235 lines
6.4 KiB
JSON
235 lines
6.4 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 0.78125,
|
|
"eval_steps": 500,
|
|
"global_step": 1000,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"entropy": 0.5879646396264434,
|
|
"epoch": 0.0390625,
|
|
"grad_norm": 2.078125,
|
|
"learning_rate": 3.828125000000001e-06,
|
|
"loss": 0.5608741760253906,
|
|
"mean_token_accuracy": 0.8732133361697197,
|
|
"num_tokens": 309908.0,
|
|
"step": 50
|
|
},
|
|
{
|
|
"entropy": 0.09950160900130868,
|
|
"epoch": 0.078125,
|
|
"grad_norm": 2.296875,
|
|
"learning_rate": 7.734375e-06,
|
|
"loss": 0.08675869941711425,
|
|
"mean_token_accuracy": 0.9764643028378487,
|
|
"num_tokens": 618662.0,
|
|
"step": 100
|
|
},
|
|
{
|
|
"entropy": 0.06861632030457258,
|
|
"epoch": 0.1171875,
|
|
"grad_norm": 0.97265625,
|
|
"learning_rate": 1.1640625000000002e-05,
|
|
"loss": 0.059808659553527835,
|
|
"mean_token_accuracy": 0.9829691171646118,
|
|
"num_tokens": 923343.0,
|
|
"step": 150
|
|
},
|
|
{
|
|
"entropy": 0.06371763135306537,
|
|
"epoch": 0.15625,
|
|
"grad_norm": 1.125,
|
|
"learning_rate": 1.5546875e-05,
|
|
"loss": 0.05767580032348633,
|
|
"mean_token_accuracy": 0.9834094086289406,
|
|
"num_tokens": 1231860.0,
|
|
"step": 200
|
|
},
|
|
{
|
|
"entropy": 0.05956988543272018,
|
|
"epoch": 0.1953125,
|
|
"grad_norm": 1.1796875,
|
|
"learning_rate": 1.9453125e-05,
|
|
"loss": 0.052296199798583985,
|
|
"mean_token_accuracy": 0.9851513150334358,
|
|
"num_tokens": 1539428.0,
|
|
"step": 250
|
|
},
|
|
{
|
|
"entropy": 0.06213963699527085,
|
|
"epoch": 0.234375,
|
|
"grad_norm": 2.515625,
|
|
"learning_rate": 1.9626736111111114e-05,
|
|
"loss": 0.057163243293762205,
|
|
"mean_token_accuracy": 0.9842589551210403,
|
|
"num_tokens": 1844184.0,
|
|
"step": 300
|
|
},
|
|
{
|
|
"entropy": 0.051620940826833245,
|
|
"epoch": 0.2734375,
|
|
"grad_norm": 0.6015625,
|
|
"learning_rate": 1.9192708333333335e-05,
|
|
"loss": 0.04214866161346435,
|
|
"mean_token_accuracy": 0.9875012004375457,
|
|
"num_tokens": 2145404.0,
|
|
"step": 350
|
|
},
|
|
{
|
|
"entropy": 0.032940625557675955,
|
|
"epoch": 0.3125,
|
|
"grad_norm": 1.3046875,
|
|
"learning_rate": 1.8758680555555557e-05,
|
|
"loss": 0.0279494047164917,
|
|
"mean_token_accuracy": 0.9905313903093338,
|
|
"num_tokens": 2454513.0,
|
|
"step": 400
|
|
},
|
|
{
|
|
"entropy": 0.03187832637690008,
|
|
"epoch": 0.3515625,
|
|
"grad_norm": 0.130859375,
|
|
"learning_rate": 1.8324652777777778e-05,
|
|
"loss": 0.024850430488586425,
|
|
"mean_token_accuracy": 0.9914047086238861,
|
|
"num_tokens": 2763494.0,
|
|
"step": 450
|
|
},
|
|
{
|
|
"entropy": 0.029858413645997645,
|
|
"epoch": 0.390625,
|
|
"grad_norm": 0.279296875,
|
|
"learning_rate": 1.7890625000000003e-05,
|
|
"loss": 0.022111225128173827,
|
|
"mean_token_accuracy": 0.9919667416810989,
|
|
"num_tokens": 3070032.0,
|
|
"step": 500
|
|
},
|
|
{
|
|
"entropy": 0.032282271897420285,
|
|
"epoch": 0.4296875,
|
|
"grad_norm": 0.099609375,
|
|
"learning_rate": 1.7456597222222224e-05,
|
|
"loss": 0.024263689517974852,
|
|
"mean_token_accuracy": 0.9916365548968316,
|
|
"num_tokens": 3371923.0,
|
|
"step": 550
|
|
},
|
|
{
|
|
"entropy": 0.02768919712398201,
|
|
"epoch": 0.46875,
|
|
"grad_norm": 0.154296875,
|
|
"learning_rate": 1.7022569444444446e-05,
|
|
"loss": 0.02068368673324585,
|
|
"mean_token_accuracy": 0.9922330856323243,
|
|
"num_tokens": 3681052.0,
|
|
"step": 600
|
|
},
|
|
{
|
|
"entropy": 0.027916706460528077,
|
|
"epoch": 0.5078125,
|
|
"grad_norm": 0.146484375,
|
|
"learning_rate": 1.6588541666666667e-05,
|
|
"loss": 0.02017390251159668,
|
|
"mean_token_accuracy": 0.9925840869545937,
|
|
"num_tokens": 3985154.0,
|
|
"step": 650
|
|
},
|
|
{
|
|
"entropy": 0.024339500861242414,
|
|
"epoch": 0.546875,
|
|
"grad_norm": 0.09521484375,
|
|
"learning_rate": 1.615451388888889e-05,
|
|
"loss": 0.017752493619918822,
|
|
"mean_token_accuracy": 0.992968330681324,
|
|
"num_tokens": 4294960.0,
|
|
"step": 700
|
|
},
|
|
{
|
|
"entropy": 0.02500485216733068,
|
|
"epoch": 0.5859375,
|
|
"grad_norm": 0.08837890625,
|
|
"learning_rate": 1.5720486111111114e-05,
|
|
"loss": 0.018282731771469118,
|
|
"mean_token_accuracy": 0.9928615567088127,
|
|
"num_tokens": 4596846.0,
|
|
"step": 750
|
|
},
|
|
{
|
|
"entropy": 0.023879703185521066,
|
|
"epoch": 0.625,
|
|
"grad_norm": 0.158203125,
|
|
"learning_rate": 1.5286458333333335e-05,
|
|
"loss": 0.017627782821655273,
|
|
"mean_token_accuracy": 0.9931773543357849,
|
|
"num_tokens": 4904237.0,
|
|
"step": 800
|
|
},
|
|
{
|
|
"entropy": 0.02569641772657633,
|
|
"epoch": 0.6640625,
|
|
"grad_norm": 0.09765625,
|
|
"learning_rate": 1.4852430555555556e-05,
|
|
"loss": 0.01919528007507324,
|
|
"mean_token_accuracy": 0.9929538157582283,
|
|
"num_tokens": 5205258.0,
|
|
"step": 850
|
|
},
|
|
{
|
|
"entropy": 0.02546143492218107,
|
|
"epoch": 0.703125,
|
|
"grad_norm": 0.1162109375,
|
|
"learning_rate": 1.4418402777777778e-05,
|
|
"loss": 0.018223344087600707,
|
|
"mean_token_accuracy": 0.9930938303470611,
|
|
"num_tokens": 5511048.0,
|
|
"step": 900
|
|
},
|
|
{
|
|
"entropy": 0.023580400524660945,
|
|
"epoch": 0.7421875,
|
|
"grad_norm": 0.1611328125,
|
|
"learning_rate": 1.3984375000000001e-05,
|
|
"loss": 0.01719343423843384,
|
|
"mean_token_accuracy": 0.9934566602110863,
|
|
"num_tokens": 5824149.0,
|
|
"step": 950
|
|
},
|
|
{
|
|
"entropy": 0.022929171579889954,
|
|
"epoch": 0.78125,
|
|
"grad_norm": 0.19921875,
|
|
"learning_rate": 1.3550347222222224e-05,
|
|
"loss": 0.016624774932861328,
|
|
"mean_token_accuracy": 0.9935434520244598,
|
|
"num_tokens": 6133395.0,
|
|
"step": 1000
|
|
}
|
|
],
|
|
"logging_steps": 50,
|
|
"max_steps": 2560,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 2,
|
|
"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": 2.885719471186084e+17,
|
|
"train_batch_size": 2,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|