tokens": 40485789.0,
      "step": 21710
    },
    {
      "entropy": 5.822904968261719,
      "epoch": 1.9176086188625927,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004632243641719469,
      "loss": 5.3087,
      "mean_token_accuracy": 0.20545696020126342,
      "num_tokens": 40494139.0,
      "step": 21715
    },
    {
      "entropy": 5.857991886138916,
      "epoch": 1.9180501589544332,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046320710855979677,
      "loss": 5.3234,
      "mean_token_accuracy": 0.2088524729013443,
      "num_tokens": 40503965.0,
      "step": 21720
    },
    {
      "entropy": 5.859075736999512,
      "epoch": 1.9184916990462733,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046318984926077637,
      "loss": 5.3339,
      "mean_token_accuracy": 0.21351005136966705,
      "num_tokens": 40513140.0,
      "step": 21725
    },
    {
      "entropy": 5.830721712112426,
      "epoch": 1.9189332391381138,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046317258627522393,
      "loss": 5.334,
      "mean_token_accuracy": 0.2193341925740242,
      "num_tokens": 40522858.0,
      "step": 21730
    },
    {
      "entropy": 5.832539844512939,
      "epoch": 1.919374779229954,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004631553196034774,
      "loss": 5.3281,
      "mean_token_accuracy": 0.21606432199478148,
      "num_tokens": 40532050.0,
      "step": 21735
    },
    {
      "entropy": 5.85909743309021,
      "epoch": 1.9198163193217943,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004631380492458753,
      "loss": 5.3193,
      "mean_token_accuracy": 0.20459819585084915,
      "num_tokens": 40540942.0,
      "step": 21740
    },
    {
      "entropy": 5.9041587829589846,
      "epoch": 1.9202578594136348,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004631207752027558,
      "loss": 5.3587,
      "mean_token_accuracy": 0.2107951134443283,
      "num_tokens": 40550158.0,
      "step": 21745
    },
    {
      "entropy": 5.971966934204102,
      "epoch": 1.920699399505475,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004631034974744573,
      "loss": 5.4672,
      "mean_token_accuracy": 0.19536066353321074,
      "num_tokens": 40559267.0,
      "step": 21750
    },
    {
      "entropy": 5.903441667556763,
      "epoch": 1.9211409395973154,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004630862160613183,
      "loss": 5.362,
      "mean_token_accuracy": 0.20459360331296922,
      "num_tokens": 40568503.0,
      "step": 21755
    },
    {
      "entropy": 5.877523279190063,
      "epoch": 1.9215824796891559,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004630689309636772,
      "loss": 5.3063,
      "mean_token_accuracy": 0.21025119721889496,
      "num_tokens": 40578464.0,
      "step": 21760
    },
    {
      "entropy": 5.955185747146606,
      "epoch": 1.9220240197809961,
      "grad_norm": 1.0,
      "learning_rate": 0.0004630516421818728,
      "loss": 5.4656,
      "mean_token_accuracy": 0.1963024139404297,
      "num_tokens": 40588338.0,
      "step": 21765
    },
    {
      "entropy": 5.934965562820435,
      "epoch": 1.9224655598728364,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046303434971624357,
      "loss": 5.3346,
      "mean_token_accuracy": 0.21062635779380798,
      "num_tokens": 40598912.0,
      "step": 21770
    },
    {
      "entropy": 5.812967967987061,
      "epoch": 1.922907099964677,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046301705356712846,
      "loss": 5.2545,
      "mean_token_accuracy": 0.2087301179766655,
      "num_tokens": 40607084.0,
      "step": 21775
    },
    {
      "entropy": 5.886069917678833,
      "epoch": 1.9233486400565172,
      "grad_norm": 1.125,
      "learning_rate": 0.00046299975373486626,
      "loss": 5.3648,
      "mean_token_accuracy": 0.2038731813430786,
      "num_tokens": 40616422.0,
      "step": 21780
    },
    {
      "entropy": 5.845412397384644,
      "epoch": 1.9237901801483575,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004629824502197956,
      "loss": 5.3053,
      "mean_token_accuracy": 0.20799794048070908,
      "num_tokens": 40626189.0,
      "step": 21785
    },
    {
      "entropy": 5.83835654258728,
      "epoch": 1.924231720240198,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046296514302225585,
      "loss": 5.2614,
      "mean_token_accuracy": 0.20808082967996597,
      "num_tokens": 40636053.0,
      "step": 21790
    },
    {
      "entropy": 5.927354049682617,
      "epoch": 1.924673260332038,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004629478321425858,
      "loss": 5.4047,
      "mean_token_accuracy": 0.19723138511180877,
      "num_tokens": 40645159.0,
      "step": 21795
    },
    {
      "entropy": 5.872242498397827,
      "epoch": 1.9251148004238785,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046293051758112457,
      "loss": 5.2596,
      "mean_token_accuracy": 0.2172661915421486,
      "num_tokens": 40653890.0,
      "step": 21800
    },
    {
      "entropy": 5.955357456207276,
      "epoch": 1.9255563405157188,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046291319933821143,
      "loss": 5.4262,
      "mean_token_accuracy": 0.19706215262413024,
      "num_tokens": 40663041.0,
      "step": 21805
    },
    {
      "entropy": 5.833741331100464,
      "epoch": 1.925997880607559,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046289587741418553,
      "loss": 5.2702,
      "mean_token_accuracy": 0.2103172242641449,
      "num_tokens": 40670972.0,
      "step": 21810
    },
    {
      "entropy": 5.877347087860107,
      "epoch": 1.9264394206993996,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046287855180938633,
      "loss": 5.3996,
      "mean_token_accuracy": 0.20953988283872604,
      "num_tokens": 40680503.0,
      "step": 21815
    },
    {
      "entropy": 5.869355058670044,
      "epoch": 1.9268809607912398,
      "grad_norm": 0.984375,
      "learning_rate": 0.00046286122252415303,
      "loss": 5.3944,
      "mean_token_accuracy": 0.2054565280675888,
      "num_tokens": 40691339.0,
      "step": 21820
    },
    {
      "entropy": 5.9580910205841064,
      "epoch": 1.92732250088308,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046284388955882526,
      "loss": 5.3525,
      "mean_token_accuracy": 0.2055668607354164,
      "num_tokens": 40699744.0,
      "step": 21825
    },
    {
      "entropy": 5.965705442428589,
      "epoch": 1.9277640409749206,
      "grad_norm": 1.0,
      "learning_rate": 0.00046282655291374257,
      "loss": 5.4008,
      "mean_token_accuracy": 0.19989280253648758,
      "num_tokens": 40709144.0,
      "step": 21830
    },
    {
      "entropy": 5.842627954483032,
      "epoch": 1.9282055810667609,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00046280921258924455,
      "loss": 5.4256,
      "mean_token_accuracy": 0.20699867308139802,
      "num_tokens": 40719513.0,
      "step": 21835
    },
    {
      "entropy": 5.872743844985962,
      "epoch": 1.9286471211586012,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046279186858567083,
      "loss": 5.3509,
      "mean_token_accuracy": 0.20761857479810714,
      "num_tokens": 40728846.0,
      "step": 21840
    },
    {
      "entropy": 5.857518672943115,
      "epoch": 1.9290886612504416,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004627745209033613,
      "loss": 5.27,
      "mean_token_accuracy": 0.21707218736410142,
      "num_tokens": 40737904.0,
      "step": 21845
    },
    {
      "entropy": 5.819572305679321,
      "epoch": 1.929530201342282,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004627571695426556,
      "loss": 5.2947,
      "mean_token_accuracy": 0.21966431438922882,
      "num_tokens": 40747328.0,
      "step": 21850
    },
    {
      "entropy": 5.912659072875977,
      "epoch": 1.9299717414341222,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046273981450389374,
      "loss": 5.3293,
      "mean_token_accuracy": 0.2070286139845848,
      "num_tokens": 40755567.0,
      "step": 21855
    },
    {
      "entropy": 5.850630807876587,
      "epoch": 1.9304132815259627,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046272245578741567,
      "loss": 5.3404,
      "mean_token_accuracy": 0.20996105968952178,
      "num_tokens": 40764684.0,
      "step": 21860
    },
    {
      "entropy": 5.876852512359619,
      "epoch": 1.9308548216178028,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004627050933935615,
      "loss": 5.3712,
      "mean_token_accuracy": 0.21365745067596437,
      "num_tokens": 40774050.0,
      "step": 21865
    },
    {
      "entropy": 5.871454763412475,
      "epoch": 1.9312963617096433,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046268772732267126,
      "loss": 5.3103,
      "mean_token_accuracy": 0.21163002997636796,
      "num_tokens": 40782722.0,
      "step": 21870
    },
    {
      "entropy": 5.886638879776001,
      "epoch": 1.9317379018014835,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004626703575750852,
      "loss": 5.3838,
      "mean_token_accuracy": 0.20145509093999864,
      "num_tokens": 40791978.0,
      "step": 21875
    },
    {
      "entropy": 5.892842483520508,
      "epoch": 1.9321794418933238,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004626529841511436,
      "loss": 5.3909,
      "mean_token_accuracy": 0.20379009544849397,
      "num_tokens": 40801855.0,
      "step": 21880
    },
    {
      "entropy": 5.903077697753906,
      "epoch": 1.9326209819851643,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046263560705118675,
      "loss": 5.4645,
      "mean_token_accuracy": 0.19780069440603257,
      "num_tokens": 40812025.0,
      "step": 21885
    },
    {
      "entropy": 5.926669549942017,
      "epoch": 1.9330625220770046,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004626182262755551,
      "loss": 5.2774,
      "mean_token_accuracy": 0.20607707649469376,
      "num_tokens": 40820758.0,
      "step": 21890
    },
    {
      "entropy": 5.917462491989136,
      "epoch": 1.9335040621688449,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046260084182458906,
      "loss": 5.3892,
      "mean_token_accuracy": 0.20371164977550507,
      "num_tokens": 40830653.0,
      "step": 21895
    },
    {
      "entropy": 5.888490962982178,
      "epoch": 1.9339456022606853,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046258345369862915,
      "loss": 5.4622,
      "mean_token_accuracy": 0.19466545432806015,
      "num_tokens": 40840370.0,
      "step": 21900
    },
    {
      "entropy": 5.878164577484131,
      "epoch": 1.9343871423525256,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004625660618980162,
      "loss": 5.3782,
      "mean_token_accuracy": 0.20062892138957977,
      "num_tokens": 40849590.0,
      "step": 21905
    },
    {
      "entropy": 5.906628751754761,
      "epoch": 1.934828682444366,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00046254866642309067,
      "loss": 5.3512,
      "mean_token_accuracy": 0.2105919674038887,
      "num_tokens": 40859271.0,
      "step": 21910
    },
    {
      "entropy": 5.885888576507568,
      "epoch": 1.9352702225362064,
      "grad_norm": 1.0,
      "learning_rate": 0.0004625312672741935,
      "loss": 5.3613,
      "mean_token_accuracy": 0.2009398505091667,
      "num_tokens": 40869272.0,
      "step": 21915
    },
    {
      "entropy": 5.956877517700195,
      "epoch": 1.9357117626280467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004625138644516655,
      "loss": 5.4494,
      "mean_token_accuracy": 0.19722941368818284,
      "num_tokens": 40880327.0,
      "step": 21920
    },
    {
      "entropy": 5.873924922943115,
      "epoch": 1.936153302719887,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004624964579558474,
      "loss": 5.3604,
      "mean_token_accuracy": 0.20672069489955902,
      "num_tokens": 40890407.0,
      "step": 21925
    },
    {
      "entropy": 5.893722009658814,
      "epoch": 1.9365948428117274,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004624790477870804,
      "loss": 5.3485,
      "mean_token_accuracy": 0.20273849219083787,
      "num_tokens": 40899306.0,
      "step": 21930
    },
    {
      "entropy": 5.911926555633545,
      "epoch": 1.9370363829035675,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004624616339457055,
      "loss": 5.4071,
      "mean_token_accuracy": 0.20775255858898162,
      "num_tokens": 40908908.0,
      "step": 21935
    },
    {
      "entropy": 5.931298780441284,
      "epoch": 1.937477922995408,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004624442164320638,
      "loss": 5.3654,
      "mean_token_accuracy": 0.20946729332208633,
      "num_tokens": 40917960.0,
      "step": 21940
    },
    {
      "entropy": 5.805299949645996,
      "epoch": 1.9379194630872483,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004624267952464965,
      "loss": 5.1944,
      "mean_token_accuracy": 0.2136712431907654,
      "num_tokens": 40926364.0,
      "step": 21945
    },
    {
      "entropy": 5.8090832233428955,
      "epoch": 1.9383610031790885,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046240937038934484,
      "loss": 5.3642,
      "mean_token_accuracy": 0.2044597715139389,
      "num_tokens": 40936558.0,
      "step": 21950
    },
    {
      "entropy": 5.841175937652588,
      "epoch": 1.938802543270929,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004623919418609502,
      "loss": 5.3497,
      "mean_token_accuracy": 0.21050939559936524,
      "num_tokens": 40945501.0,
      "step": 21955
    },
    {
      "entropy": 5.828894138336182,
      "epoch": 1.9392440833627693,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046237450966165407,
      "loss": 5.3282,
      "mean_token_accuracy": 0.20923472940921783,
      "num_tokens": 40954740.0,
      "step": 21960
    },
    {
      "entropy": 5.901995515823364,
      "epoch": 1.9396856234546096,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046235707379179775,
      "loss": 5.379,
      "mean_token_accuracy": 0.2061217874288559,
      "num_tokens": 40963748.0,
      "step": 21965
    },
    {
      "entropy": 5.956651735305786,
      "epoch": 1.94012716354645,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046233963425172306,
      "loss": 5.3655,
      "mean_token_accuracy": 0.20245062410831452,
      "num_tokens": 40973907.0,
      "step": 21970
    },
    {
      "entropy": 5.903942346572876,
      "epoch": 1.9405687036382904,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046232219104177136,
      "loss": 5.3956,
      "mean_token_accuracy": 0.19936734139919282,
      "num_tokens": 40983374.0,
      "step": 21975
    },
    {
      "entropy": 5.920079612731934,
      "epoch": 1.9410102437301306,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004623047441622845,
      "loss": 5.3908,
      "mean_token_accuracy": 0.1982327714562416,
      "num_tokens": 40993305.0,
      "step": 21980
    },
    {
      "entropy": 5.939617538452149,
      "epoch": 1.9414517838219711,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046228729361360427,
      "loss": 5.3369,
      "mean_token_accuracy": 0.19725982695817948,
      "num_tokens": 41002438.0,
      "step": 21985
    },
    {
      "entropy": 5.772862243652344,
      "epoch": 1.9418933239138114,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00046226983939607236,
      "loss": 5.3019,
      "mean_token_accuracy": 0.20948296785354614,
      "num_tokens": 41011491.0,
      "step": 21990
    },
    {
      "entropy": 5.908985710144043,
      "epoch": 1.9423348640056517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004622523815100309,
      "loss": 5.3901,
      "mean_token_accuracy": 0.1993059992790222,
      "num_tokens": 41021091.0,
      "step": 21995
    },
    {
      "entropy": 5.883903312683105,
      "epoch": 1.9427764040974922,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046223491995582187,
      "loss": 5.3328,
      "mean_token_accuracy": 0.20751825124025344,
      "num_tokens": 41030719.0,
      "step": 22000
    },
    {
      "entropy": 5.899147081375122,
      "epoch": 1.9432179441893322,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004622174547337871,
      "loss": 5.4061,
      "mean_token_accuracy": 0.2048639938235283,
      "num_tokens": 41039651.0,
      "step": 22005
    },
    {
      "entropy": 5.894042873382569,
      "epoch": 1.9436594842811727,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046219998584426885,
      "loss": 5.3031,
      "mean_token_accuracy": 0.20808445811271667,
      "num_tokens": 41047914.0,
      "step": 22010
    },
    {
      "entropy": 5.855243730545044,
      "epoch": 1.944101024373013,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004621825132876094,
      "loss": 5.4121,
      "mean_token_accuracy": 0.20523531883955,
      "num_tokens": 41056779.0,
      "step": 22015
    },
    {
      "entropy": 5.859929800033569,
      "epoch": 1.9445425644648533,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046216503706415096,
      "loss": 5.361,
      "mean_token_accuracy": 0.20483759939670562,
      "num_tokens": 41066356.0,
      "step": 22020
    },
    {
      "entropy": 5.94675235748291,
      "epoch": 1.9449841045566938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004621475571742359,
      "loss": 5.3751,
      "mean_token_accuracy": 0.21237125992774963,
      "num_tokens": 41075766.0,
      "step": 22025
    },
    {
      "entropy": 5.8732929706573485,
      "epoch": 1.945425644648534,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046213007361820663,
      "loss": 5.3639,
      "mean_token_accuracy": 0.2066869780421257,
      "num_tokens": 41085503.0,
      "step": 22030
    },
    {
      "entropy": 6.006605625152588,
      "epoch": 1.9458671847403743,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004621125863964057,
      "loss": 5.5377,
      "mean_token_accuracy": 0.18834761828184127,
      "num_tokens": 41096581.0,
      "step": 22035
    },
    {
      "entropy": 5.945880317687989,
      "epoch": 1.9463087248322148,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046209509550917553,
      "loss": 5.3451,
      "mean_token_accuracy": 0.21192478984594346,
      "num_tokens": 41105226.0,
      "step": 22040
    },
    {
      "entropy": 5.95869255065918,
      "epoch": 1.9467502649240551,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004620776009568589,
      "loss": 5.3417,
      "mean_token_accuracy": 0.20510756820440293,
      "num_tokens": 41113998.0,
      "step": 22045
    },
    {
      "entropy": 5.841848468780517,
      "epoch": 1.9471918050158954,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004620601027397984,
      "loss": 5.3684,
      "mean_token_accuracy": 0.19611532390117645,
      "num_tokens": 41123147.0,
      "step": 22050
    },
    {
      "entropy": 5.871012496948242,
      "epoch": 1.947633345107736,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046204260085833695,
      "loss": 5.3515,
      "mean_token_accuracy": 0.20897799581289292,
      "num_tokens": 41132317.0,
      "step": 22055
    },
    {
      "entropy": 5.92749056816101,
      "epoch": 1.9480748851995762,
      "grad_norm": 1.125,
      "learning_rate": 0.0004620250953128173,
      "loss": 5.3976,
      "mean_token_accuracy": 0.197231525182724,
      "num_tokens": 41140795.0,
      "step": 22060
    },
    {
      "entropy": 5.919864130020142,
      "epoch": 1.9485164252914164,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004620075861035824,
      "loss": 5.3554,
      "mean_token_accuracy": 0.2035280168056488,
      "num_tokens": 41149959.0,
      "step": 22065
    },
    {
      "entropy": 5.929309368133545,
      "epoch": 1.948957965383257,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004619900732309753,
      "loss": 5.4304,
      "mean_token_accuracy": 0.19045210033655166,
      "num_tokens": 41159388.0,
      "step": 22070
    },
    {
      "entropy": 6.005502319335937,
      "epoch": 1.949399505475097,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046197255669533886,
      "loss": 5.6157,
      "mean_token_accuracy": 0.1881886512041092,
      "num_tokens": 41168875.0,
      "step": 22075
    },
    {
      "entropy": 5.972923135757446,
      "epoch": 1.9498410455669375,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046195503649701646,
      "loss": 5.3954,
      "mean_token_accuracy": 0.2040553331375122,
      "num_tokens": 41177621.0,
      "step": 22080
    },
    {
      "entropy": 5.989239120483399,
      "epoch": 1.9502825856587778,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004619375126363513,
      "loss": 5.4358,
      "mean_token_accuracy": 0.20210410058498382,
      "num_tokens": 41185911.0,
      "step": 22085
    },
    {
      "entropy": 5.921224975585938,
      "epoch": 1.950724125750618,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046191998511368646,
      "loss": 5.3268,
      "mean_token_accuracy": 0.20728100091218948,
      "num_tokens": 41194811.0,
      "step": 22090
    },
    {
      "entropy": 5.7517650604248045,
      "epoch": 1.9511656658424585,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046190245392936554,
      "loss": 5.2578,
      "mean_token_accuracy": 0.21629572361707688,
      "num_tokens": 41203955.0,
      "step": 22095
    },
    {
      "entropy": 5.836204576492309,
      "epoch": 1.9516072059342988,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004618849190837317,
      "loss": 5.2955,
      "mean_token_accuracy": 0.2150256708264351,
      "num_tokens": 41212635.0,
      "step": 22100
    },
    {
      "entropy": 5.896378040313721,
      "epoch": 1.952048746026139,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004618673805771287,
      "loss": 5.3621,
      "mean_token_accuracy": 0.20168523192405702,
      "num_tokens": 41221443.0,
      "step": 22105
    },
    {
      "entropy": 5.883521270751953,
      "epoch": 1.9524902861179796,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004618498384098999,
      "loss": 5.3303,
      "mean_token_accuracy": 0.2048863038420677,
      "num_tokens": 41230089.0,
      "step": 22110
    },
    {
      "entropy": 5.95154218673706,
      "epoch": 1.9529318262098199,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046183229258238905,
      "loss": 5.4687,
      "mean_token_accuracy": 0.19285901486873627,
      "num_tokens": 41239569.0,
      "step": 22115
    },
    {
      "entropy": 5.9417143821716305,
      "epoch": 1.9533733663016601,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004618147430949399,
      "loss": 5.3535,
      "mean_token_accuracy": 0.2068372204899788,
      "num_tokens": 41248394.0,
      "step": 22120
    },
    {
      "entropy": 5.878524303436279,
      "epoch": 1.9538149063935006,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046179718994789624,
      "loss": 5.3575,
      "mean_token_accuracy": 0.20982825756072998,
      "num_tokens": 41258148.0,
      "step": 22125
    },
    {
      "entropy": 5.876276159286499,
      "epoch": 1.954256446485341,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046177963314160175,
      "loss": 5.2879,
      "mean_token_accuracy": 0.21348503977060318,
      "num_tokens": 41266984.0,
      "step": 22130
    },
    {
      "entropy": 5.91729097366333,
      "epoch": 1.9546979865771812,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046176207267640054,
      "loss": 5.4063,
      "mean_token_accuracy": 0.19888876527547836,
      "num_tokens": 41276119.0,
      "step": 22135
    },
    {
      "entropy": 5.7885355949401855,
      "epoch": 1.9551395266690217,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004617445085526365,
      "loss": 5.3038,
      "mean_token_accuracy": 0.21451292783021927,
      "num_tokens": 41284945.0,
      "step": 22140
    },
    {
      "entropy": 5.871628189086914,
      "epoch": 1.9555810667608617,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004617269407706537,
      "loss": 5.291,
      "mean_token_accuracy": 0.21350942850112914,
      "num_tokens": 41294255.0,
      "step": 22145
    },
    {
      "entropy": 5.896035385131836,
      "epoch": 1.9560226068527022,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004617093693307964,
      "loss": 5.3977,
      "mean_token_accuracy": 0.20890733301639558,
      "num_tokens": 41303056.0,
      "step": 22150
    },
    {
      "entropy": 5.909364032745361,
      "epoch": 1.9564641469445425,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046169179423340866,
      "loss": 5.4023,
      "mean_token_accuracy": 0.2050107091665268,
      "num_tokens": 41312444.0,
      "step": 22155
    },
    {
      "entropy": 5.9884837627410885,
      "epoch": 1.9569056870363828,
      "grad_norm": 1.09375,
      "learning_rate": 0.000461674215478835,
      "loss": 5.4919,
      "mean_token_accuracy": 0.20010593831539153,
      "num_tokens": 41322507.0,
      "step": 22160
    },
    {
      "entropy": 5.947047901153565,
      "epoch": 1.9573472271282233,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004616566330674195,
      "loss": 5.3652,
      "mean_token_accuracy": 0.2078099951148033,
      "num_tokens": 41332507.0,
      "step": 22165
    },
    {
      "entropy": 5.927550983428955,
      "epoch": 1.9577887672200636,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004616390469995068,
      "loss": 5.3223,
      "mean_token_accuracy": 0.20718599557876588,
      "num_tokens": 41342085.0,
      "step": 22170
    },
    {
      "entropy": 5.835907220840454,
      "epoch": 1.9582303073119038,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046162145727544126,
      "loss": 5.3101,
      "mean_token_accuracy": 0.21252421885728837,
      "num_tokens": 41351422.0,
      "step": 22175
    },
    {
      "entropy": 5.84026517868042,
      "epoch": 1.9586718474037443,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004616038638955675,
      "loss": 5.345,
      "mean_token_accuracy": 0.21026050448417663,
      "num_tokens": 41361337.0,
      "step": 22180
    },
    {
      "entropy": 5.82983136177063,
      "epoch": 1.9591133874955846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046158626686023014,
      "loss": 5.2765,
      "mean_token_accuracy": 0.21099374890327455,
      "num_tokens": 41370734.0,
      "step": 22185
    },
    {
      "entropy": 5.88320779800415,
      "epoch": 1.9595549275874249,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046156866616977406,
      "loss": 5.3056,
      "mean_token_accuracy": 0.20598259270191194,
      "num_tokens": 41380090.0,
      "step": 22190
    },
    {
      "entropy": 5.965227508544922,
      "epoch": 1.9599964676792654,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046155106182454384,
      "loss": 5.3792,
      "mean_token_accuracy": 0.2085856541991234,
      "num_tokens": 41388967.0,
      "step": 22195
    },
    {
      "entropy": 5.897452068328858,
      "epoch": 1.9604380077711057,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004615334538248844,
      "loss": 5.3517,
      "mean_token_accuracy": 0.2055166631937027,
      "num_tokens": 41397429.0,
      "step": 22200
    },
    {
      "entropy": 5.816228199005127,
      "epoch": 1.960879547862946,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046151584217114073,
      "loss": 5.3022,
      "mean_token_accuracy": 0.20593734234571456,
      "num_tokens": 41406420.0,
      "step": 22205
    },
    {
      "entropy": 5.8074297428131105,
      "epoch": 1.9613210879547864,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004614982268636577,
      "loss": 5.2603,
      "mean_token_accuracy": 0.22167766094207764,
      "num_tokens": 41415994.0,
      "step": 22210
    },
    {
      "entropy": 5.877642250061035,
      "epoch": 1.9617626280466265,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004614806079027806,
      "loss": 5.2613,
      "mean_token_accuracy": 0.215389646589756,
      "num_tokens": 41425744.0,
      "step": 22215
    },
    {
      "entropy": 5.904153633117676,
      "epoch": 1.962204168138467,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046146298528885435,
      "loss": 5.443,
      "mean_token_accuracy": 0.19678525477647782,
      "num_tokens": 41435452.0,
      "step": 22220
    },
    {
      "entropy": 5.941742277145385,
      "epoch": 1.9626457082303073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004614453590222243,
      "loss": 5.4029,
      "mean_token_accuracy": 0.20584166049957275,
      "num_tokens": 41444853.0,
      "step": 22225
    },
    {
      "entropy": 5.794569444656372,
      "epoch": 1.9630872483221475,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004614277291032357,
      "loss": 5.2619,
      "mean_token_accuracy": 0.2254179000854492,
      "num_tokens": 41454569.0,
      "step": 22230
    },
    {
      "entropy": 5.952146291732788,
      "epoch": 1.963528788413988,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004614100955322339,
      "loss": 5.4187,
      "mean_token_accuracy": 0.2028772696852684,
      "num_tokens": 41464159.0,
      "step": 22235
    },
    {
      "entropy": 5.957342052459717,
      "epoch": 1.9639703285058283,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004613924583095644,
      "loss": 5.4372,
      "mean_token_accuracy": 0.1950291946530342,
      "num_tokens": 41474354.0,
      "step": 22240
    },
    {
      "entropy": 5.849534130096435,
      "epoch": 1.9644118685976686,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004613748174355726,
      "loss": 5.2956,
      "mean_token_accuracy": 0.20756881237030028,
      "num_tokens": 41484053.0,
      "step": 22245
    },
    {
      "entropy": 5.944089937210083,
      "epoch": 1.964853408689509,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046135717291060415,
      "loss": 5.3526,
      "mean_token_accuracy": 0.1971958577632904,
      "num_tokens": 41493377.0,
      "step": 22250
    },
    {
      "entropy": 5.902534770965576,
      "epoch": 1.9652949487813494,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046133952473500464,
      "loss": 5.3158,
      "mean_token_accuracy": 0.21343559622764588,
      "num_tokens": 41502361.0,
      "step": 22255
    },
    {
      "entropy": 5.886488914489746,
      "epoch": 1.9657364888731896,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004613218729091198,
      "loss": 5.3554,
      "mean_token_accuracy": 0.20211567282676696,
      "num_tokens": 41511083.0,
      "step": 22260
    },
    {
      "entropy": 5.896930932998657,
      "epoch": 1.9661780289650301,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004613042174332955,
      "loss": 5.4491,
      "mean_token_accuracy": 0.19491046518087388,
      "num_tokens": 41520264.0,
      "step": 22265
    },
    {
      "entropy": 5.846308040618896,
      "epoch": 1.9666195690568704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004612865583078775,
      "loss": 5.2855,
      "mean_token_accuracy": 0.2126755714416504,
      "num_tokens": 41529805.0,
      "step": 22270
    },
    {
      "entropy": 5.827733278274536,
      "epoch": 1.9670611091487107,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004612688955332118,
      "loss": 5.2406,
      "mean_token_accuracy": 0.20872793793678285,
      "num_tokens": 41538348.0,
      "step": 22275
    },
    {
      "entropy": 5.87517786026001,
      "epoch": 1.9675026492405512,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046125122910964427,
      "loss": 5.3714,
      "mean_token_accuracy": 0.20601751357316972,
      "num_tokens": 41547433.0,
      "step": 22280
    },
    {
      "entropy": 5.931604194641113,
      "epoch": 1.9679441893323912,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004612335590375212,
      "loss": 5.3629,
      "mean_token_accuracy": 0.20532096326351165,
      "num_tokens": 41556162.0,
      "step": 22285
    },
    {
      "entropy": 5.912998962402344,
      "epoch": 1.9683857294242317,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004612158853171886,
      "loss": 5.3336,
      "mean_token_accuracy": 0.20705235749483109,
      "num_tokens": 41565415.0,
      "step": 22290
    },
    {
      "entropy": 5.832032537460327,
      "epoch": 1.968827269516072,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004611982079489928,
      "loss": 5.2932,
      "mean_token_accuracy": 0.21562063694000244,
      "num_tokens": 41574728.0,
      "step": 22295
    },
    {
      "entropy": 5.880598640441894,
      "epoch": 1.9692688096079123,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004611805269332798,
      "loss": 5.3814,
      "mean_token_accuracy": 0.19584305733442306,
      "num_tokens": 41584749.0,
      "step": 22300
    },
    {
      "entropy": 5.876179647445679,
      "epoch": 1.9697103496997528,
      "grad_norm": 1.0,
      "learning_rate": 0.00046116284227039626,
      "loss": 5.3288,
      "mean_token_accuracy": 0.2164471209049225,
      "num_tokens": 41594658.0,
      "step": 22305
    },
    {
      "entropy": 5.891310548782348,
      "epoch": 1.970151889791593,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004611451539606886,
      "loss": 5.4272,
      "mean_token_accuracy": 0.19682959616184234,
      "num_tokens": 41604548.0,
      "step": 22310
    },
    {
      "entropy": 5.8859165668487545,
      "epoch": 1.9705934298834333,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004611274620045031,
      "loss": 5.3778,
      "mean_token_accuracy": 0.20543351173400878,
      "num_tokens": 41614002.0,
      "step": 22315
    },
    {
      "entropy": 5.926422595977783,
      "epoch": 1.9710349699752738,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004611097664021866,
      "loss": 5.3785,
      "mean_token_accuracy": 0.19734903275966645,
      "num_tokens": 41623793.0,
      "step": 22320
    },
    {
      "entropy": 5.896677684783936,
      "epoch": 1.971476510067114,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046109206715408554,
      "loss": 5.3152,
      "mean_token_accuracy": 0.20300010740756988,
      "num_tokens": 41632608.0,
      "step": 22325
    },
    {
      "entropy": 5.921829748153686,
      "epoch": 1.9719180501589544,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046107436426054676,
      "loss": 5.4098,
      "mean_token_accuracy": 0.1996908262372017,
      "num_tokens": 41641814.0,
      "step": 22330
    },
    {
      "entropy": 5.9364342212677,
      "epoch": 1.9723595902507949,
      "grad_norm": 1.046875,
      "learning_rate": 0.000461056657721917,
      "loss": 5.3806,
      "mean_token_accuracy": 0.2009850949048996,
      "num_tokens": 41650643.0,
      "step": 22335
    },
    {
      "entropy": 5.912870216369629,
      "epoch": 1.9728011303426352,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046103894753854317,
      "loss": 5.3742,
      "mean_token_accuracy": 0.2038644790649414,
      "num_tokens": 41661308.0,
      "step": 22340
    },
    {
      "entropy": 5.947876596450806,
      "epoch": 1.9732426704344754,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004610212337107721,
      "loss": 5.42,
      "mean_token_accuracy": 0.196707184612751,
      "num_tokens": 41671090.0,
      "step": 22345
    },
    {
      "entropy": 5.856647539138794,
      "epoch": 1.973684210526316,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046100351623895087,
      "loss": 5.3623,
      "mean_token_accuracy": 0.2016085594892502,
      "num_tokens": 41679584.0,
      "step": 22350
    },
    {
      "entropy": 5.8358642101287845,
      "epoch": 1.974125750618156,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046098579512342654,
      "loss": 5.3486,
      "mean_token_accuracy": 0.19253745675086975,
      "num_tokens": 41689772.0,
      "step": 22355
    },
    {
      "entropy": 5.898290729522705,
      "epoch": 1.9745672907099965,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004609680703645463,
      "loss": 5.3724,
      "mean_token_accuracy": 0.20120088756084442,
      "num_tokens": 41699196.0,
      "step": 22360
    },
    {
      "entropy": 5.888697004318237,
      "epoch": 1.9750088308018368,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046095034196265726,
      "loss": 5.3224,
      "mean_token_accuracy": 0.2138876885175705,
      "num_tokens": 41707899.0,
      "step": 22365
    },
    {
      "entropy": 5.876837301254272,
      "epoch": 1.975450370893677,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046093260991810685,
      "loss": 5.381,
      "mean_token_accuracy": 0.20761237144470215,
      "num_tokens": 41716903.0,
      "step": 22370
    },
    {
      "entropy": 5.867739152908325,
      "epoch": 1.9758919109855175,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046091487423124234,
      "loss": 5.2375,
      "mean_token_accuracy": 0.21416876912117006,
      "num_tokens": 41725676.0,
      "step": 22375
    },
    {
      "entropy": 5.892343139648437,
      "epoch": 1.9763334510773578,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046089713490241117,
      "loss": 5.2567,
      "mean_token_accuracy": 0.21238938868045806,
      "num_tokens": 41734720.0,
      "step": 22380
    },
    {
      "entropy": 5.827096319198608,
      "epoch": 1.976774991169198,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004608793919319608,
      "loss": 5.3809,
      "mean_token_accuracy": 0.21058744490146636,
      "num_tokens": 41743866.0,
      "step": 22385
    },
    {
      "entropy": 5.839306926727295,
      "epoch": 1.9772165312610386,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00046086164532023894,
      "loss": 5.3423,
      "mean_token_accuracy": 0.21011044085025787,
      "num_tokens": 41754157.0,
      "step": 22390
    },
    {
      "entropy": 5.941110277175904,
      "epoch": 1.9776580713528789,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046084389506759313,
      "loss": 5.357,
      "mean_token_accuracy": 0.20468229949474334,
      "num_tokens": 41763105.0,
      "step": 22395
    },
    {
      "entropy": 5.9145894050598145,
      "epoch": 1.9780996114447191,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046082614117437103,
      "loss": 5.2743,
      "mean_token_accuracy": 0.21250310987234117,
      "num_tokens": 41772623.0,
      "step": 22400
    },
    {
      "entropy": 5.852737665176392,
      "epoch": 1.9785411515365596,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046080838364092057,
      "loss": 5.3774,
      "mean_token_accuracy": 0.2026889517903328,
      "num_tokens": 41781331.0,
      "step": 22405
    },
    {
      "entropy": 5.911617851257324,
      "epoch": 1.9789826916284,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004607906224675896,
      "loss": 5.4981,
      "mean_token_accuracy": 0.19725169241428375,
      "num_tokens": 41789727.0,
      "step": 22410
    },
    {
      "entropy": 5.930763483047485,
      "epoch": 1.9794242317202402,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004607728576547259,
      "loss": 5.3589,
      "mean_token_accuracy": 0.2052120476961136,
      "num_tokens": 41798531.0,
      "step": 22415
    },
    {
      "entropy": 5.927769041061401,
      "epoch": 1.9798657718120807,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046075508920267766,
      "loss": 5.368,
      "mean_token_accuracy": 0.2038962408900261,
      "num_tokens": 41808304.0,
      "step": 22420
    },
    {
      "entropy": 5.828134202957154,
      "epoch": 1.9803073119039207,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004607373171117929,
      "loss": 5.2486,
      "mean_token_accuracy": 0.21249572932720184,
      "num_tokens": 41817427.0,
      "step": 22425
    },
    {
      "entropy": 5.837864875793457,
      "epoch": 1.9807488519957612,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004607195413824197,
      "loss": 5.3276,
      "mean_token_accuracy": 0.21120381057262422,
      "num_tokens": 41827202.0,
      "step": 22430
    },
    {
      "entropy": 5.897235250473022,
      "epoch": 1.9811903920876015,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004607017620149062,
      "loss": 5.3035,
      "mean_token_accuracy": 0.21286925226449965,
      "num_tokens": 41836213.0,
      "step": 22435
    },
    {
      "entropy": 5.875213384628296,
      "epoch": 1.9816319321794418,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004606839790096009,
      "loss": 5.4201,
      "mean_token_accuracy": 0.19661785364151002,
      "num_tokens": 41845358.0,
      "step": 22440
    },
    {
      "entropy": 5.918174362182617,
      "epoch": 1.9820734722712823,
      "grad_norm": 1.125,
      "learning_rate": 0.0004606661923668521,
      "loss": 5.3709,
      "mean_token_accuracy": 0.20897252708673478,
      "num_tokens": 41854572.0,
      "step": 22445
    },
    {
      "entropy": 5.858848142623901,
      "epoch": 1.9825150123631226,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004606484020870081,
      "loss": 5.3265,
      "mean_token_accuracy": 0.21139038801193238,
      "num_tokens": 41863643.0,
      "step": 22450
    },
    {
      "entropy": 5.858176612854004,
      "epoch": 1.9829565524549628,
      "grad_norm": 1.125,
      "learning_rate": 0.00046063060817041747,
      "loss": 5.2724,
      "mean_token_accuracy": 0.2067800909280777,
      "num_tokens": 41873004.0,
      "step": 22455
    },
    {
      "entropy": 5.946782112121582,
      "epoch": 1.9833980925468033,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004606128106174288,
      "loss": 5.339,
      "mean_token_accuracy": 0.21116142868995666,
      "num_tokens": 41881530.0,
      "step": 22460
    },
    {
      "entropy": 5.872727966308593,
      "epoch": 1.9838396326386436,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004605950094283908,
      "loss": 5.3938,
      "mean_token_accuracy": 0.1966492637991905,
      "num_tokens": 41890342.0,
      "step": 22465
    },
    {
      "entropy": 5.849273014068603,
      "epoch": 1.9842811727304839,
      "grad_norm": 1.0,
      "learning_rate": 0.00046057720460365205,
      "loss": 5.2872,
      "mean_token_accuracy": 0.21358196139335633,
      "num_tokens": 41899274.0,
      "step": 22470
    },
    {
      "entropy": 5.912584590911865,
      "epoch": 1.9847227128223244,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046055939614356146,
      "loss": 5.3354,
      "mean_token_accuracy": 0.20744515359401702,
      "num_tokens": 41908835.0,
      "step": 22475
    },
    {
      "entropy": 5.851005363464355,
      "epoch": 1.9851642529141647,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004605415840484678,
      "loss": 5.2682,
      "mean_token_accuracy": 0.21649398058652877,
      "num_tokens": 41917900.0,
      "step": 22480
    },
    {
      "entropy": 5.903795146942139,
      "epoch": 1.985605793006005,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046052376831872007,
      "loss": 5.3446,
      "mean_token_accuracy": 0.2121477320790291,
      "num_tokens": 41926964.0,
      "step": 22485
    },
    {
      "entropy": 5.944635820388794,
      "epoch": 1.9860473330978454,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004605059489546672,
      "loss": 5.4173,
      "mean_token_accuracy": 0.2030104786157608,
      "num_tokens": 41936401.0,
      "step": 22490
    },
    {
      "entropy": 5.830485725402832,
      "epoch": 1.9864888731896855,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046048812595665824,
      "loss": 5.3039,
      "mean_token_accuracy": 0.20996571332216263,
      "num_tokens": 41944665.0,
      "step": 22495
    },
    {
      "entropy": 5.844066381454468,
      "epoch": 1.986930413281526,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004604702993250425,
      "loss": 5.3662,
      "mean_token_accuracy": 0.1958061069250107,
      "num_tokens": 41953868.0,
      "step": 22500
    },
    {
      "entropy": 5.810266590118408,
      "epoch": 1.9873719533733663,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046045246906016904,
      "loss": 5.3681,
      "mean_token_accuracy": 0.2104953870177269,
      "num_tokens": 41963013.0,
      "step": 22505
    },
    {
      "entropy": 5.902640914916992,
      "epoch": 1.9878134934652065,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004604346351623872,
      "loss": 5.363,
      "mean_token_accuracy": 0.19927488267421722,
      "num_tokens": 41972926.0,
      "step": 22510
    },
    {
      "entropy": 5.942632150650025,
      "epoch": 1.988255033557047,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046041679763204625,
      "loss": 5.3581,
      "mean_token_accuracy": 0.21053942143917084,
      "num_tokens": 41981773.0,
      "step": 22515
    },
    {
      "entropy": 5.913764953613281,
      "epoch": 1.9886965736488873,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004603989564694957,
      "loss": 5.3457,
      "mean_token_accuracy": 0.20863389521837233,
      "num_tokens": 41990893.0,
      "step": 22520
    },
    {
      "entropy": 5.876496982574463,
      "epoch": 1.9891381137407276,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004603811116750851,
      "loss": 5.3682,
      "mean_token_accuracy": 0.2028349071741104,
      "num_tokens": 42000842.0,
      "step": 22525
    },
    {
      "entropy": 5.9543054580688475,
      "epoch": 1.989579653832568,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004603632632491639,
      "loss": 5.3356,
      "mean_token_accuracy": 0.21135483980178832,
      "num_tokens": 42009430.0,
      "step": 22530
    },
    {
      "entropy": 5.913153839111328,
      "epoch": 1.9900211939244083,
      "grad_norm": 1.0,
      "learning_rate": 0.00046034541119208177,
      "loss": 5.4722,
      "mean_token_accuracy": 0.2030464231967926,
      "num_tokens": 42018985.0,
      "step": 22535
    },
    {
      "entropy": 5.905248403549194,
      "epoch": 1.9904627340162486,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046032755550418845,
      "loss": 5.3344,
      "mean_token_accuracy": 0.20350456684827806,
      "num_tokens": 42028885.0,
      "step": 22540
    },
    {
      "entropy": 5.960259628295899,
      "epoch": 1.9909042741080891,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004603096961858338,
      "loss": 5.4422,
      "mean_token_accuracy": 0.19680802673101425,
      "num_tokens": 42037838.0,
      "step": 22545
    },
    {
      "entropy": 5.859803628921509,
      "epoch": 1.9913458141999294,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046029183323736755,
      "loss": 5.3297,
      "mean_token_accuracy": 0.2057755097746849,
      "num_tokens": 42047212.0,
      "step": 22550
    },
    {
      "entropy": 5.960439300537109,
      "epoch": 1.9917873542917697,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004602739666591397,
      "loss": 5.4123,
      "mean_token_accuracy": 0.20028246194124222,
      "num_tokens": 42056680.0,
      "step": 22555
    },
    {
      "entropy": 5.956166744232178,
      "epoch": 1.9922288943836102,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046025609645150014,
      "loss": 5.4116,
      "mean_token_accuracy": 0.19626805037260056,
      "num_tokens": 42065604.0,
      "step": 22560
    },
    {
      "entropy": 5.921209716796875,
      "epoch": 1.9926704344754502,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000460238222614799,
      "loss": 5.3433,
      "mean_token_accuracy": 0.20482654571533204,
      "num_tokens": 42074672.0,
      "step": 22565
    },
    {
      "entropy": 5.894555330276489,
      "epoch": 1.9931119745672907,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004602203451493865,
      "loss": 5.3036,
      "mean_token_accuracy": 0.20232925713062286,
      "num_tokens": 42083586.0,
      "step": 22570
    },
    {
      "entropy": 5.862616872787475,
      "epoch": 1.993553514659131,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046020246405561283,
      "loss": 5.326,
      "mean_token_accuracy": 0.20374557375907898,
      "num_tokens": 42092328.0,
      "step": 22575
    },
    {
      "entropy": 5.827028703689575,
      "epoch": 1.9939950547509713,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046018457933382807,
      "loss": 5.3187,
      "mean_token_accuracy": 0.21235123425722122,
      "num_tokens": 42102154.0,
      "step": 22580
    },
    {
      "entropy": 5.936002159118653,
      "epoch": 1.9944365948428118,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046016669098438286,
      "loss": 5.3829,
      "mean_token_accuracy": 0.20435209274291993,
      "num_tokens": 42111762.0,
      "step": 22585
    },
    {
      "entropy": 5.911132431030273,
      "epoch": 1.994878134934652,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004601487990076274,
      "loss": 5.3762,
      "mean_token_accuracy": 0.20171091109514236,
      "num_tokens": 42120631.0,
      "step": 22590
    },
    {
      "entropy": 5.822972202301026,
      "epoch": 1.9953196750264923,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004601309034039123,
      "loss": 5.2363,
      "mean_token_accuracy": 0.22829282730817796,
      "num_tokens": 42129311.0,
      "step": 22595
    },
    {
      "entropy": 5.937223196029663,
      "epoch": 1.9957612151183328,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004601130041735881,
      "loss": 5.4706,
      "mean_token_accuracy": 0.19585298597812653,
      "num_tokens": 42138735.0,
      "step": 22600
    },
    {
      "entropy": 5.997986698150635,
      "epoch": 1.996202755210173,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004600951013170054,
      "loss": 5.4305,
      "mean_token_accuracy": 0.20378115326166152,
      "num_tokens": 42148416.0,
      "step": 22605
    },
    {
      "entropy": 5.9393072605133055,
      "epoch": 1.9966442953020134,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046007719483451493,
      "loss": 5.3558,
      "mean_token_accuracy": 0.2071908250451088,
      "num_tokens": 42157191.0,
      "step": 22610
    },
    {
      "entropy": 5.775449609756469,
      "epoch": 1.9970858353938539,
      "grad_norm": 1.125,
      "learning_rate": 0.00046005928472646747,
      "loss": 5.2936,
      "mean_token_accuracy": 0.22011100500822067,
      "num_tokens": 42165798.0,
      "step": 22615
    },
    {
      "entropy": 5.871996688842773,
      "epoch": 1.9975273754856941,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046004137099321386,
      "loss": 5.4053,
      "mean_token_accuracy": 0.1983458638191223,
      "num_tokens": 42175318.0,
      "step": 22620
    },
    {
      "entropy": 5.954274415969849,
      "epoch": 1.9979689155775344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046002345363510506,
      "loss": 5.4177,
      "mean_token_accuracy": 0.19837913960218428,
      "num_tokens": 42184800.0,
      "step": 22625
    },
    {
      "entropy": 5.900021648406982,
      "epoch": 1.998410455669375,
      "grad_norm": 1.0,
      "learning_rate": 0.0004600055326524921,
      "loss": 5.2444,
      "mean_token_accuracy": 0.21323984414339064,
      "num_tokens": 42194172.0,
      "step": 22630
    },
    {
      "entropy": 5.86443395614624,
      "epoch": 1.998851995761215,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045998760804572583,
      "loss": 5.2903,
      "mean_token_accuracy": 0.21045953035354614,
      "num_tokens": 42203102.0,
      "step": 22635
    },
    {
      "entropy": 5.787165307998658,
      "epoch": 1.9992935358530555,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045996967981515763,
      "loss": 5.2459,
      "mean_token_accuracy": 0.21351298987865447,
      "num_tokens": 42212421.0,
      "step": 22640
    },
    {
      "entropy": 5.909124040603638,
      "epoch": 1.9997350759448957,
      "grad_norm": 1.0,
      "learning_rate": 0.00045995174796113856,
      "loss": 5.3652,
      "mean_token_accuracy": 0.2063007727265358,
      "num_tokens": 42221528.0,
      "step": 22645
    },
    {
      "entropy": 5.87777943611145,
      "epoch": 2.000176616036736,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00045993381248402,
      "loss": 5.3305,
      "mean_token_accuracy": 0.2049049600958824,
      "num_tokens": 42230628.0,
      "step": 22650
    },
    {
      "entropy": 5.868379402160644,
      "epoch": 2.0006181561285765,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045991587338415306,
      "loss": 5.2927,
      "mean_token_accuracy": 0.20688465386629104,
      "num_tokens": 42239571.0,
      "step": 22655
    },
    {
      "entropy": 5.907155132293701,
      "epoch": 2.001059696220417,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045989793066188953,
      "loss": 5.2309,
      "mean_token_accuracy": 0.21225377470254897,
      "num_tokens": 42248010.0,
      "step": 22660
    },
    {
      "entropy": 5.803134965896606,
      "epoch": 2.001501236312257,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045987998431758063,
      "loss": 5.2375,
      "mean_token_accuracy": 0.21220990270376205,
      "num_tokens": 42257251.0,
      "step": 22665
    },
    {
      "entropy": 5.881656980514526,
      "epoch": 2.0019427764040976,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004598620343515779,
      "loss": 5.3549,
      "mean_token_accuracy": 0.20693680346012117,
      "num_tokens": 42266441.0,
      "step": 22670
    },
    {
      "entropy": 5.8739899635314945,
      "epoch": 2.0023843164959376,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045984408076423313,
      "loss": 5.2753,
      "mean_token_accuracy": 0.2062024623155594,
      "num_tokens": 42275995.0,
      "step": 22675
    },
    {
      "entropy": 5.928394842147827,
      "epoch": 2.002825856587778,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004598261235558979,
      "loss": 5.2725,
      "mean_token_accuracy": 0.21312019377946853,
      "num_tokens": 42284065.0,
      "step": 22680
    },
    {
      "entropy": 5.872630262374878,
      "epoch": 2.0032673966796186,
      "grad_norm": 0.9375,
      "learning_rate": 0.00045980816272692403,
      "loss": 5.3725,
      "mean_token_accuracy": 0.20882306694984437,
      "num_tokens": 42295309.0,
      "step": 22685
    },
    {
      "entropy": 5.960452842712402,
      "epoch": 2.0037089367714587,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045979019827766345,
      "loss": 5.2739,
      "mean_token_accuracy": 0.21062087267637253,
      "num_tokens": 42304390.0,
      "step": 22690
    },
    {
      "entropy": 5.915776681900025,
      "epoch": 2.004150476863299,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045977223020846785,
      "loss": 5.3048,
      "mean_token_accuracy": 0.21240830272436143,
      "num_tokens": 42313830.0,
      "step": 22695
    },
    {
      "entropy": 5.913122510910034,
      "epoch": 2.0045920169551397,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004597542585196895,
      "loss": 5.2997,
      "mean_token_accuracy": 0.2076733887195587,
      "num_tokens": 42323827.0,
      "step": 22700
    },
    {
      "entropy": 5.964270734786988,
      "epoch": 2.0050335570469797,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045973628321168016,
      "loss": 5.3975,
      "mean_token_accuracy": 0.20324095636606215,
      "num_tokens": 42333221.0,
      "step": 22705
    },
    {
      "entropy": 5.819288206100464,
      "epoch": 2.00547509713882,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045971830428479215,
      "loss": 5.2696,
      "mean_token_accuracy": 0.21562483608722688,
      "num_tokens": 42343036.0,
      "step": 22710
    },
    {
      "entropy": 5.854319334030151,
      "epoch": 2.0059166372306607,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045970032173937766,
      "loss": 5.2503,
      "mean_token_accuracy": 0.21470864713191987,
      "num_tokens": 42351916.0,
      "step": 22715
    },
    {
      "entropy": 5.993334007263184,
      "epoch": 2.0063581773225008,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045968233557578883,
      "loss": 5.3547,
      "mean_token_accuracy": 0.20781833529472352,
      "num_tokens": 42362412.0,
      "step": 22720
    },
    {
      "entropy": 5.867878150939942,
      "epoch": 2.0067997174143413,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004596643457943781,
      "loss": 5.2354,
      "mean_token_accuracy": 0.21803030967712403,
      "num_tokens": 42372773.0,
      "step": 22725
    },
    {
      "entropy": 5.886961364746094,
      "epoch": 2.0072412575061818,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004596463523954979,
      "loss": 5.2547,
      "mean_token_accuracy": 0.2088466003537178,
      "num_tokens": 42382009.0,
      "step": 22730
    },
    {
      "entropy": 5.882224369049072,
      "epoch": 2.007682797598022,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045962835537950063,
      "loss": 5.2758,
      "mean_token_accuracy": 0.21333387345075608,
      "num_tokens": 42390961.0,
      "step": 22735
    },
    {
      "entropy": 5.8336258888244625,
      "epoch": 2.0081243376898623,
      "grad_norm": 1.0,
      "learning_rate": 0.00045961035474673893,
      "loss": 5.2789,
      "mean_token_accuracy": 0.2099034771323204,
      "num_tokens": 42400483.0,
      "step": 22740
    },
    {
      "entropy": 5.759735345840454,
      "epoch": 2.0085658777817024,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045959235049756537,
      "loss": 5.1332,
      "mean_token_accuracy": 0.22602652311325072,
      "num_tokens": 42409285.0,
      "step": 22745
    },
    {
      "entropy": 5.9130964279174805,
      "epoch": 2.009007417873543,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045957434263233253,
      "loss": 5.3131,
      "mean_token_accuracy": 0.2079264223575592,
      "num_tokens": 42418950.0,
      "step": 22750
    },
    {
      "entropy": 5.894699907302856,
      "epoch": 2.0094489579653834,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045955633115139343,
      "loss": 5.3883,
      "mean_token_accuracy": 0.20179129242897034,
      "num_tokens": 42428285.0,
      "step": 22755
    },
    {
      "entropy": 5.87842812538147,
      "epoch": 2.0098904980572234,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004595383160551007,
      "loss": 5.2342,
      "mean_token_accuracy": 0.2104954943060875,
      "num_tokens": 42437832.0,
      "step": 22760
    },
    {
      "entropy": 5.872386884689331,
      "epoch": 2.010332038149064,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004595202973438074,
      "loss": 5.2547,
      "mean_token_accuracy": 0.2119743764400482,
      "num_tokens": 42447704.0,
      "step": 22765
    },
    {
      "entropy": 5.9236311435699465,
      "epoch": 2.0107735782409044,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004595022750178663,
      "loss": 5.3509,
      "mean_token_accuracy": 0.20961358696222304,
      "num_tokens": 42456571.0,
      "step": 22770
    },
    {
      "entropy": 5.90190806388855,
      "epoch": 2.0112151183327445,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004594842490776306,
      "loss": 5.2302,
      "mean_token_accuracy": 0.21186455190181733,
      "num_tokens": 42466522.0,
      "step": 22775
    },
    {
      "entropy": 5.840451097488403,
      "epoch": 2.011656658424585,
      "grad_norm": 1.0,
      "learning_rate": 0.00045946621952345335,
      "loss": 5.2938,
      "mean_token_accuracy": 0.20743368566036224,
      "num_tokens": 42475940.0,
      "step": 22780
    },
    {
      "entropy": 5.922690153121948,
      "epoch": 2.0120981985164255,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004594481863556879,
      "loss": 5.3625,
      "mean_token_accuracy": 0.20207197219133377,
      "num_tokens": 42486456.0,
      "step": 22785
    },
    {
      "entropy": 5.965627670288086,
      "epoch": 2.0125397386082655,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045943014957468726,
      "loss": 5.3418,
      "mean_token_accuracy": 0.21078796535730362,
      "num_tokens": 42495578.0,
      "step": 22790
    },
    {
      "entropy": 5.885344362258911,
      "epoch": 2.012981278700106,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004594121091808049,
      "loss": 5.1969,
      "mean_token_accuracy": 0.21920282989740372,
      "num_tokens": 42504532.0,
      "step": 22795
    },
    {
      "entropy": 5.872156238555908,
      "epoch": 2.0134228187919465,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045939406517439427,
      "loss": 5.2543,
      "mean_token_accuracy": 0.2080533429980278,
      "num_tokens": 42513683.0,
      "step": 22800
    },
    {
      "entropy": 5.92125301361084,
      "epoch": 2.0138643588837866,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045937601755580864,
      "loss": 5.3336,
      "mean_token_accuracy": 0.20605723708868026,
      "num_tokens": 42522717.0,
      "step": 22805
    },
    {
      "entropy": 5.932913970947266,
      "epoch": 2.014305898975627,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004593579663254018,
      "loss": 5.3516,
      "mean_token_accuracy": 0.20671084374189377,
      "num_tokens": 42531847.0,
      "step": 22810
    },
    {
      "entropy": 5.863706588745117,
      "epoch": 2.014747439067467,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004593399114835272,
      "loss": 5.2705,
      "mean_token_accuracy": 0.21442181318998338,
      "num_tokens": 42540869.0,
      "step": 22815
    },
    {
      "entropy": 5.8496442317962645,
      "epoch": 2.0151889791593076,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045932185303053857,
      "loss": 5.2563,
      "mean_token_accuracy": 0.21570290476083756,
      "num_tokens": 42550875.0,
      "step": 22820
    },
    {
      "entropy": 5.883614873886108,
      "epoch": 2.015630519251148,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004593037909667898,
      "loss": 5.2332,
      "mean_token_accuracy": 0.2173098623752594,
      "num_tokens": 42560066.0,
      "step": 22825
    },
    {
      "entropy": 5.836062431335449,
      "epoch": 2.016072059342988,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004592857252926344,
      "loss": 5.2487,
      "mean_token_accuracy": 0.2159641429781914,
      "num_tokens": 42568733.0,
      "step": 22830
    },
    {
      "entropy": 5.858053302764892,
      "epoch": 2.0165135994348287,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045926765600842653,
      "loss": 5.2856,
      "mean_token_accuracy": 0.20940887480974196,
      "num_tokens": 42577565.0,
      "step": 22835
    },
    {
      "entropy": 5.866206693649292,
      "epoch": 2.016955139526669,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045924958311452006,
      "loss": 5.3193,
      "mean_token_accuracy": 0.20851740688085557,
      "num_tokens": 42586489.0,
      "step": 22840
    },
    {
      "entropy": 5.892873382568359,
      "epoch": 2.017396679618509,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045923150661126904,
      "loss": 5.3567,
      "mean_token_accuracy": 0.20520659685134887,
      "num_tokens": 42596263.0,
      "step": 22845
    },
    {
      "entropy": 5.8912801265716555,
      "epoch": 2.0178382197103497,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004592134264990276,
      "loss": 5.232,
      "mean_token_accuracy": 0.21182719022035598,
      "num_tokens": 42605787.0,
      "step": 22850
    },
    {
      "entropy": 5.99868597984314,
      "epoch": 2.01827975980219,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045919534277814996,
      "loss": 5.4502,
      "mean_token_accuracy": 0.20145653635263444,
      "num_tokens": 42616014.0,
      "step": 22855
    },
    {
      "entropy": 5.913133955001831,
      "epoch": 2.0187212998940303,
      "grad_norm": 1.125,
      "learning_rate": 0.0004591772554489902,
      "loss": 5.3207,
      "mean_token_accuracy": 0.2100447028875351,
      "num_tokens": 42624839.0,
      "step": 22860
    },
    {
      "entropy": 5.905199146270752,
      "epoch": 2.0191628399858708,
      "grad_norm": 1.125,
      "learning_rate": 0.0004591591645119028,
      "loss": 5.3115,
      "mean_token_accuracy": 0.20897043645381927,
      "num_tokens": 42633811.0,
      "step": 22865
    },
    {
      "entropy": 5.8146604061126705,
      "epoch": 2.0196043800777113,
      "grad_norm": 1.125,
      "learning_rate": 0.00045914106996724217,
      "loss": 5.2131,
      "mean_token_accuracy": 0.21572656333446502,
      "num_tokens": 42643140.0,
      "step": 22870
    },
    {
      "entropy": 5.840064477920532,
      "epoch": 2.0200459201695513,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045912297181536257,
      "loss": 5.2172,
      "mean_token_accuracy": 0.2192785531282425,
      "num_tokens": 42651123.0,
      "step": 22875
    },
    {
      "entropy": 5.893158578872681,
      "epoch": 2.020487460261392,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045910487005661877,
      "loss": 5.2732,
      "mean_token_accuracy": 0.2179436832666397,
      "num_tokens": 42659880.0,
      "step": 22880
    },
    {
      "entropy": 5.878518772125244,
      "epoch": 2.020929000353232,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004590867646913653,
      "loss": 5.2157,
      "mean_token_accuracy": 0.22213126569986344,
      "num_tokens": 42668689.0,
      "step": 22885
    },
    {
      "entropy": 5.858295583724976,
      "epoch": 2.0213705404450724,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045906865571995666,
      "loss": 5.2611,
      "mean_token_accuracy": 0.21760706454515458,
      "num_tokens": 42678403.0,
      "step": 22890
    },
    {
      "entropy": 5.786852550506592,
      "epoch": 2.021812080536913,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004590505431427479,
      "loss": 5.2673,
      "mean_token_accuracy": 0.21441237777471542,
      "num_tokens": 42688203.0,
      "step": 22895
    },
    {
      "entropy": 5.888368701934814,
      "epoch": 2.022253620628753,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004590324269600936,
      "loss": 5.3581,
      "mean_token_accuracy": 0.20022109597921373,
      "num_tokens": 42698343.0,
      "step": 22900
    },
    {
      "entropy": 5.843066930770874,
      "epoch": 2.0226951607205934,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045901430717234875,
      "loss": 5.1911,
      "mean_token_accuracy": 0.22421941608190538,
      "num_tokens": 42707483.0,
      "step": 22905
    },
    {
      "entropy": 5.906141948699951,
      "epoch": 2.023136700812434,
      "grad_norm": 1.0,
      "learning_rate": 0.0004589961837798682,
      "loss": 5.3392,
      "mean_token_accuracy": 0.20942073464393615,
      "num_tokens": 42717820.0,
      "step": 22910
    },
    {
      "entropy": 5.895740938186646,
      "epoch": 2.023578240904274,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045897805678300714,
      "loss": 5.3087,
      "mean_token_accuracy": 0.21306860446929932,
      "num_tokens": 42726996.0,
      "step": 22915
    },
    {
      "entropy": 5.875725412368775,
      "epoch": 2.0240197809961145,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004589599261821205,
      "loss": 5.3059,
      "mean_token_accuracy": 0.2133346602320671,
      "num_tokens": 42735779.0,
      "step": 22920
    },
    {
      "entropy": 5.9298499584197994,
      "epoch": 2.024461321087955,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045894179197756364,
      "loss": 5.3848,
      "mean_token_accuracy": 0.2061198353767395,
      "num_tokens": 42745089.0,
      "step": 22925
    },
    {
      "entropy": 5.889540958404541,
      "epoch": 2.024902861179795,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045892365416969164,
      "loss": 5.3199,
      "mean_token_accuracy": 0.2028687983751297,
      "num_tokens": 42754963.0,
      "step": 22930
    },
    {
      "entropy": 5.920861864089966,
      "epoch": 2.0253444012716355,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045890551275885985,
      "loss": 5.3594,
      "mean_token_accuracy": 0.20339042395353318,
      "num_tokens": 42764856.0,
      "step": 22935
    },
    {
      "entropy": 5.9101465225219725,
      "epoch": 2.025785941363476,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045888736774542363,
      "loss": 5.3182,
      "mean_token_accuracy": 0.2119409903883934,
      "num_tokens": 42775028.0,
      "step": 22940
    },
    {
      "entropy": 5.905062532424926,
      "epoch": 2.026227481455316,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004588692191297385,
      "loss": 5.3237,
      "mean_token_accuracy": 0.21250978857278824,
      "num_tokens": 42784288.0,
      "step": 22945
    },
    {
      "entropy": 5.912501239776612,
      "epoch": 2.0266690215471566,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004588510669121599,
      "loss": 5.3852,
      "mean_token_accuracy": 0.19859351217746735,
      "num_tokens": 42793356.0,
      "step": 22950
    },
    {
      "entropy": 5.805608034133911,
      "epoch": 2.0271105616389966,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045883291109304346,
      "loss": 5.2053,
      "mean_token_accuracy": 0.2177696481347084,
      "num_tokens": 42802287.0,
      "step": 22955
    },
    {
      "entropy": 5.923625135421753,
      "epoch": 2.027552101730837,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004588147516727449,
      "loss": 5.3307,
      "mean_token_accuracy": 0.21063922196626664,
      "num_tokens": 42812206.0,
      "step": 22960
    },
    {
      "entropy": 5.9731368064880375,
      "epoch": 2.0279936418226776,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004587965886516199,
      "loss": 5.3472,
      "mean_token_accuracy": 0.2017676427960396,
      "num_tokens": 42822046.0,
      "step": 22965
    },
    {
      "entropy": 5.87266206741333,
      "epoch": 2.0284351819145177,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045877842203002417,
      "loss": 5.2739,
      "mean_token_accuracy": 0.2126902461051941,
      "num_tokens": 42831228.0,
      "step": 22970
    },
    {
      "entropy": 5.8698663234710695,
      "epoch": 2.028876722006358,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045876025180831373,
      "loss": 5.3543,
      "mean_token_accuracy": 0.2101733535528183,
      "num_tokens": 42840683.0,
      "step": 22975
    },
    {
      "entropy": 5.90928955078125,
      "epoch": 2.0293182620981987,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045874207798684446,
      "loss": 5.273,
      "mean_token_accuracy": 0.21106504946947097,
      "num_tokens": 42849731.0,
      "step": 22980
    },
    {
      "entropy": 5.940430593490601,
      "epoch": 2.0297598021900387,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045872390056597235,
      "loss": 5.3328,
      "mean_token_accuracy": 0.20695223212242125,
      "num_tokens": 42858028.0,
      "step": 22985
    },
    {
      "entropy": 5.893157339096069,
      "epoch": 2.030201342281879,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004587057195460536,
      "loss": 5.3382,
      "mean_token_accuracy": 0.20177621990442277,
      "num_tokens": 42867340.0,
      "step": 22990
    },
    {
      "entropy": 5.964124965667724,
      "epoch": 2.0306428823737197,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045868753492744424,
      "loss": 5.3521,
      "mean_token_accuracy": 0.20110375732183455,
      "num_tokens": 42877006.0,
      "step": 22995
    },
    {
      "entropy": 5.850784778594971,
      "epoch": 2.0310844224655598,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004586693467105005,
      "loss": 5.2063,
      "mean_token_accuracy": 0.2156612440943718,
      "num_tokens": 42886324.0,
      "step": 23000
    },
    {
      "entropy": 5.92107253074646,
      "epoch": 2.0315259625574003,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004586511548955788,
      "loss": 5.3068,
      "mean_token_accuracy": 0.20806549489498138,
      "num_tokens": 42896411.0,
      "step": 23005
    },
    {
      "entropy": 5.804952096939087,
      "epoch": 2.0319675026492408,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004586329594830353,
      "loss": 5.2006,
      "mean_token_accuracy": 0.21858483254909516,
      "num_tokens": 42905606.0,
      "step": 23010
    },
    {
      "entropy": 5.783320331573487,
      "epoch": 2.032409042741081,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045861476047322664,
      "loss": 5.1788,
      "mean_token_accuracy": 0.2198178991675377,
      "num_tokens": 42914226.0,
      "step": 23015
    },
    {
      "entropy": 5.882395792007446,
      "epoch": 2.0328505828329213,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045859655786650925,
      "loss": 5.3622,
      "mean_token_accuracy": 0.2007863849401474,
      "num_tokens": 42923910.0,
      "step": 23020
    },
    {
      "entropy": 5.843041944503784,
      "epoch": 2.0332921229247614,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004585783516632397,
      "loss": 5.2073,
      "mean_token_accuracy": 0.23171699792146683,
      "num_tokens": 42933203.0,
      "step": 23025
    },
    {
      "entropy": 5.90801682472229,
      "epoch": 2.033733663016602,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004585601418637747,
      "loss": 5.2788,
      "mean_token_accuracy": 0.2077164724469185,
      "num_tokens": 42942143.0,
      "step": 23030
    },
    {
      "entropy": 5.861828327178955,
      "epoch": 2.0341752031084424,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004585419284684708,
      "loss": 5.3023,
      "mean_token_accuracy": 0.21032149642705916,
      "num_tokens": 42951432.0,
      "step": 23035
    },
    {
      "entropy": 5.8482099056243895,
      "epoch": 2.0346167432002824,
      "grad_norm": 1.078125,
      "learning_rate": 0.000458523711477685,
      "loss": 5.1956,
      "mean_token_accuracy": 0.21475542336702347,
      "num_tokens": 42961033.0,
      "step": 23040
    },
    {
      "entropy": 5.879715538024902,
      "epoch": 2.035058283292123,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045850549089177406,
      "loss": 5.1871,
      "mean_token_accuracy": 0.22608125060796738,
      "num_tokens": 42971371.0,
      "step": 23045
    },
    {
      "entropy": 5.969196462631226,
      "epoch": 2.0354998233839634,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045848726671109493,
      "loss": 5.3198,
      "mean_token_accuracy": 0.19820572286844254,
      "num_tokens": 42980445.0,
      "step": 23050
    },
    {
      "entropy": 5.8610227584838865,
      "epoch": 2.0359413634758035,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045846903893600457,
      "loss": 5.2904,
      "mean_token_accuracy": 0.20875540375709534,
      "num_tokens": 42989923.0,
      "step": 23055
    },
    {
      "entropy": 5.757601356506347,
      "epoch": 2.036382903567644,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045845080756686007,
      "loss": 5.1617,
      "mean_token_accuracy": 0.219550783932209,
      "num_tokens": 42998370.0,
      "step": 23060
    },
    {
      "entropy": 5.899701642990112,
      "epoch": 2.0368244436594845,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045843257260401855,
      "loss": 5.2656,
      "mean_token_accuracy": 0.21414129436016083,
      "num_tokens": 43007641.0,
      "step": 23065
    },
    {
      "entropy": 5.846440362930298,
      "epoch": 2.0372659837513245,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045841433404783733,
      "loss": 5.1923,
      "mean_token_accuracy": 0.21901296377182006,
      "num_tokens": 43016271.0,
      "step": 23070
    },
    {
      "entropy": 5.805743741989136,
      "epoch": 2.037707523843165,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004583960918986736,
      "loss": 5.1859,
      "mean_token_accuracy": 0.21661363542079926,
      "num_tokens": 43024525.0,
      "step": 23075
    },
    {
      "entropy": 5.8223387718200685,
      "epoch": 2.0381490639350055,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045837784615688473,
      "loss": 5.2871,
      "mean_token_accuracy": 0.21378793865442275,
      "num_tokens": 43033248.0,
      "step": 23080
    },
    {
      "entropy": 6.0096031665802006,
      "epoch": 2.0385906040268456,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004583595968228281,
      "loss": 5.4614,
      "mean_token_accuracy": 0.20177992582321166,
      "num_tokens": 43042171.0,
      "step": 23085
    },
    {
      "entropy": 5.879728603363037,
      "epoch": 2.039032144118686,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045834134389686137,
      "loss": 5.2216,
      "mean_token_accuracy": 0.22028322219848634,
      "num_tokens": 43051447.0,
      "step": 23090
    },
    {
      "entropy": 5.81450252532959,
      "epoch": 2.039473684210526,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004583230873793419,
      "loss": 5.1933,
      "mean_token_accuracy": 0.2075284019112587,
      "num_tokens": 43059869.0,
      "step": 23095
    },
    {
      "entropy": 5.7618450164794925,
      "epoch": 2.0399152243023666,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045830482727062736,
      "loss": 5.1734,
      "mean_token_accuracy": 0.21728577464818954,
      "num_tokens": 43068412.0,
      "step": 23100
    },
    {
      "entropy": 5.900337553024292,
      "epoch": 2.040356764394207,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045828656357107557,
      "loss": 5.3249,
      "mean_token_accuracy": 0.20258362889289855,
      "num_tokens": 43077130.0,
      "step": 23105
    },
    {
      "entropy": 5.941557168960571,
      "epoch": 2.040798304486047,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004582682962810442,
      "loss": 5.3504,
      "mean_token_accuracy": 0.20063544660806656,
      "num_tokens": 43086304.0,
      "step": 23110
    },
    {
      "entropy": 5.913349962234497,
      "epoch": 2.0412398445778877,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004582500254008912,
      "loss": 5.2715,
      "mean_token_accuracy": 0.21482429504394532,
      "num_tokens": 43095734.0,
      "step": 23115
    },
    {
      "entropy": 5.860885715484619,
      "epoch": 2.041681384669728,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045823175093097444,
      "loss": 5.255,
      "mean_token_accuracy": 0.2117638811469078,
      "num_tokens": 43104761.0,
      "step": 23120
    },
    {
      "entropy": 5.923091793060303,
      "epoch": 2.042122924761568,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004582134728716518,
      "loss": 5.3214,
      "mean_token_accuracy": 0.20944496244192123,
      "num_tokens": 43114182.0,
      "step": 23125
    },
    {
      "entropy": 5.832463979721069,
      "epoch": 2.0425644648534087,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004581951912232814,
      "loss": 5.2694,
      "mean_token_accuracy": 0.21437161713838576,
      "num_tokens": 43124893.0,
      "step": 23130
    },
    {
      "entropy": 5.894551181793213,
      "epoch": 2.043006004945249,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004581769059862215,
      "loss": 5.2739,
      "mean_token_accuracy": 0.21186769157648086,
      "num_tokens": 43133514.0,
      "step": 23135
    },
    {
      "entropy": 5.9472887992858885,
      "epoch": 2.0434475450370893,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004581586171608301,
      "loss": 5.366,
      "mean_token_accuracy": 0.2069222465157509,
      "num_tokens": 43142082.0,
      "step": 23140
    },
    {
      "entropy": 5.8757179260253904,
      "epoch": 2.0438890851289298,
      "grad_norm": 1.0,
      "learning_rate": 0.0004581403247474657,
      "loss": 5.3086,
      "mean_token_accuracy": 0.20761488974094391,
      "num_tokens": 43151466.0,
      "step": 23145
    },
    {
      "entropy": 5.928344774246216,
      "epoch": 2.0443306252207702,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004581220287464864,
      "loss": 5.3791,
      "mean_token_accuracy": 0.19838937073946,
      "num_tokens": 43160631.0,
      "step": 23150
    },
    {
      "entropy": 5.902462577819824,
      "epoch": 2.0447721653126103,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004581037291582507,
      "loss": 5.3461,
      "mean_token_accuracy": 0.21122662723064423,
      "num_tokens": 43170968.0,
      "step": 23155
    },
    {
      "entropy": 5.880038213729859,
      "epoch": 2.045213705404451,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004580854259831171,
      "loss": 5.2705,
      "mean_token_accuracy": 0.20816658437252045,
      "num_tokens": 43179661.0,
      "step": 23160
    },
    {
      "entropy": 5.806720542907715,
      "epoch": 2.045655245496291,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004580671192214441,
      "loss": 5.1291,
      "mean_token_accuracy": 0.22906865179538727,
      "num_tokens": 43188994.0,
      "step": 23165
    },
    {
      "entropy": 5.97022442817688,
      "epoch": 2.0460967855881314,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045804880887359036,
      "loss": 5.3974,
      "mean_token_accuracy": 0.20026474744081496,
      "num_tokens": 43199001.0,
      "step": 23170
    },
    {
      "entropy": 5.896676683425904,
      "epoch": 2.046538325679972,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004580304949399146,
      "loss": 5.2541,
      "mean_token_accuracy": 0.21216757595539093,
      "num_tokens": 43208113.0,
      "step": 23175
    },
    {
      "entropy": 5.838364839553833,
      "epoch": 2.046979865771812,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045801217742077544,
      "loss": 5.2629,
      "mean_token_accuracy": 0.2162158265709877,
      "num_tokens": 43216983.0,
      "step": 23180
    },
    {
      "entropy": 5.87812442779541,
      "epoch": 2.0474214058636524,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004579938563165319,
      "loss": 5.2582,
      "mean_token_accuracy": 0.21697108447551727,
      "num_tokens": 43225965.0,
      "step": 23185
    },
    {
      "entropy": 5.883720636367798,
      "epoch": 2.047862945955493,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004579755316275427,
      "loss": 5.1993,
      "mean_token_accuracy": 0.2154235154390335,
      "num_tokens": 43235078.0,
      "step": 23190
    },
    {
      "entropy": 5.854296731948852,
      "epoch": 2.048304486047333,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000457957203354167,
      "loss": 5.312,
      "mean_token_accuracy": 0.21079981178045273,
      "num_tokens": 43244398.0,
      "step": 23195
    },
    {
      "entropy": 5.898467397689819,
      "epoch": 2.0487460261391734,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045793887149676366,
      "loss": 5.3475,
      "mean_token_accuracy": 0.20842875689268112,
      "num_tokens": 43254638.0,
      "step": 23200
    },
    {
      "entropy": 5.853589296340942,
      "epoch": 2.049187566231014,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000457920536055692,
      "loss": 5.2556,
      "mean_token_accuracy": 0.20489853322505952,
      "num_tokens": 43263287.0,
      "step": 23205
    },
    {
      "entropy": 5.824107694625854,
      "epoch": 2.049629106322854,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045790219703131097,
      "loss": 5.1991,
      "mean_token_accuracy": 0.21740050315856935,
      "num_tokens": 43272921.0,
      "step": 23210
    },
    {
      "entropy": 5.834905529022217,
      "epoch": 2.0500706464146945,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045788385442397994,
      "loss": 5.2603,
      "mean_token_accuracy": 0.21566871106624602,
      "num_tokens": 43283423.0,
      "step": 23215
    },
    {
      "entropy": 5.955960464477539,
      "epoch": 2.050512186506535,
      "grad_norm": 1.0,
      "learning_rate": 0.0004578655082340582,
      "loss": 5.3722,
      "mean_token_accuracy": 0.2022867813706398,
      "num_tokens": 43293031.0,
      "step": 23220
    },
    {
      "entropy": 5.95057225227356,
      "epoch": 2.050953726598375,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004578471584619051,
      "loss": 5.3189,
      "mean_token_accuracy": 0.21219599545001983,
      "num_tokens": 43302775.0,
      "step": 23225
    },
    {
      "entropy": 5.794325494766236,
      "epoch": 2.0513952666902155,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004578288051078803,
      "loss": 5.2276,
      "mean_token_accuracy": 0.219016931951046,
      "num_tokens": 43312341.0,
      "step": 23230
    },
    {
      "entropy": 5.892101144790649,
      "epoch": 2.0518368067820556,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045781044817234305,
      "loss": 5.2703,
      "mean_token_accuracy": 0.20737400501966477,
      "num_tokens": 43321358.0,
      "step": 23235
    },
    {
      "entropy": 5.87251033782959,
      "epoch": 2.052278346873896,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045779208765565324,
      "loss": 5.2762,
      "mean_token_accuracy": 0.21060893982648848,
      "num_tokens": 43331839.0,
      "step": 23240
    },
    {
      "entropy": 5.8935997009277346,
      "epoch": 2.0527198869657366,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004577737235581702,
      "loss": 5.3401,
      "mean_token_accuracy": 0.20013512820005416,
      "num_tokens": 43340384.0,
      "step": 23245
    },
    {
      "entropy": 5.974007654190063,
      "epoch": 2.0531614270575766,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045775535588025404,
      "loss": 5.3282,
      "mean_token_accuracy": 0.20916105359792708,
      "num_tokens": 43349264.0,
      "step": 23250
    },
    {
      "entropy": 5.876816987991333,
      "epoch": 2.053602967149417,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004577369846222643,
      "loss": 5.3145,
      "mean_token_accuracy": 0.21036041229963304,
      "num_tokens": 43358598.0,
      "step": 23255
    },
    {
      "entropy": 5.8475799560546875,
      "epoch": 2.0540445072412576,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004577186097845609,
      "loss": 5.2732,
      "mean_token_accuracy": 0.20657162964344025,
      "num_tokens": 43368357.0,
      "step": 23260
    },
    {
      "entropy": 5.932975196838379,
      "epoch": 2.0544860473330977,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045770023136750405,
      "loss": 5.3351,
      "mean_token_accuracy": 0.21048603802919388,
      "num_tokens": 43378143.0,
      "step": 23265
    },
    {
      "entropy": 5.830256748199463,
      "epoch": 2.054927587424938,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004576818493714534,
      "loss": 5.2482,
      "mean_token_accuracy": 0.22011639922857285,
      "num_tokens": 43387010.0,
      "step": 23270
    },
    {
      "entropy": 5.770633840560913,
      "epoch": 2.0553691275167787,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004576634637967693,
      "loss": 5.1353,
      "mean_token_accuracy": 0.21366823017597197,
      "num_tokens": 43395864.0,
      "step": 23275
    },
    {
      "entropy": 5.874236440658569,
      "epoch": 2.0558106676086187,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004576450746438118,
      "loss": 5.238,
      "mean_token_accuracy": 0.21384260058403015,
      "num_tokens": 43404662.0,
      "step": 23280
    },
    {
      "entropy": 5.947582578659057,
      "epoch": 2.0562522077004592,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004576266819129412,
      "loss": 5.2634,
      "mean_token_accuracy": 0.21163098365068436,
      "num_tokens": 43413927.0,
      "step": 23285
    },
    {
      "entropy": 5.932430458068848,
      "epoch": 2.0566937477922997,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004576082856045177,
      "loss": 5.4263,
      "mean_token_accuracy": 0.1927229180932045,
      "num_tokens": 43424197.0,
      "step": 23290
    },
    {
      "entropy": 5.950639772415161,
      "epoch": 2.05713528788414,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045758988571890176,
      "loss": 5.3221,
      "mean_token_accuracy": 0.20873454362154006,
      "num_tokens": 43432630.0,
      "step": 23295
    },
    {
      "entropy": 5.867905569076538,
      "epoch": 2.0575768279759803,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045757148225645383,
      "loss": 5.3066,
      "mean_token_accuracy": 0.2055157884955406,
      "num_tokens": 43442186.0,
      "step": 23300
    },
    {
      "entropy": 5.831416416168213,
      "epoch": 2.0580183680678203,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045755307521753443,
      "loss": 5.2489,
      "mean_token_accuracy": 0.21740500181913375,
      "num_tokens": 43451927.0,
      "step": 23305
    },
    {
      "entropy": 5.938334369659424,
      "epoch": 2.058459908159661,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045753466460250405,
      "loss": 5.3021,
      "mean_token_accuracy": 0.2072618693113327,
      "num_tokens": 43460879.0,
      "step": 23310
    },
    {
      "entropy": 5.87259202003479,
      "epoch": 2.0589014482515013,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004575162504117234,
      "loss": 5.3632,
      "mean_token_accuracy": 0.20011452138423919,
      "num_tokens": 43469918.0,
      "step": 23315
    },
    {
      "entropy": 5.879549407958985,
      "epoch": 2.0593429883433414,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004574978326455532,
      "loss": 5.2864,
      "mean_token_accuracy": 0.20705562978982925,
      "num_tokens": 43478779.0,
      "step": 23320
    },
    {
      "entropy": 5.918912124633789,
      "epoch": 2.059784528435182,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004574794113043543,
      "loss": 5.3546,
      "mean_token_accuracy": 0.20848302245140077,
      "num_tokens": 43488128.0,
      "step": 23325
    },
    {
      "entropy": 5.86453948020935,
      "epoch": 2.0602260685270224,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004574609863884875,
      "loss": 5.2933,
      "mean_token_accuracy": 0.2075021594762802,
      "num_tokens": 43497567.0,
      "step": 23330
    },
    {
      "entropy": 5.893348693847656,
      "epoch": 2.0606676086188624,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045744255789831377,
      "loss": 5.3112,
      "mean_token_accuracy": 0.20413670539855958,
      "num_tokens": 43507360.0,
      "step": 23335
    },
    {
      "entropy": 5.879449033737183,
      "epoch": 2.061109148710703,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045742412583419403,
      "loss": 5.3127,
      "mean_token_accuracy": 0.20965750366449357,
      "num_tokens": 43517601.0,
      "step": 23340
    },
    {
      "entropy": 5.8456395149230955,
      "epoch": 2.0615506888025434,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045740569019648945,
      "loss": 5.2136,
      "mean_token_accuracy": 0.2215661346912384,
      "num_tokens": 43526348.0,
      "step": 23345
    },
    {
      "entropy": 5.860084772109985,
      "epoch": 2.0619922288943835,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004573872509855612,
      "loss": 5.2259,
      "mean_token_accuracy": 0.2159516081213951,
      "num_tokens": 43536106.0,
      "step": 23350
    },
    {
      "entropy": 5.958714532852173,
      "epoch": 2.062433768986224,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004573688082017703,
      "loss": 5.3625,
      "mean_token_accuracy": 0.20386123806238174,
      "num_tokens": 43545110.0,
      "step": 23355
    },
    {
      "entropy": 5.907396459579468,
      "epoch": 2.0628753090780645,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045735036184547825,
      "loss": 5.2947,
      "mean_token_accuracy": 0.203805835545063,
      "num_tokens": 43554226.0,
      "step": 23360
    },
    {
      "entropy": 5.851353502273559,
      "epoch": 2.0633168491699045,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045733191191704633,
      "loss": 5.2987,
      "mean_token_accuracy": 0.21612913608551027,
      "num_tokens": 43563184.0,
      "step": 23365
    },
    {
      "entropy": 5.859892988204956,
      "epoch": 2.063758389261745,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004573134584168359,
      "loss": 5.28,
      "mean_token_accuracy": 0.22314720451831818,
      "num_tokens": 43572595.0,
      "step": 23370
    },
    {
      "entropy": 5.91445026397705,
      "epoch": 2.064199929353585,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004572950013452086,
      "loss": 5.3716,
      "mean_token_accuracy": 0.19807493388652803,
      "num_tokens": 43582211.0,
      "step": 23375
    },
    {
      "entropy": 5.9036365985870365,
      "epoch": 2.0646414694454256,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004572765407025257,
      "loss": 5.2272,
      "mean_token_accuracy": 0.2168590322136879,
      "num_tokens": 43591060.0,
      "step": 23380
    },
    {
      "entropy": 5.975190877914429,
      "epoch": 2.065083009537266,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004572580764891493,
      "loss": 5.3986,
      "mean_token_accuracy": 0.20219420790672302,
      "num_tokens": 43600057.0,
      "step": 23385
    },
    {
      "entropy": 5.89881591796875,
      "epoch": 2.065524549629106,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004572396087054406,
      "loss": 5.1932,
      "mean_token_accuracy": 0.21935433149337769,
      "num_tokens": 43608609.0,
      "step": 23390
    },
    {
      "entropy": 5.852277374267578,
      "epoch": 2.0659660897209466,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004572211373517618,
      "loss": 5.3267,
      "mean_token_accuracy": 0.20932533144950866,
      "num_tokens": 43618751.0,
      "step": 23395
    },
    {
      "entropy": 5.926204824447632,
      "epoch": 2.066407629812787,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045720266242847446,
      "loss": 5.4134,
      "mean_token_accuracy": 0.20356207489967346,
      "num_tokens": 43628934.0,
      "step": 23400
    },
    {
      "entropy": 5.960275554656983,
      "epoch": 2.066849169904627,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045718418393594055,
      "loss": 5.2958,
      "mean_token_accuracy": 0.21271106153726577,
      "num_tokens": 43638512.0,
      "step": 23405
    },
    {
      "entropy": 5.8644452571868895,
      "epoch": 2.0672907099964677,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045716570187452214,
      "loss": 5.2471,
      "mean_token_accuracy": 0.20868025869131088,
      "num_tokens": 43647708.0,
      "step": 23410
    },
    {
      "entropy": 5.898849296569824,
      "epoch": 2.067732250088308,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004571472162445813,
      "loss": 5.3225,
      "mean_token_accuracy": 0.21653203368186952,
      "num_tokens": 43657096.0,
      "step": 23415
    },
    {
      "entropy": 5.916511583328247,
      "epoch": 2.0681737901801482,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045712872704647994,
      "loss": 5.3291,
      "mean_token_accuracy": 0.2047939881682396,
      "num_tokens": 43667410.0,
      "step": 23420
    },
    {
      "entropy": 5.86441388130188,
      "epoch": 2.0686153302719887,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004571102342805805,
      "loss": 5.2766,
      "mean_token_accuracy": 0.21467859596014022,
      "num_tokens": 43675800.0,
      "step": 23425
    },
    {
      "entropy": 5.853344249725342,
      "epoch": 2.0690568703638292,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004570917379472451,
      "loss": 5.2842,
      "mean_token_accuracy": 0.21802525073289872,
      "num_tokens": 43684802.0,
      "step": 23430
    },
    {
      "entropy": 5.852845191955566,
      "epoch": 2.0694984104556693,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045707323804683616,
      "loss": 5.2881,
      "mean_token_accuracy": 0.21087391823530197,
      "num_tokens": 43693968.0,
      "step": 23435
    },
    {
      "entropy": 5.971442413330078,
      "epoch": 2.06993995054751,
      "grad_norm": 1.15625,
      "learning_rate": 0.000457054734579716,
      "loss": 5.3038,
      "mean_token_accuracy": 0.20274436473846436,
      "num_tokens": 43703177.0,
      "step": 23440
    },
    {
      "entropy": 5.871698904037475,
      "epoch": 2.07038149063935,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045703622754624716,
      "loss": 5.3242,
      "mean_token_accuracy": 0.20798488110303878,
      "num_tokens": 43712894.0,
      "step": 23445
    },
    {
      "entropy": 5.898809242248535,
      "epoch": 2.0708230307311903,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045701771694679213,
      "loss": 5.2313,
      "mean_token_accuracy": 0.2106014609336853,
      "num_tokens": 43722591.0,
      "step": 23450
    },
    {
      "entropy": 5.831428050994873,
      "epoch": 2.071264570823031,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045699920278171355,
      "loss": 5.2211,
      "mean_token_accuracy": 0.22353032380342483,
      "num_tokens": 43732251.0,
      "step": 23455
    },
    {
      "entropy": 5.955831289291382,
      "epoch": 2.071706110914871,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004569806850513741,
      "loss": 5.3866,
      "mean_token_accuracy": 0.20358723849058152,
      "num_tokens": 43741061.0,
      "step": 23460
    },
    {
      "entropy": 5.960063028335571,
      "epoch": 2.0721476510067114,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045696216375613647,
      "loss": 5.344,
      "mean_token_accuracy": 0.20664846152067184,
      "num_tokens": 43750147.0,
      "step": 23465
    },
    {
      "entropy": 5.944277334213257,
      "epoch": 2.072589191098552,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004569436388963636,
      "loss": 5.3813,
      "mean_token_accuracy": 0.19914156794548035,
      "num_tokens": 43760960.0,
      "step": 23470
    },
    {
      "entropy": 5.876435232162476,
      "epoch": 2.073030731190392,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004569251104724184,
      "loss": 5.29,
      "mean_token_accuracy": 0.21085326373577118,
      "num_tokens": 43771306.0,
      "step": 23475
    },
    {
      "entropy": 5.827397775650025,
      "epoch": 2.0734722712822324,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004569065784846636,
      "loss": 5.2162,
      "mean_token_accuracy": 0.21310461759567262,
      "num_tokens": 43780520.0,
      "step": 23480
    },
    {
      "entropy": 5.924309492111206,
      "epoch": 2.073913811374073,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004568880429334624,
      "loss": 5.3474,
      "mean_token_accuracy": 0.20316143929958344,
      "num_tokens": 43789472.0,
      "step": 23485
    },
    {
      "entropy": 5.944648361206054,
      "epoch": 2.074355351465913,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045686950381917805,
      "loss": 5.3329,
      "mean_token_accuracy": 0.20691990852355957,
      "num_tokens": 43798728.0,
      "step": 23490
    },
    {
      "entropy": 5.900263738632202,
      "epoch": 2.0747968915577535,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045685096114217336,
      "loss": 5.2694,
      "mean_token_accuracy": 0.21408158540725708,
      "num_tokens": 43807959.0,
      "step": 23495
    },
    {
      "entropy": 5.9307873249053955,
      "epoch": 2.075238431649594,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004568324149028119,
      "loss": 5.2792,
      "mean_token_accuracy": 0.21293805688619613,
      "num_tokens": 43816922.0,
      "step": 23500
    },
    {
      "entropy": 5.865668439865113,
      "epoch": 2.075679971741434,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004568138651014567,
      "loss": 5.2798,
      "mean_token_accuracy": 0.21039648205041886,
      "num_tokens": 43827195.0,
      "step": 23505
    },
    {
      "entropy": 5.8485218524932865,
      "epoch": 2.0761215118332745,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004567953117384714,
      "loss": 5.2377,
      "mean_token_accuracy": 0.2087216168642044,
      "num_tokens": 43836289.0,
      "step": 23510
    },
    {
      "entropy": 5.925881385803223,
      "epoch": 2.0765630519251146,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004567767548142193,
      "loss": 5.3305,
      "mean_token_accuracy": 0.1997263640165329,
      "num_tokens": 43845918.0,
      "step": 23515
    },
    {
      "entropy": 5.949883651733399,
      "epoch": 2.077004592016955,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045675819432906394,
      "loss": 5.3831,
      "mean_token_accuracy": 0.20143934786319734,
      "num_tokens": 43853907.0,
      "step": 23520
    },
    {
      "entropy": 5.926950120925904,
      "epoch": 2.0774461321087956,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045673963028336896,
      "loss": 5.2985,
      "mean_token_accuracy": 0.2107247456908226,
      "num_tokens": 43862992.0,
      "step": 23525
    },
    {
      "entropy": 5.944992971420288,
      "epoch": 2.0778876722006356,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045672106267749795,
      "loss": 5.3076,
      "mean_token_accuracy": 0.20400300472974778,
      "num_tokens": 43871686.0,
      "step": 23530
    },
    {
      "entropy": 5.861617517471314,
      "epoch": 2.078329212292476,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004567024915118148,
      "loss": 5.2705,
      "mean_token_accuracy": 0.21350021064281463,
      "num_tokens": 43881302.0,
      "step": 23535
    },
    {
      "entropy": 5.98052225112915,
      "epoch": 2.0787707523843166,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045668391678668296,
      "loss": 5.3715,
      "mean_token_accuracy": 0.1990432098507881,
      "num_tokens": 43889982.0,
      "step": 23540
    },
    {
      "entropy": 5.944049739837647,
      "epoch": 2.0792122924761567,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045666533850246665,
      "loss": 5.3208,
      "mean_token_accuracy": 0.21239465922117234,
      "num_tokens": 43898385.0,
      "step": 23545
    },
    {
      "entropy": 5.903633546829224,
      "epoch": 2.079653832567997,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004566467566595297,
      "loss": 5.2903,
      "mean_token_accuracy": 0.20155944675207138,
      "num_tokens": 43907882.0,
      "step": 23550
    },
    {
      "entropy": 5.994525814056397,
      "epoch": 2.0800953726598377,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045662817125823605,
      "loss": 5.3112,
      "mean_token_accuracy": 0.20893317013978957,
      "num_tokens": 43916222.0,
      "step": 23555
    },
    {
      "entropy": 5.881528663635254,
      "epoch": 2.0805369127516777,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045660958229894985,
      "loss": 5.2574,
      "mean_token_accuracy": 0.2135061115026474,
      "num_tokens": 43925181.0,
      "step": 23560
    },
    {
      "entropy": 5.7951517581939695,
      "epoch": 2.0809784528435182,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045659098978203524,
      "loss": 5.3031,
      "mean_token_accuracy": 0.21031264066696168,
      "num_tokens": 43933973.0,
      "step": 23565
    },
    {
      "entropy": 5.808804893493653,
      "epoch": 2.0814199929353587,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004565723937078563,
      "loss": 5.291,
      "mean_token_accuracy": 0.2105001538991928,
      "num_tokens": 43943297.0,
      "step": 23570
    },
    {
      "entropy": 5.924958324432373,
      "epoch": 2.081861533027199,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045655379407677754,
      "loss": 5.325,
      "mean_token_accuracy": 0.21227133572101592,
      "num_tokens": 43952679.0,
      "step": 23575
    },
    {
      "entropy": 5.91604175567627,
      "epoch": 2.0823030731190393,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004565351908891632,
      "loss": 5.3078,
      "mean_token_accuracy": 0.20379597544670106,
      "num_tokens": 43961629.0,
      "step": 23580
    },
    {
      "entropy": 5.919429159164428,
      "epoch": 2.0827446132108793,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045651658414537765,
      "loss": 5.3176,
      "mean_token_accuracy": 0.20624782890081406,
      "num_tokens": 43970880.0,
      "step": 23585
    },
    {
      "entropy": 5.86760778427124,
      "epoch": 2.08318615330272,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004564979738457855,
      "loss": 5.243,
      "mean_token_accuracy": 0.21680566668510437,
      "num_tokens": 43980449.0,
      "step": 23590
    },
    {
      "entropy": 5.806300783157349,
      "epoch": 2.0836276933945603,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045647935999075127,
      "loss": 5.268,
      "mean_token_accuracy": 0.2086215451359749,
      "num_tokens": 43988716.0,
      "step": 23595
    },
    {
      "entropy": 5.899968242645263,
      "epoch": 2.0840692334864004,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045646074258063956,
      "loss": 5.3184,
      "mean_token_accuracy": 0.2169812336564064,
      "num_tokens": 43998464.0,
      "step": 23600
    },
    {
      "entropy": 5.960954332351685,
      "epoch": 2.084510773578241,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004564421216158152,
      "loss": 5.3207,
      "mean_token_accuracy": 0.21247612088918685,
      "num_tokens": 44006993.0,
      "step": 23605
    },
    {
      "entropy": 5.959644079208374,
      "epoch": 2.0849523136700814,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004564234970966428,
      "loss": 5.3605,
      "mean_token_accuracy": 0.20201978981494903,
      "num_tokens": 44015801.0,
      "step": 23610
    },
    {
      "entropy": 5.814644622802734,
      "epoch": 2.0853938537619214,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004564048690234872,
      "loss": 5.2859,
      "mean_token_accuracy": 0.20764555484056474,
      "num_tokens": 44025069.0,
      "step": 23615
    },
    {
      "entropy": 5.928499841690064,
      "epoch": 2.085835393853762,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045638623739671343,
      "loss": 5.4187,
      "mean_token_accuracy": 0.19506964981555938,
      "num_tokens": 44034700.0,
      "step": 23620
    },
    {
      "entropy": 5.920329284667969,
      "epoch": 2.0862769339456024,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045636760221668646,
      "loss": 5.2299,
      "mean_token_accuracy": 0.21019828021526338,
      "num_tokens": 44044417.0,
      "step": 23625
    },
    {
      "entropy": 5.955547571182251,
      "epoch": 2.0867184740374425,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004563489634837713,
      "loss": 5.2978,
      "mean_token_accuracy": 0.2106105163693428,
      "num_tokens": 44053617.0,
      "step": 23630
    },
    {
      "entropy": 5.844348382949829,
      "epoch": 2.087160014129283,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045633032119833314,
      "loss": 5.2449,
      "mean_token_accuracy": 0.20604305118322372,
      "num_tokens": 44062738.0,
      "step": 23635
    },
    {
      "entropy": 5.903376245498658,
      "epoch": 2.0876015542211235,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000456311675360737,
      "loss": 5.3416,
      "mean_token_accuracy": 0.19985350221395493,
      "num_tokens": 44072154.0,
      "step": 23640
    },
    {
      "entropy": 5.9602282524108885,
      "epoch": 2.0880430943129635,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004562930259713483,
      "loss": 5.4065,
      "mean_token_accuracy": 0.20159982442855834,
      "num_tokens": 44081239.0,
      "step": 23645
    },
    {
      "entropy": 5.894659423828125,
      "epoch": 2.088484634404804,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045627437303053236,
      "loss": 5.2513,
      "mean_token_accuracy": 0.21409705579280852,
      "num_tokens": 44090866.0,
      "step": 23650
    },
    {
      "entropy": 5.928786420822144,
      "epoch": 2.088926174496644,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004562557165386546,
      "loss": 5.4022,
      "mean_token_accuracy": 0.2024052545428276,
      "num_tokens": 44100225.0,
      "step": 23655
    },
    {
      "entropy": 5.900218963623047,
      "epoch": 2.0893677145884846,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004562370564960804,
      "loss": 5.3129,
      "mean_token_accuracy": 0.20917170196771623,
      "num_tokens": 44110139.0,
      "step": 23660
    },
    {
      "entropy": 5.883913373947143,
      "epoch": 2.089809254680325,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045621839290317526,
      "loss": 5.3128,
      "mean_token_accuracy": 0.2127288267016411,
      "num_tokens": 44118687.0,
      "step": 23665
    },
    {
      "entropy": 5.863697481155396,
      "epoch": 2.090250794772165,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045619972576030495,
      "loss": 5.295,
      "mean_token_accuracy": 0.20769568234682084,
      "num_tokens": 44127377.0,
      "step": 23670
    },
    {
      "entropy": 5.851103830337524,
      "epoch": 2.0906923348640056,
      "grad_norm": 1.015625,
      "learning_rate": 0.000456181055067835,
      "loss": 5.322,
      "mean_token_accuracy": 0.2059984490275383,
      "num_tokens": 44136583.0,
      "step": 23675
    },
    {
      "entropy": 5.866923856735229,
      "epoch": 2.091133874955846,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004561623808261313,
      "loss": 5.3066,
      "mean_token_accuracy": 0.20673833340406417,
      "num_tokens": 44145467.0,
      "step": 23680
    },
    {
      "entropy": 5.860962438583374,
      "epoch": 2.091575415047686,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004561437030355595,
      "loss": 5.2872,
      "mean_token_accuracy": 0.21174018681049347,
      "num_tokens": 44154932.0,
      "step": 23685
    },
    {
      "entropy": 5.835395765304566,
      "epoch": 2.0920169551395267,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004561250216964857,
      "loss": 5.2177,
      "mean_token_accuracy": 0.216437166929245,
      "num_tokens": 44162791.0,
      "step": 23690
    },
    {
      "entropy": 5.955163192749024,
      "epoch": 2.092458495231367,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004561063368092757,
      "loss": 5.3477,
      "mean_token_accuracy": 0.21063789576292039,
      "num_tokens": 44171993.0,
      "step": 23695
    },
    {
      "entropy": 5.9274516105651855,
      "epoch": 2.0929000353232072,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045608764837429543,
      "loss": 5.3483,
      "mean_token_accuracy": 0.207102532684803,
      "num_tokens": 44180807.0,
      "step": 23700
    },
    {
      "entropy": 5.9769903182983395,
      "epoch": 2.0933415754150477,
      "grad_norm": 1.0,
      "learning_rate": 0.00045606895639191126,
      "loss": 5.3712,
      "mean_token_accuracy": 0.20448126792907714,
      "num_tokens": 44191764.0,
      "step": 23705
    },
    {
      "entropy": 5.8805944442749025,
      "epoch": 2.0937831155068882,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045605026086248916,
      "loss": 5.3658,
      "mean_token_accuracy": 0.205495585501194,
      "num_tokens": 44201165.0,
      "step": 23710
    },
    {
      "entropy": 5.906459140777588,
      "epoch": 2.0942246555987283,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004560315617863954,
      "loss": 5.362,
      "mean_token_accuracy": 0.20707882344722747,
      "num_tokens": 44210709.0,
      "step": 23715
    },
    {
      "entropy": 5.90650110244751,
      "epoch": 2.0946661956905688,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004560128591639964,
      "loss": 5.322,
      "mean_token_accuracy": 0.20687424540519714,
      "num_tokens": 44219827.0,
      "step": 23720
    },
    {
      "entropy": 5.9476954460144045,
      "epoch": 2.095107735782409,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004559941529956584,
      "loss": 5.3852,
      "mean_token_accuracy": 0.20104854106903075,
      "num_tokens": 44229498.0,
      "step": 23725
    },
    {
      "entropy": 5.884517669677734,
      "epoch": 2.0955492758742493,
      "grad_norm": 1.0,
      "learning_rate": 0.0004559754432817479,
      "loss": 5.2882,
      "mean_token_accuracy": 0.20786643475294114,
      "num_tokens": 44238844.0,
      "step": 23730
    },
    {
      "entropy": 5.85141773223877,
      "epoch": 2.09599081596609,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045595673002263125,
      "loss": 5.2867,
      "mean_token_accuracy": 0.21021698862314225,
      "num_tokens": 44247512.0,
      "step": 23735
    },
    {
      "entropy": 5.870313405990601,
      "epoch": 2.09643235605793,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004559380132186753,
      "loss": 5.233,
      "mean_token_accuracy": 0.21559084057807923,
      "num_tokens": 44256485.0,
      "step": 23740
    },
    {
      "entropy": 5.785613679885865,
      "epoch": 2.0968738961497704,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045591929287024655,
      "loss": 5.1362,
      "mean_token_accuracy": 0.2268437996506691,
      "num_tokens": 44265095.0,
      "step": 23745
    },
    {
      "entropy": 5.952377986907959,
      "epoch": 2.097315436241611,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004559005689777117,
      "loss": 5.3197,
      "mean_token_accuracy": 0.20878552198410033,
      "num_tokens": 44274582.0,
      "step": 23750
    },
    {
      "entropy": 5.964988803863525,
      "epoch": 2.097756976333451,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004558818415414377,
      "loss": 5.3585,
      "mean_token_accuracy": 0.20715403109788894,
      "num_tokens": 44285752.0,
      "step": 23755
    },
    {
      "entropy": 5.932619190216064,
      "epoch": 2.0981985164252914,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004558631105617912,
      "loss": 5.3519,
      "mean_token_accuracy": 0.20505301356315614,
      "num_tokens": 44295315.0,
      "step": 23760
    },
    {
      "entropy": 5.882716035842895,
      "epoch": 2.098640056517132,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045584437603913923,
      "loss": 5.4127,
      "mean_token_accuracy": 0.20295526087284088,
      "num_tokens": 44305151.0,
      "step": 23765
    },
    {
      "entropy": 5.892424249649048,
      "epoch": 2.099081596608972,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045582563797384886,
      "loss": 5.3798,
      "mean_token_accuracy": 0.2029080092906952,
      "num_tokens": 44314462.0,
      "step": 23770
    },
    {
      "entropy": 5.89073224067688,
      "epoch": 2.0995231367008125,
      "grad_norm": 1.109375,
      "learning_rate": 0.000455806896366287,
      "loss": 5.1972,
      "mean_token_accuracy": 0.2278853550553322,
      "num_tokens": 44322782.0,
      "step": 23775
    },
    {
      "entropy": 5.886151266098023,
      "epoch": 2.099964676792653,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004557881512168209,
      "loss": 5.2521,
      "mean_token_accuracy": 0.21490204632282256,
      "num_tokens": 44331588.0,
      "step": 23780
    },
    {
      "entropy": 5.861545515060425,
      "epoch": 2.100406216884493,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004557694025258177,
      "loss": 5.297,
      "mean_token_accuracy": 0.218393275141716,
      "num_tokens": 44340997.0,
      "step": 23785
    },
    {
      "entropy": 5.890779113769531,
      "epoch": 2.1008477569763335,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004557506502936448,
      "loss": 5.3388,
      "mean_token_accuracy": 0.20636630952358245,
      "num_tokens": 44350679.0,
      "step": 23790
    },
    {
      "entropy": 5.870981740951538,
      "epoch": 2.1012892970681736,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045573189452066936,
      "loss": 5.2414,
      "mean_token_accuracy": 0.20990516096353531,
      "num_tokens": 44359934.0,
      "step": 23795
    },
    {
      "entropy": 5.852165365219117,
      "epoch": 2.101730837160014,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045571313520725897,
      "loss": 5.2844,
      "mean_token_accuracy": 0.2032035231590271,
      "num_tokens": 44369104.0,
      "step": 23800
    },
    {
      "entropy": 5.898313808441162,
      "epoch": 2.1021723772518546,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045569437235378103,
      "loss": 5.2945,
      "mean_token_accuracy": 0.211219222843647,
      "num_tokens": 44379645.0,
      "step": 23805
    },
    {
      "entropy": 5.959102010726928,
      "epoch": 2.1026139173436946,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004556756059606031,
      "loss": 5.3457,
      "mean_token_accuracy": 0.20944539457559586,
      "num_tokens": 44389276.0,
      "step": 23810
    },
    {
      "entropy": 5.868195581436157,
      "epoch": 2.103055457435535,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004556568360280928,
      "loss": 5.2925,
      "mean_token_accuracy": 0.20601432025432587,
      "num_tokens": 44399114.0,
      "step": 23815
    },
    {
      "entropy": 5.74198055267334,
      "epoch": 2.1034969975273756,
      "grad_norm": 1.21875,
      "learning_rate": 0.00045563806255661784,
      "loss": 5.0541,
      "mean_token_accuracy": 0.236716628074646,
      "num_tokens": 44406917.0,
      "step": 23820
    },
    {
      "entropy": 5.911471319198609,
      "epoch": 2.1039385376192157,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004556192855465459,
      "loss": 5.3171,
      "mean_token_accuracy": 0.19750374257564546,
      "num_tokens": 44416678.0,
      "step": 23825
    },
    {
      "entropy": 5.739448165893554,
      "epoch": 2.104380077711056,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004556005049982449,
      "loss": 5.2176,
      "mean_token_accuracy": 0.21464794725179673,
      "num_tokens": 44425427.0,
      "step": 23830
    },
    {
      "entropy": 5.9034843921661375,
      "epoch": 2.1048216178028967,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045558172091208284,
      "loss": 5.2744,
      "mean_token_accuracy": 0.21139759868383406,
      "num_tokens": 44434267.0,
      "step": 23835
    },
    {
      "entropy": 5.930653715133667,
      "epoch": 2.1052631578947367,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045556293328842746,
      "loss": 5.2926,
      "mean_token_accuracy": 0.21125019043684007,
      "num_tokens": 44444037.0,
      "step": 23840
    },
    {
      "entropy": 5.889728116989136,
      "epoch": 2.1057046979865772,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004555441421276469,
      "loss": 5.2307,
      "mean_token_accuracy": 0.2160533383488655,
      "num_tokens": 44453707.0,
      "step": 23845
    },
    {
      "entropy": 5.87437105178833,
      "epoch": 2.1061462380784177,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045552534743010934,
      "loss": 5.2888,
      "mean_token_accuracy": 0.20803144723176956,
      "num_tokens": 44464357.0,
      "step": 23850
    },
    {
      "entropy": 5.936022567749023,
      "epoch": 2.1065877781702578,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045550654919618283,
      "loss": 5.4028,
      "mean_token_accuracy": 0.1988999828696251,
      "num_tokens": 44473803.0,
      "step": 23855
    },
    {
      "entropy": 5.94085168838501,
      "epoch": 2.1070293182620983,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004554877474262357,
      "loss": 5.3156,
      "mean_token_accuracy": 0.20876477062702178,
      "num_tokens": 44482830.0,
      "step": 23860
    },
    {
      "entropy": 5.863091659545899,
      "epoch": 2.1074708583539383,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004554689421206362,
      "loss": 5.2874,
      "mean_token_accuracy": 0.2077044203877449,
      "num_tokens": 44493045.0,
      "step": 23865
    },
    {
      "entropy": 5.886468124389649,
      "epoch": 2.107912398445779,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004554501332797529,
      "loss": 5.2631,
      "mean_token_accuracy": 0.21491543352603912,
      "num_tokens": 44503304.0,
      "step": 23870
    },
    {
      "entropy": 5.9362359046936035,
      "epoch": 2.1083539385376193,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045543132090395403,
      "loss": 5.3624,
      "mean_token_accuracy": 0.2076067328453064,
      "num_tokens": 44512337.0,
      "step": 23875
    },
    {
      "entropy": 5.8820459842681885,
      "epoch": 2.1087954786294594,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004554125049936082,
      "loss": 5.2974,
      "mean_token_accuracy": 0.2137009918689728,
      "num_tokens": 44521881.0,
      "step": 23880
    },
    {
      "entropy": 5.847499990463257,
      "epoch": 2.1092370187213,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000455393685549084,
      "loss": 5.3205,
      "mean_token_accuracy": 0.20926243662834168,
      "num_tokens": 44531351.0,
      "step": 23885
    },
    {
      "entropy": 5.9422526359558105,
      "epoch": 2.1096785588131404,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004553748625707501,
      "loss": 5.341,
      "mean_token_accuracy": 0.20805851221084595,
      "num_tokens": 44540592.0,
      "step": 23890
    },
    {
      "entropy": 5.85393443107605,
      "epoch": 2.1101200989049804,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045535603605897516,
      "loss": 5.253,
      "mean_token_accuracy": 0.21550757437944412,
      "num_tokens": 44550019.0,
      "step": 23895
    },
    {
      "entropy": 5.859655952453613,
      "epoch": 2.110561638996821,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004553372060141282,
      "loss": 5.2711,
      "mean_token_accuracy": 0.21352385729551315,
      "num_tokens": 44559423.0,
      "step": 23900
    },
    {
      "entropy": 5.809151220321655,
      "epoch": 2.1110031790886614,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004553183724365778,
      "loss": 5.2789,
      "mean_token_accuracy": 0.2176601141691208,
      "num_tokens": 44568792.0,
      "step": 23905
    },
    {
      "entropy": 5.8898368835449215,
      "epoch": 2.1114447191805015,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004552995353266931,
      "loss": 5.3005,
      "mean_token_accuracy": 0.2047635018825531,
      "num_tokens": 44578191.0,
      "step": 23910
    },
    {
      "entropy": 5.89700927734375,
      "epoch": 2.111886259272342,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004552806946848431,
      "loss": 5.2472,
      "mean_token_accuracy": 0.21541159152984618,
      "num_tokens": 44587402.0,
      "step": 23915
    },
    {
      "entropy": 5.892294073104859,
      "epoch": 2.1123277993641825,
      "grad_norm": 1.125,
      "learning_rate": 0.00045526185051139676,
      "loss": 5.4001,
      "mean_token_accuracy": 0.1966974601149559,
      "num_tokens": 44596241.0,
      "step": 23920
    },
    {
      "entropy": 5.860474920272827,
      "epoch": 2.1127693394560225,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045524300280672326,
      "loss": 5.1832,
      "mean_token_accuracy": 0.21655723303556443,
      "num_tokens": 44604796.0,
      "step": 23925
    },
    {
      "entropy": 5.904183626174927,
      "epoch": 2.113210879547863,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045522415157119185,
      "loss": 5.3388,
      "mean_token_accuracy": 0.20518684834241868,
      "num_tokens": 44615208.0,
      "step": 23930
    },
    {
      "entropy": 5.888106870651245,
      "epoch": 2.113652419639703,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004552052968051719,
      "loss": 5.3297,
      "mean_token_accuracy": 0.2135383367538452,
      "num_tokens": 44623937.0,
      "step": 23935
    },
    {
      "entropy": 5.874234342575074,
      "epoch": 2.1140939597315436,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004551864385090326,
      "loss": 5.2459,
      "mean_token_accuracy": 0.21329738050699235,
      "num_tokens": 44634116.0,
      "step": 23940
    },
    {
      "entropy": 5.931789445877075,
      "epoch": 2.114535499823384,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004551675766831434,
      "loss": 5.3283,
      "mean_token_accuracy": 0.20871264338493348,
      "num_tokens": 44643436.0,
      "step": 23945
    },
    {
      "entropy": 5.93265700340271,
      "epoch": 2.114977039915224,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045514871132787394,
      "loss": 5.3589,
      "mean_token_accuracy": 0.20905428230762482,
      "num_tokens": 44654077.0,
      "step": 23950
    },
    {
      "entropy": 5.929782104492188,
      "epoch": 2.1154185800070646,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045512984244359364,
      "loss": 5.3321,
      "mean_token_accuracy": 0.20821423083543777,
      "num_tokens": 44663785.0,
      "step": 23955
    },
    {
      "entropy": 5.7986588954925535,
      "epoch": 2.115860120098905,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045511097003067216,
      "loss": 5.1766,
      "mean_token_accuracy": 0.2192887246608734,
      "num_tokens": 44672750.0,
      "step": 23960
    },
    {
      "entropy": 5.83315076828003,
      "epoch": 2.116301660190745,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045509209408947914,
      "loss": 5.2325,
      "mean_token_accuracy": 0.22007079273462296,
      "num_tokens": 44681814.0,
      "step": 23965
    },
    {
      "entropy": 5.837835454940796,
      "epoch": 2.1167432002825857,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004550732146203845,
      "loss": 5.2804,
      "mean_token_accuracy": 0.21167240887880326,
      "num_tokens": 44690926.0,
      "step": 23970
    },
    {
      "entropy": 5.964428472518921,
      "epoch": 2.117184740374426,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045505433162375806,
      "loss": 5.3964,
      "mean_token_accuracy": 0.2062529668211937,
      "num_tokens": 44700625.0,
      "step": 23975
    },
    {
      "entropy": 5.9191807270050045,
      "epoch": 2.117626280466266,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004550354450999696,
      "loss": 5.3378,
      "mean_token_accuracy": 0.21039126217365264,
      "num_tokens": 44710703.0,
      "step": 23980
    },
    {
      "entropy": 5.89006028175354,
      "epoch": 2.1180678205581067,
      "grad_norm": 1.125,
      "learning_rate": 0.0004550165550493891,
      "loss": 5.2458,
      "mean_token_accuracy": 0.21144918948411942,
      "num_tokens": 44719412.0,
      "step": 23985
    },
    {
      "entropy": 5.969325304031372,
      "epoch": 2.118509360649947,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045499766147238675,
      "loss": 5.3802,
      "mean_token_accuracy": 0.2084307700395584,
      "num_tokens": 44729161.0,
      "step": 23990
    },
    {
      "entropy": 6.003976058959961,
      "epoch": 2.1189509007417873,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045497876436933257,
      "loss": 5.3664,
      "mean_token_accuracy": 0.2052161544561386,
      "num_tokens": 44738985.0,
      "step": 23995
    },
    {
      "entropy": 5.8369025707244875,
      "epoch": 2.1193924408336278,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045495986374059674,
      "loss": 5.2828,
      "mean_token_accuracy": 0.21516386717557906,
      "num_tokens": 44748070.0,
      "step": 24000
    },
    {
      "epoch": 2.1193924408336278,
      "eval_entropy": 5.6000492876275345,
      "eval_loss": 5.37293815612793,
      "eval_mean_token_accuracy": 0.21494417624761133,
      "eval_num_tokens": 44748070.0,
      "eval_runtime": 20.1103,
      "eval_samples_per_second": 1756.066,
      "eval_steps_per_second": 219.539,
      "step": 24000
    },
    {
      "entropy": 5.840093088150025,
      "epoch": 2.119833980925468,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004549409595865496,
      "loss": 5.245,
      "mean_token_accuracy": 0.21080335974693298,
      "num_tokens": 44756607.0,
      "step": 24005
    },
    {
      "entropy": 5.908659410476685,
      "epoch": 2.1202755210173083,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045492205190756133,
      "loss": 5.3401,
      "mean_token_accuracy": 0.21164225786924362,
      "num_tokens": 44765435.0,
      "step": 24010
    },
    {
      "entropy": 5.837518882751465,
      "epoch": 2.120717061109149,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004549031407040024,
      "loss": 5.2953,
      "mean_token_accuracy": 0.2108205586671829,
      "num_tokens": 44775589.0,
      "step": 24015
    },
    {
      "entropy": 5.82951250076294,
      "epoch": 2.121158601200989,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004548842259762432,
      "loss": 5.3095,
      "mean_token_accuracy": 0.20778565406799315,
      "num_tokens": 44785692.0,
      "step": 24020
    },
    {
      "entropy": 5.839303016662598,
      "epoch": 2.1216001412928294,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004548653077246544,
      "loss": 5.2111,
      "mean_token_accuracy": 0.2144887626171112,
      "num_tokens": 44794074.0,
      "step": 24025
    },
    {
      "entropy": 5.919186305999756,
      "epoch": 2.12204168138467,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004548463859496065,
      "loss": 5.3064,
      "mean_token_accuracy": 0.2060460314154625,
      "num_tokens": 44803959.0,
      "step": 24030
    },
    {
      "entropy": 5.827262878417969,
      "epoch": 2.12248322147651,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045482746065147014,
      "loss": 5.2547,
      "mean_token_accuracy": 0.21491232961416246,
      "num_tokens": 44813294.0,
      "step": 24035
    },
    {
      "entropy": 5.909788274765015,
      "epoch": 2.1229247615683504,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004548085318306162,
      "loss": 5.2704,
      "mean_token_accuracy": 0.2111326724290848,
      "num_tokens": 44822788.0,
      "step": 24040
    },
    {
      "entropy": 5.879832172393799,
      "epoch": 2.123366301660191,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045478959948741526,
      "loss": 5.2203,
      "mean_token_accuracy": 0.21731253564357758,
      "num_tokens": 44832236.0,
      "step": 24045
    },
    {
      "entropy": 5.890810585021972,
      "epoch": 2.123807841752031,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004547706636222384,
      "loss": 5.3052,
      "mean_token_accuracy": 0.21453931629657746,
      "num_tokens": 44841258.0,
      "step": 24050
    },
    {
      "entropy": 5.879990100860596,
      "epoch": 2.1242493818438715,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004547517242354565,
      "loss": 5.2852,
      "mean_token_accuracy": 0.21117878407239915,
      "num_tokens": 44850365.0,
      "step": 24055
    },
    {
      "entropy": 5.933914661407471,
      "epoch": 2.124690921935712,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004547327813274405,
      "loss": 5.3128,
      "mean_token_accuracy": 0.20933755487203598,
      "num_tokens": 44860451.0,
      "step": 24060
    },
    {
      "entropy": 5.914949178695679,
      "epoch": 2.125132462027552,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004547138348985615,
      "loss": 5.229,
      "mean_token_accuracy": 0.21384336054325104,
      "num_tokens": 44868815.0,
      "step": 24065
    },
    {
      "entropy": 5.891973161697388,
      "epoch": 2.1255740021193925,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045469488494919076,
      "loss": 5.3474,
      "mean_token_accuracy": 0.20765818804502487,
      "num_tokens": 44878046.0,
      "step": 24070
    },
    {
      "entropy": 5.852796030044556,
      "epoch": 2.126015542211233,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004546759314796994,
      "loss": 5.2491,
      "mean_token_accuracy": 0.21183056086301805,
      "num_tokens": 44886915.0,
      "step": 24075
    },
    {
      "entropy": 5.947250699996948,
      "epoch": 2.126457082303073,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004546569744904586,
      "loss": 5.2978,
      "mean_token_accuracy": 0.21096168011426925,
      "num_tokens": 44895629.0,
      "step": 24080
    },
    {
      "entropy": 5.915944147109985,
      "epoch": 2.1268986223949136,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045463801398184004,
      "loss": 5.4305,
      "mean_token_accuracy": 0.19379859417676926,
      "num_tokens": 44905913.0,
      "step": 24085
    },
    {
      "entropy": 5.990835237503052,
      "epoch": 2.1273401624867536,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004546190499542149,
      "loss": 5.3184,
      "mean_token_accuracy": 0.20539964288473128,
      "num_tokens": 44915329.0,
      "step": 24090
    },
    {
      "entropy": 5.945349216461182,
      "epoch": 2.127781702578594,
      "grad_norm": 1.125,
      "learning_rate": 0.0004546000824079547,
      "loss": 5.2814,
      "mean_token_accuracy": 0.21771723330020903,
      "num_tokens": 44923810.0,
      "step": 24095
    },
    {
      "entropy": 5.89575457572937,
      "epoch": 2.1282232426704346,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000454581111343431,
      "loss": 5.3159,
      "mean_token_accuracy": 0.20633020401000976,
      "num_tokens": 44932375.0,
      "step": 24100
    },
    {
      "entropy": 5.936436319351197,
      "epoch": 2.1286647827622747,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004545621367610155,
      "loss": 5.2982,
      "mean_token_accuracy": 0.20895180851221085,
      "num_tokens": 44941341.0,
      "step": 24105
    },
    {
      "entropy": 5.955826473236084,
      "epoch": 2.129106322854115,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004545431586610799,
      "loss": 5.3346,
      "mean_token_accuracy": 0.21046103835105895,
      "num_tokens": 44951221.0,
      "step": 24110
    },
    {
      "entropy": 5.829993438720703,
      "epoch": 2.1295478629459557,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004545241770439958,
      "loss": 5.2231,
      "mean_token_accuracy": 0.22138845771551133,
      "num_tokens": 44960254.0,
      "step": 24115
    },
    {
      "entropy": 5.931124496459961,
      "epoch": 2.1299894030377957,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004545051919101353,
      "loss": 5.4155,
      "mean_token_accuracy": 0.20018528252840043,
      "num_tokens": 44969023.0,
      "step": 24120
    },
    {
      "entropy": 5.86586709022522,
      "epoch": 2.130430943129636,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045448620325987013,
      "loss": 5.2894,
      "mean_token_accuracy": 0.20908062011003495,
      "num_tokens": 44978942.0,
      "step": 24125
    },
    {
      "entropy": 5.921107578277588,
      "epoch": 2.1308724832214767,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045446721109357243,
      "loss": 5.2214,
      "mean_token_accuracy": 0.2165845513343811,
      "num_tokens": 44987794.0,
      "step": 24130
    },
    {
      "entropy": 5.948254728317261,
      "epoch": 2.1313140233133168,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045444821541161404,
      "loss": 5.3669,
      "mean_token_accuracy": 0.20377086102962494,
      "num_tokens": 44995781.0,
      "step": 24135
    },
    {
      "entropy": 5.889995145797729,
      "epoch": 2.1317555634051573,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004544292162143671,
      "loss": 5.3078,
      "mean_token_accuracy": 0.20938256084918977,
      "num_tokens": 45004731.0,
      "step": 24140
    },
    {
      "entropy": 5.9272013187408445,
      "epoch": 2.1321971034969973,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004544102135022039,
      "loss": 5.3361,
      "mean_token_accuracy": 0.20272944867610931,
      "num_tokens": 45014312.0,
      "step": 24145
    },
    {
      "entropy": 5.949671792984009,
      "epoch": 2.132638643588838,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004543912072754967,
      "loss": 5.3983,
      "mean_token_accuracy": 0.19977204501628876,
      "num_tokens": 45023903.0,
      "step": 24150
    },
    {
      "entropy": 5.915448331832886,
      "epoch": 2.1330801836806783,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004543721975346177,
      "loss": 5.3078,
      "mean_token_accuracy": 0.20177028626203536,
      "num_tokens": 45033330.0,
      "step": 24155
    },
    {
      "entropy": 5.796429586410523,
      "epoch": 2.1335217237725184,
      "grad_norm": 1.0,
      "learning_rate": 0.00045435318427993946,
      "loss": 5.1934,
      "mean_token_accuracy": 0.21853149682283401,
      "num_tokens": 45042862.0,
      "step": 24160
    },
    {
      "entropy": 5.894321537017822,
      "epoch": 2.133963263864359,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004543341675118343,
      "loss": 5.2621,
      "mean_token_accuracy": 0.2157619684934616,
      "num_tokens": 45052198.0,
      "step": 24165
    },
    {
      "entropy": 5.954981088638306,
      "epoch": 2.1344048039561994,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045431514723067487,
      "loss": 5.3183,
      "mean_token_accuracy": 0.21114950031042098,
      "num_tokens": 45060625.0,
      "step": 24170
    },
    {
      "entropy": 5.871349906921386,
      "epoch": 2.1348463440480394,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045429612343683354,
      "loss": 5.2435,
      "mean_token_accuracy": 0.2122351974248886,
      "num_tokens": 45069557.0,
      "step": 24175
    },
    {
      "entropy": 5.893993139266968,
      "epoch": 2.13528788413988,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045427709613068317,
      "loss": 5.3464,
      "mean_token_accuracy": 0.20821600705385207,
      "num_tokens": 45079281.0,
      "step": 24180
    },
    {
      "entropy": 5.951321220397949,
      "epoch": 2.1357294242317204,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004542580653125965,
      "loss": 5.3411,
      "mean_token_accuracy": 0.20824411809444426,
      "num_tokens": 45089363.0,
      "step": 24185
    },
    {
      "entropy": 5.839923429489136,
      "epoch": 2.1361709643235605,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045423903098294624,
      "loss": 5.2353,
      "mean_token_accuracy": 0.2098870813846588,
      "num_tokens": 45098735.0,
      "step": 24190
    },
    {
      "entropy": 5.921649742126465,
      "epoch": 2.136612504415401,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045421999314210535,
      "loss": 5.4091,
      "mean_token_accuracy": 0.20201925039291382,
      "num_tokens": 45109173.0,
      "step": 24195
    },
    {
      "entropy": 5.909635829925537,
      "epoch": 2.1370540445072415,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004542009517904466,
      "loss": 5.4144,
      "mean_token_accuracy": 0.1985637828707695,
      "num_tokens": 45119153.0,
      "step": 24200
    },
    {
      "entropy": 5.8969038963317875,
      "epoch": 2.1374955845990815,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004541819069283432,
      "loss": 5.2721,
      "mean_token_accuracy": 0.21434079706668854,
      "num_tokens": 45128219.0,
      "step": 24205
    },
    {
      "entropy": 5.911312007904053,
      "epoch": 2.137937124690922,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004541628585561682,
      "loss": 5.3395,
      "mean_token_accuracy": 0.206658037006855,
      "num_tokens": 45137777.0,
      "step": 24210
    },
    {
      "entropy": 5.834524297714234,
      "epoch": 2.1383786647827625,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004541438066742946,
      "loss": 5.2711,
      "mean_token_accuracy": 0.214461150765419,
      "num_tokens": 45147553.0,
      "step": 24215
    },
    {
      "entropy": 5.907449769973755,
      "epoch": 2.1388202048746026,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004541247512830958,
      "loss": 5.3231,
      "mean_token_accuracy": 0.20177627354860306,
      "num_tokens": 45157443.0,
      "step": 24220
    },
    {
      "entropy": 5.983698511123658,
      "epoch": 2.139261744966443,
      "grad_norm": 1.0,
      "learning_rate": 0.000454105692382945,
      "loss": 5.3731,
      "mean_token_accuracy": 0.2040179491043091,
      "num_tokens": 45167279.0,
      "step": 24225
    },
    {
      "entropy": 5.987958002090454,
      "epoch": 2.139703285058283,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045408662997421546,
      "loss": 5.3972,
      "mean_token_accuracy": 0.20381687134504317,
      "num_tokens": 45176515.0,
      "step": 24230
    },
    {
      "entropy": 5.981200933456421,
      "epoch": 2.1401448251501236,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004540675640572807,
      "loss": 5.3407,
      "mean_token_accuracy": 0.20446030497550965,
      "num_tokens": 45185501.0,
      "step": 24235
    },
    {
      "entropy": 5.940980339050293,
      "epoch": 2.140586365241964,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004540484946325143,
      "loss": 5.3981,
      "mean_token_accuracy": 0.2035575717687607,
      "num_tokens": 45194881.0,
      "step": 24240
    },
    {
      "entropy": 5.922618722915649,
      "epoch": 2.141027905333804,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045402942170028965,
      "loss": 5.3016,
      "mean_token_accuracy": 0.21147134751081467,
      "num_tokens": 45204463.0,
      "step": 24245
    },
    {
      "entropy": 5.915569210052491,
      "epoch": 2.1414694454256447,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045401034526098053,
      "loss": 5.3208,
      "mean_token_accuracy": 0.20379938781261445,
      "num_tokens": 45214302.0,
      "step": 24250
    },
    {
      "entropy": 5.920975065231323,
      "epoch": 2.141910985517485,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004539912653149604,
      "loss": 5.3472,
      "mean_token_accuracy": 0.20329059958457946,
      "num_tokens": 45224565.0,
      "step": 24255
    },
    {
      "entropy": 5.831735324859619,
      "epoch": 2.142352525609325,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004539721818626033,
      "loss": 5.2301,
      "mean_token_accuracy": 0.22341244518756867,
      "num_tokens": 45234574.0,
      "step": 24260
    },
    {
      "entropy": 5.797258806228638,
      "epoch": 2.1427940657011657,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004539530949042829,
      "loss": 5.213,
      "mean_token_accuracy": 0.21867974549531938,
      "num_tokens": 45242707.0,
      "step": 24265
    },
    {
      "entropy": 5.9111264705657955,
      "epoch": 2.143235605793006,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004539340044403731,
      "loss": 5.4577,
      "mean_token_accuracy": 0.19683025628328324,
      "num_tokens": 45252873.0,
      "step": 24270
    },
    {
      "entropy": 5.904419612884522,
      "epoch": 2.1436771458848463,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000453914910471248,
      "loss": 5.3181,
      "mean_token_accuracy": 0.20716434717178345,
      "num_tokens": 45262833.0,
      "step": 24275
    },
    {
      "entropy": 5.88218412399292,
      "epoch": 2.1441186859766868,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004538958129972815,
      "loss": 5.2941,
      "mean_token_accuracy": 0.21218712031841278,
      "num_tokens": 45271798.0,
      "step": 24280
    },
    {
      "entropy": 5.882066345214843,
      "epoch": 2.144560226068527,
      "grad_norm": 1.0,
      "learning_rate": 0.0004538767120188479,
      "loss": 5.2506,
      "mean_token_accuracy": 0.21435187011957169,
      "num_tokens": 45281470.0,
      "step": 24285
    },
    {
      "entropy": 5.934221267700195,
      "epoch": 2.1450017661603673,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004538576075363211,
      "loss": 5.2922,
      "mean_token_accuracy": 0.2095536097884178,
      "num_tokens": 45290685.0,
      "step": 24290
    },
    {
      "entropy": 5.903287601470947,
      "epoch": 2.145443306252208,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045383849955007554,
      "loss": 5.2919,
      "mean_token_accuracy": 0.20858779102563857,
      "num_tokens": 45299631.0,
      "step": 24295
    },
    {
      "entropy": 5.934343481063843,
      "epoch": 2.145884846344048,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045381938806048544,
      "loss": 5.3976,
      "mean_token_accuracy": 0.20383385866880416,
      "num_tokens": 45309871.0,
      "step": 24300
    },
    {
      "entropy": 5.84241509437561,
      "epoch": 2.1463263864358884,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004538002730679253,
      "loss": 5.1652,
      "mean_token_accuracy": 0.22025881409645082,
      "num_tokens": 45319233.0,
      "step": 24305
    },
    {
      "entropy": 5.813080883026123,
      "epoch": 2.146767926527729,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00045378115457276946,
      "loss": 5.2671,
      "mean_token_accuracy": 0.21412792056798935,
      "num_tokens": 45329467.0,
      "step": 24310
    },
    {
      "entropy": 5.916565704345703,
      "epoch": 2.147209466619569,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045376203257539244,
      "loss": 5.2989,
      "mean_token_accuracy": 0.2064487800002098,
      "num_tokens": 45339124.0,
      "step": 24315
    },
    {
      "entropy": 5.94030179977417,
      "epoch": 2.1476510067114094,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045374290707616893,
      "loss": 5.4022,
      "mean_token_accuracy": 0.19933412969112396,
      "num_tokens": 45348973.0,
      "step": 24320
    },
    {
      "entropy": 5.962058353424072,
      "epoch": 2.14809254680325,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004537237780754735,
      "loss": 5.3328,
      "mean_token_accuracy": 0.20377595573663712,
      "num_tokens": 45359161.0,
      "step": 24325
    },
    {
      "entropy": 5.935665798187256,
      "epoch": 2.14853408689509,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004537046455736809,
      "loss": 5.329,
      "mean_token_accuracy": 0.21389723867177962,
      "num_tokens": 45369019.0,
      "step": 24330
    },
    {
      "entropy": 5.875147485733033,
      "epoch": 2.1489756269869305,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004536855095711659,
      "loss": 5.2599,
      "mean_token_accuracy": 0.21821779012680054,
      "num_tokens": 45378491.0,
      "step": 24335
    },
    {
      "entropy": 5.8914107322692875,
      "epoch": 2.149417167078771,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004536663700683035,
      "loss": 5.3364,
      "mean_token_accuracy": 0.20960796773433685,
      "num_tokens": 45387651.0,
      "step": 24340
    },
    {
      "entropy": 5.941328382492065,
      "epoch": 2.149858707170611,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004536472270654685,
      "loss": 5.2897,
      "mean_token_accuracy": 0.21224784553050996,
      "num_tokens": 45396524.0,
      "step": 24345
    },
    {
      "entropy": 5.9586982250213625,
      "epoch": 2.1503002472624515,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004536280805630358,
      "loss": 5.3598,
      "mean_token_accuracy": 0.209202179312706,
      "num_tokens": 45404849.0,
      "step": 24350
    },
    {
      "entropy": 5.910099649429322,
      "epoch": 2.150741787354292,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045360893056138076,
      "loss": 5.2743,
      "mean_token_accuracy": 0.21451979726552964,
      "num_tokens": 45413628.0,
      "step": 24355
    },
    {
      "entropy": 5.88285117149353,
      "epoch": 2.151183327446132,
      "grad_norm": 1.125,
      "learning_rate": 0.00045358977706087827,
      "loss": 5.2805,
      "mean_token_accuracy": 0.2138541266322136,
      "num_tokens": 45423190.0,
      "step": 24360
    },
    {
      "entropy": 5.816536855697632,
      "epoch": 2.1516248675379726,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004535706200619036,
      "loss": 5.2784,
      "mean_token_accuracy": 0.213318532705307,
      "num_tokens": 45432075.0,
      "step": 24365
    },
    {
      "entropy": 5.905173301696777,
      "epoch": 2.1520664076298126,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045355145956483205,
      "loss": 5.3989,
      "mean_token_accuracy": 0.2014447495341301,
      "num_tokens": 45442627.0,
      "step": 24370
    },
    {
      "entropy": 5.8832025051116945,
      "epoch": 2.152507947721653,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045353229557003896,
      "loss": 5.2673,
      "mean_token_accuracy": 0.21728233247995377,
      "num_tokens": 45452818.0,
      "step": 24375
    },
    {
      "entropy": 5.8580810546875,
      "epoch": 2.1529494878134936,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004535131280778997,
      "loss": 5.2678,
      "mean_token_accuracy": 0.21723175048828125,
      "num_tokens": 45461964.0,
      "step": 24380
    },
    {
      "entropy": 5.853044319152832,
      "epoch": 2.1533910279053337,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004534939570887898,
      "loss": 5.2763,
      "mean_token_accuracy": 0.21086964011192322,
      "num_tokens": 45471700.0,
      "step": 24385
    },
    {
      "entropy": 5.911209630966186,
      "epoch": 2.153832567997174,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004534747826030848,
      "loss": 5.2594,
      "mean_token_accuracy": 0.20904617607593537,
      "num_tokens": 45481291.0,
      "step": 24390
    },
    {
      "entropy": 5.8785828113555905,
      "epoch": 2.1542741080890146,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004534556046211604,
      "loss": 5.3133,
      "mean_token_accuracy": 0.21149694919586182,
      "num_tokens": 45490317.0,
      "step": 24395
    },
    {
      "entropy": 5.9012034893035885,
      "epoch": 2.1547156481808547,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045343642314339206,
      "loss": 5.3569,
      "mean_token_accuracy": 0.20782990753650665,
      "num_tokens": 45499688.0,
      "step": 24400
    },
    {
      "entropy": 5.9482227802276615,
      "epoch": 2.155157188272695,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004534172381701558,
      "loss": 5.3079,
      "mean_token_accuracy": 0.2127251073718071,
      "num_tokens": 45508951.0,
      "step": 24405
    },
    {
      "entropy": 5.969006061553955,
      "epoch": 2.1555987283645357,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004533980497018272,
      "loss": 5.3491,
      "mean_token_accuracy": 0.21004294008016586,
      "num_tokens": 45518955.0,
      "step": 24410
    },
    {
      "entropy": 5.871236658096313,
      "epoch": 2.1560402684563758,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004533788577387824,
      "loss": 5.3287,
      "mean_token_accuracy": 0.2022874802350998,
      "num_tokens": 45527869.0,
      "step": 24415
    },
    {
      "entropy": 5.859797620773316,
      "epoch": 2.1564818085482163,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004533596622813971,
      "loss": 5.2316,
      "mean_token_accuracy": 0.22090501338243484,
      "num_tokens": 45537250.0,
      "step": 24420
    },
    {
      "entropy": 5.885863399505615,
      "epoch": 2.1569233486400563,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004533404633300475,
      "loss": 5.2608,
      "mean_token_accuracy": 0.2091406211256981,
      "num_tokens": 45547204.0,
      "step": 24425
    },
    {
      "entropy": 5.851839160919189,
      "epoch": 2.157364888731897,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045332126088510965,
      "loss": 5.27,
      "mean_token_accuracy": 0.20369155555963517,
      "num_tokens": 45555464.0,
      "step": 24430
    },
    {
      "entropy": 5.9216224193573,
      "epoch": 2.1578064288237373,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004533020549469597,
      "loss": 5.2419,
      "mean_token_accuracy": 0.21041091084480285,
      "num_tokens": 45564797.0,
      "step": 24435
    },
    {
      "entropy": 5.9219653606414795,
      "epoch": 2.1582479689155774,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004532828455159739,
      "loss": 5.3354,
      "mean_token_accuracy": 0.20181020647287368,
      "num_tokens": 45573317.0,
      "step": 24440
    },
    {
      "entropy": 5.792671012878418,
      "epoch": 2.158689509007418,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004532636325925286,
      "loss": 5.2378,
      "mean_token_accuracy": 0.21440215408802032,
      "num_tokens": 45582838.0,
      "step": 24445
    },
    {
      "entropy": 5.901033687591553,
      "epoch": 2.1591310490992583,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045324441617700014,
      "loss": 5.1783,
      "mean_token_accuracy": 0.2169872060418129,
      "num_tokens": 45592031.0,
      "step": 24450
    },
    {
      "entropy": 5.918382692337036,
      "epoch": 2.1595725891910984,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004532251962697649,
      "loss": 5.3899,
      "mean_token_accuracy": 0.20761216729879378,
      "num_tokens": 45601353.0,
      "step": 24455
    },
    {
      "entropy": 5.858606624603271,
      "epoch": 2.160014129282939,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045320597287119943,
      "loss": 5.2656,
      "mean_token_accuracy": 0.20914039611816407,
      "num_tokens": 45611064.0,
      "step": 24460
    },
    {
      "entropy": 5.893822145462036,
      "epoch": 2.1604556693747794,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004531867459816803,
      "loss": 5.2934,
      "mean_token_accuracy": 0.20908606946468353,
      "num_tokens": 45620704.0,
      "step": 24465
    },
    {
      "entropy": 5.9575080394744875,
      "epoch": 2.1608972094666195,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045316751560158414,
      "loss": 5.3544,
      "mean_token_accuracy": 0.20919648110866546,
      "num_tokens": 45629967.0,
      "step": 24470
    },
    {
      "entropy": 5.922767400741577,
      "epoch": 2.16133874955846,
      "grad_norm": 1.25,
      "learning_rate": 0.0004531482817312877,
      "loss": 5.3619,
      "mean_token_accuracy": 0.19891661554574966,
      "num_tokens": 45639104.0,
      "step": 24475
    },
    {
      "entropy": 5.93939094543457,
      "epoch": 2.1617802896503004,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045312904437116774,
      "loss": 5.4344,
      "mean_token_accuracy": 0.19522229135036467,
      "num_tokens": 45649217.0,
      "step": 24480
    },
    {
      "entropy": 5.978772830963135,
      "epoch": 2.1622218297421405,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004531098035216011,
      "loss": 5.3878,
      "mean_token_accuracy": 0.20174179822206498,
      "num_tokens": 45658969.0,
      "step": 24485
    },
    {
      "entropy": 5.801855945587159,
      "epoch": 2.162663369833981,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045309055918296465,
      "loss": 5.2146,
      "mean_token_accuracy": 0.21612147092819214,
      "num_tokens": 45668427.0,
      "step": 24490
    },
    {
      "entropy": 5.809905576705932,
      "epoch": 2.1631049099258215,
      "grad_norm": 1.0,
      "learning_rate": 0.0004530713113556355,
      "loss": 5.1604,
      "mean_token_accuracy": 0.21866879016160964,
      "num_tokens": 45676675.0,
      "step": 24495
    },
    {
      "entropy": 5.827397584915161,
      "epoch": 2.1635464500176615,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045305206003999057,
      "loss": 5.1987,
      "mean_token_accuracy": 0.21538830250501634,
      "num_tokens": 45685756.0,
      "step": 24500
    },
    {
      "entropy": 5.885875606536866,
      "epoch": 2.163987990109502,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004530328052364071,
      "loss": 5.3073,
      "mean_token_accuracy": 0.21500690877437592,
      "num_tokens": 45694770.0,
      "step": 24505
    },
    {
      "entropy": 5.865076971054077,
      "epoch": 2.164429530201342,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004530135469452622,
      "loss": 5.2033,
      "mean_token_accuracy": 0.21871818006038665,
      "num_tokens": 45703940.0,
      "step": 24510
    },
    {
      "entropy": 5.887880182266235,
      "epoch": 2.1648710702931826,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004529942851669333,
      "loss": 5.3443,
      "mean_token_accuracy": 0.20593252927064895,
      "num_tokens": 45713875.0,
      "step": 24515
    },
    {
      "entropy": 5.930170011520386,
      "epoch": 2.165312610385023,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004529750199017974,
      "loss": 5.3522,
      "mean_token_accuracy": 0.1997353181242943,
      "num_tokens": 45723870.0,
      "step": 24520
    },
    {
      "entropy": 5.891964721679687,
      "epoch": 2.165754150476863,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004529557511502322,
      "loss": 5.2898,
      "mean_token_accuracy": 0.2156054347753525,
      "num_tokens": 45733269.0,
      "step": 24525
    },
    {
      "entropy": 5.912575960159302,
      "epoch": 2.1661956905687036,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045293647891261494,
      "loss": 5.3112,
      "mean_token_accuracy": 0.20392105281352996,
      "num_tokens": 45743208.0,
      "step": 24530
    },
    {
      "entropy": 5.927594232559204,
      "epoch": 2.166637230660544,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004529172031893233,
      "loss": 5.2794,
      "mean_token_accuracy": 0.21576096266508102,
      "num_tokens": 45752411.0,
      "step": 24535
    },
    {
      "entropy": 5.822519731521607,
      "epoch": 2.167078770752384,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045289792398073487,
      "loss": 5.2951,
      "mean_token_accuracy": 0.21168990880250932,
      "num_tokens": 45761770.0,
      "step": 24540
    },
    {
      "entropy": 5.892732286453247,
      "epoch": 2.1675203108442247,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045287864128722724,
      "loss": 5.266,
      "mean_token_accuracy": 0.2090234711766243,
      "num_tokens": 45771227.0,
      "step": 24545
    },
    {
      "entropy": 5.861612939834595,
      "epoch": 2.167961850936065,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004528593551091782,
      "loss": 5.2215,
      "mean_token_accuracy": 0.21584538072347642,
      "num_tokens": 45780558.0,
      "step": 24550
    },
    {
      "entropy": 5.8548243045806885,
      "epoch": 2.1684033910279052,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045284006544696554,
      "loss": 5.3329,
      "mean_token_accuracy": 0.20803416520357132,
      "num_tokens": 45790078.0,
      "step": 24555
    },
    {
      "entropy": 5.882325220108032,
      "epoch": 2.1688449311197457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045282077230096715,
      "loss": 5.2546,
      "mean_token_accuracy": 0.21524593383073806,
      "num_tokens": 45799962.0,
      "step": 24560
    },
    {
      "entropy": 5.817895078659058,
      "epoch": 2.169286471211586,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045280147567156097,
      "loss": 5.1613,
      "mean_token_accuracy": 0.2302367866039276,
      "num_tokens": 45809599.0,
      "step": 24565
    },
    {
      "entropy": 5.8085283756256105,
      "epoch": 2.1697280113034263,
      "grad_norm": 1.109375,
      "learning_rate": 0.000452782175559125,
      "loss": 5.2447,
      "mean_token_accuracy": 0.22017981112003326,
      "num_tokens": 45818318.0,
      "step": 24570
    },
    {
      "entropy": 5.892239046096802,
      "epoch": 2.170169551395267,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004527628719640374,
      "loss": 5.3492,
      "mean_token_accuracy": 0.20504026412963866,
      "num_tokens": 45828083.0,
      "step": 24575
    },
    {
      "entropy": 5.935414838790893,
      "epoch": 2.170611091487107,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045274356488667614,
      "loss": 5.3017,
      "mean_token_accuracy": 0.21164270639419555,
      "num_tokens": 45838441.0,
      "step": 24580
    },
    {
      "entropy": 5.934444761276245,
      "epoch": 2.1710526315789473,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045272425432741953,
      "loss": 5.3014,
      "mean_token_accuracy": 0.20749692767858505,
      "num_tokens": 45848047.0,
      "step": 24585
    },
    {
      "entropy": 5.8710541248321535,
      "epoch": 2.171494171670788,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045270494028664586,
      "loss": 5.3091,
      "mean_token_accuracy": 0.20810611099004744,
      "num_tokens": 45857675.0,
      "step": 24590
    },
    {
      "entropy": 5.8301526546478275,
      "epoch": 2.171935711762628,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004526856227647335,
      "loss": 5.2426,
      "mean_token_accuracy": 0.21092915683984756,
      "num_tokens": 45866506.0,
      "step": 24595
    },
    {
      "entropy": 5.9587842464447025,
      "epoch": 2.1723772518544684,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004526663017620609,
      "loss": 5.3155,
      "mean_token_accuracy": 0.2150665834546089,
      "num_tokens": 45875955.0,
      "step": 24600
    },
    {
      "entropy": 5.936282682418823,
      "epoch": 2.172818791946309,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004526469772790063,
      "loss": 5.2992,
      "mean_token_accuracy": 0.2054099589586258,
      "num_tokens": 45884518.0,
      "step": 24605
    },
    {
      "entropy": 5.855802059173584,
      "epoch": 2.173260332038149,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004526276493159487,
      "loss": 5.2583,
      "mean_token_accuracy": 0.2131984516978264,
      "num_tokens": 45894006.0,
      "step": 24610
    },
    {
      "entropy": 5.805609273910522,
      "epoch": 2.1737018721299894,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004526083178732663,
      "loss": 5.2151,
      "mean_token_accuracy": 0.21467824131250382,
      "num_tokens": 45902739.0,
      "step": 24615
    },
    {
      "entropy": 5.7868249893188475,
      "epoch": 2.17414341222183,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045258898295133797,
      "loss": 5.1477,
      "mean_token_accuracy": 0.22033517062664032,
      "num_tokens": 45911418.0,
      "step": 24620
    },
    {
      "entropy": 5.86443076133728,
      "epoch": 2.17458495231367,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045256964455054255,
      "loss": 5.3076,
      "mean_token_accuracy": 0.20846038460731506,
      "num_tokens": 45921135.0,
      "step": 24625
    },
    {
      "entropy": 5.882087326049804,
      "epoch": 2.1750264924055105,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045255030267125874,
      "loss": 5.2392,
      "mean_token_accuracy": 0.21579915434122085,
      "num_tokens": 45930276.0,
      "step": 24630
    },
    {
      "entropy": 5.908164930343628,
      "epoch": 2.175468032497351,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045253095731386553,
      "loss": 5.2635,
      "mean_token_accuracy": 0.20852722823619843,
      "num_tokens": 45939466.0,
      "step": 24635
    },
    {
      "entropy": 5.874197101593017,
      "epoch": 2.175909572589191,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045251160847874174,
      "loss": 5.2183,
      "mean_token_accuracy": 0.21040224730968476,
      "num_tokens": 45948158.0,
      "step": 24640
    },
    {
      "entropy": 5.911040735244751,
      "epoch": 2.1763511126810315,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045249225616626656,
      "loss": 5.4035,
      "mean_token_accuracy": 0.19876639544963837,
      "num_tokens": 45957759.0,
      "step": 24645
    },
    {
      "entropy": 5.907107543945313,
      "epoch": 2.1767926527728716,
      "grad_norm": 1.09375,
      "learning_rate": 0.000452472900376819,
      "loss": 5.2749,
      "mean_token_accuracy": 0.20810548663139344,
      "num_tokens": 45966689.0,
      "step": 24650
    },
    {
      "entropy": 5.90785174369812,
      "epoch": 2.177234192864712,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045245354111077827,
      "loss": 5.3534,
      "mean_token_accuracy": 0.20499788969755173,
      "num_tokens": 45975618.0,
      "step": 24655
    },
    {
      "entropy": 5.877581405639648,
      "epoch": 2.1776757329565526,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045243417836852355,
      "loss": 5.2893,
      "mean_token_accuracy": 0.21044217497110368,
      "num_tokens": 45984500.0,
      "step": 24660
    },
    {
      "entropy": 5.919665718078614,
      "epoch": 2.1781172730483926,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004524148121504343,
      "loss": 5.2821,
      "mean_token_accuracy": 0.20708625465631486,
      "num_tokens": 45994471.0,
      "step": 24665
    },
    {
      "entropy": 6.013688278198242,
      "epoch": 2.178558813140233,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045239544245688967,
      "loss": 5.3607,
      "mean_token_accuracy": 0.20263609439134597,
      "num_tokens": 46003757.0,
      "step": 24670
    },
    {
      "entropy": 5.843350505828857,
      "epoch": 2.1790003532320736,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004523760692882693,
      "loss": 5.207,
      "mean_token_accuracy": 0.22185277789831162,
      "num_tokens": 46013512.0,
      "step": 24675
    },
    {
      "entropy": 5.8821147918701175,
      "epoch": 2.1794418933239137,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004523566926449526,
      "loss": 5.2902,
      "mean_token_accuracy": 0.21643492728471755,
      "num_tokens": 46022867.0,
      "step": 24680
    },
    {
      "entropy": 5.897031164169311,
      "epoch": 2.179883433415754,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045233731252731905,
      "loss": 5.3253,
      "mean_token_accuracy": 0.2071937397122383,
      "num_tokens": 46031799.0,
      "step": 24685
    },
    {
      "entropy": 5.886275148391723,
      "epoch": 2.1803249735075947,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045231792893574846,
      "loss": 5.3355,
      "mean_token_accuracy": 0.2045131057500839,
      "num_tokens": 46041917.0,
      "step": 24690
    },
    {
      "entropy": 5.919997787475586,
      "epoch": 2.1807665135994347,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045229854187062054,
      "loss": 5.3231,
      "mean_token_accuracy": 0.21723242104053497,
      "num_tokens": 46051629.0,
      "step": 24695
    },
    {
      "entropy": 5.8439795017242435,
      "epoch": 2.1812080536912752,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004522791513323149,
      "loss": 5.217,
      "mean_token_accuracy": 0.2257873997092247,
      "num_tokens": 46060840.0,
      "step": 24700
    },
    {
      "entropy": 5.8255692481994625,
      "epoch": 2.1816495937831153,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045225975732121165,
      "loss": 5.2464,
      "mean_token_accuracy": 0.21378356367349624,
      "num_tokens": 46069774.0,
      "step": 24705
    },
    {
      "entropy": 5.941590452194214,
      "epoch": 2.182091133874956,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045224035983769054,
      "loss": 5.3464,
      "mean_token_accuracy": 0.20414550006389617,
      "num_tokens": 46079377.0,
      "step": 24710
    },
    {
      "entropy": 5.8948118686676025,
      "epoch": 2.1825326739667963,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045222095888213155,
      "loss": 5.2737,
      "mean_token_accuracy": 0.21850760281085968,
      "num_tokens": 46088190.0,
      "step": 24715
    },
    {
      "entropy": 5.88330717086792,
      "epoch": 2.1829742140586363,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045220155445491465,
      "loss": 5.3242,
      "mean_token_accuracy": 0.2182814046740532,
      "num_tokens": 46097487.0,
      "step": 24720
    },
    {
      "entropy": 5.988511943817139,
      "epoch": 2.183415754150477,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004521821465564203,
      "loss": 5.3989,
      "mean_token_accuracy": 0.20368327498435973,
      "num_tokens": 46106554.0,
      "step": 24725
    },
    {
      "entropy": 5.910604333877563,
      "epoch": 2.1838572942423173,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004521627351870283,
      "loss": 5.3665,
      "mean_token_accuracy": 0.19943333566188812,
      "num_tokens": 46115665.0,
      "step": 24730
    },
    {
      "entropy": 5.875160169601441,
      "epoch": 2.1842988343341574,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004521433203471192,
      "loss": 5.2427,
      "mean_token_accuracy": 0.21412675380706786,
      "num_tokens": 46124863.0,
      "step": 24735
    },
    {
      "entropy": 5.914878892898559,
      "epoch": 2.184740374425998,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045212390203707315,
      "loss": 5.2673,
      "mean_token_accuracy": 0.20714907199144364,
      "num_tokens": 46133986.0,
      "step": 24740
    },
    {
      "entropy": 5.867431211471557,
      "epoch": 2.1851819145178384,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004521044802572705,
      "loss": 5.2326,
      "mean_token_accuracy": 0.2173826053738594,
      "num_tokens": 46142234.0,
      "step": 24745
    },
    {
      "entropy": 5.871627235412598,
      "epoch": 2.1856234546096784,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045208505500809196,
      "loss": 5.3164,
      "mean_token_accuracy": 0.21288973242044448,
      "num_tokens": 46150976.0,
      "step": 24750
    },
    {
      "entropy": 5.9165575981140135,
      "epoch": 2.186064994701519,
      "grad_norm": 1.125,
      "learning_rate": 0.00045206562628991785,
      "loss": 5.3212,
      "mean_token_accuracy": 0.20821150541305541,
      "num_tokens": 46160541.0,
      "step": 24755
    },
    {
      "entropy": 5.983079338073731,
      "epoch": 2.1865065347933594,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045204619410312875,
      "loss": 5.3598,
      "mean_token_accuracy": 0.2055536165833473,
      "num_tokens": 46170958.0,
      "step": 24760
    },
    {
      "entropy": 5.911781120300293,
      "epoch": 2.1869480748851995,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004520267584481055,
      "loss": 5.3002,
      "mean_token_accuracy": 0.21306820958852768,
      "num_tokens": 46180293.0,
      "step": 24765
    },
    {
      "entropy": 5.854779386520386,
      "epoch": 2.18738961497704,
      "grad_norm": 1.125,
      "learning_rate": 0.00045200731932522873,
      "loss": 5.2644,
      "mean_token_accuracy": 0.21739444881677628,
      "num_tokens": 46189264.0,
      "step": 24770
    },
    {
      "entropy": 5.958702898025512,
      "epoch": 2.1878311550688805,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004519878767348792,
      "loss": 5.3999,
      "mean_token_accuracy": 0.20089254528284073,
      "num_tokens": 46199417.0,
      "step": 24775
    },
    {
      "entropy": 5.87711820602417,
      "epoch": 2.1882726951607205,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004519684306774378,
      "loss": 5.2395,
      "mean_token_accuracy": 0.21633054614067077,
      "num_tokens": 46209360.0,
      "step": 24780
    },
    {
      "entropy": 5.8455109119415285,
      "epoch": 2.188714235252561,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045194898115328555,
      "loss": 5.2822,
      "mean_token_accuracy": 0.20362259894609452,
      "num_tokens": 46218414.0,
      "step": 24785
    },
    {
      "entropy": 5.867178297042846,
      "epoch": 2.189155775344401,
      "grad_norm": 1.125,
      "learning_rate": 0.0004519295281628033,
      "loss": 5.2841,
      "mean_token_accuracy": 0.20989346206188203,
      "num_tokens": 46227526.0,
      "step": 24790
    },
    {
      "entropy": 5.910853910446167,
      "epoch": 2.1895973154362416,
      "grad_norm": 1.125,
      "learning_rate": 0.0004519100717063723,
      "loss": 5.2326,
      "mean_token_accuracy": 0.2142106845974922,
      "num_tokens": 46236041.0,
      "step": 24795
    },
    {
      "entropy": 5.8417548656463625,
      "epoch": 2.190038855528082,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045189061178437363,
      "loss": 5.2359,
      "mean_token_accuracy": 0.2107262760400772,
      "num_tokens": 46244557.0,
      "step": 24800
    },
    {
      "entropy": 5.863319206237793,
      "epoch": 2.190480395619922,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004518711483971885,
      "loss": 5.2374,
      "mean_token_accuracy": 0.21817210614681243,
      "num_tokens": 46253957.0,
      "step": 24805
    },
    {
      "entropy": 5.948736715316772,
      "epoch": 2.1909219357117626,
      "grad_norm": 1.203125,
      "learning_rate": 0.000451851681545198,
      "loss": 5.3347,
      "mean_token_accuracy": 0.20921835005283357,
      "num_tokens": 46263110.0,
      "step": 24810
    },
    {
      "entropy": 5.872492361068725,
      "epoch": 2.191363475803603,
      "grad_norm": 1.125,
      "learning_rate": 0.00045183221122878375,
      "loss": 5.3469,
      "mean_token_accuracy": 0.20458832383155823,
      "num_tokens": 46272163.0,
      "step": 24815
    },
    {
      "entropy": 5.835750770568848,
      "epoch": 2.191805015895443,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045181273744832707,
      "loss": 5.3153,
      "mean_token_accuracy": 0.2052254855632782,
      "num_tokens": 46281097.0,
      "step": 24820
    },
    {
      "entropy": 5.933538150787354,
      "epoch": 2.1922465559872837,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004517932602042094,
      "loss": 5.2769,
      "mean_token_accuracy": 0.20738780945539476,
      "num_tokens": 46291552.0,
      "step": 24825
    },
    {
      "entropy": 5.8951469421386715,
      "epoch": 2.192688096079124,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045177377949681235,
      "loss": 5.3115,
      "mean_token_accuracy": 0.21112328469753266,
      "num_tokens": 46300661.0,
      "step": 24830
    },
    {
      "entropy": 5.8536769390106205,
      "epoch": 2.1931296361709642,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045175429532651743,
      "loss": 5.2531,
      "mean_token_accuracy": 0.21318843215703964,
      "num_tokens": 46309705.0,
      "step": 24835
    },
    {
      "entropy": 5.842858219146729,
      "epoch": 2.1935711762628047,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004517348076937064,
      "loss": 5.2745,
      "mean_token_accuracy": 0.21584783345460892,
      "num_tokens": 46319701.0,
      "step": 24840
    },
    {
      "entropy": 5.804854965209961,
      "epoch": 2.194012716354645,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045171531659876104,
      "loss": 5.2619,
      "mean_token_accuracy": 0.21641617715358735,
      "num_tokens": 46328195.0,
      "step": 24845
    },
    {
      "entropy": 5.920786809921265,
      "epoch": 2.1944542564464853,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004516958220420631,
      "loss": 5.3088,
      "mean_token_accuracy": 0.21326993256807328,
      "num_tokens": 46337518.0,
      "step": 24850
    },
    {
      "entropy": 5.860372734069824,
      "epoch": 2.194895796538326,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004516763240239945,
      "loss": 5.235,
      "mean_token_accuracy": 0.2159891277551651,
      "num_tokens": 46346406.0,
      "step": 24855
    },
    {
      "entropy": 5.843536233901977,
      "epoch": 2.195337336630166,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004516568225449372,
      "loss": 5.2314,
      "mean_token_accuracy": 0.21313668191432952,
      "num_tokens": 46355284.0,
      "step": 24860
    },
    {
      "entropy": 5.902785110473633,
      "epoch": 2.1957788767220063,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045163731760527326,
      "loss": 5.3466,
      "mean_token_accuracy": 0.2091919496655464,
      "num_tokens": 46364982.0,
      "step": 24865
    },
    {
      "entropy": 5.814175987243653,
      "epoch": 2.196220416813847,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004516178092053847,
      "loss": 5.1796,
      "mean_token_accuracy": 0.22120079398155212,
      "num_tokens": 46373923.0,
      "step": 24870
    },
    {
      "entropy": 5.854439735412598,
      "epoch": 2.196661956905687,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004515982973456537,
      "loss": 5.2101,
      "mean_token_accuracy": 0.21815921664237975,
      "num_tokens": 46382194.0,
      "step": 24875
    },
    {
      "entropy": 5.829131126403809,
      "epoch": 2.1971034969975274,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004515787820264624,
      "loss": 5.2868,
      "mean_token_accuracy": 0.20817655622959136,
      "num_tokens": 46391463.0,
      "step": 24880
    },
    {
      "entropy": 5.891905784606934,
      "epoch": 2.197545037089368,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045155926324819335,
      "loss": 5.3097,
      "mean_token_accuracy": 0.21281156688928604,
      "num_tokens": 46401235.0,
      "step": 24885
    },
    {
      "entropy": 5.895998191833496,
      "epoch": 2.197986577181208,
      "grad_norm": 1.125,
      "learning_rate": 0.00045153974101122866,
      "loss": 5.3426,
      "mean_token_accuracy": 0.20992677211761473,
      "num_tokens": 46410631.0,
      "step": 24890
    },
    {
      "entropy": 5.933129453659058,
      "epoch": 2.1984281172730484,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045152021531595085,
      "loss": 5.3483,
      "mean_token_accuracy": 0.20004705488681793,
      "num_tokens": 46420467.0,
      "step": 24895
    },
    {
      "entropy": 5.877530336380005,
      "epoch": 2.198869657364889,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004515006861627424,
      "loss": 5.2475,
      "mean_token_accuracy": 0.21700648814439774,
      "num_tokens": 46430276.0,
      "step": 24900
    },
    {
      "entropy": 5.972809648513794,
      "epoch": 2.199311197456729,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045148115355198596,
      "loss": 5.3574,
      "mean_token_accuracy": 0.21151867508888245,
      "num_tokens": 46439227.0,
      "step": 24905
    },
    {
      "entropy": 5.8662396430969235,
      "epoch": 2.1997527375485695,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000451461617484064,
      "loss": 5.3092,
      "mean_token_accuracy": 0.20609249621629716,
      "num_tokens": 46449043.0,
      "step": 24910
    },
    {
      "entropy": 5.8734776973724365,
      "epoch": 2.20019427764041,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004514420779593594,
      "loss": 5.2975,
      "mean_token_accuracy": 0.20904383957386016,
      "num_tokens": 46458332.0,
      "step": 24915
    },
    {
      "entropy": 5.971198177337646,
      "epoch": 2.20063581773225,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045142253497825477,
      "loss": 5.2397,
      "mean_token_accuracy": 0.21337376832962035,
      "num_tokens": 46467208.0,
      "step": 24920
    },
    {
      "entropy": 5.8712622165679935,
      "epoch": 2.2010773578240905,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045140298854113306,
      "loss": 5.2608,
      "mean_token_accuracy": 0.21790449917316437,
      "num_tokens": 46476715.0,
      "step": 24925
    },
    {
      "entropy": 5.7847902297973635,
      "epoch": 2.2015188979159306,
      "grad_norm": 1.125,
      "learning_rate": 0.00045138343864837703,
      "loss": 5.208,
      "mean_token_accuracy": 0.22131199687719344,
      "num_tokens": 46484783.0,
      "step": 24930
    },
    {
      "entropy": 5.83673677444458,
      "epoch": 2.201960438007771,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004513638853003699,
      "loss": 5.1931,
      "mean_token_accuracy": 0.21337351500988005,
      "num_tokens": 46493616.0,
      "step": 24935
    },
    {
      "entropy": 5.920994091033935,
      "epoch": 2.2024019780996116,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045134432849749446,
      "loss": 5.2948,
      "mean_token_accuracy": 0.21216038912534713,
      "num_tokens": 46502779.0,
      "step": 24940
    },
    {
      "entropy": 5.855118370056152,
      "epoch": 2.2028435181914516,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045132476824013396,
      "loss": 5.2946,
      "mean_token_accuracy": 0.21107901632785797,
      "num_tokens": 46511506.0,
      "step": 24945
    },
    {
      "entropy": 5.926555442810058,
      "epoch": 2.203285058283292,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004513052045286715,
      "loss": 5.3223,
      "mean_token_accuracy": 0.20706935822963715,
      "num_tokens": 46520219.0,
      "step": 24950
    },
    {
      "entropy": 5.947846698760986,
      "epoch": 2.2037265983751326,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004512856373634904,
      "loss": 5.2831,
      "mean_token_accuracy": 0.21450922787189483,
      "num_tokens": 46528737.0,
      "step": 24955
    },
    {
      "entropy": 5.925307989120483,
      "epoch": 2.2041681384669727,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045126606674497394,
      "loss": 5.374,
      "mean_token_accuracy": 0.20536865592002868,
      "num_tokens": 46538802.0,
      "step": 24960
    },
    {
      "entropy": 5.87228307723999,
      "epoch": 2.204609678558813,
      "grad_norm": 1.125,
      "learning_rate": 0.0004512464926735054,
      "loss": 5.2969,
      "mean_token_accuracy": 0.21601083278656005,
      "num_tokens": 46548369.0,
      "step": 24965
    },
    {
      "entropy": 5.873328018188476,
      "epoch": 2.2050512186506537,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045122691514946835,
      "loss": 5.3408,
      "mean_token_accuracy": 0.207840633392334,
      "num_tokens": 46557876.0,
      "step": 24970
    },
    {
      "entropy": 5.839500808715821,
      "epoch": 2.2054927587424937,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004512073341732463,
      "loss": 5.2873,
      "mean_token_accuracy": 0.22232381552457808,
      "num_tokens": 46566747.0,
      "step": 24975
    },
    {
      "entropy": 5.923096609115601,
      "epoch": 2.2059342988343342,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004511877497452227,
      "loss": 5.2837,
      "mean_token_accuracy": 0.21339815557003022,
      "num_tokens": 46576196.0,
      "step": 24980
    },
    {
      "entropy": 5.970819520950317,
      "epoch": 2.2063758389261743,
      "grad_norm": 1.0,
      "learning_rate": 0.0004511681618657813,
      "loss": 5.3767,
      "mean_token_accuracy": 0.20547797977924348,
      "num_tokens": 46585734.0,
      "step": 24985
    },
    {
      "entropy": 5.813763380050659,
      "epoch": 2.206817379018015,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045114857053530586,
      "loss": 5.2274,
      "mean_token_accuracy": 0.21588525772094727,
      "num_tokens": 46595931.0,
      "step": 24990
    },
    {
      "entropy": 5.910294151306152,
      "epoch": 2.2072589191098553,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00045112897575418006,
      "loss": 5.2965,
      "mean_token_accuracy": 0.20620531141757964,
      "num_tokens": 46604688.0,
      "step": 24995
    },
    {
      "entropy": 5.848907613754273,
      "epoch": 2.2077004592016953,
      "grad_norm": 1.125,
      "learning_rate": 0.0004511093775227878,
      "loss": 5.2798,
      "mean_token_accuracy": 0.21882594972848893,
      "num_tokens": 46613860.0,
      "step": 25000
    },
    {
      "entropy": 5.929471445083618,
      "epoch": 2.208141999293536,
      "grad_norm": 1.0625,
      "learning_rate": 0.000451089775841513,
      "loss": 5.3365,
      "mean_token_accuracy": 0.20733083933591842,
      "num_tokens": 46623041.0,
      "step": 25005
    },
    {
      "entropy": 5.898074197769165,
      "epoch": 2.2085835393853763,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045107017071073963,
      "loss": 5.3134,
      "mean_token_accuracy": 0.20037408620119096,
      "num_tokens": 46633025.0,
      "step": 25010
    },
    {
      "entropy": 5.954243850708008,
      "epoch": 2.2090250794772164,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004510505621308517,
      "loss": 5.3108,
      "mean_token_accuracy": 0.20995305925607682,
      "num_tokens": 46641826.0,
      "step": 25015
    },
    {
      "entropy": 5.968498039245605,
      "epoch": 2.209466619569057,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045103095010223354,
      "loss": 5.395,
      "mean_token_accuracy": 0.20329425781965255,
      "num_tokens": 46652028.0,
      "step": 25020
    },
    {
      "entropy": 5.86537013053894,
      "epoch": 2.2099081596608974,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045101133462526903,
      "loss": 5.3911,
      "mean_token_accuracy": 0.20344689041376113,
      "num_tokens": 46660763.0,
      "step": 25025
    },
    {
      "entropy": 5.909679746627807,
      "epoch": 2.2103496997527374,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004509917157003427,
      "loss": 5.3289,
      "mean_token_accuracy": 0.20585323572158815,
      "num_tokens": 46670213.0,
      "step": 25030
    },
    {
      "entropy": 5.984507894515991,
      "epoch": 2.210791239844578,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004509720933278386,
      "loss": 5.3113,
      "mean_token_accuracy": 0.21056326180696489,
      "num_tokens": 46679658.0,
      "step": 25035
    },
    {
      "entropy": 5.881831407546997,
      "epoch": 2.2112327799364184,
      "grad_norm": 1.125,
      "learning_rate": 0.0004509524675081414,
      "loss": 5.2593,
      "mean_token_accuracy": 0.21351644545793533,
      "num_tokens": 46687920.0,
      "step": 25040
    },
    {
      "entropy": 5.897279882431031,
      "epoch": 2.2116743200282585,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004509328382416354,
      "loss": 5.3571,
      "mean_token_accuracy": 0.20236642360687257,
      "num_tokens": 46696779.0,
      "step": 25045
    },
    {
      "entropy": 5.918566751480102,
      "epoch": 2.212115860120099,
      "grad_norm": 1.125,
      "learning_rate": 0.0004509132055287051,
      "loss": 5.3122,
      "mean_token_accuracy": 0.19762982279062272,
      "num_tokens": 46706568.0,
      "step": 25050
    },
    {
      "entropy": 5.9651261329650875,
      "epoch": 2.2125574002119395,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004508935693697352,
      "loss": 5.3682,
      "mean_token_accuracy": 0.21093694418668746,
      "num_tokens": 46715369.0,
      "step": 25055
    },
    {
      "entropy": 5.8661106586456295,
      "epoch": 2.2129989403037795,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004508739297651103,
      "loss": 5.2658,
      "mean_token_accuracy": 0.21148535013198852,
      "num_tokens": 46724867.0,
      "step": 25060
    },
    {
      "entropy": 5.866259765625,
      "epoch": 2.21344048039562,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045085428671521514,
      "loss": 5.3114,
      "mean_token_accuracy": 0.21071349531412126,
      "num_tokens": 46734611.0,
      "step": 25065
    },
    {
      "entropy": 5.935286378860473,
      "epoch": 2.21388202048746,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045083464022043456,
      "loss": 5.3203,
      "mean_token_accuracy": 0.20520623475313188,
      "num_tokens": 46744031.0,
      "step": 25070
    },
    {
      "entropy": 5.873317766189575,
      "epoch": 2.2143235605793006,
      "grad_norm": 1.125,
      "learning_rate": 0.00045081499028115337,
      "loss": 5.2467,
      "mean_token_accuracy": 0.22226097583770751,
      "num_tokens": 46753379.0,
      "step": 25075
    },
    {
      "entropy": 5.963942050933838,
      "epoch": 2.214765100671141,
      "grad_norm": 1.125,
      "learning_rate": 0.0004507953368977565,
      "loss": 5.3827,
      "mean_token_accuracy": 0.19638515263795853,
      "num_tokens": 46762625.0,
      "step": 25080
    },
    {
      "entropy": 5.924911308288574,
      "epoch": 2.215206640762981,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045077568007062893,
      "loss": 5.3529,
      "mean_token_accuracy": 0.21058709174394608,
      "num_tokens": 46772577.0,
      "step": 25085
    },
    {
      "entropy": 5.929365015029907,
      "epoch": 2.2156481808548216,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045075601980015573,
      "loss": 5.3221,
      "mean_token_accuracy": 0.21401261538267136,
      "num_tokens": 46782226.0,
      "step": 25090
    },
    {
      "entropy": 5.872507476806641,
      "epoch": 2.216089720946662,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045073635608672215,
      "loss": 5.2176,
      "mean_token_accuracy": 0.21572758108377457,
      "num_tokens": 46789988.0,
      "step": 25095
    },
    {
      "entropy": 5.910791254043579,
      "epoch": 2.216531261038502,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045071668893071324,
      "loss": 5.3543,
      "mean_token_accuracy": 0.19882809668779372,
      "num_tokens": 46799448.0,
      "step": 25100
    },
    {
      "entropy": 5.93634033203125,
      "epoch": 2.2169728011303427,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045069701833251434,
      "loss": 5.3458,
      "mean_token_accuracy": 0.20672561824321747,
      "num_tokens": 46809829.0,
      "step": 25105
    },
    {
      "entropy": 5.8733679294586185,
      "epoch": 2.217414341222183,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045067734429251086,
      "loss": 5.2877,
      "mean_token_accuracy": 0.21472548097372054,
      "num_tokens": 46817975.0,
      "step": 25110
    },
    {
      "entropy": 5.829145097732544,
      "epoch": 2.2178558813140232,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045065766681108793,
      "loss": 5.243,
      "mean_token_accuracy": 0.2141059309244156,
      "num_tokens": 46827350.0,
      "step": 25115
    },
    {
      "entropy": 5.910316228866577,
      "epoch": 2.2182974214058637,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004506379858886314,
      "loss": 5.303,
      "mean_token_accuracy": 0.21683756709098817,
      "num_tokens": 46835944.0,
      "step": 25120
    },
    {
      "entropy": 5.8803784370422365,
      "epoch": 2.2187389614977038,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045061830152552646,
      "loss": 5.2655,
      "mean_token_accuracy": 0.21124572157859803,
      "num_tokens": 46844843.0,
      "step": 25125
    },
    {
      "entropy": 5.902810001373291,
      "epoch": 2.2191805015895443,
      "grad_norm": 1.1875,
      "learning_rate": 0.000450598613722159,
      "loss": 5.3852,
      "mean_token_accuracy": 0.19811780899763107,
      "num_tokens": 46854188.0,
      "step": 25130
    },
    {
      "entropy": 5.913742113113403,
      "epoch": 2.2196220416813848,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004505789224789145,
      "loss": 5.3385,
      "mean_token_accuracy": 0.20688226521015168,
      "num_tokens": 46863279.0,
      "step": 25135
    },
    {
      "entropy": 5.931808853149414,
      "epoch": 2.220063581773225,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045055922779617874,
      "loss": 5.2956,
      "mean_token_accuracy": 0.20689380913972855,
      "num_tokens": 46873336.0,
      "step": 25140
    },
    {
      "entropy": 5.912406778335571,
      "epoch": 2.2205051218650653,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004505395296743376,
      "loss": 5.2988,
      "mean_token_accuracy": 0.2146769717335701,
      "num_tokens": 46882497.0,
      "step": 25145
    },
    {
      "entropy": 5.846728801727295,
      "epoch": 2.220946661956906,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00045051982811377696,
      "loss": 5.2445,
      "mean_token_accuracy": 0.2113554820418358,
      "num_tokens": 46891260.0,
      "step": 25150
    },
    {
      "entropy": 5.896357488632202,
      "epoch": 2.221388202048746,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045050012311488265,
      "loss": 5.2155,
      "mean_token_accuracy": 0.21709609925746917,
      "num_tokens": 46899969.0,
      "step": 25155
    },
    {
      "entropy": 5.842054843902588,
      "epoch": 2.2218297421405864,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045048041467804067,
      "loss": 5.2331,
      "mean_token_accuracy": 0.21707316786050795,
      "num_tokens": 46909206.0,
      "step": 25160
    },
    {
      "entropy": 5.862126064300537,
      "epoch": 2.222271282232427,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045046070280363725,
      "loss": 5.2447,
      "mean_token_accuracy": 0.21472883969545364,
      "num_tokens": 46918423.0,
      "step": 25165
    },
    {
      "entropy": 5.918490266799926,
      "epoch": 2.222712822324267,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004504409874920585,
      "loss": 5.2886,
      "mean_token_accuracy": 0.20576518774032593,
      "num_tokens": 46927817.0,
      "step": 25170
    },
    {
      "entropy": 5.988033962249756,
      "epoch": 2.2231543624161074,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004504212687436905,
      "loss": 5.3587,
      "mean_token_accuracy": 0.20495160669088364,
      "num_tokens": 46936880.0,
      "step": 25175
    },
    {
      "entropy": 5.996985864639282,
      "epoch": 2.223595902507948,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004504015465589196,
      "loss": 5.2994,
      "mean_token_accuracy": 0.1994699567556381,
      "num_tokens": 46946722.0,
      "step": 25180
    },
    {
      "entropy": 5.881836557388306,
      "epoch": 2.224037442599788,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004503818209381322,
      "loss": 5.2676,
      "mean_token_accuracy": 0.21085475087165834,
      "num_tokens": 46956660.0,
      "step": 25185
    },
    {
      "entropy": 5.929487466812134,
      "epoch": 2.2244789826916285,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004503620918817147,
      "loss": 5.3206,
      "mean_token_accuracy": 0.19917973428964614,
      "num_tokens": 46965823.0,
      "step": 25190
    },
    {
      "entropy": 5.926318836212158,
      "epoch": 2.224920522783469,
      "grad_norm": 1.125,
      "learning_rate": 0.00045034235939005353,
      "loss": 5.2955,
      "mean_token_accuracy": 0.2085857257246971,
      "num_tokens": 46974933.0,
      "step": 25195
    },
    {
      "entropy": 5.936947011947632,
      "epoch": 2.225362062875309,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045032262346353526,
      "loss": 5.3779,
      "mean_token_accuracy": 0.20409031063318253,
      "num_tokens": 46983506.0,
      "step": 25200
    },
    {
      "entropy": 5.879003047943115,
      "epoch": 2.2258036029671495,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004503028841025465,
      "loss": 5.276,
      "mean_token_accuracy": 0.20933781564235687,
      "num_tokens": 46992234.0,
      "step": 25205
    },
    {
      "entropy": 5.8367249965667725,
      "epoch": 2.2262451430589896,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045028314130747397,
      "loss": 5.2185,
      "mean_token_accuracy": 0.21828076839447022,
      "num_tokens": 47002249.0,
      "step": 25210
    },
    {
      "entropy": 5.859875059127807,
      "epoch": 2.22668668315083,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045026339507870435,
      "loss": 5.3495,
      "mean_token_accuracy": 0.21007241904735566,
      "num_tokens": 47010721.0,
      "step": 25215
    },
    {
      "entropy": 5.956480169296265,
      "epoch": 2.2271282232426706,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045024364541662457,
      "loss": 5.3518,
      "mean_token_accuracy": 0.204175928235054,
      "num_tokens": 47019675.0,
      "step": 25220
    },
    {
      "entropy": 5.9527867317199705,
      "epoch": 2.2275697633345106,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004502238923216213,
      "loss": 5.2812,
      "mean_token_accuracy": 0.21272652447223664,
      "num_tokens": 47028293.0,
      "step": 25225
    },
    {
      "entropy": 5.965755987167358,
      "epoch": 2.228011303426351,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004502041357940817,
      "loss": 5.3419,
      "mean_token_accuracy": 0.2071531221270561,
      "num_tokens": 47037605.0,
      "step": 25230
    },
    {
      "entropy": 5.868108034133911,
      "epoch": 2.2284528435181916,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045018437583439275,
      "loss": 5.2762,
      "mean_token_accuracy": 0.21650639474391936,
      "num_tokens": 47045889.0,
      "step": 25235
    },
    {
      "entropy": 5.867903137207032,
      "epoch": 2.2288943836100317,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045016461244294154,
      "loss": 5.322,
      "mean_token_accuracy": 0.2104552283883095,
      "num_tokens": 47055896.0,
      "step": 25240
    },
    {
      "entropy": 5.935128831863404,
      "epoch": 2.229335923701872,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045014484562011514,
      "loss": 5.3473,
      "mean_token_accuracy": 0.20300930440425874,
      "num_tokens": 47064980.0,
      "step": 25245
    },
    {
      "entropy": 5.853549432754517,
      "epoch": 2.2297774637937127,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004501250753663008,
      "loss": 5.2202,
      "mean_token_accuracy": 0.21265652030706406,
      "num_tokens": 47074045.0,
      "step": 25250
    },
    {
      "entropy": 5.849943590164185,
      "epoch": 2.2302190038855527,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004501053016818859,
      "loss": 5.2379,
      "mean_token_accuracy": 0.21710557788610457,
      "num_tokens": 47083439.0,
      "step": 25255
    },
    {
      "entropy": 5.870268821716309,
      "epoch": 2.230660543977393,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045008552456725766,
      "loss": 5.276,
      "mean_token_accuracy": 0.21208076030015946,
      "num_tokens": 47091977.0,
      "step": 25260
    },
    {
      "entropy": 5.877220201492309,
      "epoch": 2.2311020840692333,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004500657440228037,
      "loss": 5.3058,
      "mean_token_accuracy": 0.20941203236579894,
      "num_tokens": 47102197.0,
      "step": 25265
    },
    {
      "entropy": 5.966434955596924,
      "epoch": 2.2315436241610738,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045004596004891124,
      "loss": 5.3353,
      "mean_token_accuracy": 0.2138953238725662,
      "num_tokens": 47111563.0,
      "step": 25270
    },
    {
      "entropy": 5.888266372680664,
      "epoch": 2.2319851642529143,
      "grad_norm": 1.125,
      "learning_rate": 0.00045002617264596807,
      "loss": 5.3041,
      "mean_token_accuracy": 0.21606009900569917,
      "num_tokens": 47120272.0,
      "step": 25275
    },
    {
      "entropy": 5.863013458251953,
      "epoch": 2.2324267043447543,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004500063818143617,
      "loss": 5.2776,
      "mean_token_accuracy": 0.21193020939826965,
      "num_tokens": 47131023.0,
      "step": 25280
    },
    {
      "entropy": 5.828634023666382,
      "epoch": 2.232868244436595,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004499865875544797,
      "loss": 5.2548,
      "mean_token_accuracy": 0.21766403317451477,
      "num_tokens": 47139968.0,
      "step": 25285
    },
    {
      "entropy": 5.859275817871094,
      "epoch": 2.2333097845284353,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004499667898667101,
      "loss": 5.258,
      "mean_token_accuracy": 0.21532260477542878,
      "num_tokens": 47149843.0,
      "step": 25290
    },
    {
      "entropy": 5.895445394515991,
      "epoch": 2.2337513246202754,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044994698875144064,
      "loss": 5.3087,
      "mean_token_accuracy": 0.2108255445957184,
      "num_tokens": 47159025.0,
      "step": 25295
    },
    {
      "entropy": 5.904364681243896,
      "epoch": 2.234192864712116,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004499271842090591,
      "loss": 5.3293,
      "mean_token_accuracy": 0.2063928499817848,
      "num_tokens": 47168644.0,
      "step": 25300
    },
    {
      "entropy": 5.87594690322876,
      "epoch": 2.2346344048039564,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044990737623995357,
      "loss": 5.2394,
      "mean_token_accuracy": 0.2168074816465378,
      "num_tokens": 47178051.0,
      "step": 25305
    },
    {
      "entropy": 5.957010126113891,
      "epoch": 2.2350759448957964,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004498875648445119,
      "loss": 5.3934,
      "mean_token_accuracy": 0.20169862508773803,
      "num_tokens": 47187855.0,
      "step": 25310
    },
    {
      "entropy": 5.852364349365234,
      "epoch": 2.235517484987637,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004498677500231224,
      "loss": 5.3207,
      "mean_token_accuracy": 0.2070738285779953,
      "num_tokens": 47196700.0,
      "step": 25315
    },
    {
      "entropy": 5.932506465911866,
      "epoch": 2.2359590250794774,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004498479317761731,
      "loss": 5.3812,
      "mean_token_accuracy": 0.2044564366340637,
      "num_tokens": 47205202.0,
      "step": 25320
    },
    {
      "entropy": 5.848980379104614,
      "epoch": 2.2364005651713175,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004498281101040522,
      "loss": 5.2853,
      "mean_token_accuracy": 0.2187192052602768,
      "num_tokens": 47214803.0,
      "step": 25325
    },
    {
      "entropy": 5.972981214523315,
      "epoch": 2.236842105263158,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044980828500714814,
      "loss": 5.3826,
      "mean_token_accuracy": 0.20191094726324083,
      "num_tokens": 47224641.0,
      "step": 25330
    },
    {
      "entropy": 5.915935611724853,
      "epoch": 2.2372836453549985,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004497884564858492,
      "loss": 5.339,
      "mean_token_accuracy": 0.1996872067451477,
      "num_tokens": 47233905.0,
      "step": 25335
    },
    {
      "entropy": 5.903089284896851,
      "epoch": 2.2377251854468385,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044976862454054376,
      "loss": 5.2588,
      "mean_token_accuracy": 0.21022669225931168,
      "num_tokens": 47242209.0,
      "step": 25340
    },
    {
      "entropy": 5.9644183158874515,
      "epoch": 2.238166725538679,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004497487891716204,
      "loss": 5.3399,
      "mean_token_accuracy": 0.21070777326822282,
      "num_tokens": 47252609.0,
      "step": 25345
    },
    {
      "entropy": 5.962730026245117,
      "epoch": 2.238608265630519,
      "grad_norm": 1.125,
      "learning_rate": 0.00044972895037946755,
      "loss": 5.3551,
      "mean_token_accuracy": 0.20537231415510177,
      "num_tokens": 47261868.0,
      "step": 25350
    },
    {
      "entropy": 5.935928535461426,
      "epoch": 2.2390498057223596,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044970910816447403,
      "loss": 5.2872,
      "mean_token_accuracy": 0.21497126966714858,
      "num_tokens": 47271404.0,
      "step": 25355
    },
    {
      "entropy": 5.922522735595703,
      "epoch": 2.2394913458142,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044968926252702835,
      "loss": 5.3435,
      "mean_token_accuracy": 0.2059139981865883,
      "num_tokens": 47280669.0,
      "step": 25360
    },
    {
      "entropy": 5.938412523269653,
      "epoch": 2.23993288590604,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004496694134675195,
      "loss": 5.3536,
      "mean_token_accuracy": 0.2070578634738922,
      "num_tokens": 47290096.0,
      "step": 25365
    },
    {
      "entropy": 5.964613246917724,
      "epoch": 2.2403744259978806,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004496495609863361,
      "loss": 5.3442,
      "mean_token_accuracy": 0.20849495381116867,
      "num_tokens": 47298767.0,
      "step": 25370
    },
    {
      "entropy": 5.834093189239502,
      "epoch": 2.240815966089721,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044962970508386716,
      "loss": 5.2142,
      "mean_token_accuracy": 0.2193813592195511,
      "num_tokens": 47307950.0,
      "step": 25375
    },
    {
      "entropy": 5.842846775054932,
      "epoch": 2.241257506181561,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004496098457605016,
      "loss": 5.3171,
      "mean_token_accuracy": 0.21000723540782928,
      "num_tokens": 47317963.0,
      "step": 25380
    },
    {
      "entropy": 5.89036283493042,
      "epoch": 2.2416990462734017,
      "grad_norm": 1.125,
      "learning_rate": 0.0004495899830166284,
      "loss": 5.3154,
      "mean_token_accuracy": 0.2095501020550728,
      "num_tokens": 47326728.0,
      "step": 25385
    },
    {
      "entropy": 6.011849403381348,
      "epoch": 2.242140586365242,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004495701168526368,
      "loss": 5.3265,
      "mean_token_accuracy": 0.20927484929561616,
      "num_tokens": 47336386.0,
      "step": 25390
    },
    {
      "entropy": 5.95103268623352,
      "epoch": 2.242582126457082,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044955024726891585,
      "loss": 5.3366,
      "mean_token_accuracy": 0.2021625816822052,
      "num_tokens": 47345958.0,
      "step": 25395
    },
    {
      "entropy": 5.902404022216797,
      "epoch": 2.2430236665489227,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004495303742658549,
      "loss": 5.3458,
      "mean_token_accuracy": 0.20588893741369246,
      "num_tokens": 47355758.0,
      "step": 25400
    },
    {
      "entropy": 5.965249490737915,
      "epoch": 2.2434652066407628,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004495104978438431,
      "loss": 5.3252,
      "mean_token_accuracy": 0.21320498287677764,
      "num_tokens": 47365186.0,
      "step": 25405
    },
    {
      "entropy": 5.91492018699646,
      "epoch": 2.2439067467326033,
      "grad_norm": 1.40625,
      "learning_rate": 0.00044949061800327,
      "loss": 5.3093,
      "mean_token_accuracy": 0.21295351833105086,
      "num_tokens": 47374443.0,
      "step": 25410
    },
    {
      "entropy": 5.8478227138519285,
      "epoch": 2.2443482868244438,
      "grad_norm": 1.125,
      "learning_rate": 0.0004494707347445248,
      "loss": 5.2356,
      "mean_token_accuracy": 0.21066135615110398,
      "num_tokens": 47383369.0,
      "step": 25415
    },
    {
      "entropy": 5.853169679641724,
      "epoch": 2.244789826916284,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004494508480679972,
      "loss": 5.2627,
      "mean_token_accuracy": 0.2113923028111458,
      "num_tokens": 47393104.0,
      "step": 25420
    },
    {
      "entropy": 5.840248966217041,
      "epoch": 2.2452313670081243,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044943095797407676,
      "loss": 5.2599,
      "mean_token_accuracy": 0.2124545991420746,
      "num_tokens": 47401954.0,
      "step": 25425
    },
    {
      "entropy": 5.949639463424683,
      "epoch": 2.245672907099965,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000449411064463153,
      "loss": 5.2631,
      "mean_token_accuracy": 0.2147961214184761,
      "num_tokens": 47411463.0,
      "step": 25430
    },
    {
      "entropy": 5.9110359191894535,
      "epoch": 2.246114447191805,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044939116753561564,
      "loss": 5.2016,
      "mean_token_accuracy": 0.2140165835618973,
      "num_tokens": 47421405.0,
      "step": 25435
    },
    {
      "entropy": 5.8378636837005615,
      "epoch": 2.2465559872836454,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004493712671918546,
      "loss": 5.3247,
      "mean_token_accuracy": 0.2112680584192276,
      "num_tokens": 47430268.0,
      "step": 25440
    },
    {
      "entropy": 5.9168859958648685,
      "epoch": 2.246997527375486,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004493513634322595,
      "loss": 5.3532,
      "mean_token_accuracy": 0.20273826718330384,
      "num_tokens": 47439571.0,
      "step": 25445
    },
    {
      "entropy": 5.908704090118408,
      "epoch": 2.247439067467326,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004493314562572204,
      "loss": 5.3002,
      "mean_token_accuracy": 0.20778064131736756,
      "num_tokens": 47449085.0,
      "step": 25450
    },
    {
      "entropy": 5.846994256973266,
      "epoch": 2.2478806075591664,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004493115456671272,
      "loss": 5.2136,
      "mean_token_accuracy": 0.21585794240236283,
      "num_tokens": 47458740.0,
      "step": 25455
    },
    {
      "entropy": 5.9260420322418215,
      "epoch": 2.248322147651007,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044929163166236997,
      "loss": 5.3499,
      "mean_token_accuracy": 0.20734633058309554,
      "num_tokens": 47467458.0,
      "step": 25460
    },
    {
      "entropy": 5.906539440155029,
      "epoch": 2.248763687742847,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004492717142433388,
      "loss": 5.3123,
      "mean_token_accuracy": 0.20516202747821807,
      "num_tokens": 47476845.0,
      "step": 25465
    },
    {
      "entropy": 5.882449293136597,
      "epoch": 2.2492052278346875,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044925179341042385,
      "loss": 5.3097,
      "mean_token_accuracy": 0.20874747037887573,
      "num_tokens": 47485632.0,
      "step": 25470
    },
    {
      "entropy": 5.944317388534546,
      "epoch": 2.249646767926528,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044923186916401536,
      "loss": 5.3056,
      "mean_token_accuracy": 0.21193400025367737,
      "num_tokens": 47494681.0,
      "step": 25475
    },
    {
      "entropy": 5.867374753952026,
      "epoch": 2.250088308018368,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004492119415045036,
      "loss": 5.3371,
      "mean_token_accuracy": 0.21191453486680983,
      "num_tokens": 47504482.0,
      "step": 25480
    },
    {
      "entropy": 5.869369745254517,
      "epoch": 2.2505298481102085,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004491920104322791,
      "loss": 5.3113,
      "mean_token_accuracy": 0.21246003955602646,
      "num_tokens": 47513601.0,
      "step": 25485
    },
    {
      "entropy": 5.910813283920288,
      "epoch": 2.2509713882020486,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004491720759477321,
      "loss": 5.2678,
      "mean_token_accuracy": 0.21182847023010254,
      "num_tokens": 47522811.0,
      "step": 25490
    },
    {
      "entropy": 5.912567615509033,
      "epoch": 2.251412928293889,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004491521380512533,
      "loss": 5.2812,
      "mean_token_accuracy": 0.2175702065229416,
      "num_tokens": 47532471.0,
      "step": 25495
    },
    {
      "entropy": 5.828784799575805,
      "epoch": 2.2518544683857296,
      "grad_norm": 1.203125,
      "learning_rate": 0.000449132196743233,
      "loss": 5.2998,
      "mean_token_accuracy": 0.21837954074144364,
      "num_tokens": 47541336.0,
      "step": 25500
    },
    {
      "entropy": 5.864141893386841,
      "epoch": 2.2522960084775696,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004491122520240621,
      "loss": 5.2223,
      "mean_token_accuracy": 0.21477721333503724,
      "num_tokens": 47551446.0,
      "step": 25505
    },
    {
      "entropy": 5.889760589599609,
      "epoch": 2.25273754856941,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004490923038941312,
      "loss": 5.287,
      "mean_token_accuracy": 0.21306386291980745,
      "num_tokens": 47560596.0,
      "step": 25510
    },
    {
      "entropy": 5.958892393112182,
      "epoch": 2.2531790886612506,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000449072352353831,
      "loss": 5.3854,
      "mean_token_accuracy": 0.20812500715255738,
      "num_tokens": 47570612.0,
      "step": 25515
    },
    {
      "entropy": 5.92105450630188,
      "epoch": 2.2536206287530907,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004490523974035525,
      "loss": 5.3114,
      "mean_token_accuracy": 0.2044833332300186,
      "num_tokens": 47580234.0,
      "step": 25520
    },
    {
      "entropy": 5.953958892822266,
      "epoch": 2.254062168844931,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004490324390436866,
      "loss": 5.3402,
      "mean_token_accuracy": 0.2086621806025505,
      "num_tokens": 47589777.0,
      "step": 25525
    },
    {
      "entropy": 5.9193359375,
      "epoch": 2.254503708936771,
      "grad_norm": 1.21875,
      "learning_rate": 0.000449012477274624,
      "loss": 5.2469,
      "mean_token_accuracy": 0.21461936533451081,
      "num_tokens": 47597951.0,
      "step": 25530
    },
    {
      "entropy": 5.904752635955811,
      "epoch": 2.2549452490286117,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044899251209675606,
      "loss": 5.352,
      "mean_token_accuracy": 0.20868321657180786,
      "num_tokens": 47606054.0,
      "step": 25535
    },
    {
      "entropy": 5.992991876602173,
      "epoch": 2.255386789120452,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004489725435104737,
      "loss": 5.3342,
      "mean_token_accuracy": 0.21043829768896102,
      "num_tokens": 47614704.0,
      "step": 25540
    },
    {
      "entropy": 5.888143825531006,
      "epoch": 2.2558283292122923,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004489525715161682,
      "loss": 5.3159,
      "mean_token_accuracy": 0.2094903200864792,
      "num_tokens": 47623189.0,
      "step": 25545
    },
    {
      "entropy": 5.8683079242706295,
      "epoch": 2.2562698693041328,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044893259611423076,
      "loss": 5.281,
      "mean_token_accuracy": 0.20954559892416,
      "num_tokens": 47633831.0,
      "step": 25550
    },
    {
      "entropy": 5.979770135879517,
      "epoch": 2.2567114093959733,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004489126173050527,
      "loss": 5.2176,
      "mean_token_accuracy": 0.21380435079336166,
      "num_tokens": 47642155.0,
      "step": 25555
    },
    {
      "entropy": 6.064649438858032,
      "epoch": 2.2571529494878133,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004488926350890253,
      "loss": 5.4504,
      "mean_token_accuracy": 0.19683613181114196,
      "num_tokens": 47651042.0,
      "step": 25560
    },
    {
      "entropy": 5.95261516571045,
      "epoch": 2.257594489579654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004488726494665402,
      "loss": 5.3637,
      "mean_token_accuracy": 0.20659175515174866,
      "num_tokens": 47661424.0,
      "step": 25565
    },
    {
      "entropy": 5.918775033950806,
      "epoch": 2.2580360296714943,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044885266043798866,
      "loss": 5.3001,
      "mean_token_accuracy": 0.20935991108417512,
      "num_tokens": 47671603.0,
      "step": 25570
    },
    {
      "entropy": 5.903744792938232,
      "epoch": 2.2584775697633344,
      "grad_norm": 1.25,
      "learning_rate": 0.0004488326680037624,
      "loss": 5.301,
      "mean_token_accuracy": 0.20374856442213057,
      "num_tokens": 47679861.0,
      "step": 25575
    },
    {
      "entropy": 5.78634729385376,
      "epoch": 2.258919109855175,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004488126721642531,
      "loss": 5.195,
      "mean_token_accuracy": 0.21474214494228364,
      "num_tokens": 47689702.0,
      "step": 25580
    },
    {
      "entropy": 5.812030076980591,
      "epoch": 2.2593606499470154,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044879267291985243,
      "loss": 5.2884,
      "mean_token_accuracy": 0.2032901331782341,
      "num_tokens": 47699261.0,
      "step": 25585
    },
    {
      "entropy": 5.962048864364624,
      "epoch": 2.2598021900388554,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044877267027095203,
      "loss": 5.3635,
      "mean_token_accuracy": 0.20693973153829576,
      "num_tokens": 47709079.0,
      "step": 25590
    },
    {
      "entropy": 5.875109338760376,
      "epoch": 2.260243730130696,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004487526642179439,
      "loss": 5.2452,
      "mean_token_accuracy": 0.21441958844661713,
      "num_tokens": 47718641.0,
      "step": 25595
    },
    {
      "entropy": 5.996947765350342,
      "epoch": 2.2606852702225364,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044873265476121993,
      "loss": 5.3722,
      "mean_token_accuracy": 0.20505607426166533,
      "num_tokens": 47729156.0,
      "step": 25600
    },
    {
      "entropy": 5.91913628578186,
      "epoch": 2.2611268103143765,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044871264190117196,
      "loss": 5.3121,
      "mean_token_accuracy": 0.20826969146728516,
      "num_tokens": 47738424.0,
      "step": 25605
    },
    {
      "entropy": 5.9685125827789305,
      "epoch": 2.261568350406217,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004486926256381922,
      "loss": 5.4466,
      "mean_token_accuracy": 0.19461868405342103,
      "num_tokens": 47749032.0,
      "step": 25610
    },
    {
      "entropy": 5.977568626403809,
      "epoch": 2.2620098904980575,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004486726059726727,
      "loss": 5.3256,
      "mean_token_accuracy": 0.2057436317205429,
      "num_tokens": 47758979.0,
      "step": 25615
    },
    {
      "entropy": 5.892540216445923,
      "epoch": 2.2624514305898975,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004486525829050056,
      "loss": 5.3005,
      "mean_token_accuracy": 0.20875338912010194,
      "num_tokens": 47768752.0,
      "step": 25620
    },
    {
      "entropy": 5.929640245437622,
      "epoch": 2.262892970681738,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004486325564355831,
      "loss": 5.3552,
      "mean_token_accuracy": 0.20305759459733963,
      "num_tokens": 47778175.0,
      "step": 25625
    },
    {
      "entropy": 5.866891765594483,
      "epoch": 2.263334510773578,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004486125265647977,
      "loss": 5.2947,
      "mean_token_accuracy": 0.21630572229623796,
      "num_tokens": 47787864.0,
      "step": 25630
    },
    {
      "entropy": 5.955856084823608,
      "epoch": 2.2637760508654186,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004485924932930416,
      "loss": 5.3574,
      "mean_token_accuracy": 0.19902995526790618,
      "num_tokens": 47796912.0,
      "step": 25635
    },
    {
      "entropy": 5.969979858398437,
      "epoch": 2.264217590957259,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004485724566207073,
      "loss": 5.3896,
      "mean_token_accuracy": 0.2024100422859192,
      "num_tokens": 47807452.0,
      "step": 25640
    },
    {
      "entropy": 6.00482211112976,
      "epoch": 2.264659131049099,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044855241654818723,
      "loss": 5.4097,
      "mean_token_accuracy": 0.20293576568365096,
      "num_tokens": 47817575.0,
      "step": 25645
    },
    {
      "entropy": 5.956986618041992,
      "epoch": 2.2651006711409396,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004485323730758741,
      "loss": 5.3141,
      "mean_token_accuracy": 0.21868909746408463,
      "num_tokens": 47826177.0,
      "step": 25650
    },
    {
      "entropy": 5.873044681549072,
      "epoch": 2.26554221123278,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004485123262041604,
      "loss": 5.2547,
      "mean_token_accuracy": 0.20949017405509948,
      "num_tokens": 47835622.0,
      "step": 25655
    },
    {
      "entropy": 5.871547794342041,
      "epoch": 2.26598375132462,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004484922759334389,
      "loss": 5.2971,
      "mean_token_accuracy": 0.21253005564212799,
      "num_tokens": 47845552.0,
      "step": 25660
    },
    {
      "entropy": 5.9136628150939945,
      "epoch": 2.2664252914164607,
      "grad_norm": 1.25,
      "learning_rate": 0.0004484722222641024,
      "loss": 5.2477,
      "mean_token_accuracy": 0.20884334594011306,
      "num_tokens": 47853820.0,
      "step": 25665
    },
    {
      "entropy": 5.895097112655639,
      "epoch": 2.2668668315083007,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004484521651965438,
      "loss": 5.2528,
      "mean_token_accuracy": 0.21776512116193772,
      "num_tokens": 47863416.0,
      "step": 25670
    },
    {
      "entropy": 5.919981908798218,
      "epoch": 2.267308371600141,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044843210473115584,
      "loss": 5.3251,
      "mean_token_accuracy": 0.21240510940551757,
      "num_tokens": 47871407.0,
      "step": 25675
    },
    {
      "entropy": 5.927334308624268,
      "epoch": 2.2677499116919817,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004484120408683316,
      "loss": 5.276,
      "mean_token_accuracy": 0.20964734852313996,
      "num_tokens": 47880742.0,
      "step": 25680
    },
    {
      "entropy": 5.8885444641113285,
      "epoch": 2.2681914517838218,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044839197360846416,
      "loss": 5.3543,
      "mean_token_accuracy": 0.20122861117124557,
      "num_tokens": 47889573.0,
      "step": 25685
    },
    {
      "entropy": 5.901750230789185,
      "epoch": 2.2686329918756623,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044837190295194654,
      "loss": 5.2868,
      "mean_token_accuracy": 0.21235938370227814,
      "num_tokens": 47898855.0,
      "step": 25690
    },
    {
      "entropy": 5.921929597854614,
      "epoch": 2.2690745319675028,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044835182889917195,
      "loss": 5.3224,
      "mean_token_accuracy": 0.20815297216176987,
      "num_tokens": 47908981.0,
      "step": 25695
    },
    {
      "entropy": 5.908852291107178,
      "epoch": 2.269516072059343,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004483317514505336,
      "loss": 5.3144,
      "mean_token_accuracy": 0.2115495041012764,
      "num_tokens": 47918313.0,
      "step": 25700
    },
    {
      "entropy": 5.861187314987182,
      "epoch": 2.2699576121511833,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004483116706064248,
      "loss": 5.2471,
      "mean_token_accuracy": 0.21142636239528656,
      "num_tokens": 47927838.0,
      "step": 25705
    },
    {
      "entropy": 5.905887460708618,
      "epoch": 2.270399152243024,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004482915863672389,
      "loss": 5.2902,
      "mean_token_accuracy": 0.20914349853992462,
      "num_tokens": 47937774.0,
      "step": 25710
    },
    {
      "entropy": 5.943370914459228,
      "epoch": 2.270840692334864,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004482714987333695,
      "loss": 5.2514,
      "mean_token_accuracy": 0.2146176055073738,
      "num_tokens": 47947091.0,
      "step": 25715
    },
    {
      "entropy": 5.8821768283844,
      "epoch": 2.2712822324267044,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044825140770520997,
      "loss": 5.2402,
      "mean_token_accuracy": 0.21849192529916764,
      "num_tokens": 47955715.0,
      "step": 25720
    },
    {
      "entropy": 5.741145372390747,
      "epoch": 2.271723772518545,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044823131328315387,
      "loss": 5.1576,
      "mean_token_accuracy": 0.23379732221364974,
      "num_tokens": 47965991.0,
      "step": 25725
    },
    {
      "entropy": 5.924877166748047,
      "epoch": 2.272165312610385,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004482112154675948,
      "loss": 5.3498,
      "mean_token_accuracy": 0.20598774254322053,
      "num_tokens": 47975610.0,
      "step": 25730
    },
    {
      "entropy": 5.876081752777099,
      "epoch": 2.2726068527022254,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044819111425892667,
      "loss": 5.2691,
      "mean_token_accuracy": 0.21684676706790923,
      "num_tokens": 47984977.0,
      "step": 25735
    },
    {
      "entropy": 5.863063144683838,
      "epoch": 2.273048392794066,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044817100965754297,
      "loss": 5.2424,
      "mean_token_accuracy": 0.21109579503536224,
      "num_tokens": 47994205.0,
      "step": 25740
    },
    {
      "entropy": 5.877716207504273,
      "epoch": 2.273489932885906,
      "grad_norm": 1.125,
      "learning_rate": 0.00044815090166383775,
      "loss": 5.3187,
      "mean_token_accuracy": 0.21038877815008164,
      "num_tokens": 48003715.0,
      "step": 25745
    },
    {
      "entropy": 5.9276769161224365,
      "epoch": 2.2739314729777464,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004481307902782048,
      "loss": 5.4015,
      "mean_token_accuracy": 0.20905213952064514,
      "num_tokens": 48013200.0,
      "step": 25750
    },
    {
      "entropy": 5.955729198455811,
      "epoch": 2.274373013069587,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004481106755010382,
      "loss": 5.3283,
      "mean_token_accuracy": 0.20823585689067842,
      "num_tokens": 48022366.0,
      "step": 25755
    },
    {
      "entropy": 5.931550931930542,
      "epoch": 2.274814553161427,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044809055733273183,
      "loss": 5.347,
      "mean_token_accuracy": 0.20925108939409257,
      "num_tokens": 48031968.0,
      "step": 25760
    },
    {
      "entropy": 5.894931507110596,
      "epoch": 2.2752560932532675,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044807043577367996,
      "loss": 5.2863,
      "mean_token_accuracy": 0.2153531491756439,
      "num_tokens": 48041069.0,
      "step": 25765
    },
    {
      "entropy": 5.795672035217285,
      "epoch": 2.2756976333451076,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044805031082427666,
      "loss": 5.1953,
      "mean_token_accuracy": 0.21762861013412477,
      "num_tokens": 48050550.0,
      "step": 25770
    },
    {
      "entropy": 5.922796010971069,
      "epoch": 2.276139173436948,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044803018248491615,
      "loss": 5.2726,
      "mean_token_accuracy": 0.20829552114009858,
      "num_tokens": 48059971.0,
      "step": 25775
    },
    {
      "entropy": 5.890300512313843,
      "epoch": 2.2765807135287885,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004480100507559928,
      "loss": 5.1586,
      "mean_token_accuracy": 0.21602099090814592,
      "num_tokens": 48068566.0,
      "step": 25780
    },
    {
      "entropy": 5.7922278881073,
      "epoch": 2.2770222536206286,
      "grad_norm": 1.0625,
      "learning_rate": 0.000447989915637901,
      "loss": 5.2006,
      "mean_token_accuracy": 0.21922920793294906,
      "num_tokens": 48078352.0,
      "step": 25785
    },
    {
      "entropy": 5.849106931686402,
      "epoch": 2.277463793712469,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004479697771310351,
      "loss": 5.3018,
      "mean_token_accuracy": 0.2083287462592125,
      "num_tokens": 48087089.0,
      "step": 25790
    },
    {
      "entropy": 5.9186639308929445,
      "epoch": 2.2779053338043096,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044794963523578965,
      "loss": 5.3621,
      "mean_token_accuracy": 0.20507449954748153,
      "num_tokens": 48096762.0,
      "step": 25795
    },
    {
      "entropy": 5.919962739944458,
      "epoch": 2.2783468738961496,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044792948995255917,
      "loss": 5.2812,
      "mean_token_accuracy": 0.21261240243911744,
      "num_tokens": 48105994.0,
      "step": 25800
    },
    {
      "entropy": 5.888382148742676,
      "epoch": 2.27878841398799,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004479093412817383,
      "loss": 5.2587,
      "mean_token_accuracy": 0.21734432876110077,
      "num_tokens": 48115148.0,
      "step": 25805
    },
    {
      "entropy": 5.841091680526733,
      "epoch": 2.27922995407983,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044788918922372196,
      "loss": 5.3449,
      "mean_token_accuracy": 0.20284795761108398,
      "num_tokens": 48124696.0,
      "step": 25810
    },
    {
      "entropy": 5.868496656417847,
      "epoch": 2.2796714941716707,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004478690337789046,
      "loss": 5.2343,
      "mean_token_accuracy": 0.21560890972614288,
      "num_tokens": 48134165.0,
      "step": 25815
    },
    {
      "entropy": 5.968175220489502,
      "epoch": 2.280113034263511,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044784887494768113,
      "loss": 5.2773,
      "mean_token_accuracy": 0.20464987605810164,
      "num_tokens": 48144080.0,
      "step": 25820
    },
    {
      "entropy": 5.8811687469482425,
      "epoch": 2.2805545743553512,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044782871273044654,
      "loss": 5.3186,
      "mean_token_accuracy": 0.21887168139219285,
      "num_tokens": 48156161.0,
      "step": 25825
    },
    {
      "entropy": 5.89983115196228,
      "epoch": 2.2809961144471917,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004478085471275958,
      "loss": 5.3069,
      "mean_token_accuracy": 0.20691936910152436,
      "num_tokens": 48165317.0,
      "step": 25830
    },
    {
      "entropy": 5.8984099388122555,
      "epoch": 2.2814376545390322,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004477883781395239,
      "loss": 5.2647,
      "mean_token_accuracy": 0.21246760487556457,
      "num_tokens": 48173639.0,
      "step": 25835
    },
    {
      "entropy": 5.8816565990448,
      "epoch": 2.2818791946308723,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044776820576662593,
      "loss": 5.283,
      "mean_token_accuracy": 0.21220593750476838,
      "num_tokens": 48183596.0,
      "step": 25840
    },
    {
      "entropy": 5.985248136520386,
      "epoch": 2.282320734722713,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004477480300092971,
      "loss": 5.3315,
      "mean_token_accuracy": 0.21500713378190994,
      "num_tokens": 48194269.0,
      "step": 25845
    },
    {
      "entropy": 5.915961074829101,
      "epoch": 2.2827622748145533,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044772785086793257,
      "loss": 5.3138,
      "mean_token_accuracy": 0.20815593749284744,
      "num_tokens": 48203180.0,
      "step": 25850
    },
    {
      "entropy": 5.898174190521241,
      "epoch": 2.2832038149063933,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044770766834292765,
      "loss": 5.3275,
      "mean_token_accuracy": 0.21373473405838012,
      "num_tokens": 48212768.0,
      "step": 25855
    },
    {
      "entropy": 5.915253353118897,
      "epoch": 2.283645354998234,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004476874824346777,
      "loss": 5.3488,
      "mean_token_accuracy": 0.20545073449611664,
      "num_tokens": 48221620.0,
      "step": 25860
    },
    {
      "entropy": 5.917410564422608,
      "epoch": 2.2840868950900743,
      "grad_norm": 1.125,
      "learning_rate": 0.00044766729314357827,
      "loss": 5.3288,
      "mean_token_accuracy": 0.20083833932876588,
      "num_tokens": 48231379.0,
      "step": 25865
    },
    {
      "entropy": 5.9638396263122555,
      "epoch": 2.2845284351819144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004476471004700248,
      "loss": 5.3471,
      "mean_token_accuracy": 0.20801860839128494,
      "num_tokens": 48241827.0,
      "step": 25870
    },
    {
      "entropy": 5.978330278396607,
      "epoch": 2.284969975273755,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004476269044144127,
      "loss": 5.3073,
      "mean_token_accuracy": 0.206272092461586,
      "num_tokens": 48250747.0,
      "step": 25875
    },
    {
      "entropy": 5.931081247329712,
      "epoch": 2.2854115153655954,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004476067049771378,
      "loss": 5.2985,
      "mean_token_accuracy": 0.20079461336135865,
      "num_tokens": 48260909.0,
      "step": 25880
    },
    {
      "entropy": 5.9064287662506105,
      "epoch": 2.2858530554574354,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004475865021585957,
      "loss": 5.3647,
      "mean_token_accuracy": 0.19917707443237304,
      "num_tokens": 48270974.0,
      "step": 25885
    },
    {
      "entropy": 5.884914541244507,
      "epoch": 2.286294595549276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004475662959591822,
      "loss": 5.2302,
      "mean_token_accuracy": 0.21621289253234863,
      "num_tokens": 48279466.0,
      "step": 25890
    },
    {
      "entropy": 5.915259218215942,
      "epoch": 2.2867361356411164,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000447546086379293,
      "loss": 5.3452,
      "mean_token_accuracy": 0.20903617590665818,
      "num_tokens": 48289830.0,
      "step": 25895
    },
    {
      "entropy": 5.91605019569397,
      "epoch": 2.2871776757329565,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004475258734193242,
      "loss": 5.3006,
      "mean_token_accuracy": 0.21248300671577453,
      "num_tokens": 48299493.0,
      "step": 25900
    },
    {
      "entropy": 5.925859689712524,
      "epoch": 2.287619215824797,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004475056570796717,
      "loss": 5.3196,
      "mean_token_accuracy": 0.21409946382045747,
      "num_tokens": 48308366.0,
      "step": 25905
    },
    {
      "entropy": 5.910337591171265,
      "epoch": 2.288060755916637,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004474854373607314,
      "loss": 5.2834,
      "mean_token_accuracy": 0.20431629717350006,
      "num_tokens": 48317760.0,
      "step": 25910
    },
    {
      "entropy": 5.890132856369019,
      "epoch": 2.2885022960084775,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004474652142628995,
      "loss": 5.3346,
      "mean_token_accuracy": 0.20410245954990386,
      "num_tokens": 48325847.0,
      "step": 25915
    },
    {
      "entropy": 5.9430078029632565,
      "epoch": 2.288943836100318,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004474449877865722,
      "loss": 5.3304,
      "mean_token_accuracy": 0.2001043140888214,
      "num_tokens": 48334938.0,
      "step": 25920
    },
    {
      "entropy": 5.903149366378784,
      "epoch": 2.289385376192158,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004474247579321456,
      "loss": 5.3239,
      "mean_token_accuracy": 0.2027680292725563,
      "num_tokens": 48344414.0,
      "step": 25925
    },
    {
      "entropy": 5.902489852905274,
      "epoch": 2.2898269162839986,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000447404524700016,
      "loss": 5.2523,
      "mean_token_accuracy": 0.20890749841928483,
      "num_tokens": 48353711.0,
      "step": 25930
    },
    {
      "entropy": 5.819829511642456,
      "epoch": 2.290268456375839,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004473842880905799,
      "loss": 5.2066,
      "mean_token_accuracy": 0.22657289505004882,
      "num_tokens": 48362907.0,
      "step": 25935
    },
    {
      "entropy": 5.915345096588135,
      "epoch": 2.290709996467679,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044736404810423366,
      "loss": 5.2713,
      "mean_token_accuracy": 0.21852019280195237,
      "num_tokens": 48372450.0,
      "step": 25940
    },
    {
      "entropy": 5.842714166641235,
      "epoch": 2.2911515365595196,
      "grad_norm": 1.125,
      "learning_rate": 0.0004473438047413737,
      "loss": 5.2266,
      "mean_token_accuracy": 0.2167292207479477,
      "num_tokens": 48381521.0,
      "step": 25945
    },
    {
      "entropy": 5.887005186080932,
      "epoch": 2.2915930766513597,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004473235580023966,
      "loss": 5.2517,
      "mean_token_accuracy": 0.21939159333705902,
      "num_tokens": 48390635.0,
      "step": 25950
    },
    {
      "entropy": 5.887283658981323,
      "epoch": 2.2920346167432,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044730330788769903,
      "loss": 5.3246,
      "mean_token_accuracy": 0.20345648676156997,
      "num_tokens": 48400199.0,
      "step": 25955
    },
    {
      "entropy": 5.862120914459228,
      "epoch": 2.2924761568350407,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044728305439767764,
      "loss": 5.289,
      "mean_token_accuracy": 0.21052179336547852,
      "num_tokens": 48409266.0,
      "step": 25960
    },
    {
      "entropy": 5.927556467056275,
      "epoch": 2.2929176969268807,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004472627975327293,
      "loss": 5.3299,
      "mean_token_accuracy": 0.20798514932394027,
      "num_tokens": 48419195.0,
      "step": 25965
    },
    {
      "entropy": 5.97014627456665,
      "epoch": 2.2933592370187212,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004472425372932506,
      "loss": 5.3744,
      "mean_token_accuracy": 0.205880506336689,
      "num_tokens": 48430637.0,
      "step": 25970
    },
    {
      "entropy": 5.794405126571656,
      "epoch": 2.2938007771105617,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004472222736796386,
      "loss": 5.1628,
      "mean_token_accuracy": 0.21999103128910064,
      "num_tokens": 48439929.0,
      "step": 25975
    },
    {
      "entropy": 5.959406042098999,
      "epoch": 2.294242317202402,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004472020066922902,
      "loss": 5.4216,
      "mean_token_accuracy": 0.20152697712183,
      "num_tokens": 48449398.0,
      "step": 25980
    },
    {
      "entropy": 5.979047536849976,
      "epoch": 2.2946838572942423,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004471817363316024,
      "loss": 5.2966,
      "mean_token_accuracy": 0.21314818859100343,
      "num_tokens": 48458843.0,
      "step": 25985
    },
    {
      "entropy": 5.9526755809783936,
      "epoch": 2.295125397386083,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004471614625979723,
      "loss": 5.2409,
      "mean_token_accuracy": 0.21272154301404952,
      "num_tokens": 48468111.0,
      "step": 25990
    },
    {
      "entropy": 5.882488059997558,
      "epoch": 2.295566937477923,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004471411854917972,
      "loss": 5.3028,
      "mean_token_accuracy": 0.21438165307044982,
      "num_tokens": 48477501.0,
      "step": 25995
    },
    {
      "entropy": 6.024596929550171,
      "epoch": 2.2960084775697633,
      "grad_norm": 1.09375,
      "learning_rate": 0.000447120905013474,
      "loss": 5.44,
      "mean_token_accuracy": 0.19502519816160202,
      "num_tokens": 48487576.0,
      "step": 26000
    },
    {
      "entropy": 5.89276614189148,
      "epoch": 2.296450017661604,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004471006211634002,
      "loss": 5.2236,
      "mean_token_accuracy": 0.22616962045431138,
      "num_tokens": 48496230.0,
      "step": 26005
    },
    {
      "entropy": 5.959626388549805,
      "epoch": 2.296891557753444,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044708033394197323,
      "loss": 5.3099,
      "mean_token_accuracy": 0.20844869464635848,
      "num_tokens": 48505871.0,
      "step": 26010
    },
    {
      "entropy": 5.901160430908203,
      "epoch": 2.2973330978452844,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044706004334959024,
      "loss": 5.3114,
      "mean_token_accuracy": 0.2045974016189575,
      "num_tokens": 48515298.0,
      "step": 26015
    },
    {
      "entropy": 5.828893661499023,
      "epoch": 2.297774637937125,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004470397493866489,
      "loss": 5.3616,
      "mean_token_accuracy": 0.20081913471221924,
      "num_tokens": 48524573.0,
      "step": 26020
    },
    {
      "entropy": 5.954742431640625,
      "epoch": 2.298216178028965,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004470194520535467,
      "loss": 5.3858,
      "mean_token_accuracy": 0.20398117303848268,
      "num_tokens": 48533244.0,
      "step": 26025
    },
    {
      "entropy": 5.976409244537353,
      "epoch": 2.2986577181208054,
      "grad_norm": 1.0,
      "learning_rate": 0.0004469991513506814,
      "loss": 5.2495,
      "mean_token_accuracy": 0.2126106098294258,
      "num_tokens": 48542464.0,
      "step": 26030
    },
    {
      "entropy": 5.865573120117188,
      "epoch": 2.299099258212646,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004469788472784504,
      "loss": 5.3244,
      "mean_token_accuracy": 0.21156174093484878,
      "num_tokens": 48551947.0,
      "step": 26035
    },
    {
      "entropy": 5.892144346237183,
      "epoch": 2.299540798304486,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044695853983725163,
      "loss": 5.2394,
      "mean_token_accuracy": 0.2179814860224724,
      "num_tokens": 48561287.0,
      "step": 26040
    },
    {
      "entropy": 5.940769100189209,
      "epoch": 2.2999823383963265,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004469382290274829,
      "loss": 5.3803,
      "mean_token_accuracy": 0.20299497991800308,
      "num_tokens": 48570800.0,
      "step": 26045
    },
    {
      "entropy": 5.924418830871582,
      "epoch": 2.3004238784881665,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044691791484954195,
      "loss": 5.2126,
      "mean_token_accuracy": 0.2202403575181961,
      "num_tokens": 48579705.0,
      "step": 26050
    },
    {
      "entropy": 5.889727973937989,
      "epoch": 2.300865418580007,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000446897597303827,
      "loss": 5.3412,
      "mean_token_accuracy": 0.2094974100589752,
      "num_tokens": 48588793.0,
      "step": 26055
    },
    {
      "entropy": 5.894078111648559,
      "epoch": 2.3013069586718475,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004468772763907357,
      "loss": 5.188,
      "mean_token_accuracy": 0.21566384434700012,
      "num_tokens": 48598361.0,
      "step": 26060
    },
    {
      "entropy": 5.9139727592468265,
      "epoch": 2.3017484987636876,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004468569521106664,
      "loss": 5.3145,
      "mean_token_accuracy": 0.20633782893419267,
      "num_tokens": 48607703.0,
      "step": 26065
    },
    {
      "entropy": 5.889852142333984,
      "epoch": 2.302190038855528,
      "grad_norm": 1.125,
      "learning_rate": 0.0004468366244640171,
      "loss": 5.2644,
      "mean_token_accuracy": 0.21064090579748154,
      "num_tokens": 48616345.0,
      "step": 26070
    },
    {
      "entropy": 5.980006980895996,
      "epoch": 2.3026315789473686,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00044681629345118615,
      "loss": 5.3295,
      "mean_token_accuracy": 0.2057981863617897,
      "num_tokens": 48624907.0,
      "step": 26075
    },
    {
      "entropy": 5.897442245483399,
      "epoch": 2.3030731190392086,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004467959590725716,
      "loss": 5.262,
      "mean_token_accuracy": 0.21115820109844208,
      "num_tokens": 48634766.0,
      "step": 26080
    },
    {
      "entropy": 5.91874737739563,
      "epoch": 2.303514659131049,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044677562132857204,
      "loss": 5.3042,
      "mean_token_accuracy": 0.20883141010999678,
      "num_tokens": 48644321.0,
      "step": 26085
    },
    {
      "entropy": 5.8903004169464115,
      "epoch": 2.303956199222889,
      "grad_norm": 1.21875,
      "learning_rate": 0.00044675528021958567,
      "loss": 5.312,
      "mean_token_accuracy": 0.21108170002698898,
      "num_tokens": 48653507.0,
      "step": 26090
    },
    {
      "entropy": 5.924942922592163,
      "epoch": 2.3043977393147297,
      "grad_norm": 1.078125,
      "learning_rate": 0.000446734935746011,
      "loss": 5.3224,
      "mean_token_accuracy": 0.2184051901102066,
      "num_tokens": 48662186.0,
      "step": 26095
    },
    {
      "entropy": 5.869546031951904,
      "epoch": 2.30483927940657,
      "grad_norm": 1.25,
      "learning_rate": 0.0004467145879082467,
      "loss": 5.2877,
      "mean_token_accuracy": 0.21554356813430786,
      "num_tokens": 48671605.0,
      "step": 26100
    },
    {
      "entropy": 5.82264199256897,
      "epoch": 2.3052808194984102,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004466942367066912,
      "loss": 5.2211,
      "mean_token_accuracy": 0.21485772579908372,
      "num_tokens": 48679873.0,
      "step": 26105
    },
    {
      "entropy": 5.875986337661743,
      "epoch": 2.3057223595902507,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044667388214174336,
      "loss": 5.2074,
      "mean_token_accuracy": 0.2163578301668167,
      "num_tokens": 48688678.0,
      "step": 26110
    },
    {
      "entropy": 5.903117370605469,
      "epoch": 2.3061638996820912,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004466535242138018,
      "loss": 5.3039,
      "mean_token_accuracy": 0.2125939905643463,
      "num_tokens": 48697794.0,
      "step": 26115
    },
    {
      "entropy": 5.9491369247436525,
      "epoch": 2.3066054397739313,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044663316292326526,
      "loss": 5.4288,
      "mean_token_accuracy": 0.2034354031085968,
      "num_tokens": 48706859.0,
      "step": 26120
    },
    {
      "entropy": 5.871827459335327,
      "epoch": 2.307046979865772,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044661279827053267,
      "loss": 5.2511,
      "mean_token_accuracy": 0.22039943486452102,
      "num_tokens": 48716627.0,
      "step": 26125
    },
    {
      "entropy": 5.94290862083435,
      "epoch": 2.3074885199576123,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044659243025600297,
      "loss": 5.3204,
      "mean_token_accuracy": 0.2023373529314995,
      "num_tokens": 48726514.0,
      "step": 26130
    },
    {
      "entropy": 5.884207439422608,
      "epoch": 2.3079300600494523,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044657205888007525,
      "loss": 5.2866,
      "mean_token_accuracy": 0.21925854682922363,
      "num_tokens": 48735546.0,
      "step": 26135
    },
    {
      "entropy": 5.79543981552124,
      "epoch": 2.308371600141293,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004465516841431483,
      "loss": 5.2103,
      "mean_token_accuracy": 0.21801828145980834,
      "num_tokens": 48744719.0,
      "step": 26140
    },
    {
      "entropy": 5.971226692199707,
      "epoch": 2.3088131402331333,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004465313060456216,
      "loss": 5.2802,
      "mean_token_accuracy": 0.21436866670846938,
      "num_tokens": 48754201.0,
      "step": 26145
    },
    {
      "entropy": 5.951769828796387,
      "epoch": 2.3092546803249734,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044651092458789405,
      "loss": 5.3033,
      "mean_token_accuracy": 0.2043192520737648,
      "num_tokens": 48763552.0,
      "step": 26150
    },
    {
      "entropy": 5.84910364151001,
      "epoch": 2.309696220416814,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004464905397703651,
      "loss": 5.2791,
      "mean_token_accuracy": 0.2093621864914894,
      "num_tokens": 48772494.0,
      "step": 26155
    },
    {
      "entropy": 5.858048295974731,
      "epoch": 2.3101377605086544,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044647015159343396,
      "loss": 5.2995,
      "mean_token_accuracy": 0.20694191455841066,
      "num_tokens": 48782121.0,
      "step": 26160
    },
    {
      "entropy": 5.913721656799316,
      "epoch": 2.3105793006004944,
      "grad_norm": 1.234375,
      "learning_rate": 0.00044644976005750003,
      "loss": 5.2368,
      "mean_token_accuracy": 0.2198876455426216,
      "num_tokens": 48792040.0,
      "step": 26165
    },
    {
      "entropy": 5.891841840744019,
      "epoch": 2.311020840692335,
      "grad_norm": 1.125,
      "learning_rate": 0.0004464293651629629,
      "loss": 5.261,
      "mean_token_accuracy": 0.2119569733738899,
      "num_tokens": 48801429.0,
      "step": 26170
    },
    {
      "entropy": 5.940471649169922,
      "epoch": 2.3114623807841754,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044640896691022195,
      "loss": 5.3195,
      "mean_token_accuracy": 0.2093986988067627,
      "num_tokens": 48811185.0,
      "step": 26175
    },
    {
      "entropy": 5.980248498916626,
      "epoch": 2.3119039208760155,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004463885652996768,
      "loss": 5.3611,
      "mean_token_accuracy": 0.2037942513823509,
      "num_tokens": 48820691.0,
      "step": 26180
    },
    {
      "entropy": 5.964035177230835,
      "epoch": 2.312345460967856,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044636816033172716,
      "loss": 5.3715,
      "mean_token_accuracy": 0.20366516411304475,
      "num_tokens": 48829960.0,
      "step": 26185
    },
    {
      "entropy": 5.895146179199219,
      "epoch": 2.312787001059696,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004463477520067728,
      "loss": 5.2499,
      "mean_token_accuracy": 0.21204532384872438,
      "num_tokens": 48839107.0,
      "step": 26190
    },
    {
      "entropy": 5.9575910568237305,
      "epoch": 2.3132285411515365,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004463273403252133,
      "loss": 5.3508,
      "mean_token_accuracy": 0.2099679559469223,
      "num_tokens": 48848430.0,
      "step": 26195
    },
    {
      "entropy": 5.867958211898804,
      "epoch": 2.313670081243377,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004463069252874487,
      "loss": 5.2424,
      "mean_token_accuracy": 0.21630232334136962,
      "num_tokens": 48857869.0,
      "step": 26200
    },
    {
      "entropy": 5.959478378295898,
      "epoch": 2.314111621335217,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044628650689387887,
      "loss": 5.3443,
      "mean_token_accuracy": 0.20582959800958633,
      "num_tokens": 48867730.0,
      "step": 26205
    },
    {
      "entropy": 5.962307596206665,
      "epoch": 2.3145531614270576,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00044626608514490384,
      "loss": 5.3177,
      "mean_token_accuracy": 0.20978791266679764,
      "num_tokens": 48876615.0,
      "step": 26210
    },
    {
      "entropy": 5.934621667861938,
      "epoch": 2.314994701518898,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004462456600409235,
      "loss": 5.3276,
      "mean_token_accuracy": 0.21311743706464767,
      "num_tokens": 48885732.0,
      "step": 26215
    },
    {
      "entropy": 5.945902109146118,
      "epoch": 2.315436241610738,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004462252315823381,
      "loss": 5.2594,
      "mean_token_accuracy": 0.2124856159090996,
      "num_tokens": 48894372.0,
      "step": 26220
    },
    {
      "entropy": 5.9715066909790036,
      "epoch": 2.3158777817025786,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004462047997695478,
      "loss": 5.3978,
      "mean_token_accuracy": 0.2053362786769867,
      "num_tokens": 48904321.0,
      "step": 26225
    },
    {
      "entropy": 5.874537229537964,
      "epoch": 2.3163193217944187,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044618436460295294,
      "loss": 5.328,
      "mean_token_accuracy": 0.20437328964471818,
      "num_tokens": 48914215.0,
      "step": 26230
    },
    {
      "entropy": 5.889763832092285,
      "epoch": 2.316760861886259,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004461639260829535,
      "loss": 5.281,
      "mean_token_accuracy": 0.2028007388114929,
      "num_tokens": 48923383.0,
      "step": 26235
    },
    {
      "entropy": 5.960892152786255,
      "epoch": 2.3172024019780997,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004461434842099503,
      "loss": 5.3207,
      "mean_token_accuracy": 0.20639059990644454,
      "num_tokens": 48933560.0,
      "step": 26240
    },
    {
      "entropy": 5.837919044494629,
      "epoch": 2.3176439420699397,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004461230389843435,
      "loss": 5.2407,
      "mean_token_accuracy": 0.21555689424276353,
      "num_tokens": 48943111.0,
      "step": 26245
    },
    {
      "entropy": 5.88907470703125,
      "epoch": 2.3180854821617802,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004461025904065338,
      "loss": 5.2866,
      "mean_token_accuracy": 0.21421474665403367,
      "num_tokens": 48952470.0,
      "step": 26250
    },
    {
      "entropy": 5.950889253616333,
      "epoch": 2.3185270222536207,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044608213847692154,
      "loss": 5.3453,
      "mean_token_accuracy": 0.20455030649900435,
      "num_tokens": 48961273.0,
      "step": 26255
    },
    {
      "entropy": 5.9744789600372314,
      "epoch": 2.318968562345461,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004460616831959075,
      "loss": 5.3756,
      "mean_token_accuracy": 0.20767430067062378,
      "num_tokens": 48971078.0,
      "step": 26260
    },
    {
      "entropy": 5.938797855377198,
      "epoch": 2.3194101024373013,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004460412245638924,
      "loss": 5.3388,
      "mean_token_accuracy": 0.21122367531061173,
      "num_tokens": 48980711.0,
      "step": 26265
    },
    {
      "entropy": 5.935947322845459,
      "epoch": 2.3198516425291418,
      "grad_norm": 1.09375,
      "learning_rate": 0.000446020762581277,
      "loss": 5.2563,
      "mean_token_accuracy": 0.21150058656930923,
      "num_tokens": 48990195.0,
      "step": 26270
    },
    {
      "entropy": 5.8040759563446045,
      "epoch": 2.320293182620982,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044600029724846206,
      "loss": 5.15,
      "mean_token_accuracy": 0.2249244526028633,
      "num_tokens": 49000271.0,
      "step": 26275
    },
    {
      "entropy": 5.876610994338989,
      "epoch": 2.3207347227128223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004459798285658487,
      "loss": 5.2848,
      "mean_token_accuracy": 0.2026217296719551,
      "num_tokens": 49009624.0,
      "step": 26280
    },
    {
      "entropy": 5.8926310539245605,
      "epoch": 2.321176262804663,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044595935653383756,
      "loss": 5.2677,
      "mean_token_accuracy": 0.21139924526214598,
      "num_tokens": 49018845.0,
      "step": 26285
    },
    {
      "entropy": 5.843112373352051,
      "epoch": 2.321617802896503,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044593888115282995,
      "loss": 5.229,
      "mean_token_accuracy": 0.2147248864173889,
      "num_tokens": 49028352.0,
      "step": 26290
    },
    {
      "entropy": 5.889179420471192,
      "epoch": 2.3220593429883434,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044591840242322686,
      "loss": 5.3083,
      "mean_token_accuracy": 0.21021538376808166,
      "num_tokens": 49038180.0,
      "step": 26295
    },
    {
      "entropy": 5.894275569915772,
      "epoch": 2.322500883080184,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004458979203454295,
      "loss": 5.2238,
      "mean_token_accuracy": 0.2269373804330826,
      "num_tokens": 49048038.0,
      "step": 26300
    },
    {
      "entropy": 5.911482334136963,
      "epoch": 2.322942423172024,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044587743491983903,
      "loss": 5.2979,
      "mean_token_accuracy": 0.21373720616102218,
      "num_tokens": 49057304.0,
      "step": 26305
    },
    {
      "entropy": 5.915894842147827,
      "epoch": 2.3233839632638644,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004458569461468568,
      "loss": 5.3971,
      "mean_token_accuracy": 0.20674056857824324,
      "num_tokens": 49067071.0,
      "step": 26310
    },
    {
      "entropy": 5.865229368209839,
      "epoch": 2.323825503355705,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044583645402688414,
      "loss": 5.3305,
      "mean_token_accuracy": 0.2091332569718361,
      "num_tokens": 49075814.0,
      "step": 26315
    },
    {
      "entropy": 5.980800628662109,
      "epoch": 2.324267043447545,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004458159585603226,
      "loss": 5.402,
      "mean_token_accuracy": 0.20223182886838914,
      "num_tokens": 49085831.0,
      "step": 26320
    },
    {
      "entropy": 5.985586261749267,
      "epoch": 2.3247085835393855,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044579545974757353,
      "loss": 5.3144,
      "mean_token_accuracy": 0.20553084909915925,
      "num_tokens": 49095893.0,
      "step": 26325
    },
    {
      "entropy": 5.854241371154785,
      "epoch": 2.3251501236312255,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004457749575890385,
      "loss": 5.2765,
      "mean_token_accuracy": 0.2071152910590172,
      "num_tokens": 49104810.0,
      "step": 26330
    },
    {
      "entropy": 5.874606037139893,
      "epoch": 2.325591663723066,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004457544520851192,
      "loss": 5.3381,
      "mean_token_accuracy": 0.20311879068613053,
      "num_tokens": 49114108.0,
      "step": 26335
    },
    {
      "entropy": 5.94635796546936,
      "epoch": 2.3260332038149065,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044573394323621735,
      "loss": 5.2883,
      "mean_token_accuracy": 0.2100869044661522,
      "num_tokens": 49123371.0,
      "step": 26340
    },
    {
      "entropy": 5.985779094696045,
      "epoch": 2.3264747439067466,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004457134310427346,
      "loss": 5.3304,
      "mean_token_accuracy": 0.2075203150510788,
      "num_tokens": 49132056.0,
      "step": 26345
    },
    {
      "entropy": 5.931562328338623,
      "epoch": 2.326916283998587,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004456929155050729,
      "loss": 5.3258,
      "mean_token_accuracy": 0.206938473880291,
      "num_tokens": 49141316.0,
      "step": 26350
    },
    {
      "entropy": 5.8523143291473385,
      "epoch": 2.3273578240904276,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044567239662363393,
      "loss": 5.2719,
      "mean_token_accuracy": 0.2166082426905632,
      "num_tokens": 49151337.0,
      "step": 26355
    },
    {
      "entropy": 5.958336544036865,
      "epoch": 2.3277993641822676,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004456518743988199,
      "loss": 5.3066,
      "mean_token_accuracy": 0.20598851144313812,
      "num_tokens": 49160970.0,
      "step": 26360
    },
    {
      "entropy": 5.935576772689819,
      "epoch": 2.328240904274108,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004456313488310326,
      "loss": 5.2792,
      "mean_token_accuracy": 0.2192042887210846,
      "num_tokens": 49170727.0,
      "step": 26365
    },
    {
      "entropy": 5.898619794845581,
      "epoch": 2.328682444365948,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044561081992067425,
      "loss": 5.2957,
      "mean_token_accuracy": 0.21167370080947875,
      "num_tokens": 49179485.0,
      "step": 26370
    },
    {
      "entropy": 5.903181552886963,
      "epoch": 2.3291239844577887,
      "grad_norm": 1.078125,
      "learning_rate": 0.000445590287668147,
      "loss": 5.3931,
      "mean_token_accuracy": 0.2014477387070656,
      "num_tokens": 49189339.0,
      "step": 26375
    },
    {
      "entropy": 5.890514945983886,
      "epoch": 2.329565524549629,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044556975207385305,
      "loss": 5.2362,
      "mean_token_accuracy": 0.21826662868261337,
      "num_tokens": 49198387.0,
      "step": 26380
    },
    {
      "entropy": 5.932420825958252,
      "epoch": 2.3300070646414692,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044554921313819465,
      "loss": 5.3376,
      "mean_token_accuracy": 0.2069243609905243,
      "num_tokens": 49207518.0,
      "step": 26385
    },
    {
      "entropy": 5.832262420654297,
      "epoch": 2.3304486047333097,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004455286708615742,
      "loss": 5.2288,
      "mean_token_accuracy": 0.21882445514202117,
      "num_tokens": 49217031.0,
      "step": 26390
    },
    {
      "entropy": 5.973717975616455,
      "epoch": 2.3308901448251502,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000445508125244394,
      "loss": 5.3063,
      "mean_token_accuracy": 0.21102532297372817,
      "num_tokens": 49225500.0,
      "step": 26395
    },
    {
      "entropy": 5.912530755996704,
      "epoch": 2.3313316849169903,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044548757628705654,
      "loss": 5.2322,
      "mean_token_accuracy": 0.2207881659269333,
      "num_tokens": 49235088.0,
      "step": 26400
    },
    {
      "entropy": 5.84417462348938,
      "epoch": 2.3317732250088308,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004454670239899646,
      "loss": 5.3429,
      "mean_token_accuracy": 0.19991041272878646,
      "num_tokens": 49245597.0,
      "step": 26405
    },
    {
      "entropy": 5.827473831176758,
      "epoch": 2.3322147651006713,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044544646835352047,
      "loss": 5.1475,
      "mean_token_accuracy": 0.22434647381305695,
      "num_tokens": 49254284.0,
      "step": 26410
    },
    {
      "entropy": 5.904017591476441,
      "epoch": 2.3326563051925113,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000445425909378127,
      "loss": 5.3749,
      "mean_token_accuracy": 0.20371006727218627,
      "num_tokens": 49264226.0,
      "step": 26415
    },
    {
      "entropy": 5.859605073928833,
      "epoch": 2.333097845284352,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044540534706418697,
      "loss": 5.3265,
      "mean_token_accuracy": 0.20770861357450485,
      "num_tokens": 49274022.0,
      "step": 26420
    },
    {
      "entropy": 5.953948211669922,
      "epoch": 2.3335393853761923,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000445384781412103,
      "loss": 5.3061,
      "mean_token_accuracy": 0.20780505686998368,
      "num_tokens": 49283376.0,
      "step": 26425
    },
    {
      "entropy": 5.95822057723999,
      "epoch": 2.3339809254680324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000445364212422278,
      "loss": 5.3445,
      "mean_token_accuracy": 0.20842400193214417,
      "num_tokens": 49292994.0,
      "step": 26430
    },
    {
      "entropy": 5.990125274658203,
      "epoch": 2.334422465559873,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004453436400951152,
      "loss": 5.414,
      "mean_token_accuracy": 0.21070096492767335,
      "num_tokens": 49303599.0,
      "step": 26435
    },
    {
      "entropy": 6.002164125442505,
      "epoch": 2.3348640056517134,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004453230644310172,
      "loss": 5.3829,
      "mean_token_accuracy": 0.20825206637382507,
      "num_tokens": 49313397.0,
      "step": 26440
    },
    {
      "entropy": 5.943878889083862,
      "epoch": 2.3353055457435534,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044530248543038725,
      "loss": 5.3043,
      "mean_token_accuracy": 0.21043455004692077,
      "num_tokens": 49323288.0,
      "step": 26445
    },
    {
      "entropy": 5.71579065322876,
      "epoch": 2.335747085835394,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044528190309362853,
      "loss": 5.1433,
      "mean_token_accuracy": 0.2238108515739441,
      "num_tokens": 49332153.0,
      "step": 26450
    },
    {
      "entropy": 5.880008697509766,
      "epoch": 2.3361886259272344,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004452613174211441,
      "loss": 5.3063,
      "mean_token_accuracy": 0.20676835775375366,
      "num_tokens": 49340773.0,
      "step": 26455
    },
    {
      "entropy": 5.866214990615845,
      "epoch": 2.3366301660190745,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044524072841333736,
      "loss": 5.3217,
      "mean_token_accuracy": 0.20998685210943221,
      "num_tokens": 49349863.0,
      "step": 26460
    },
    {
      "entropy": 5.8846845626831055,
      "epoch": 2.337071706110915,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004452201360706115,
      "loss": 5.1621,
      "mean_token_accuracy": 0.22079674452543258,
      "num_tokens": 49358518.0,
      "step": 26465
    },
    {
      "entropy": 5.891619682312012,
      "epoch": 2.337513246202755,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004451995403933701,
      "loss": 5.2496,
      "mean_token_accuracy": 0.2173308476805687,
      "num_tokens": 49367509.0,
      "step": 26470
    },
    {
      "entropy": 5.885852527618408,
      "epoch": 2.3379547862945955,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004451789413820164,
      "loss": 5.3115,
      "mean_token_accuracy": 0.21425034403800963,
      "num_tokens": 49377157.0,
      "step": 26475
    },
    {
      "entropy": 5.943954420089722,
      "epoch": 2.338396326386436,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004451583390369541,
      "loss": 5.3486,
      "mean_token_accuracy": 0.20654962211847305,
      "num_tokens": 49386580.0,
      "step": 26480
    },
    {
      "entropy": 5.85311861038208,
      "epoch": 2.338837866478276,
      "grad_norm": 1.125,
      "learning_rate": 0.0004451377333585866,
      "loss": 5.2346,
      "mean_token_accuracy": 0.21684196591377258,
      "num_tokens": 49395894.0,
      "step": 26485
    },
    {
      "entropy": 5.826545286178589,
      "epoch": 2.3392794065701166,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044511712434731774,
      "loss": 5.2746,
      "mean_token_accuracy": 0.2159070134162903,
      "num_tokens": 49405475.0,
      "step": 26490
    },
    {
      "entropy": 5.9277873039245605,
      "epoch": 2.339720946661957,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004450965120035511,
      "loss": 5.297,
      "mean_token_accuracy": 0.2140595495700836,
      "num_tokens": 49414982.0,
      "step": 26495
    },
    {
      "entropy": 5.861628770828247,
      "epoch": 2.340162486753797,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004450758963276907,
      "loss": 5.2433,
      "mean_token_accuracy": 0.21881699711084365,
      "num_tokens": 49425502.0,
      "step": 26500
    },
    {
      "entropy": 5.864910364151001,
      "epoch": 2.3406040268456376,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044505527732014,
      "loss": 5.2322,
      "mean_token_accuracy": 0.211099337041378,
      "num_tokens": 49435278.0,
      "step": 26505
    },
    {
      "entropy": 5.856877279281616,
      "epoch": 2.3410455669374777,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004450346549813033,
      "loss": 5.2796,
      "mean_token_accuracy": 0.21108997315168382,
      "num_tokens": 49444505.0,
      "step": 26510
    },
    {
      "entropy": 5.871217346191406,
      "epoch": 2.341487107029318,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044501402931158433,
      "loss": 5.2587,
      "mean_token_accuracy": 0.21675732880830764,
      "num_tokens": 49453485.0,
      "step": 26515
    },
    {
      "entropy": 5.868362808227539,
      "epoch": 2.3419286471211587,
      "grad_norm": 1.125,
      "learning_rate": 0.0004449934003113872,
      "loss": 5.223,
      "mean_token_accuracy": 0.21355945467948914,
      "num_tokens": 49462319.0,
      "step": 26520
    },
    {
      "entropy": 5.894816637039185,
      "epoch": 2.3423701872129987,
      "grad_norm": 1.125,
      "learning_rate": 0.0004449727679811161,
      "loss": 5.2638,
      "mean_token_accuracy": 0.21920399963855744,
      "num_tokens": 49470949.0,
      "step": 26525
    },
    {
      "entropy": 5.891136217117309,
      "epoch": 2.342811727304839,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004449521323211751,
      "loss": 5.2676,
      "mean_token_accuracy": 0.21321358233690263,
      "num_tokens": 49480269.0,
      "step": 26530
    },
    {
      "entropy": 5.891949939727783,
      "epoch": 2.3432532673966797,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004449314933319685,
      "loss": 5.276,
      "mean_token_accuracy": 0.21341665536165239,
      "num_tokens": 49489364.0,
      "step": 26535
    },
    {
      "entropy": 5.95717420578003,
      "epoch": 2.3436948074885198,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044491085101390065,
      "loss": 5.2907,
      "mean_token_accuracy": 0.21524744033813475,
      "num_tokens": 49499360.0,
      "step": 26540
    },
    {
      "entropy": 5.87869930267334,
      "epoch": 2.3441363475803603,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004448902053673758,
      "loss": 5.2151,
      "mean_token_accuracy": 0.21302966475486756,
      "num_tokens": 49507770.0,
      "step": 26545
    },
    {
      "entropy": 5.7908093452453615,
      "epoch": 2.3445778876722008,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044486955639279845,
      "loss": 5.3046,
      "mean_token_accuracy": 0.2130023419857025,
      "num_tokens": 49516824.0,
      "step": 26550
    },
    {
      "entropy": 5.848095989227295,
      "epoch": 2.345019427764041,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004448489040905731,
      "loss": 5.288,
      "mean_token_accuracy": 0.2065812900662422,
      "num_tokens": 49525176.0,
      "step": 26555
    },
    {
      "entropy": 5.945406532287597,
      "epoch": 2.3454609678558813,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004448282484611043,
      "loss": 5.3914,
      "mean_token_accuracy": 0.20405150800943375,
      "num_tokens": 49535701.0,
      "step": 26560
    },
    {
      "entropy": 5.916968822479248,
      "epoch": 2.345902507947722,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044480758950479673,
      "loss": 5.3,
      "mean_token_accuracy": 0.2253708213567734,
      "num_tokens": 49545473.0,
      "step": 26565
    },
    {
      "entropy": 5.832232427597046,
      "epoch": 2.346344048039562,
      "grad_norm": 1.125,
      "learning_rate": 0.00044478692722205507,
      "loss": 5.2189,
      "mean_token_accuracy": 0.21153826713562013,
      "num_tokens": 49553955.0,
      "step": 26570
    },
    {
      "entropy": 5.882525444030762,
      "epoch": 2.3467855881314024,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000444766261613284,
      "loss": 5.3385,
      "mean_token_accuracy": 0.20648249983787537,
      "num_tokens": 49563305.0,
      "step": 26575
    },
    {
      "entropy": 5.914910078048706,
      "epoch": 2.347227128223243,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00044474559267888847,
      "loss": 5.3204,
      "mean_token_accuracy": 0.2100345715880394,
      "num_tokens": 49571841.0,
      "step": 26580
    },
    {
      "entropy": 5.906492614746094,
      "epoch": 2.347668668315083,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004447249204192733,
      "loss": 5.2841,
      "mean_token_accuracy": 0.2114686354994774,
      "num_tokens": 49582862.0,
      "step": 26585
    },
    {
      "entropy": 5.954437923431397,
      "epoch": 2.3481102084069234,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004447042448348435,
      "loss": 5.344,
      "mean_token_accuracy": 0.20796673893928527,
      "num_tokens": 49592410.0,
      "step": 26590
    },
    {
      "entropy": 5.903286790847778,
      "epoch": 2.348551748498764,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044468356592600405,
      "loss": 5.2978,
      "mean_token_accuracy": 0.217125041782856,
      "num_tokens": 49600582.0,
      "step": 26595
    },
    {
      "entropy": 5.8548342227935795,
      "epoch": 2.348993288590604,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044466288369316006,
      "loss": 5.2796,
      "mean_token_accuracy": 0.21384060382843018,
      "num_tokens": 49610165.0,
      "step": 26600
    },
    {
      "entropy": 5.838822698593139,
      "epoch": 2.3494348286824445,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044464219813671664,
      "loss": 5.2901,
      "mean_token_accuracy": 0.22041625827550887,
      "num_tokens": 49619231.0,
      "step": 26605
    },
    {
      "entropy": 5.875089073181153,
      "epoch": 2.3498763687742845,
      "grad_norm": 1.15625,
      "learning_rate": 0.000444621509257079,
      "loss": 5.1968,
      "mean_token_accuracy": 0.21730110347270964,
      "num_tokens": 49627899.0,
      "step": 26610
    },
    {
      "entropy": 5.93037896156311,
      "epoch": 2.350317908866125,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044460081705465247,
      "loss": 5.3638,
      "mean_token_accuracy": 0.1980777531862259,
      "num_tokens": 49637866.0,
      "step": 26615
    },
    {
      "entropy": 5.964282608032226,
      "epoch": 2.3507594489579655,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044458012152984253,
      "loss": 5.3392,
      "mean_token_accuracy": 0.2110477700829506,
      "num_tokens": 49648273.0,
      "step": 26620
    },
    {
      "entropy": 5.955967378616333,
      "epoch": 2.3512009890498056,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004445594226830544,
      "loss": 5.3139,
      "mean_token_accuracy": 0.20588536113500594,
      "num_tokens": 49657944.0,
      "step": 26625
    },
    {
      "entropy": 5.939783000946045,
      "epoch": 2.351642529141646,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044453872051469345,
      "loss": 5.3178,
      "mean_token_accuracy": 0.21336936950683594,
      "num_tokens": 49667158.0,
      "step": 26630
    },
    {
      "entropy": 5.8636696338653564,
      "epoch": 2.3520840692334866,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044451801502516555,
      "loss": 5.2274,
      "mean_token_accuracy": 0.21531042009592055,
      "num_tokens": 49675535.0,
      "step": 26635
    },
    {
      "entropy": 5.888688945770264,
      "epoch": 2.3525256093253266,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004444973062148761,
      "loss": 5.3121,
      "mean_token_accuracy": 0.20918356031179428,
      "num_tokens": 49684922.0,
      "step": 26640
    },
    {
      "entropy": 5.8577906608581545,
      "epoch": 2.352967149417167,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004444765940842308,
      "loss": 5.2419,
      "mean_token_accuracy": 0.21934637576341628,
      "num_tokens": 49695302.0,
      "step": 26645
    },
    {
      "entropy": 5.891083431243897,
      "epoch": 2.3534086895090076,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004444558786336355,
      "loss": 5.3084,
      "mean_token_accuracy": 0.21281784772872925,
      "num_tokens": 49704668.0,
      "step": 26650
    },
    {
      "entropy": 5.853655529022217,
      "epoch": 2.3538502296008477,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004444351598634958,
      "loss": 5.2457,
      "mean_token_accuracy": 0.21667134761810303,
      "num_tokens": 49713704.0,
      "step": 26655
    },
    {
      "entropy": 5.905678462982178,
      "epoch": 2.354291769692688,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004444144377742178,
      "loss": 5.3267,
      "mean_token_accuracy": 0.2111634835600853,
      "num_tokens": 49723743.0,
      "step": 26660
    },
    {
      "entropy": 5.967500114440918,
      "epoch": 2.354733309784528,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004443937123662071,
      "loss": 5.3906,
      "mean_token_accuracy": 0.20422385782003402,
      "num_tokens": 49734760.0,
      "step": 26665
    },
    {
      "entropy": 5.878413343429566,
      "epoch": 2.3551748498763687,
      "grad_norm": 2.8125,
      "learning_rate": 0.0004443729836398702,
      "loss": 5.2663,
      "mean_token_accuracy": 0.2070953667163849,
      "num_tokens": 49744836.0,
      "step": 26670
    },
    {
      "entropy": 5.8912896633148195,
      "epoch": 2.355616389968209,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044435225159561263,
      "loss": 5.2528,
      "mean_token_accuracy": 0.2180324152112007,
      "num_tokens": 49753646.0,
      "step": 26675
    },
    {
      "entropy": 5.88151707649231,
      "epoch": 2.3560579300600493,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004443315162338409,
      "loss": 5.2941,
      "mean_token_accuracy": 0.21318313479423523,
      "num_tokens": 49762576.0,
      "step": 26680
    },
    {
      "entropy": 5.943893098831177,
      "epoch": 2.3564994701518898,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000444310777554961,
      "loss": 5.2926,
      "mean_token_accuracy": 0.20822067260742189,
      "num_tokens": 49772614.0,
      "step": 26685
    },
    {
      "entropy": 5.859329748153686,
      "epoch": 2.3569410102437303,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004442900355593793,
      "loss": 5.2595,
      "mean_token_accuracy": 0.21090115755796432,
      "num_tokens": 49781005.0,
      "step": 26690
    },
    {
      "entropy": 5.828287792205811,
      "epoch": 2.3573825503355703,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044426929024750204,
      "loss": 5.2705,
      "mean_token_accuracy": 0.21130183190107346,
      "num_tokens": 49790639.0,
      "step": 26695
    },
    {
      "entropy": 5.938632488250732,
      "epoch": 2.357824090427411,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004442485416197357,
      "loss": 5.3764,
      "mean_token_accuracy": 0.20766025334596633,
      "num_tokens": 49800010.0,
      "step": 26700
    },
    {
      "entropy": 5.926017236709595,
      "epoch": 2.3582656305192513,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044422778967648675,
      "loss": 5.2738,
      "mean_token_accuracy": 0.20750439614057542,
      "num_tokens": 49808510.0,
      "step": 26705
    },
    {
      "entropy": 5.874882411956787,
      "epoch": 2.3587071706110914,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044420703441816153,
      "loss": 5.3053,
      "mean_token_accuracy": 0.2100327119231224,
      "num_tokens": 49818017.0,
      "step": 26710
    },
    {
      "entropy": 5.845489645004273,
      "epoch": 2.359148710702932,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004441862758451668,
      "loss": 5.279,
      "mean_token_accuracy": 0.21044786274433136,
      "num_tokens": 49826797.0,
      "step": 26715
    },
    {
      "entropy": 5.910222911834717,
      "epoch": 2.3595902507947724,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044416551395790916,
      "loss": 5.2785,
      "mean_token_accuracy": 0.20705964863300325,
      "num_tokens": 49835632.0,
      "step": 26720
    },
    {
      "entropy": 5.903456354141236,
      "epoch": 2.3600317908866124,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004441447487567953,
      "loss": 5.3059,
      "mean_token_accuracy": 0.21210643649101257,
      "num_tokens": 49844770.0,
      "step": 26725
    },
    {
      "entropy": 5.877901554107666,
      "epoch": 2.360473330978453,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044412398024223194,
      "loss": 5.3167,
      "mean_token_accuracy": 0.21286253929138182,
      "num_tokens": 49854170.0,
      "step": 26730
    },
    {
      "entropy": 5.969999217987061,
      "epoch": 2.3609148710702934,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004441032084146261,
      "loss": 5.338,
      "mean_token_accuracy": 0.20706886053085327,
      "num_tokens": 49863541.0,
      "step": 26735
    },
    {
      "entropy": 5.947088289260864,
      "epoch": 2.3613564111621335,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044408243327438446,
      "loss": 5.2531,
      "mean_token_accuracy": 0.2140100210905075,
      "num_tokens": 49872556.0,
      "step": 26740
    },
    {
      "entropy": 5.900307989120483,
      "epoch": 2.361797951253974,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044406165482191424,
      "loss": 5.3271,
      "mean_token_accuracy": 0.21021255254745483,
      "num_tokens": 49882188.0,
      "step": 26745
    },
    {
      "entropy": 5.9232886791229244,
      "epoch": 2.362239491345814,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004440408730576223,
      "loss": 5.3914,
      "mean_token_accuracy": 0.20298184603452682,
      "num_tokens": 49891980.0,
      "step": 26750
    },
    {
      "entropy": 5.940526247024536,
      "epoch": 2.3626810314376545,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044402008798191575,
      "loss": 5.2782,
      "mean_token_accuracy": 0.2152009680867195,
      "num_tokens": 49900783.0,
      "step": 26755
    },
    {
      "entropy": 5.854488849639893,
      "epoch": 2.363122571529495,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004439992995952019,
      "loss": 5.317,
      "mean_token_accuracy": 0.21335611790418624,
      "num_tokens": 49910525.0,
      "step": 26760
    },
    {
      "entropy": 5.903363609313965,
      "epoch": 2.363564111621335,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044397850789788774,
      "loss": 5.3084,
      "mean_token_accuracy": 0.20832739770412445,
      "num_tokens": 49919286.0,
      "step": 26765
    },
    {
      "entropy": 5.925664663314819,
      "epoch": 2.3640056517131756,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004439577128903808,
      "loss": 5.2568,
      "mean_token_accuracy": 0.21320044696331025,
      "num_tokens": 49927882.0,
      "step": 26770
    },
    {
      "entropy": 5.95701413154602,
      "epoch": 2.364447191805016,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004439369145730884,
      "loss": 5.307,
      "mean_token_accuracy": 0.20887205749750137,
      "num_tokens": 49936289.0,
      "step": 26775
    },
    {
      "entropy": 5.903356981277466,
      "epoch": 2.364888731896856,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004439161129464179,
      "loss": 5.3161,
      "mean_token_accuracy": 0.19926695972681047,
      "num_tokens": 49945356.0,
      "step": 26780
    },
    {
      "entropy": 5.908850574493409,
      "epoch": 2.3653302719886966,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044389530801077675,
      "loss": 5.2252,
      "mean_token_accuracy": 0.2146916702389717,
      "num_tokens": 49954216.0,
      "step": 26785
    },
    {
      "entropy": 5.960436105728149,
      "epoch": 2.365771812080537,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004438744997665726,
      "loss": 5.3547,
      "mean_token_accuracy": 0.20617856830358505,
      "num_tokens": 49963179.0,
      "step": 26790
    },
    {
      "entropy": 5.909926271438598,
      "epoch": 2.366213352172377,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004438536882142131,
      "loss": 5.3587,
      "mean_token_accuracy": 0.20382177382707595,
      "num_tokens": 49973108.0,
      "step": 26795
    },
    {
      "entropy": 5.912693452835083,
      "epoch": 2.3666548922642177,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004438328733541059,
      "loss": 5.2506,
      "mean_token_accuracy": 0.21181951016187667,
      "num_tokens": 49982003.0,
      "step": 26800
    },
    {
      "entropy": 5.870848941802978,
      "epoch": 2.3670964323560577,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004438120551866586,
      "loss": 5.2925,
      "mean_token_accuracy": 0.20937933623790742,
      "num_tokens": 49992034.0,
      "step": 26805
    },
    {
      "entropy": 5.828803634643554,
      "epoch": 2.367537972447898,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004437912337122792,
      "loss": 5.291,
      "mean_token_accuracy": 0.21592638045549392,
      "num_tokens": 50001379.0,
      "step": 26810
    },
    {
      "entropy": 5.945402050018311,
      "epoch": 2.3679795125397387,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004437704089313756,
      "loss": 5.3812,
      "mean_token_accuracy": 0.20421904921531678,
      "num_tokens": 50010721.0,
      "step": 26815
    },
    {
      "entropy": 5.942852115631103,
      "epoch": 2.3684210526315788,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044374958084435564,
      "loss": 5.363,
      "mean_token_accuracy": 0.210134793817997,
      "num_tokens": 50019571.0,
      "step": 26820
    },
    {
      "entropy": 5.925140047073365,
      "epoch": 2.3688625927234193,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044372874945162734,
      "loss": 5.3012,
      "mean_token_accuracy": 0.21074575036764145,
      "num_tokens": 50028045.0,
      "step": 26825
    },
    {
      "entropy": 5.942230081558227,
      "epoch": 2.3693041328152598,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044370791475359877,
      "loss": 5.2767,
      "mean_token_accuracy": 0.2164221227169037,
      "num_tokens": 50036769.0,
      "step": 26830
    },
    {
      "entropy": 5.896571350097656,
      "epoch": 2.3697456729071,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044368707675067814,
      "loss": 5.3065,
      "mean_token_accuracy": 0.2126169517636299,
      "num_tokens": 50045973.0,
      "step": 26835
    },
    {
      "entropy": 5.834485340118408,
      "epoch": 2.3701872129989403,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004436662354432736,
      "loss": 5.2297,
      "mean_token_accuracy": 0.2074983149766922,
      "num_tokens": 50055243.0,
      "step": 26840
    },
    {
      "entropy": 5.943623208999634,
      "epoch": 2.370628753090781,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044364539083179347,
      "loss": 5.3003,
      "mean_token_accuracy": 0.21723183244466782,
      "num_tokens": 50064573.0,
      "step": 26845
    },
    {
      "entropy": 5.822544050216675,
      "epoch": 2.371070293182621,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000443624542916646,
      "loss": 5.1537,
      "mean_token_accuracy": 0.2214773789048195,
      "num_tokens": 50073578.0,
      "step": 26850
    },
    {
      "entropy": 5.8752302646636965,
      "epoch": 2.3715118332744614,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004436036916982397,
      "loss": 5.2996,
      "mean_token_accuracy": 0.2183082014322281,
      "num_tokens": 50083620.0,
      "step": 26855
    },
    {
      "entropy": 5.857642459869385,
      "epoch": 2.371953373366302,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004435828371769829,
      "loss": 5.2327,
      "mean_token_accuracy": 0.2220786392688751,
      "num_tokens": 50092050.0,
      "step": 26860
    },
    {
      "entropy": 5.918768882751465,
      "epoch": 2.372394913458142,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044356197935328426,
      "loss": 5.3242,
      "mean_token_accuracy": 0.20807345062494279,
      "num_tokens": 50102425.0,
      "step": 26865
    },
    {
      "entropy": 5.88598861694336,
      "epoch": 2.3728364535499824,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044354111822755225,
      "loss": 5.226,
      "mean_token_accuracy": 0.21028743386268617,
      "num_tokens": 50112338.0,
      "step": 26870
    },
    {
      "entropy": 5.932241296768188,
      "epoch": 2.373277993641823,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044352025380019563,
      "loss": 5.2958,
      "mean_token_accuracy": 0.21172044426202774,
      "num_tokens": 50121302.0,
      "step": 26875
    },
    {
      "entropy": 5.817448234558105,
      "epoch": 2.373719533733663,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044349938607162305,
      "loss": 5.2381,
      "mean_token_accuracy": 0.21841894388198851,
      "num_tokens": 50129820.0,
      "step": 26880
    },
    {
      "entropy": 5.915347385406494,
      "epoch": 2.3741610738255035,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044347851504224336,
      "loss": 5.3644,
      "mean_token_accuracy": 0.20760912746191024,
      "num_tokens": 50138860.0,
      "step": 26885
    },
    {
      "entropy": 5.894362831115723,
      "epoch": 2.3746026139173435,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044345764071246536,
      "loss": 5.2826,
      "mean_token_accuracy": 0.2128836765885353,
      "num_tokens": 50148257.0,
      "step": 26890
    },
    {
      "entropy": 5.845790719985962,
      "epoch": 2.375044154009184,
      "grad_norm": 1.125,
      "learning_rate": 0.00044343676308269796,
      "loss": 5.2711,
      "mean_token_accuracy": 0.2171759709715843,
      "num_tokens": 50157380.0,
      "step": 26895
    },
    {
      "entropy": 5.970784759521484,
      "epoch": 2.3754856941010245,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044341588215335016,
      "loss": 5.3054,
      "mean_token_accuracy": 0.20860047340393068,
      "num_tokens": 50167806.0,
      "step": 26900
    },
    {
      "entropy": 5.986636066436768,
      "epoch": 2.3759272341928646,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044339499792483104,
      "loss": 5.3353,
      "mean_token_accuracy": 0.20293398052453995,
      "num_tokens": 50176763.0,
      "step": 26905
    },
    {
      "entropy": 5.816183471679688,
      "epoch": 2.376368774284705,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004433741103975498,
      "loss": 5.3014,
      "mean_token_accuracy": 0.21069140285253524,
      "num_tokens": 50186735.0,
      "step": 26910
    },
    {
      "entropy": 5.8311268329620365,
      "epoch": 2.3768103143765456,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044335321957191535,
      "loss": 5.2535,
      "mean_token_accuracy": 0.21647219210863114,
      "num_tokens": 50195206.0,
      "step": 26915
    },
    {
      "entropy": 5.972170972824097,
      "epoch": 2.3772518544683856,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044333232544833707,
      "loss": 5.3435,
      "mean_token_accuracy": 0.20145729631185533,
      "num_tokens": 50203706.0,
      "step": 26920
    },
    {
      "entropy": 5.909353971481323,
      "epoch": 2.377693394560226,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004433114280272243,
      "loss": 5.3159,
      "mean_token_accuracy": 0.20624017119407653,
      "num_tokens": 50213007.0,
      "step": 26925
    },
    {
      "entropy": 5.921986865997314,
      "epoch": 2.3781349346520666,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004432905273089864,
      "loss": 5.2271,
      "mean_token_accuracy": 0.21782964169979097,
      "num_tokens": 50223155.0,
      "step": 26930
    },
    {
      "entropy": 5.795108938217163,
      "epoch": 2.3785764747439067,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044326962329403275,
      "loss": 5.161,
      "mean_token_accuracy": 0.22062039226293564,
      "num_tokens": 50231788.0,
      "step": 26935
    },
    {
      "entropy": 5.8474769115448,
      "epoch": 2.379018014835747,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044324871598277294,
      "loss": 5.2554,
      "mean_token_accuracy": 0.21597664207220077,
      "num_tokens": 50241557.0,
      "step": 26940
    },
    {
      "entropy": 5.905368709564209,
      "epoch": 2.379459554927587,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044322780537561643,
      "loss": 5.3213,
      "mean_token_accuracy": 0.21200947612524032,
      "num_tokens": 50252141.0,
      "step": 26945
    },
    {
      "entropy": 5.933778619766235,
      "epoch": 2.3799010950194277,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004432068914729729,
      "loss": 5.358,
      "mean_token_accuracy": 0.2091843694448471,
      "num_tokens": 50261361.0,
      "step": 26950
    },
    {
      "entropy": 5.907635068893432,
      "epoch": 2.380342635111268,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00044318597427525197,
      "loss": 5.3747,
      "mean_token_accuracy": 0.21069760769605636,
      "num_tokens": 50270193.0,
      "step": 26955
    },
    {
      "entropy": 5.977096796035767,
      "epoch": 2.3807841752031083,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044316505378286356,
      "loss": 5.3233,
      "mean_token_accuracy": 0.20597683787345886,
      "num_tokens": 50280217.0,
      "step": 26960
    },
    {
      "entropy": 5.882652139663696,
      "epoch": 2.3812257152949488,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004431441299962173,
      "loss": 5.2369,
      "mean_token_accuracy": 0.21186361312866211,
      "num_tokens": 50289610.0,
      "step": 26965
    },
    {
      "entropy": 5.848551607131958,
      "epoch": 2.3816672553867893,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004431232029157232,
      "loss": 5.2106,
      "mean_token_accuracy": 0.21853208541870117,
      "num_tokens": 50298734.0,
      "step": 26970
    },
    {
      "entropy": 5.852073717117309,
      "epoch": 2.3821087954786293,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004431022725417912,
      "loss": 5.2288,
      "mean_token_accuracy": 0.21615064442157744,
      "num_tokens": 50307679.0,
      "step": 26975
    },
    {
      "entropy": 5.928052520751953,
      "epoch": 2.38255033557047,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004430813388748312,
      "loss": 5.2182,
      "mean_token_accuracy": 0.21318549811840057,
      "num_tokens": 50315983.0,
      "step": 26980
    },
    {
      "entropy": 5.951604557037354,
      "epoch": 2.3829918756623103,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00044306040191525347,
      "loss": 5.3343,
      "mean_token_accuracy": 0.2079811379313469,
      "num_tokens": 50325222.0,
      "step": 26985
    },
    {
      "entropy": 5.83290319442749,
      "epoch": 2.3834334157541504,
      "grad_norm": 1.1875,
      "learning_rate": 0.000443039461663468,
      "loss": 5.2398,
      "mean_token_accuracy": 0.21699571907520293,
      "num_tokens": 50333417.0,
      "step": 26990
    },
    {
      "entropy": 5.878643608093261,
      "epoch": 2.383874955845991,
      "grad_norm": 1.265625,
      "learning_rate": 0.00044301851811988504,
      "loss": 5.2488,
      "mean_token_accuracy": 0.22254746854305268,
      "num_tokens": 50342836.0,
      "step": 26995
    },
    {
      "entropy": 5.94852123260498,
      "epoch": 2.3843164959378313,
      "grad_norm": 1.171875,
      "learning_rate": 0.00044299757128491487,
      "loss": 5.3788,
      "mean_token_accuracy": 0.20585856586694717,
      "num_tokens": 50351988.0,
      "step": 27000
    },
    {
      "epoch": 2.3843164959378313,
      "eval_entropy": 5.618579518889517,
      "eval_loss": 5.360966682434082,
      "eval_mean_token_accuracy": 0.2160217850480655,
      "eval_num_tokens": 50351988.0,
      "eval_runtime": 20.2811,
      "eval_samples_per_second": 1741.276,
      "eval_steps_per_second": 217.69,
      "step": 27000
    }
  ],
  "logging_steps": 5,
  "max_steps": 113230,
  "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": 1.105320446631936e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}