{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.865800865800866,
"eval_steps": 15,
"global_step": 165,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.017316017316017316,
"eval_loss": 0.8123326897621155,
"eval_runtime": 21.4215,
"eval_samples_per_second": 4.528,
"eval_steps_per_second": 0.607,
"step": 1
},
{
"epoch": 0.05194805194805195,
"grad_norm": 1.44769287109375,
"learning_rate": 3e-05,
"loss": 3.0463,
"step": 3
},
{
"epoch": 0.1038961038961039,
"grad_norm": 0.9673851728439331,
"learning_rate": 6e-05,
"loss": 2.883,
"step": 6
},
{
"epoch": 0.15584415584415584,
"grad_norm": 1.2237026691436768,
"learning_rate": 9e-05,
"loss": 3.0048,
"step": 9
},
{
"epoch": 0.2077922077922078,
"grad_norm": 1.0807669162750244,
"learning_rate": 9.9962857531815e-05,
"loss": 2.8812,
"step": 12
},
{
"epoch": 0.2597402597402597,
"grad_norm": 0.8111446499824524,
"learning_rate": 9.976801044672608e-05,
"loss": 2.8254,
"step": 15
},
{
"epoch": 0.2597402597402597,
"eval_loss": 0.7221285700798035,
"eval_runtime": 21.8495,
"eval_samples_per_second": 4.439,
"eval_steps_per_second": 0.595,
"step": 15
},
{
"epoch": 0.3116883116883117,
"grad_norm": 0.7993184924125671,
"learning_rate": 9.940682350363912e-05,
"loss": 2.5783,
"step": 18
},
{
"epoch": 0.36363636363636365,
"grad_norm": 0.88717120885849,
"learning_rate": 9.888050389939172e-05,
"loss": 3.0288,
"step": 21
},
{
"epoch": 0.4155844155844156,
"grad_norm": 0.9446913003921509,
"learning_rate": 9.819081075450014e-05,
"loss": 2.6752,
"step": 24
},
{
"epoch": 0.4675324675324675,
"grad_norm": 0.7598215937614441,
"learning_rate": 9.734004923364257e-05,
"loss": 2.6636,
"step": 27
},
{
"epoch": 0.5194805194805194,
"grad_norm": 0.9138664603233337,
"learning_rate": 9.63310628410961e-05,
"loss": 2.6451,
"step": 30
},
{
"epoch": 0.5194805194805194,
"eval_loss": 0.6947051286697388,
"eval_runtime": 21.8569,
"eval_samples_per_second": 4.438,
"eval_steps_per_second": 0.595,
"step": 30
},
{
"epoch": 0.5714285714285714,
"grad_norm": 0.7083372473716736,
"learning_rate": 9.516722391687902e-05,
"loss": 2.5942,
"step": 33
},
{
"epoch": 0.6233766233766234,
"grad_norm": 0.6996327042579651,
"learning_rate": 9.38524223653626e-05,
"loss": 2.648,
"step": 36
},
{
"epoch": 0.6753246753246753,
"grad_norm": 0.7574095726013184,
"learning_rate": 9.239105265402525e-05,
"loss": 2.5124,
"step": 39
},
{
"epoch": 0.7272727272727273,
"grad_norm": 0.7147637605667114,
"learning_rate": 9.078799912580304e-05,
"loss": 2.652,
"step": 42
},
{
"epoch": 0.7792207792207793,
"grad_norm": 0.8316633701324463,
"learning_rate": 8.904861967412703e-05,
"loss": 2.5992,
"step": 45
},
{
"epoch": 0.7792207792207793,
"eval_loss": 0.6810405254364014,
"eval_runtime": 21.865,
"eval_samples_per_second": 4.436,
"eval_steps_per_second": 0.595,
"step": 45
},
{
"epoch": 0.8311688311688312,
"grad_norm": 0.7336857318878174,
"learning_rate": 8.717872783521047e-05,
"loss": 2.4763,
"step": 48
},
{
"epoch": 0.8831168831168831,
"grad_norm": 0.6515845060348511,
"learning_rate": 8.518457335743926e-05,
"loss": 2.5329,
"step": 51
},
{
"epoch": 0.935064935064935,
"grad_norm": 0.7598447203636169,
"learning_rate": 8.307282131280804e-05,
"loss": 2.6017,
"step": 54
},
{
"epoch": 0.987012987012987,
"grad_norm": 0.6881215572357178,
"learning_rate": 8.085052982021847e-05,
"loss": 2.5885,
"step": 57
},
{
"epoch": 1.0432900432900434,
"grad_norm": 0.6885347962379456,
"learning_rate": 7.85251264550948e-05,
"loss": 2.3153,
"step": 60
},
{
"epoch": 1.0432900432900434,
"eval_loss": 0.6708703637123108,
"eval_runtime": 21.8845,
"eval_samples_per_second": 4.432,
"eval_steps_per_second": 0.594,
"step": 60
},
{
"epoch": 1.0952380952380953,
"grad_norm": 0.8033732175827026,
"learning_rate": 7.610438342416319e-05,
"loss": 2.1058,
"step": 63
},
{
"epoch": 1.1471861471861473,
"grad_norm": 0.7723002433776855,
"learning_rate": 7.359639158836828e-05,
"loss": 2.0976,
"step": 66
},
{
"epoch": 1.199134199134199,
"grad_norm": 0.8295950889587402,
"learning_rate": 7.10095334207501e-05,
"loss": 2.1156,
"step": 69
},
{
"epoch": 1.251082251082251,
"grad_norm": 0.8891462683677673,
"learning_rate": 6.835245498966461e-05,
"loss": 2.3097,
"step": 72
},
{
"epoch": 1.303030303030303,
"grad_norm": 0.8947754502296448,
"learning_rate": 6.563403706098833e-05,
"loss": 2.2014,
"step": 75
},
{
"epoch": 1.303030303030303,
"eval_loss": 0.6706818342208862,
"eval_runtime": 21.8884,
"eval_samples_per_second": 4.432,
"eval_steps_per_second": 0.594,
"step": 75
},
{
"epoch": 1.354978354978355,
"grad_norm": 0.8432238101959229,
"learning_rate": 6.286336541589224e-05,
"loss": 2.2449,
"step": 78
},
{
"epoch": 1.406926406926407,
"grad_norm": 0.9027056097984314,
"learning_rate": 6.004970048339226e-05,
"loss": 2.2785,
"step": 81
},
{
"epoch": 1.4588744588744589,
"grad_norm": 0.977626621723175,
"learning_rate": 5.7202446389173223e-05,
"loss": 2.1978,
"step": 84
},
{
"epoch": 1.5108225108225108,
"grad_norm": 1.0718011856079102,
"learning_rate": 5.433111952413495e-05,
"loss": 2.1774,
"step": 87
},
{
"epoch": 1.5627705627705628,
"grad_norm": 1.034270167350769,
"learning_rate": 5.144531673771363e-05,
"loss": 2.1726,
"step": 90
},
{
"epoch": 1.5627705627705628,
"eval_loss": 0.6731178164482117,
"eval_runtime": 21.8631,
"eval_samples_per_second": 4.437,
"eval_steps_per_second": 0.595,
"step": 90
},
{
"epoch": 1.6147186147186146,
"grad_norm": 0.9706316590309143,
"learning_rate": 4.855468326228638e-05,
"loss": 2.0843,
"step": 93
},
{
"epoch": 1.6666666666666665,
"grad_norm": 1.0962471961975098,
"learning_rate": 4.566888047586507e-05,
"loss": 2.0797,
"step": 96
},
{
"epoch": 1.7186147186147185,
"grad_norm": 0.9773306846618652,
"learning_rate": 4.27975536108268e-05,
"loss": 2.1115,
"step": 99
},
{
"epoch": 1.7705627705627704,
"grad_norm": 0.9265680909156799,
"learning_rate": 3.9950299516607766e-05,
"loss": 2.1697,
"step": 102
},
{
"epoch": 1.8225108225108224,
"grad_norm": 1.1050467491149902,
"learning_rate": 3.713663458410779e-05,
"loss": 2.157,
"step": 105
},
{
"epoch": 1.8225108225108224,
"eval_loss": 0.6672056317329407,
"eval_runtime": 21.8494,
"eval_samples_per_second": 4.439,
"eval_steps_per_second": 0.595,
"step": 105
},
{
"epoch": 1.8744588744588744,
"grad_norm": 1.011114239692688,
"learning_rate": 3.43659629390117e-05,
"loss": 2.2031,
"step": 108
},
{
"epoch": 1.9264069264069263,
"grad_norm": 0.9879550933837891,
"learning_rate": 3.16475450103354e-05,
"loss": 2.1106,
"step": 111
},
{
"epoch": 1.9783549783549783,
"grad_norm": 1.0455985069274902,
"learning_rate": 2.899046657924992e-05,
"loss": 2.0031,
"step": 114
},
{
"epoch": 2.034632034632035,
"grad_norm": 1.1132123470306396,
"learning_rate": 2.6403608411631742e-05,
"loss": 1.843,
"step": 117
},
{
"epoch": 2.0865800865800868,
"grad_norm": 1.035132646560669,
"learning_rate": 2.389561657583681e-05,
"loss": 1.6831,
"step": 120
},
{
"epoch": 2.0865800865800868,
"eval_loss": 0.6776129603385925,
"eval_runtime": 21.8631,
"eval_samples_per_second": 4.437,
"eval_steps_per_second": 0.595,
"step": 120
},
{
"epoch": 2.1385281385281387,
"grad_norm": 1.176239252090454,
"learning_rate": 2.1474873544905205e-05,
"loss": 1.7413,
"step": 123
},
{
"epoch": 2.1904761904761907,
"grad_norm": 1.239511489868164,
"learning_rate": 1.914947017978153e-05,
"loss": 1.8036,
"step": 126
},
{
"epoch": 2.242424242424242,
"grad_norm": 1.0849404335021973,
"learning_rate": 1.692717868719195e-05,
"loss": 1.9443,
"step": 129
},
{
"epoch": 2.2943722943722946,
"grad_norm": 1.1872035264968872,
"learning_rate": 1.4815426642560754e-05,
"loss": 1.7633,
"step": 132
},
{
"epoch": 2.346320346320346,
"grad_norm": 1.1865441799163818,
"learning_rate": 1.2821272164789544e-05,
"loss": 1.8487,
"step": 135
},
{
"epoch": 2.346320346320346,
"eval_loss": 0.6821556687355042,
"eval_runtime": 21.8619,
"eval_samples_per_second": 4.437,
"eval_steps_per_second": 0.595,
"step": 135
},
{
"epoch": 2.398268398268398,
"grad_norm": 1.1166489124298096,
"learning_rate": 1.0951380325872979e-05,
"loss": 1.5943,
"step": 138
},
{
"epoch": 2.45021645021645,
"grad_norm": 1.1581447124481201,
"learning_rate": 9.212000874196953e-06,
"loss": 1.7557,
"step": 141
},
{
"epoch": 2.502164502164502,
"grad_norm": 1.1575648784637451,
"learning_rate": 7.60894734597476e-06,
"loss": 1.6507,
"step": 144
},
{
"epoch": 2.554112554112554,
"grad_norm": 1.119139313697815,
"learning_rate": 6.147577634637414e-06,
"loss": 1.7355,
"step": 147
},
{
"epoch": 2.606060606060606,
"grad_norm": 1.2103753089904785,
"learning_rate": 4.832776083120982e-06,
"loss": 1.6898,
"step": 150
},
{
"epoch": 2.606060606060606,
"eval_loss": 0.6833479404449463,
"eval_runtime": 21.8503,
"eval_samples_per_second": 4.439,
"eval_steps_per_second": 0.595,
"step": 150
},
{
"epoch": 2.658008658008658,
"grad_norm": 1.2263282537460327,
"learning_rate": 3.668937158903901e-06,
"loss": 1.7352,
"step": 153
},
{
"epoch": 2.70995670995671,
"grad_norm": 1.3095903396606445,
"learning_rate": 2.6599507663574384e-06,
"loss": 1.8122,
"step": 156
},
{
"epoch": 2.761904761904762,
"grad_norm": 1.2558013200759888,
"learning_rate": 1.8091892454998594e-06,
"loss": 1.6398,
"step": 159
},
{
"epoch": 2.813852813852814,
"grad_norm": 1.2757107019424438,
"learning_rate": 1.1194961006082972e-06,
"loss": 1.7014,
"step": 162
},
{
"epoch": 2.865800865800866,
"grad_norm": 1.1844847202301025,
"learning_rate": 5.931764963608866e-07,
"loss": 1.8026,
"step": 165
},
{
"epoch": 2.865800865800866,
"eval_loss": 0.6837310194969177,
"eval_runtime": 43.4558,
"eval_samples_per_second": 2.232,
"eval_steps_per_second": 0.299,
"step": 165
}
],
"logging_steps": 3,
"max_steps": 173,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 15,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 8.481582471551386e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}