463 lines
11 KiB
JSON
463 lines
11 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.026214395526076496,
|
|
"eval_steps": 3000,
|
|
"global_step": 3000,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.00043690659210127495,
|
|
"grad_norm": 1.8826903104782104,
|
|
"learning_rate": 2.4500000000000003e-05,
|
|
"loss": 9.762939453125,
|
|
"step": 50
|
|
},
|
|
{
|
|
"epoch": 0.0008738131842025499,
|
|
"grad_norm": 0.9995923042297363,
|
|
"learning_rate": 4.9500000000000004e-05,
|
|
"loss": 8.305597534179688,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.001310719776303825,
|
|
"grad_norm": 0.9061496257781982,
|
|
"learning_rate": 7.45e-05,
|
|
"loss": 7.173495483398438,
|
|
"step": 150
|
|
},
|
|
{
|
|
"epoch": 0.0017476263684050998,
|
|
"grad_norm": 1.6024847030639648,
|
|
"learning_rate": 9.95e-05,
|
|
"loss": 6.624603271484375,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.002184532960506375,
|
|
"grad_norm": 1.2123334407806396,
|
|
"learning_rate": 0.0001245,
|
|
"loss": 6.243399658203125,
|
|
"step": 250
|
|
},
|
|
{
|
|
"epoch": 0.00262143955260765,
|
|
"grad_norm": 1.0432322025299072,
|
|
"learning_rate": 0.0001495,
|
|
"loss": 5.9879296875,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.0030583461447089245,
|
|
"grad_norm": 0.9280146360397339,
|
|
"learning_rate": 0.00017449999999999999,
|
|
"loss": 5.776607666015625,
|
|
"step": 350
|
|
},
|
|
{
|
|
"epoch": 0.0034952527368101996,
|
|
"grad_norm": 1.0289900302886963,
|
|
"learning_rate": 0.00019950000000000002,
|
|
"loss": 5.5920703125,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.003932159328911475,
|
|
"grad_norm": 1.0593407154083252,
|
|
"learning_rate": 0.0002245,
|
|
"loss": 5.437271728515625,
|
|
"step": 450
|
|
},
|
|
{
|
|
"epoch": 0.00436906592101275,
|
|
"grad_norm": 1.0122662782669067,
|
|
"learning_rate": 0.0002495,
|
|
"loss": 5.287777709960937,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.004805972513114025,
|
|
"grad_norm": 0.7662743926048279,
|
|
"learning_rate": 0.0002745,
|
|
"loss": 5.155453491210937,
|
|
"step": 550
|
|
},
|
|
{
|
|
"epoch": 0.0052428791052153,
|
|
"grad_norm": 0.7941539287567139,
|
|
"learning_rate": 0.0002995,
|
|
"loss": 5.030631713867187,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 0.005679785697316574,
|
|
"grad_norm": 0.6195730566978455,
|
|
"learning_rate": 0.00032450000000000003,
|
|
"loss": 4.909389953613282,
|
|
"step": 650
|
|
},
|
|
{
|
|
"epoch": 0.006116692289417849,
|
|
"grad_norm": 0.6587674617767334,
|
|
"learning_rate": 0.0003495,
|
|
"loss": 4.786273498535156,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 0.006553598881519124,
|
|
"grad_norm": 0.7233771681785583,
|
|
"learning_rate": 0.0003745,
|
|
"loss": 4.6648342895507815,
|
|
"step": 750
|
|
},
|
|
{
|
|
"epoch": 0.006990505473620399,
|
|
"grad_norm": 0.5489601492881775,
|
|
"learning_rate": 0.0003995,
|
|
"loss": 4.555048522949218,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 0.007427412065721674,
|
|
"grad_norm": 0.6154086589813232,
|
|
"learning_rate": 0.0004245,
|
|
"loss": 4.461051330566407,
|
|
"step": 850
|
|
},
|
|
{
|
|
"epoch": 0.00786431865782295,
|
|
"grad_norm": 0.41333407163619995,
|
|
"learning_rate": 0.00044950000000000003,
|
|
"loss": 4.374936828613281,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 0.008301225249924224,
|
|
"grad_norm": 0.5036063194274902,
|
|
"learning_rate": 0.0004745,
|
|
"loss": 4.3033209228515625,
|
|
"step": 950
|
|
},
|
|
{
|
|
"epoch": 0.0087381318420255,
|
|
"grad_norm": 0.3687891364097595,
|
|
"learning_rate": 0.0004995,
|
|
"loss": 4.233715209960938,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 0.009175038434126774,
|
|
"grad_norm": 0.5433846116065979,
|
|
"learning_rate": 0.0005245,
|
|
"loss": 4.186253967285157,
|
|
"step": 1050
|
|
},
|
|
{
|
|
"epoch": 0.00961194502622805,
|
|
"grad_norm": 0.439351350069046,
|
|
"learning_rate": 0.0005495,
|
|
"loss": 4.139383850097656,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 0.010048851618329325,
|
|
"grad_norm": 0.36215880513191223,
|
|
"learning_rate": 0.0005745,
|
|
"loss": 4.08899658203125,
|
|
"step": 1150
|
|
},
|
|
{
|
|
"epoch": 0.0104857582104306,
|
|
"grad_norm": 0.3970701992511749,
|
|
"learning_rate": 0.0005995000000000001,
|
|
"loss": 4.061238403320313,
|
|
"step": 1200
|
|
},
|
|
{
|
|
"epoch": 0.010922664802531873,
|
|
"grad_norm": 0.3410918712615967,
|
|
"learning_rate": 0.0006245000000000001,
|
|
"loss": 4.026336669921875,
|
|
"step": 1250
|
|
},
|
|
{
|
|
"epoch": 0.011359571394633148,
|
|
"grad_norm": 0.3047930598258972,
|
|
"learning_rate": 0.0006495,
|
|
"loss": 3.9928851318359375,
|
|
"step": 1300
|
|
},
|
|
{
|
|
"epoch": 0.011796477986734423,
|
|
"grad_norm": 0.2819700241088867,
|
|
"learning_rate": 0.0006745,
|
|
"loss": 3.9584146118164063,
|
|
"step": 1350
|
|
},
|
|
{
|
|
"epoch": 0.012233384578835698,
|
|
"grad_norm": 0.29095014929771423,
|
|
"learning_rate": 0.0006995,
|
|
"loss": 3.9291363525390626,
|
|
"step": 1400
|
|
},
|
|
{
|
|
"epoch": 0.012670291170936973,
|
|
"grad_norm": 0.2987957298755646,
|
|
"learning_rate": 0.0007245000000000001,
|
|
"loss": 3.9090225219726564,
|
|
"step": 1450
|
|
},
|
|
{
|
|
"epoch": 0.013107197763038248,
|
|
"grad_norm": 0.28419050574302673,
|
|
"learning_rate": 0.0007495000000000001,
|
|
"loss": 3.890744934082031,
|
|
"step": 1500
|
|
},
|
|
{
|
|
"epoch": 0.013544104355139523,
|
|
"grad_norm": 0.3039053976535797,
|
|
"learning_rate": 0.0007745,
|
|
"loss": 3.861073303222656,
|
|
"step": 1550
|
|
},
|
|
{
|
|
"epoch": 0.013981010947240798,
|
|
"grad_norm": 0.25676754117012024,
|
|
"learning_rate": 0.0007995,
|
|
"loss": 3.8428253173828124,
|
|
"step": 1600
|
|
},
|
|
{
|
|
"epoch": 0.014417917539342073,
|
|
"grad_norm": 0.2573896646499634,
|
|
"learning_rate": 0.0008245,
|
|
"loss": 3.8260653686523436,
|
|
"step": 1650
|
|
},
|
|
{
|
|
"epoch": 0.014854824131443348,
|
|
"grad_norm": 0.272127240896225,
|
|
"learning_rate": 0.0008495000000000001,
|
|
"loss": 3.8006341552734373,
|
|
"step": 1700
|
|
},
|
|
{
|
|
"epoch": 0.015291730723544623,
|
|
"grad_norm": 0.23205944895744324,
|
|
"learning_rate": 0.0008745000000000001,
|
|
"loss": 3.7862649536132813,
|
|
"step": 1750
|
|
},
|
|
{
|
|
"epoch": 0.0157286373156459,
|
|
"grad_norm": 0.22847425937652588,
|
|
"learning_rate": 0.0008995,
|
|
"loss": 3.771934814453125,
|
|
"step": 1800
|
|
},
|
|
{
|
|
"epoch": 0.016165543907747174,
|
|
"grad_norm": 0.2651786208152771,
|
|
"learning_rate": 0.0009245,
|
|
"loss": 3.76009521484375,
|
|
"step": 1850
|
|
},
|
|
{
|
|
"epoch": 0.01660245049984845,
|
|
"grad_norm": 0.24903950095176697,
|
|
"learning_rate": 0.0009495,
|
|
"loss": 3.7454238891601563,
|
|
"step": 1900
|
|
},
|
|
{
|
|
"epoch": 0.017039357091949724,
|
|
"grad_norm": 0.24524764716625214,
|
|
"learning_rate": 0.0009745000000000001,
|
|
"loss": 3.735928955078125,
|
|
"step": 1950
|
|
},
|
|
{
|
|
"epoch": 0.017476263684051,
|
|
"grad_norm": 0.22325558960437775,
|
|
"learning_rate": 0.0009995000000000002,
|
|
"loss": 3.7209417724609377,
|
|
"step": 2000
|
|
},
|
|
{
|
|
"epoch": 0.017913170276152274,
|
|
"grad_norm": 0.23860357701778412,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.7122769165039062,
|
|
"step": 2050
|
|
},
|
|
{
|
|
"epoch": 0.01835007686825355,
|
|
"grad_norm": 0.2162218689918518,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.690401611328125,
|
|
"step": 2100
|
|
},
|
|
{
|
|
"epoch": 0.018786983460354824,
|
|
"grad_norm": 0.2171812504529953,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.6825027465820312,
|
|
"step": 2150
|
|
},
|
|
{
|
|
"epoch": 0.0192238900524561,
|
|
"grad_norm": 0.22677843272686005,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.6679034423828125,
|
|
"step": 2200
|
|
},
|
|
{
|
|
"epoch": 0.019660796644557374,
|
|
"grad_norm": 0.23934350907802582,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.6494000244140623,
|
|
"step": 2250
|
|
},
|
|
{
|
|
"epoch": 0.02009770323665865,
|
|
"grad_norm": 0.20819289982318878,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.640907897949219,
|
|
"step": 2300
|
|
},
|
|
{
|
|
"epoch": 0.020534609828759924,
|
|
"grad_norm": 0.22890810668468475,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.6309326171875,
|
|
"step": 2350
|
|
},
|
|
{
|
|
"epoch": 0.0209715164208612,
|
|
"grad_norm": 0.2085026353597641,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.6261746215820314,
|
|
"step": 2400
|
|
},
|
|
{
|
|
"epoch": 0.021408423012962474,
|
|
"grad_norm": 0.22136099636554718,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.614878845214844,
|
|
"step": 2450
|
|
},
|
|
{
|
|
"epoch": 0.021845329605063746,
|
|
"grad_norm": 0.21210630238056183,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.599817810058594,
|
|
"step": 2500
|
|
},
|
|
{
|
|
"epoch": 0.02228223619716502,
|
|
"grad_norm": 0.24638274312019348,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.591536865234375,
|
|
"step": 2550
|
|
},
|
|
{
|
|
"epoch": 0.022719142789266296,
|
|
"grad_norm": 0.20501184463500977,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.58053955078125,
|
|
"step": 2600
|
|
},
|
|
{
|
|
"epoch": 0.02315604938136757,
|
|
"grad_norm": 0.2101815640926361,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.5824459838867186,
|
|
"step": 2650
|
|
},
|
|
{
|
|
"epoch": 0.023592955973468846,
|
|
"grad_norm": 0.22906431555747986,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.564821472167969,
|
|
"step": 2700
|
|
},
|
|
{
|
|
"epoch": 0.02402986256557012,
|
|
"grad_norm": 0.20708167552947998,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.5566305541992187,
|
|
"step": 2750
|
|
},
|
|
{
|
|
"epoch": 0.024466769157671396,
|
|
"grad_norm": 0.21956811845302582,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.5507748413085936,
|
|
"step": 2800
|
|
},
|
|
{
|
|
"epoch": 0.02490367574977267,
|
|
"grad_norm": 0.24124488234519958,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.5426095581054686,
|
|
"step": 2850
|
|
},
|
|
{
|
|
"epoch": 0.025340582341873946,
|
|
"grad_norm": 0.20921742916107178,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.531602783203125,
|
|
"step": 2900
|
|
},
|
|
{
|
|
"epoch": 0.02577748893397522,
|
|
"grad_norm": 0.19601280987262726,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.5225894165039064,
|
|
"step": 2950
|
|
},
|
|
{
|
|
"epoch": 0.026214395526076496,
|
|
"grad_norm": 0.20243392884731293,
|
|
"learning_rate": 0.001,
|
|
"loss": 3.5183840942382814,
|
|
"step": 3000
|
|
},
|
|
{
|
|
"epoch": 0.026214395526076496,
|
|
"eval_loss": 3.518667221069336,
|
|
"eval_runtime": 5.8945,
|
|
"eval_samples_per_second": 165.579,
|
|
"eval_steps_per_second": 10.349,
|
|
"step": 3000
|
|
}
|
|
],
|
|
"logging_steps": 50,
|
|
"max_steps": 114440,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 1,
|
|
"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": 3.56339612123136e+17,
|
|
"train_batch_size": 16,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|