1035 lines
27 KiB
JSON
1035 lines
27 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.010999890001099988,
|
|
"eval_steps": 500,
|
|
"global_step": 500,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 10.692012214660645,
|
|
"epoch": 0.00010999890001099989,
|
|
"grad_norm": 13.5625,
|
|
"learning_rate": 2e-06,
|
|
"loss": 10.7924,
|
|
"mean_token_accuracy": 0.0,
|
|
"num_tokens": 7960.0,
|
|
"step": 5
|
|
},
|
|
{
|
|
"entropy": 10.691990089416503,
|
|
"epoch": 0.00021999780002199978,
|
|
"grad_norm": 12.875,
|
|
"learning_rate": 4.5e-06,
|
|
"loss": 10.7105,
|
|
"mean_token_accuracy": 0.0001712328754365444,
|
|
"num_tokens": 15356.0,
|
|
"step": 10
|
|
},
|
|
{
|
|
"entropy": 10.691377449035645,
|
|
"epoch": 0.00032999670003299966,
|
|
"grad_norm": 10.4375,
|
|
"learning_rate": 7e-06,
|
|
"loss": 10.492,
|
|
"mean_token_accuracy": 0.014663392491638661,
|
|
"num_tokens": 22887.0,
|
|
"step": 15
|
|
},
|
|
{
|
|
"entropy": 10.685822200775146,
|
|
"epoch": 0.00043999560004399956,
|
|
"grad_norm": 6.90625,
|
|
"learning_rate": 9.5e-06,
|
|
"loss": 10.2232,
|
|
"mean_token_accuracy": 0.03656770326197147,
|
|
"num_tokens": 31706.0,
|
|
"step": 20
|
|
},
|
|
{
|
|
"entropy": 10.64632453918457,
|
|
"epoch": 0.0005499945000549994,
|
|
"grad_norm": 4.65625,
|
|
"learning_rate": 1.2e-05,
|
|
"loss": 9.9293,
|
|
"mean_token_accuracy": 0.04357381202280521,
|
|
"num_tokens": 40130.0,
|
|
"step": 25
|
|
},
|
|
{
|
|
"entropy": 10.607057476043702,
|
|
"epoch": 0.0006599934000659993,
|
|
"grad_norm": 3.65625,
|
|
"learning_rate": 1.4500000000000002e-05,
|
|
"loss": 9.7483,
|
|
"mean_token_accuracy": 0.040078548900783065,
|
|
"num_tokens": 48745.0,
|
|
"step": 30
|
|
},
|
|
{
|
|
"entropy": 10.614838027954102,
|
|
"epoch": 0.0007699923000769992,
|
|
"grad_norm": 3.0625,
|
|
"learning_rate": 1.7000000000000003e-05,
|
|
"loss": 9.7073,
|
|
"mean_token_accuracy": 0.03877733238041401,
|
|
"num_tokens": 58411.0,
|
|
"step": 35
|
|
},
|
|
{
|
|
"entropy": 10.61097354888916,
|
|
"epoch": 0.0008799912000879991,
|
|
"grad_norm": 2.8125,
|
|
"learning_rate": 1.95e-05,
|
|
"loss": 9.6595,
|
|
"mean_token_accuracy": 0.039670221880078314,
|
|
"num_tokens": 66944.0,
|
|
"step": 40
|
|
},
|
|
{
|
|
"entropy": 10.603102111816407,
|
|
"epoch": 0.000989990100098999,
|
|
"grad_norm": 2.6875,
|
|
"learning_rate": 2.2e-05,
|
|
"loss": 9.5761,
|
|
"mean_token_accuracy": 0.037658361718058586,
|
|
"num_tokens": 74600.0,
|
|
"step": 45
|
|
},
|
|
{
|
|
"entropy": 10.603211784362793,
|
|
"epoch": 0.0010999890001099988,
|
|
"grad_norm": 2.65625,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.5233,
|
|
"mean_token_accuracy": 0.0407950758934021,
|
|
"num_tokens": 82355.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 10.586796092987061,
|
|
"epoch": 0.0012099879001209988,
|
|
"grad_norm": 2.671875,
|
|
"learning_rate": 2.7e-05,
|
|
"loss": 9.5392,
|
|
"mean_token_accuracy": 0.03949462212622166,
|
|
"num_tokens": 90949.0,
|
|
"step": 55
|
|
},
|
|
{
|
|
"entropy": 10.55664939880371,
|
|
"epoch": 0.0013199868001319986,
|
|
"grad_norm": 2.46875,
|
|
"learning_rate": 2.95e-05,
|
|
"loss": 9.3671,
|
|
"mean_token_accuracy": 0.056805190071463584,
|
|
"num_tokens": 98932.0,
|
|
"step": 60
|
|
},
|
|
{
|
|
"entropy": 10.420584201812744,
|
|
"epoch": 0.0014299857001429986,
|
|
"grad_norm": 2.453125,
|
|
"learning_rate": 3.2e-05,
|
|
"loss": 9.3337,
|
|
"mean_token_accuracy": 0.053294488415122035,
|
|
"num_tokens": 107729.0,
|
|
"step": 65
|
|
},
|
|
{
|
|
"entropy": 10.395568752288819,
|
|
"epoch": 0.0015399846001539984,
|
|
"grad_norm": 2.625,
|
|
"learning_rate": 3.4500000000000005e-05,
|
|
"loss": 9.2159,
|
|
"mean_token_accuracy": 0.059437951818108556,
|
|
"num_tokens": 116586.0,
|
|
"step": 70
|
|
},
|
|
{
|
|
"entropy": 10.401200771331787,
|
|
"epoch": 0.0016499835001649984,
|
|
"grad_norm": 2.375,
|
|
"learning_rate": 3.7e-05,
|
|
"loss": 9.1694,
|
|
"mean_token_accuracy": 0.05521665029227733,
|
|
"num_tokens": 124571.0,
|
|
"step": 75
|
|
},
|
|
{
|
|
"entropy": 10.463058757781983,
|
|
"epoch": 0.0017599824001759982,
|
|
"grad_norm": 2.421875,
|
|
"learning_rate": 3.95e-05,
|
|
"loss": 9.088,
|
|
"mean_token_accuracy": 0.05925002098083496,
|
|
"num_tokens": 133009.0,
|
|
"step": 80
|
|
},
|
|
{
|
|
"entropy": 10.312766933441162,
|
|
"epoch": 0.001869981300186998,
|
|
"grad_norm": 2.375,
|
|
"learning_rate": 4.2000000000000004e-05,
|
|
"loss": 8.9279,
|
|
"mean_token_accuracy": 0.060802001133561136,
|
|
"num_tokens": 141158.0,
|
|
"step": 85
|
|
},
|
|
{
|
|
"entropy": 10.33760175704956,
|
|
"epoch": 0.001979980200197998,
|
|
"grad_norm": 2.390625,
|
|
"learning_rate": 4.45e-05,
|
|
"loss": 8.9016,
|
|
"mean_token_accuracy": 0.0594728983938694,
|
|
"num_tokens": 148882.0,
|
|
"step": 90
|
|
},
|
|
{
|
|
"entropy": 10.32020616531372,
|
|
"epoch": 0.002089979100208998,
|
|
"grad_norm": 2.328125,
|
|
"learning_rate": 4.7000000000000004e-05,
|
|
"loss": 8.8083,
|
|
"mean_token_accuracy": 0.061231246963143346,
|
|
"num_tokens": 156679.0,
|
|
"step": 95
|
|
},
|
|
{
|
|
"entropy": 10.197657775878906,
|
|
"epoch": 0.0021999780002199976,
|
|
"grad_norm": 2.078125,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.7248,
|
|
"mean_token_accuracy": 0.05933593772351742,
|
|
"num_tokens": 164165.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 10.180267143249512,
|
|
"epoch": 0.002309976900230998,
|
|
"grad_norm": 1.984375,
|
|
"learning_rate": 5.2e-05,
|
|
"loss": 8.6021,
|
|
"mean_token_accuracy": 0.06621009036898613,
|
|
"num_tokens": 172516.0,
|
|
"step": 105
|
|
},
|
|
{
|
|
"entropy": 10.091108131408692,
|
|
"epoch": 0.0024199758002419977,
|
|
"grad_norm": 1.8359375,
|
|
"learning_rate": 5.45e-05,
|
|
"loss": 8.4934,
|
|
"mean_token_accuracy": 0.06615843437612057,
|
|
"num_tokens": 181663.0,
|
|
"step": 110
|
|
},
|
|
{
|
|
"entropy": 10.013619709014893,
|
|
"epoch": 0.0025299747002529975,
|
|
"grad_norm": 1.8125,
|
|
"learning_rate": 5.7e-05,
|
|
"loss": 8.3895,
|
|
"mean_token_accuracy": 0.06955270171165466,
|
|
"num_tokens": 189346.0,
|
|
"step": 115
|
|
},
|
|
{
|
|
"entropy": 9.867325115203858,
|
|
"epoch": 0.0026399736002639972,
|
|
"grad_norm": 1.53125,
|
|
"learning_rate": 5.9499999999999996e-05,
|
|
"loss": 8.351,
|
|
"mean_token_accuracy": 0.0660489596426487,
|
|
"num_tokens": 197631.0,
|
|
"step": 120
|
|
},
|
|
{
|
|
"entropy": 9.711495304107666,
|
|
"epoch": 0.002749972500274997,
|
|
"grad_norm": 1.6484375,
|
|
"learning_rate": 6.2e-05,
|
|
"loss": 8.1405,
|
|
"mean_token_accuracy": 0.07187777906656265,
|
|
"num_tokens": 206443.0,
|
|
"step": 125
|
|
},
|
|
{
|
|
"entropy": 9.555363941192628,
|
|
"epoch": 0.0028599714002859973,
|
|
"grad_norm": 2.265625,
|
|
"learning_rate": 6.450000000000001e-05,
|
|
"loss": 7.9835,
|
|
"mean_token_accuracy": 0.07658364996314049,
|
|
"num_tokens": 213629.0,
|
|
"step": 130
|
|
},
|
|
{
|
|
"entropy": 9.406402778625488,
|
|
"epoch": 0.002969970300296997,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 6.7e-05,
|
|
"loss": 7.9644,
|
|
"mean_token_accuracy": 0.06791270598769188,
|
|
"num_tokens": 221871.0,
|
|
"step": 135
|
|
},
|
|
{
|
|
"entropy": 9.185676288604736,
|
|
"epoch": 0.003079969200307997,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 6.950000000000001e-05,
|
|
"loss": 7.8391,
|
|
"mean_token_accuracy": 0.0753303386271,
|
|
"num_tokens": 229725.0,
|
|
"step": 140
|
|
},
|
|
{
|
|
"entropy": 9.08232479095459,
|
|
"epoch": 0.0031899681003189967,
|
|
"grad_norm": 1.3203125,
|
|
"learning_rate": 7.2e-05,
|
|
"loss": 7.8271,
|
|
"mean_token_accuracy": 0.06317906640470028,
|
|
"num_tokens": 237438.0,
|
|
"step": 145
|
|
},
|
|
{
|
|
"entropy": 8.92593355178833,
|
|
"epoch": 0.003299967000329997,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.7499,
|
|
"mean_token_accuracy": 0.0731364518404007,
|
|
"num_tokens": 245377.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 8.75550880432129,
|
|
"epoch": 0.0034099659003409967,
|
|
"grad_norm": 1.5703125,
|
|
"learning_rate": 7.7e-05,
|
|
"loss": 7.7059,
|
|
"mean_token_accuracy": 0.07082148268818855,
|
|
"num_tokens": 254308.0,
|
|
"step": 155
|
|
},
|
|
{
|
|
"entropy": 8.569324111938476,
|
|
"epoch": 0.0035199648003519965,
|
|
"grad_norm": 2.171875,
|
|
"learning_rate": 7.950000000000001e-05,
|
|
"loss": 7.5912,
|
|
"mean_token_accuracy": 0.0794397659599781,
|
|
"num_tokens": 261964.0,
|
|
"step": 160
|
|
},
|
|
{
|
|
"entropy": 8.406212520599365,
|
|
"epoch": 0.0036299637003629963,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 8.2e-05,
|
|
"loss": 7.5302,
|
|
"mean_token_accuracy": 0.08629432320594788,
|
|
"num_tokens": 270225.0,
|
|
"step": 165
|
|
},
|
|
{
|
|
"entropy": 8.392980098724365,
|
|
"epoch": 0.003739962600373996,
|
|
"grad_norm": 1.8515625,
|
|
"learning_rate": 8.450000000000001e-05,
|
|
"loss": 7.4554,
|
|
"mean_token_accuracy": 0.08913237303495407,
|
|
"num_tokens": 277628.0,
|
|
"step": 170
|
|
},
|
|
{
|
|
"entropy": 8.240210437774659,
|
|
"epoch": 0.0038499615003849963,
|
|
"grad_norm": 1.203125,
|
|
"learning_rate": 8.7e-05,
|
|
"loss": 7.6321,
|
|
"mean_token_accuracy": 0.07898878455162048,
|
|
"num_tokens": 286126.0,
|
|
"step": 175
|
|
},
|
|
{
|
|
"entropy": 8.138569593429565,
|
|
"epoch": 0.003959960400395996,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 8.95e-05,
|
|
"loss": 7.4222,
|
|
"mean_token_accuracy": 0.08429618775844575,
|
|
"num_tokens": 293439.0,
|
|
"step": 180
|
|
},
|
|
{
|
|
"entropy": 8.08548469543457,
|
|
"epoch": 0.004069959300406996,
|
|
"grad_norm": 1.734375,
|
|
"learning_rate": 9.2e-05,
|
|
"loss": 7.4368,
|
|
"mean_token_accuracy": 0.08340740129351616,
|
|
"num_tokens": 302336.0,
|
|
"step": 185
|
|
},
|
|
{
|
|
"entropy": 8.068483829498291,
|
|
"epoch": 0.004179958200417996,
|
|
"grad_norm": 1.5078125,
|
|
"learning_rate": 9.45e-05,
|
|
"loss": 7.4136,
|
|
"mean_token_accuracy": 0.07967994846403599,
|
|
"num_tokens": 309605.0,
|
|
"step": 190
|
|
},
|
|
{
|
|
"entropy": 8.027423477172851,
|
|
"epoch": 0.0042899571004289955,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 9.7e-05,
|
|
"loss": 7.4703,
|
|
"mean_token_accuracy": 0.07479370459914207,
|
|
"num_tokens": 317722.0,
|
|
"step": 195
|
|
},
|
|
{
|
|
"entropy": 7.9662737369537355,
|
|
"epoch": 0.004399956000439995,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 7.3471,
|
|
"mean_token_accuracy": 0.08859865292906761,
|
|
"num_tokens": 325221.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 7.9728742122650145,
|
|
"epoch": 0.004509954900450995,
|
|
"grad_norm": 1.4765625,
|
|
"learning_rate": 0.000102,
|
|
"loss": 7.439,
|
|
"mean_token_accuracy": 0.07891935519874096,
|
|
"num_tokens": 333548.0,
|
|
"step": 205
|
|
},
|
|
{
|
|
"entropy": 7.8386619091033936,
|
|
"epoch": 0.004619953800461996,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 0.00010449999999999999,
|
|
"loss": 7.321,
|
|
"mean_token_accuracy": 0.08747617453336716,
|
|
"num_tokens": 341064.0,
|
|
"step": 210
|
|
},
|
|
{
|
|
"entropy": 7.8650328636169435,
|
|
"epoch": 0.0047299527004729955,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.000107,
|
|
"loss": 7.265,
|
|
"mean_token_accuracy": 0.08942043855786323,
|
|
"num_tokens": 349166.0,
|
|
"step": 215
|
|
},
|
|
{
|
|
"entropy": 7.779710388183593,
|
|
"epoch": 0.004839951600483995,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.0001095,
|
|
"loss": 7.369,
|
|
"mean_token_accuracy": 0.07782569229602813,
|
|
"num_tokens": 358483.0,
|
|
"step": 220
|
|
},
|
|
{
|
|
"entropy": 7.808506011962891,
|
|
"epoch": 0.004949950500494995,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.000112,
|
|
"loss": 7.2969,
|
|
"mean_token_accuracy": 0.08983959630131721,
|
|
"num_tokens": 366738.0,
|
|
"step": 225
|
|
},
|
|
{
|
|
"entropy": 7.680256462097168,
|
|
"epoch": 0.005059949400505995,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.0001145,
|
|
"loss": 7.2971,
|
|
"mean_token_accuracy": 0.08871394842863083,
|
|
"num_tokens": 374438.0,
|
|
"step": 230
|
|
},
|
|
{
|
|
"entropy": 7.785244131088257,
|
|
"epoch": 0.005169948300516995,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.00011700000000000001,
|
|
"loss": 7.251,
|
|
"mean_token_accuracy": 0.09088906422257423,
|
|
"num_tokens": 382721.0,
|
|
"step": 235
|
|
},
|
|
{
|
|
"entropy": 7.663246393203735,
|
|
"epoch": 0.0052799472005279945,
|
|
"grad_norm": 1.4296875,
|
|
"learning_rate": 0.00011949999999999999,
|
|
"loss": 7.2833,
|
|
"mean_token_accuracy": 0.08662524111568928,
|
|
"num_tokens": 391631.0,
|
|
"step": 240
|
|
},
|
|
{
|
|
"entropy": 7.757506418228149,
|
|
"epoch": 0.005389946100538994,
|
|
"grad_norm": 2.46875,
|
|
"learning_rate": 0.000122,
|
|
"loss": 7.2387,
|
|
"mean_token_accuracy": 0.08566439598798752,
|
|
"num_tokens": 400938.0,
|
|
"step": 245
|
|
},
|
|
{
|
|
"entropy": 7.7211596965789795,
|
|
"epoch": 0.005499945000549994,
|
|
"grad_norm": 1.5078125,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 7.3308,
|
|
"mean_token_accuracy": 0.07669527009129525,
|
|
"num_tokens": 409462.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 7.738055324554443,
|
|
"epoch": 0.005609943900560995,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.000127,
|
|
"loss": 7.3223,
|
|
"mean_token_accuracy": 0.08668528124690056,
|
|
"num_tokens": 416854.0,
|
|
"step": 255
|
|
},
|
|
{
|
|
"entropy": 7.6736632823944095,
|
|
"epoch": 0.0057199428005719945,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.0001295,
|
|
"loss": 7.2296,
|
|
"mean_token_accuracy": 0.08230746909976006,
|
|
"num_tokens": 425008.0,
|
|
"step": 260
|
|
},
|
|
{
|
|
"entropy": 7.753271055221558,
|
|
"epoch": 0.005829941700582994,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.000132,
|
|
"loss": 7.3013,
|
|
"mean_token_accuracy": 0.07934577167034149,
|
|
"num_tokens": 433724.0,
|
|
"step": 265
|
|
},
|
|
{
|
|
"entropy": 7.711584377288818,
|
|
"epoch": 0.005939940600593994,
|
|
"grad_norm": 1.359375,
|
|
"learning_rate": 0.00013450000000000002,
|
|
"loss": 7.3422,
|
|
"mean_token_accuracy": 0.08288544788956642,
|
|
"num_tokens": 442054.0,
|
|
"step": 270
|
|
},
|
|
{
|
|
"entropy": 7.599883699417115,
|
|
"epoch": 0.006049939500604994,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00013700000000000002,
|
|
"loss": 7.2328,
|
|
"mean_token_accuracy": 0.08908804506063461,
|
|
"num_tokens": 448870.0,
|
|
"step": 275
|
|
},
|
|
{
|
|
"entropy": 7.610891246795655,
|
|
"epoch": 0.006159938400615994,
|
|
"grad_norm": 1.1484375,
|
|
"learning_rate": 0.0001395,
|
|
"loss": 7.1407,
|
|
"mean_token_accuracy": 0.08629654720425606,
|
|
"num_tokens": 457319.0,
|
|
"step": 280
|
|
},
|
|
{
|
|
"entropy": 7.5924468517303465,
|
|
"epoch": 0.0062699373006269935,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.00014199999999999998,
|
|
"loss": 7.3073,
|
|
"mean_token_accuracy": 0.08258003666996956,
|
|
"num_tokens": 466429.0,
|
|
"step": 285
|
|
},
|
|
{
|
|
"entropy": 7.619415283203125,
|
|
"epoch": 0.006379936200637993,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0001445,
|
|
"loss": 7.1771,
|
|
"mean_token_accuracy": 0.08537527620792389,
|
|
"num_tokens": 473054.0,
|
|
"step": 290
|
|
},
|
|
{
|
|
"entropy": 7.554811334609985,
|
|
"epoch": 0.006489935100648993,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.000147,
|
|
"loss": 7.1241,
|
|
"mean_token_accuracy": 0.0913820244371891,
|
|
"num_tokens": 480863.0,
|
|
"step": 295
|
|
},
|
|
{
|
|
"entropy": 7.666982746124267,
|
|
"epoch": 0.006599934000659994,
|
|
"grad_norm": 1.296875,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 7.2601,
|
|
"mean_token_accuracy": 0.08177951537072659,
|
|
"num_tokens": 489619.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 7.500078439712524,
|
|
"epoch": 0.0067099329006709936,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.000152,
|
|
"loss": 7.0637,
|
|
"mean_token_accuracy": 0.0889226533472538,
|
|
"num_tokens": 497064.0,
|
|
"step": 305
|
|
},
|
|
{
|
|
"entropy": 7.493754625320435,
|
|
"epoch": 0.006819931800681993,
|
|
"grad_norm": 1.34375,
|
|
"learning_rate": 0.00015450000000000001,
|
|
"loss": 7.2238,
|
|
"mean_token_accuracy": 0.08511324226856232,
|
|
"num_tokens": 505855.0,
|
|
"step": 310
|
|
},
|
|
{
|
|
"entropy": 7.635987758636475,
|
|
"epoch": 0.006929930700692993,
|
|
"grad_norm": 1.453125,
|
|
"learning_rate": 0.000157,
|
|
"loss": 7.233,
|
|
"mean_token_accuracy": 0.07769860029220581,
|
|
"num_tokens": 514225.0,
|
|
"step": 315
|
|
},
|
|
{
|
|
"entropy": 7.500319290161133,
|
|
"epoch": 0.007039929600703993,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.0001595,
|
|
"loss": 7.1509,
|
|
"mean_token_accuracy": 0.09244054853916168,
|
|
"num_tokens": 522286.0,
|
|
"step": 320
|
|
},
|
|
{
|
|
"entropy": 7.594728899002075,
|
|
"epoch": 0.007149928500714993,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.000162,
|
|
"loss": 7.1788,
|
|
"mean_token_accuracy": 0.08274376317858696,
|
|
"num_tokens": 531181.0,
|
|
"step": 325
|
|
},
|
|
{
|
|
"entropy": 7.4382788181304935,
|
|
"epoch": 0.0072599274007259925,
|
|
"grad_norm": 1.25,
|
|
"learning_rate": 0.00016450000000000001,
|
|
"loss": 7.2554,
|
|
"mean_token_accuracy": 0.08583186492323876,
|
|
"num_tokens": 540046.0,
|
|
"step": 330
|
|
},
|
|
{
|
|
"entropy": 7.552954006195068,
|
|
"epoch": 0.007369926300736992,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.00016700000000000002,
|
|
"loss": 7.1868,
|
|
"mean_token_accuracy": 0.09165607839822769,
|
|
"num_tokens": 547512.0,
|
|
"step": 335
|
|
},
|
|
{
|
|
"entropy": 7.612061357498169,
|
|
"epoch": 0.007479925200747992,
|
|
"grad_norm": 1.375,
|
|
"learning_rate": 0.00016950000000000003,
|
|
"loss": 7.1953,
|
|
"mean_token_accuracy": 0.08806446865200997,
|
|
"num_tokens": 555114.0,
|
|
"step": 340
|
|
},
|
|
{
|
|
"entropy": 7.450941181182861,
|
|
"epoch": 0.007589924100758993,
|
|
"grad_norm": 1.3515625,
|
|
"learning_rate": 0.00017199999999999998,
|
|
"loss": 7.1963,
|
|
"mean_token_accuracy": 0.08417715430259705,
|
|
"num_tokens": 564719.0,
|
|
"step": 345
|
|
},
|
|
{
|
|
"entropy": 7.624731874465942,
|
|
"epoch": 0.007699923000769993,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 7.2362,
|
|
"mean_token_accuracy": 0.08734179884195328,
|
|
"num_tokens": 572263.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 7.427968072891235,
|
|
"epoch": 0.007809921900780992,
|
|
"grad_norm": 1.2421875,
|
|
"learning_rate": 0.000177,
|
|
"loss": 7.0413,
|
|
"mean_token_accuracy": 0.09226703494787217,
|
|
"num_tokens": 580518.0,
|
|
"step": 355
|
|
},
|
|
{
|
|
"entropy": 7.48348913192749,
|
|
"epoch": 0.007919920800791992,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0001795,
|
|
"loss": 7.0992,
|
|
"mean_token_accuracy": 0.08984480872750282,
|
|
"num_tokens": 589765.0,
|
|
"step": 360
|
|
},
|
|
{
|
|
"entropy": 7.411487674713134,
|
|
"epoch": 0.008029919700802992,
|
|
"grad_norm": 1.3359375,
|
|
"learning_rate": 0.000182,
|
|
"loss": 7.0331,
|
|
"mean_token_accuracy": 0.09457142502069474,
|
|
"num_tokens": 597874.0,
|
|
"step": 365
|
|
},
|
|
{
|
|
"entropy": 7.312029266357422,
|
|
"epoch": 0.008139918600813992,
|
|
"grad_norm": 1.1171875,
|
|
"learning_rate": 0.0001845,
|
|
"loss": 7.0479,
|
|
"mean_token_accuracy": 0.09531447216868401,
|
|
"num_tokens": 605258.0,
|
|
"step": 370
|
|
},
|
|
{
|
|
"entropy": 7.525022172927857,
|
|
"epoch": 0.008249917500824992,
|
|
"grad_norm": 1.1640625,
|
|
"learning_rate": 0.000187,
|
|
"loss": 7.0944,
|
|
"mean_token_accuracy": 0.09678534865379333,
|
|
"num_tokens": 614059.0,
|
|
"step": 375
|
|
},
|
|
{
|
|
"entropy": 7.381738662719727,
|
|
"epoch": 0.008359916400835991,
|
|
"grad_norm": 1.234375,
|
|
"learning_rate": 0.0001895,
|
|
"loss": 7.1123,
|
|
"mean_token_accuracy": 0.09476015195250512,
|
|
"num_tokens": 622509.0,
|
|
"step": 380
|
|
},
|
|
{
|
|
"entropy": 7.439949369430542,
|
|
"epoch": 0.008469915300846991,
|
|
"grad_norm": 1.28125,
|
|
"learning_rate": 0.000192,
|
|
"loss": 7.0818,
|
|
"mean_token_accuracy": 0.08750179558992385,
|
|
"num_tokens": 630524.0,
|
|
"step": 385
|
|
},
|
|
{
|
|
"entropy": 7.464513826370239,
|
|
"epoch": 0.008579914200857991,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.0001945,
|
|
"loss": 7.1312,
|
|
"mean_token_accuracy": 0.08964811265468597,
|
|
"num_tokens": 639629.0,
|
|
"step": 390
|
|
},
|
|
{
|
|
"entropy": 7.40151834487915,
|
|
"epoch": 0.00868991310086899,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 0.00019700000000000002,
|
|
"loss": 7.134,
|
|
"mean_token_accuracy": 0.08695418313145638,
|
|
"num_tokens": 648281.0,
|
|
"step": 395
|
|
},
|
|
{
|
|
"entropy": 7.293155288696289,
|
|
"epoch": 0.00879991200087999,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 6.9677,
|
|
"mean_token_accuracy": 0.09901286512613297,
|
|
"num_tokens": 655925.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 7.2998762130737305,
|
|
"epoch": 0.00890991090089099,
|
|
"grad_norm": 1.390625,
|
|
"learning_rate": 0.000202,
|
|
"loss": 6.8856,
|
|
"mean_token_accuracy": 0.10275417193770409,
|
|
"num_tokens": 663195.0,
|
|
"step": 405
|
|
},
|
|
{
|
|
"entropy": 7.407110023498535,
|
|
"epoch": 0.00901990980090199,
|
|
"grad_norm": 1.2578125,
|
|
"learning_rate": 0.00020449999999999998,
|
|
"loss": 7.0436,
|
|
"mean_token_accuracy": 0.08666960522532463,
|
|
"num_tokens": 670935.0,
|
|
"step": 410
|
|
},
|
|
{
|
|
"entropy": 7.310142374038696,
|
|
"epoch": 0.009129908700912992,
|
|
"grad_norm": 1.078125,
|
|
"learning_rate": 0.000207,
|
|
"loss": 7.0698,
|
|
"mean_token_accuracy": 0.08886591419577598,
|
|
"num_tokens": 679995.0,
|
|
"step": 415
|
|
},
|
|
{
|
|
"entropy": 7.335082101821899,
|
|
"epoch": 0.009239907600923991,
|
|
"grad_norm": 1.3671875,
|
|
"learning_rate": 0.0002095,
|
|
"loss": 7.0261,
|
|
"mean_token_accuracy": 0.09914991110563279,
|
|
"num_tokens": 688511.0,
|
|
"step": 420
|
|
},
|
|
{
|
|
"entropy": 7.4029449939727785,
|
|
"epoch": 0.009349906500934991,
|
|
"grad_norm": 1.4453125,
|
|
"learning_rate": 0.000212,
|
|
"loss": 7.0199,
|
|
"mean_token_accuracy": 0.09210473299026489,
|
|
"num_tokens": 696346.0,
|
|
"step": 425
|
|
},
|
|
{
|
|
"entropy": 7.221910095214843,
|
|
"epoch": 0.009459905400945991,
|
|
"grad_norm": 1.1015625,
|
|
"learning_rate": 0.0002145,
|
|
"loss": 7.0386,
|
|
"mean_token_accuracy": 0.09151387587189674,
|
|
"num_tokens": 705295.0,
|
|
"step": 430
|
|
},
|
|
{
|
|
"entropy": 7.411863374710083,
|
|
"epoch": 0.00956990430095699,
|
|
"grad_norm": 1.15625,
|
|
"learning_rate": 0.00021700000000000002,
|
|
"loss": 7.1125,
|
|
"mean_token_accuracy": 0.08886803835630416,
|
|
"num_tokens": 712631.0,
|
|
"step": 435
|
|
},
|
|
{
|
|
"entropy": 7.294147968292236,
|
|
"epoch": 0.00967990320096799,
|
|
"grad_norm": 1.421875,
|
|
"learning_rate": 0.0002195,
|
|
"loss": 7.0025,
|
|
"mean_token_accuracy": 0.09658410027623177,
|
|
"num_tokens": 720914.0,
|
|
"step": 440
|
|
},
|
|
{
|
|
"entropy": 7.393889141082764,
|
|
"epoch": 0.00978990210097899,
|
|
"grad_norm": 1.1875,
|
|
"learning_rate": 0.000222,
|
|
"loss": 7.0301,
|
|
"mean_token_accuracy": 0.09346665740013123,
|
|
"num_tokens": 729552.0,
|
|
"step": 445
|
|
},
|
|
{
|
|
"entropy": 7.315341663360596,
|
|
"epoch": 0.00989990100098999,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 7.0847,
|
|
"mean_token_accuracy": 0.0887609712779522,
|
|
"num_tokens": 738361.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 7.165237855911255,
|
|
"epoch": 0.01000989990100099,
|
|
"grad_norm": 1.1953125,
|
|
"learning_rate": 0.00022700000000000002,
|
|
"loss": 6.9777,
|
|
"mean_token_accuracy": 0.09182030335068703,
|
|
"num_tokens": 746743.0,
|
|
"step": 455
|
|
},
|
|
{
|
|
"entropy": 7.342606496810913,
|
|
"epoch": 0.01011989880101199,
|
|
"grad_norm": 1.2734375,
|
|
"learning_rate": 0.00022950000000000002,
|
|
"loss": 6.9741,
|
|
"mean_token_accuracy": 0.09438188895583152,
|
|
"num_tokens": 754410.0,
|
|
"step": 460
|
|
},
|
|
{
|
|
"entropy": 7.2022956848144535,
|
|
"epoch": 0.01022989770102299,
|
|
"grad_norm": 1.3984375,
|
|
"learning_rate": 0.00023200000000000003,
|
|
"loss": 6.9393,
|
|
"mean_token_accuracy": 0.09460799917578697,
|
|
"num_tokens": 761878.0,
|
|
"step": 465
|
|
},
|
|
{
|
|
"entropy": 7.261647701263428,
|
|
"epoch": 0.01033989660103399,
|
|
"grad_norm": 1.40625,
|
|
"learning_rate": 0.00023449999999999998,
|
|
"loss": 6.9743,
|
|
"mean_token_accuracy": 0.08888068273663521,
|
|
"num_tokens": 770217.0,
|
|
"step": 470
|
|
},
|
|
{
|
|
"entropy": 7.156740999221801,
|
|
"epoch": 0.01044989550104499,
|
|
"grad_norm": 1.21875,
|
|
"learning_rate": 0.000237,
|
|
"loss": 6.8641,
|
|
"mean_token_accuracy": 0.10069640651345253,
|
|
"num_tokens": 777915.0,
|
|
"step": 475
|
|
},
|
|
{
|
|
"entropy": 7.317181253433228,
|
|
"epoch": 0.010559894401055989,
|
|
"grad_norm": 1.109375,
|
|
"learning_rate": 0.0002395,
|
|
"loss": 7.0903,
|
|
"mean_token_accuracy": 0.08290443792939187,
|
|
"num_tokens": 788183.0,
|
|
"step": 480
|
|
},
|
|
{
|
|
"entropy": 7.3058023929595945,
|
|
"epoch": 0.010669893301066989,
|
|
"grad_norm": 1.171875,
|
|
"learning_rate": 0.000242,
|
|
"loss": 7.0258,
|
|
"mean_token_accuracy": 0.08370587825775147,
|
|
"num_tokens": 796512.0,
|
|
"step": 485
|
|
},
|
|
{
|
|
"entropy": 7.1996331214904785,
|
|
"epoch": 0.010779892201077989,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.0002445,
|
|
"loss": 6.9689,
|
|
"mean_token_accuracy": 0.09842925891280174,
|
|
"num_tokens": 804157.0,
|
|
"step": 490
|
|
},
|
|
{
|
|
"entropy": 7.272369623184204,
|
|
"epoch": 0.010889891101088988,
|
|
"grad_norm": 1.3125,
|
|
"learning_rate": 0.000247,
|
|
"loss": 6.9559,
|
|
"mean_token_accuracy": 0.08966508060693741,
|
|
"num_tokens": 812315.0,
|
|
"step": 495
|
|
},
|
|
{
|
|
"entropy": 7.174358892440796,
|
|
"epoch": 0.010999890001099988,
|
|
"grad_norm": 1.4375,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 6.9047,
|
|
"mean_token_accuracy": 0.10042067542672158,
|
|
"num_tokens": 820819.0,
|
|
"step": 500
|
|
}
|
|
],
|
|
"logging_steps": 5,
|
|
"max_steps": 4000,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"save_steps": 500,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": false
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 1526957973504000.0,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|