309 lines
8.2 KiB
JSON
309 lines
8.2 KiB
JSON
|
{
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 0.9977046671767407,
|
||
|
|
"eval_steps": 100,
|
||
|
|
"global_step": 163,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.030604437643458302,
|
||
|
|
"grad_norm": 3.2639463232867727,
|
||
|
|
"learning_rate": 5.882352941176471e-06,
|
||
|
|
"loss": 1.3614,
|
||
|
|
"mean_token_accuracy": 0.6603095468373033,
|
||
|
|
"step": 5
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.061208875286916604,
|
||
|
|
"grad_norm": 1.8845571147500495,
|
||
|
|
"learning_rate": 1.1764705882352942e-05,
|
||
|
|
"loss": 1.2294,
|
||
|
|
"mean_token_accuracy": 0.6804112665679274,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09181331293037491,
|
||
|
|
"grad_norm": 0.9660031323201645,
|
||
|
|
"learning_rate": 1.7647058823529414e-05,
|
||
|
|
"loss": 1.1178,
|
||
|
|
"mean_token_accuracy": 0.6980193088248918,
|
||
|
|
"step": 15
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12241775057383321,
|
||
|
|
"grad_norm": 0.7758901578343,
|
||
|
|
"learning_rate": 1.9979171608653926e-05,
|
||
|
|
"loss": 1.0706,
|
||
|
|
"mean_token_accuracy": 0.7049760930682207,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1530221882172915,
|
||
|
|
"grad_norm": 0.6623960496251887,
|
||
|
|
"learning_rate": 1.9852201067560607e-05,
|
||
|
|
"loss": 1.0232,
|
||
|
|
"mean_token_accuracy": 0.7151910787493595,
|
||
|
|
"step": 25
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18362662586074982,
|
||
|
|
"grad_norm": 0.5861717419194806,
|
||
|
|
"learning_rate": 1.961129783872301e-05,
|
||
|
|
"loss": 0.9929,
|
||
|
|
"mean_token_accuracy": 0.7206187026715785,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21423106350420812,
|
||
|
|
"grad_norm": 0.570254952876122,
|
||
|
|
"learning_rate": 1.92592477719385e-05,
|
||
|
|
"loss": 0.9757,
|
||
|
|
"mean_token_accuracy": 0.7239615372510362,
|
||
|
|
"step": 35
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24483550114766642,
|
||
|
|
"grad_norm": 0.576438285009795,
|
||
|
|
"learning_rate": 1.880012203973536e-05,
|
||
|
|
"loss": 0.978,
|
||
|
|
"mean_token_accuracy": 0.7227768523027247,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2754399387911247,
|
||
|
|
"grad_norm": 0.54308496589203,
|
||
|
|
"learning_rate": 1.8239230057575542e-05,
|
||
|
|
"loss": 0.9573,
|
||
|
|
"mean_token_accuracy": 0.7269145541418254,
|
||
|
|
"step": 45
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.306044376434583,
|
||
|
|
"grad_norm": 0.5406620878458456,
|
||
|
|
"learning_rate": 1.7583058084785626e-05,
|
||
|
|
"loss": 0.9455,
|
||
|
|
"mean_token_accuracy": 0.7292135642290933,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3366488140780413,
|
||
|
|
"grad_norm": 0.5715247560846312,
|
||
|
|
"learning_rate": 1.683919421624611e-05,
|
||
|
|
"loss": 0.939,
|
||
|
|
"mean_token_accuracy": 0.7302346068090824,
|
||
|
|
"step": 55
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.36725325172149964,
|
||
|
|
"grad_norm": 0.5637735221725771,
|
||
|
|
"learning_rate": 1.6016240632249224e-05,
|
||
|
|
"loss": 0.9221,
|
||
|
|
"mean_token_accuracy": 0.734274136733944,
|
||
|
|
"step": 60
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3978576893649579,
|
||
|
|
"grad_norm": 0.5292344211321623,
|
||
|
|
"learning_rate": 1.512371412128424e-05,
|
||
|
|
"loss": 0.9052,
|
||
|
|
"mean_token_accuracy": 0.7389625183344621,
|
||
|
|
"step": 65
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.42846212700841624,
|
||
|
|
"grad_norm": 0.5252029559150836,
|
||
|
|
"learning_rate": 1.417193602612317e-05,
|
||
|
|
"loss": 0.9067,
|
||
|
|
"mean_token_accuracy": 0.7376815073153835,
|
||
|
|
"step": 70
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4590665646518745,
|
||
|
|
"grad_norm": 0.49710871969384784,
|
||
|
|
"learning_rate": 1.3171912885891063e-05,
|
||
|
|
"loss": 0.9115,
|
||
|
|
"mean_token_accuracy": 0.7359336540056602,
|
||
|
|
"step": 75
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.48967100229533284,
|
||
|
|
"grad_norm": 0.4957192093232857,
|
||
|
|
"learning_rate": 1.2135209154397962e-05,
|
||
|
|
"loss": 0.8917,
|
||
|
|
"mean_token_accuracy": 0.7414595493585209,
|
||
|
|
"step": 80
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5202754399387911,
|
||
|
|
"grad_norm": 0.5181596146926827,
|
||
|
|
"learning_rate": 1.1073813466641633e-05,
|
||
|
|
"loss": 0.9137,
|
||
|
|
"mean_token_accuracy": 0.7351079904890452,
|
||
|
|
"step": 85
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5508798775822494,
|
||
|
|
"grad_norm": 0.4883315332109265,
|
||
|
|
"learning_rate": 1e-05,
|
||
|
|
"loss": 0.8955,
|
||
|
|
"mean_token_accuracy": 0.7403773241174643,
|
||
|
|
"step": 90
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5814843152257078,
|
||
|
|
"grad_norm": 0.44231558881936617,
|
||
|
|
"learning_rate": 8.92618653335837e-06,
|
||
|
|
"loss": 0.9043,
|
||
|
|
"mean_token_accuracy": 0.7382262563051478,
|
||
|
|
"step": 95
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.612088752869166,
|
||
|
|
"grad_norm": 0.4540383775813619,
|
||
|
|
"learning_rate": 7.86479084560204e-06,
|
||
|
|
"loss": 0.9033,
|
||
|
|
"mean_token_accuracy": 0.7379618006575994,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.612088752869166,
|
||
|
|
"eval_loss": 0.9143875241279602,
|
||
|
|
"eval_mean_token_accuracy": 0.733248744327197,
|
||
|
|
"eval_runtime": 1.6272,
|
||
|
|
"eval_samples_per_second": 76.82,
|
||
|
|
"eval_steps_per_second": 2.458,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6426931905126243,
|
||
|
|
"grad_norm": 0.4556814538792029,
|
||
|
|
"learning_rate": 6.8280871141089415e-06,
|
||
|
|
"loss": 0.9009,
|
||
|
|
"mean_token_accuracy": 0.738005902941335,
|
||
|
|
"step": 105
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6732976281560826,
|
||
|
|
"grad_norm": 0.426504190091701,
|
||
|
|
"learning_rate": 5.828063973876834e-06,
|
||
|
|
"loss": 0.8931,
|
||
|
|
"mean_token_accuracy": 0.7399447865390983,
|
||
|
|
"step": 110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.703902065799541,
|
||
|
|
"grad_norm": 0.4470507695319038,
|
||
|
|
"learning_rate": 4.876285878715764e-06,
|
||
|
|
"loss": 0.8833,
|
||
|
|
"mean_token_accuracy": 0.7425298350188995,
|
||
|
|
"step": 115
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7345065034429993,
|
||
|
|
"grad_norm": 0.4306972155966625,
|
||
|
|
"learning_rate": 3.983759367750772e-06,
|
||
|
|
"loss": 0.8797,
|
||
|
|
"mean_token_accuracy": 0.7437409773046375,
|
||
|
|
"step": 120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7651109410864575,
|
||
|
|
"grad_norm": 0.4397210730509649,
|
||
|
|
"learning_rate": 3.1608057837538976e-06,
|
||
|
|
"loss": 0.9016,
|
||
|
|
"mean_token_accuracy": 0.7381355060905159,
|
||
|
|
"step": 125
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7957153787299158,
|
||
|
|
"grad_norm": 0.4281976846605365,
|
||
|
|
"learning_rate": 2.416941915214377e-06,
|
||
|
|
"loss": 0.883,
|
||
|
|
"mean_token_accuracy": 0.7420921944084118,
|
||
|
|
"step": 130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8263198163733741,
|
||
|
|
"grad_norm": 0.4084845117699037,
|
||
|
|
"learning_rate": 1.7607699424244583e-06,
|
||
|
|
"loss": 0.8891,
|
||
|
|
"mean_token_accuracy": 0.741522209985835,
|
||
|
|
"step": 135
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8569242540168325,
|
||
|
|
"grad_norm": 0.3912447309104287,
|
||
|
|
"learning_rate": 1.1998779602646438e-06,
|
||
|
|
"loss": 0.8779,
|
||
|
|
"mean_token_accuracy": 0.7436454928950815,
|
||
|
|
"step": 140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8875286916602907,
|
||
|
|
"grad_norm": 0.3799616199939088,
|
||
|
|
"learning_rate": 7.40752228061502e-07,
|
||
|
|
"loss": 0.8814,
|
||
|
|
"mean_token_accuracy": 0.7431115073707326,
|
||
|
|
"step": 145
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.918133129303749,
|
||
|
|
"grad_norm": 0.39425482521397076,
|
||
|
|
"learning_rate": 3.887021612769937e-07,
|
||
|
|
"loss": 0.8939,
|
||
|
|
"mean_token_accuracy": 0.7401144059050032,
|
||
|
|
"step": 150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9487375669472073,
|
||
|
|
"grad_norm": 0.4036052314719457,
|
||
|
|
"learning_rate": 1.4779893243939358e-07,
|
||
|
|
"loss": 0.8868,
|
||
|
|
"mean_token_accuracy": 0.7419396353528682,
|
||
|
|
"step": 155
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9793420045906657,
|
||
|
|
"grad_norm": 0.40481688584953807,
|
||
|
|
"learning_rate": 2.082839134607828e-08,
|
||
|
|
"loss": 0.8825,
|
||
|
|
"mean_token_accuracy": 0.7423017742571384,
|
||
|
|
"step": 160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9977046671767407,
|
||
|
|
"mean_token_accuracy": 0.742763755133944,
|
||
|
|
"step": 163,
|
||
|
|
"total_flos": 34627368517632.0,
|
||
|
|
"train_loss": 0.9493660312488766,
|
||
|
|
"train_runtime": 483.9234,
|
||
|
|
"train_samples_per_second": 43.207,
|
||
|
|
"train_steps_per_second": 0.337
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 5,
|
||
|
|
"max_steps": 163,
|
||
|
|
"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": false,
|
||
|
|
"should_training_stop": false
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 34627368517632.0,
|
||
|
|
"train_batch_size": 2,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|