DeepScaleR-1.5B-L36-RollE3 / trainer_state.json
ZMC2019's picture
Model save
8c7403e verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 357,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.04201680672268908,
"grad_norm": 4.70634130932114,
"learning_rate": 6.944444444444445e-06,
"loss": 0.9136,
"step": 5
},
{
"epoch": 0.08403361344537816,
"grad_norm": 1.8906913115297757,
"learning_rate": 1.388888888888889e-05,
"loss": 0.766,
"step": 10
},
{
"epoch": 0.12605042016806722,
"grad_norm": 0.6826181059075677,
"learning_rate": 2.0833333333333336e-05,
"loss": 0.6325,
"step": 15
},
{
"epoch": 0.16806722689075632,
"grad_norm": 0.33363228311628446,
"learning_rate": 2.777777777777778e-05,
"loss": 0.5511,
"step": 20
},
{
"epoch": 0.21008403361344538,
"grad_norm": 0.2907608801770254,
"learning_rate": 3.472222222222222e-05,
"loss": 0.5244,
"step": 25
},
{
"epoch": 0.25210084033613445,
"grad_norm": 0.2733826152956853,
"learning_rate": 4.166666666666667e-05,
"loss": 0.4913,
"step": 30
},
{
"epoch": 0.29411764705882354,
"grad_norm": 0.2732334536953152,
"learning_rate": 4.8611111111111115e-05,
"loss": 0.4751,
"step": 35
},
{
"epoch": 0.33613445378151263,
"grad_norm": 0.2585751886114213,
"learning_rate": 4.937694704049845e-05,
"loss": 0.4596,
"step": 40
},
{
"epoch": 0.37815126050420167,
"grad_norm": 0.2397885612111248,
"learning_rate": 4.85981308411215e-05,
"loss": 0.4349,
"step": 45
},
{
"epoch": 0.42016806722689076,
"grad_norm": 0.21116743482172115,
"learning_rate": 4.781931464174455e-05,
"loss": 0.4326,
"step": 50
},
{
"epoch": 0.46218487394957986,
"grad_norm": 0.14451656745477998,
"learning_rate": 4.7040498442367604e-05,
"loss": 0.4157,
"step": 55
},
{
"epoch": 0.5042016806722689,
"grad_norm": 0.14127541762263016,
"learning_rate": 4.6261682242990654e-05,
"loss": 0.4122,
"step": 60
},
{
"epoch": 0.5462184873949579,
"grad_norm": 0.671598848272524,
"learning_rate": 4.548286604361371e-05,
"loss": 0.4149,
"step": 65
},
{
"epoch": 0.5882352941176471,
"grad_norm": 0.13463214509481336,
"learning_rate": 4.470404984423676e-05,
"loss": 0.3959,
"step": 70
},
{
"epoch": 0.6302521008403361,
"grad_norm": 0.1561144783705333,
"learning_rate": 4.392523364485982e-05,
"loss": 0.4038,
"step": 75
},
{
"epoch": 0.6722689075630253,
"grad_norm": 0.135660513281463,
"learning_rate": 4.314641744548287e-05,
"loss": 0.4012,
"step": 80
},
{
"epoch": 0.7142857142857143,
"grad_norm": 0.1329272951870047,
"learning_rate": 4.236760124610592e-05,
"loss": 0.3952,
"step": 85
},
{
"epoch": 0.7563025210084033,
"grad_norm": 0.12595448081633448,
"learning_rate": 4.1588785046728974e-05,
"loss": 0.3893,
"step": 90
},
{
"epoch": 0.7983193277310925,
"grad_norm": 0.1293818852034363,
"learning_rate": 4.0809968847352024e-05,
"loss": 0.4,
"step": 95
},
{
"epoch": 0.8403361344537815,
"grad_norm": 0.1469536727483663,
"learning_rate": 4.003115264797508e-05,
"loss": 0.3929,
"step": 100
},
{
"epoch": 0.8823529411764706,
"grad_norm": 0.12410046786523299,
"learning_rate": 3.925233644859813e-05,
"loss": 0.388,
"step": 105
},
{
"epoch": 0.9243697478991597,
"grad_norm": 0.13619892881270407,
"learning_rate": 3.847352024922119e-05,
"loss": 0.3821,
"step": 110
},
{
"epoch": 0.9663865546218487,
"grad_norm": 0.11885167396488784,
"learning_rate": 3.769470404984424e-05,
"loss": 0.3946,
"step": 115
},
{
"epoch": 1.0084033613445378,
"grad_norm": 0.14287918831630414,
"learning_rate": 3.691588785046729e-05,
"loss": 0.3801,
"step": 120
},
{
"epoch": 1.050420168067227,
"grad_norm": 0.1285290919486453,
"learning_rate": 3.6137071651090344e-05,
"loss": 0.3652,
"step": 125
},
{
"epoch": 1.092436974789916,
"grad_norm": 0.12470234407748469,
"learning_rate": 3.5358255451713394e-05,
"loss": 0.3587,
"step": 130
},
{
"epoch": 1.134453781512605,
"grad_norm": 0.11633316420488911,
"learning_rate": 3.457943925233645e-05,
"loss": 0.357,
"step": 135
},
{
"epoch": 1.1764705882352942,
"grad_norm": 0.12243208920648133,
"learning_rate": 3.38006230529595e-05,
"loss": 0.3534,
"step": 140
},
{
"epoch": 1.2184873949579833,
"grad_norm": 0.12086774297684207,
"learning_rate": 3.302180685358255e-05,
"loss": 0.3603,
"step": 145
},
{
"epoch": 1.2605042016806722,
"grad_norm": 0.12226080249060899,
"learning_rate": 3.224299065420561e-05,
"loss": 0.355,
"step": 150
},
{
"epoch": 1.3025210084033614,
"grad_norm": 0.11411438458299594,
"learning_rate": 3.146417445482866e-05,
"loss": 0.3602,
"step": 155
},
{
"epoch": 1.3445378151260505,
"grad_norm": 0.11464926522938973,
"learning_rate": 3.0685358255451714e-05,
"loss": 0.3518,
"step": 160
},
{
"epoch": 1.3865546218487395,
"grad_norm": 0.12001429741902232,
"learning_rate": 2.9906542056074764e-05,
"loss": 0.355,
"step": 165
},
{
"epoch": 1.4285714285714286,
"grad_norm": 0.11341054683006553,
"learning_rate": 2.9127725856697818e-05,
"loss": 0.3546,
"step": 170
},
{
"epoch": 1.4705882352941178,
"grad_norm": 0.11534598808870551,
"learning_rate": 2.834890965732087e-05,
"loss": 0.3481,
"step": 175
},
{
"epoch": 1.5126050420168067,
"grad_norm": 0.11555410704104117,
"learning_rate": 2.7570093457943924e-05,
"loss": 0.3566,
"step": 180
},
{
"epoch": 1.5546218487394958,
"grad_norm": 0.11882277392875285,
"learning_rate": 2.6791277258566978e-05,
"loss": 0.3503,
"step": 185
},
{
"epoch": 1.596638655462185,
"grad_norm": 0.10912064576299817,
"learning_rate": 2.601246105919003e-05,
"loss": 0.3467,
"step": 190
},
{
"epoch": 1.638655462184874,
"grad_norm": 0.11671548577730292,
"learning_rate": 2.5233644859813084e-05,
"loss": 0.3518,
"step": 195
},
{
"epoch": 1.680672268907563,
"grad_norm": 0.11709095347100043,
"learning_rate": 2.4454828660436138e-05,
"loss": 0.3514,
"step": 200
},
{
"epoch": 1.7226890756302522,
"grad_norm": 0.4547280409942813,
"learning_rate": 2.367601246105919e-05,
"loss": 0.3545,
"step": 205
},
{
"epoch": 1.7647058823529411,
"grad_norm": 0.12116945588708594,
"learning_rate": 2.2897196261682244e-05,
"loss": 0.3471,
"step": 210
},
{
"epoch": 1.8067226890756303,
"grad_norm": 0.11202328680044989,
"learning_rate": 2.2118380062305298e-05,
"loss": 0.3501,
"step": 215
},
{
"epoch": 1.8487394957983194,
"grad_norm": 0.11989297797083326,
"learning_rate": 2.133956386292835e-05,
"loss": 0.3511,
"step": 220
},
{
"epoch": 1.8907563025210083,
"grad_norm": 0.11344656402005421,
"learning_rate": 2.05607476635514e-05,
"loss": 0.3427,
"step": 225
},
{
"epoch": 1.9327731092436975,
"grad_norm": 0.11459966187379855,
"learning_rate": 1.9781931464174454e-05,
"loss": 0.35,
"step": 230
},
{
"epoch": 1.9747899159663866,
"grad_norm": 0.10470258572360276,
"learning_rate": 1.9003115264797507e-05,
"loss": 0.3459,
"step": 235
},
{
"epoch": 2.0168067226890756,
"grad_norm": 0.11916183310406858,
"learning_rate": 1.822429906542056e-05,
"loss": 0.3366,
"step": 240
},
{
"epoch": 2.0588235294117645,
"grad_norm": 0.1385981052541961,
"learning_rate": 1.7445482866043614e-05,
"loss": 0.3251,
"step": 245
},
{
"epoch": 2.100840336134454,
"grad_norm": 0.1235803846831308,
"learning_rate": 1.6666666666666667e-05,
"loss": 0.3237,
"step": 250
},
{
"epoch": 2.142857142857143,
"grad_norm": 0.11353299337591315,
"learning_rate": 1.588785046728972e-05,
"loss": 0.3198,
"step": 255
},
{
"epoch": 2.184873949579832,
"grad_norm": 0.11174702397768707,
"learning_rate": 1.5109034267912772e-05,
"loss": 0.3191,
"step": 260
},
{
"epoch": 2.226890756302521,
"grad_norm": 0.1137863668401699,
"learning_rate": 1.4330218068535826e-05,
"loss": 0.3229,
"step": 265
},
{
"epoch": 2.26890756302521,
"grad_norm": 0.12072471085734401,
"learning_rate": 1.3551401869158877e-05,
"loss": 0.324,
"step": 270
},
{
"epoch": 2.310924369747899,
"grad_norm": 0.11123684767070798,
"learning_rate": 1.277258566978193e-05,
"loss": 0.3214,
"step": 275
},
{
"epoch": 2.3529411764705883,
"grad_norm": 0.11295726491194695,
"learning_rate": 1.1993769470404986e-05,
"loss": 0.318,
"step": 280
},
{
"epoch": 2.3949579831932772,
"grad_norm": 0.1106796291907045,
"learning_rate": 1.1214953271028037e-05,
"loss": 0.3235,
"step": 285
},
{
"epoch": 2.4369747899159666,
"grad_norm": 0.11033037272448094,
"learning_rate": 1.043613707165109e-05,
"loss": 0.3227,
"step": 290
},
{
"epoch": 2.4789915966386555,
"grad_norm": 0.10873033497026395,
"learning_rate": 9.657320872274144e-06,
"loss": 0.3215,
"step": 295
},
{
"epoch": 2.5210084033613445,
"grad_norm": 0.11267308869650684,
"learning_rate": 8.878504672897196e-06,
"loss": 0.3236,
"step": 300
},
{
"epoch": 2.5630252100840334,
"grad_norm": 0.10662033804378954,
"learning_rate": 8.099688473520249e-06,
"loss": 0.3185,
"step": 305
},
{
"epoch": 2.6050420168067228,
"grad_norm": 0.10419898622414428,
"learning_rate": 7.3208722741433015e-06,
"loss": 0.3201,
"step": 310
},
{
"epoch": 2.6470588235294117,
"grad_norm": 0.1194071123415722,
"learning_rate": 6.542056074766355e-06,
"loss": 0.3181,
"step": 315
},
{
"epoch": 2.689075630252101,
"grad_norm": 0.10315644198712422,
"learning_rate": 5.763239875389408e-06,
"loss": 0.319,
"step": 320
},
{
"epoch": 2.73109243697479,
"grad_norm": 0.10169341589102616,
"learning_rate": 4.9844236760124615e-06,
"loss": 0.3155,
"step": 325
},
{
"epoch": 2.773109243697479,
"grad_norm": 0.10841897463334976,
"learning_rate": 4.205607476635514e-06,
"loss": 0.3221,
"step": 330
},
{
"epoch": 2.815126050420168,
"grad_norm": 0.11304762877258454,
"learning_rate": 3.426791277258567e-06,
"loss": 0.3224,
"step": 335
},
{
"epoch": 2.857142857142857,
"grad_norm": 0.10152209760807467,
"learning_rate": 2.64797507788162e-06,
"loss": 0.3117,
"step": 340
},
{
"epoch": 2.899159663865546,
"grad_norm": 0.10952489428700839,
"learning_rate": 1.8691588785046728e-06,
"loss": 0.3218,
"step": 345
},
{
"epoch": 2.9411764705882355,
"grad_norm": 0.1024965907199079,
"learning_rate": 1.0903426791277259e-06,
"loss": 0.3206,
"step": 350
},
{
"epoch": 2.9831932773109244,
"grad_norm": 0.1048143744280903,
"learning_rate": 3.1152647975077885e-07,
"loss": 0.3265,
"step": 355
},
{
"epoch": 3.0,
"step": 357,
"total_flos": 1.2726481677031834e+17,
"train_loss": 0.3811905702956918,
"train_runtime": 3910.3734,
"train_samples_per_second": 1.455,
"train_steps_per_second": 0.091
}
],
"logging_steps": 5,
"max_steps": 357,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.2726481677031834e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}