345 lines
8.3 KiB
JSON
345 lines
8.3 KiB
JSON
|
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 1.469850121993726,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 2109,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.0006971070059254096,
|
||
|
|
"grad_norm": 5.6915082931518555,
|
||
|
|
"learning_rate": 0.0,
|
||
|
|
"loss": 1.5972,
|
||
|
|
"step": 1
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03485535029627048,
|
||
|
|
"grad_norm": 3.605487108230591,
|
||
|
|
"learning_rate": 1.9999434046461045e-05,
|
||
|
|
"loss": 1.348,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06971070059254096,
|
||
|
|
"grad_norm": 3.076913595199585,
|
||
|
|
"learning_rate": 1.996249692618611e-05,
|
||
|
|
"loss": 1.1239,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10456605088881143,
|
||
|
|
"grad_norm": 3.1423285007476807,
|
||
|
|
"learning_rate": 1.9868053167196865e-05,
|
||
|
|
"loss": 1.0576,
|
||
|
|
"step": 150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13942140118508192,
|
||
|
|
"grad_norm": 2.7687857151031494,
|
||
|
|
"learning_rate": 1.971664792831919e-05,
|
||
|
|
"loss": 1.0202,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17427675148135238,
|
||
|
|
"grad_norm": 2.9235846996307373,
|
||
|
|
"learning_rate": 1.9509155167802316e-05,
|
||
|
|
"loss": 0.9927,
|
||
|
|
"step": 250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20913210177762287,
|
||
|
|
"grad_norm": 2.6419451236724854,
|
||
|
|
"learning_rate": 1.9246772598559302e-05,
|
||
|
|
"loss": 0.9707,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24398745207389333,
|
||
|
|
"grad_norm": 2.7948801517486572,
|
||
|
|
"learning_rate": 1.8931014774594656e-05,
|
||
|
|
"loss": 0.9556,
|
||
|
|
"step": 350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27884280237016384,
|
||
|
|
"grad_norm": 2.7692606449127197,
|
||
|
|
"learning_rate": 1.8563704348526337e-05,
|
||
|
|
"loss": 0.9504,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3136981526664343,
|
||
|
|
"grad_norm": 2.7550599575042725,
|
||
|
|
"learning_rate": 1.8146961550666525e-05,
|
||
|
|
"loss": 0.9416,
|
||
|
|
"step": 450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34855350296270476,
|
||
|
|
"grad_norm": 2.628190517425537,
|
||
|
|
"learning_rate": 1.7683191950391142e-05,
|
||
|
|
"loss": 0.9246,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3834088532589753,
|
||
|
|
"grad_norm": 2.7417476177215576,
|
||
|
|
"learning_rate": 1.717507257044331e-05,
|
||
|
|
"loss": 0.9055,
|
||
|
|
"step": 550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.41826420355524574,
|
||
|
|
"grad_norm": 2.638144016265869,
|
||
|
|
"learning_rate": 1.6625536434323358e-05,
|
||
|
|
"loss": 0.9031,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4531195538515162,
|
||
|
|
"grad_norm": 2.6510379314422607,
|
||
|
|
"learning_rate": 1.6037755635962587e-05,
|
||
|
|
"loss": 0.9003,
|
||
|
|
"step": 650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.48797490414778666,
|
||
|
|
"grad_norm": 2.6764113903045654,
|
||
|
|
"learning_rate": 1.5415123029408046e-05,
|
||
|
|
"loss": 0.8829,
|
||
|
|
"step": 700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5228302544440572,
|
||
|
|
"grad_norm": 2.693927764892578,
|
||
|
|
"learning_rate": 1.4761232644210963e-05,
|
||
|
|
"loss": 0.8813,
|
||
|
|
"step": 750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5576856047403277,
|
||
|
|
"grad_norm": 2.5667612552642822,
|
||
|
|
"learning_rate": 1.4079858939567557e-05,
|
||
|
|
"loss": 0.8739,
|
||
|
|
"step": 800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5925409550365981,
|
||
|
|
"grad_norm": 2.4323246479034424,
|
||
|
|
"learning_rate": 1.3374935016963595e-05,
|
||
|
|
"loss": 0.8617,
|
||
|
|
"step": 850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6273963053328686,
|
||
|
|
"grad_norm": 2.5843608379364014,
|
||
|
|
"learning_rate": 1.2650529917086232e-05,
|
||
|
|
"loss": 0.8637,
|
||
|
|
"step": 900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6622516556291391,
|
||
|
|
"grad_norm": 2.498666286468506,
|
||
|
|
"learning_rate": 1.1910825132052356e-05,
|
||
|
|
"loss": 0.8522,
|
||
|
|
"step": 950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6971070059254095,
|
||
|
|
"grad_norm": 2.595188617706299,
|
||
|
|
"learning_rate": 1.1160090468532266e-05,
|
||
|
|
"loss": 0.8607,
|
||
|
|
"step": 1000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.73196235622168,
|
||
|
|
"grad_norm": 2.4613826274871826,
|
||
|
|
"learning_rate": 1.0402659401094154e-05,
|
||
|
|
"loss": 0.8424,
|
||
|
|
"step": 1050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7668177065179506,
|
||
|
|
"grad_norm": 2.597670078277588,
|
||
|
|
"learning_rate": 9.642904058037667e-06,
|
||
|
|
"loss": 0.8387,
|
||
|
|
"step": 1100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.801673056814221,
|
||
|
|
"grad_norm": 2.4956326484680176,
|
||
|
|
"learning_rate": 8.885209984106072e-06,
|
||
|
|
"loss": 0.83,
|
||
|
|
"step": 1150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8365284071104915,
|
||
|
|
"grad_norm": 2.4177091121673584,
|
||
|
|
"learning_rate": 8.133950825754511e-06,
|
||
|
|
"loss": 0.84,
|
||
|
|
"step": 1200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8713837574067619,
|
||
|
|
"grad_norm": 2.42484712600708,
|
||
|
|
"learning_rate": 7.393463085098886e-06,
|
||
|
|
"loss": 0.8304,
|
||
|
|
"step": 1250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9062391077030324,
|
||
|
|
"grad_norm": 2.5396852493286133,
|
||
|
|
"learning_rate": 6.6680210882734805e-06,
|
||
|
|
"loss": 0.8315,
|
||
|
|
"step": 1300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9410944579993029,
|
||
|
|
"grad_norm": 2.5951361656188965,
|
||
|
|
"learning_rate": 5.961812312687689e-06,
|
||
|
|
"loss": 0.8229,
|
||
|
|
"step": 1350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9759498082955733,
|
||
|
|
"grad_norm": 2.4725124835968018,
|
||
|
|
"learning_rate": 5.278913215600714e-06,
|
||
|
|
"loss": 0.815,
|
||
|
|
"step": 1400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0104566050888812,
|
||
|
|
"grad_norm": 2.5333492755889893,
|
||
|
|
"learning_rate": 4.623265703539146e-06,
|
||
|
|
"loss": 0.7775,
|
||
|
|
"step": 1450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0453119553851515,
|
||
|
|
"grad_norm": 2.530529499053955,
|
||
|
|
"learning_rate": 3.998654378383361e-06,
|
||
|
|
"loss": 0.693,
|
||
|
|
"step": 1500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.080167305681422,
|
||
|
|
"grad_norm": 2.500720262527466,
|
||
|
|
"learning_rate": 3.408684691465355e-06,
|
||
|
|
"loss": 0.7003,
|
||
|
|
"step": 1550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1150226559776926,
|
||
|
|
"grad_norm": 2.5019266605377197,
|
||
|
|
"learning_rate": 2.85676213177945e-06,
|
||
|
|
"loss": 0.6936,
|
||
|
|
"step": 1600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.149878006273963,
|
||
|
|
"grad_norm": 2.41332745552063,
|
||
|
|
"learning_rate": 2.3460725684379002e-06,
|
||
|
|
"loss": 0.699,
|
||
|
|
"step": 1650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1847333565702336,
|
||
|
|
"grad_norm": 2.46690034866333,
|
||
|
|
"learning_rate": 1.8795638608410016e-06,
|
||
|
|
"loss": 0.6873,
|
||
|
|
"step": 1700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.219588706866504,
|
||
|
|
"grad_norm": 2.2801692485809326,
|
||
|
|
"learning_rate": 1.4599288427134283e-06,
|
||
|
|
"loss": 0.6863,
|
||
|
|
"step": 1750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2544440571627744,
|
||
|
|
"grad_norm": 2.466907501220703,
|
||
|
|
"learning_rate": 1.0895897782283305e-06,
|
||
|
|
"loss": 0.6915,
|
||
|
|
"step": 1800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.289299407459045,
|
||
|
|
"grad_norm": 2.507089376449585,
|
||
|
|
"learning_rate": 7.706843799431985e-07,
|
||
|
|
"loss": 0.6878,
|
||
|
|
"step": 1850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3241547577553154,
|
||
|
|
"grad_norm": 2.3262147903442383,
|
||
|
|
"learning_rate": 5.050534692564358e-07,
|
||
|
|
"loss": 0.6811,
|
||
|
|
"step": 1900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.359010108051586,
|
||
|
|
"grad_norm": 2.4431302547454834,
|
||
|
|
"learning_rate": 2.94230350612239e-07,
|
||
|
|
"loss": 0.678,
|
||
|
|
"step": 1950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3938654583478565,
|
||
|
|
"grad_norm": 2.3482253551483154,
|
||
|
|
"learning_rate": 1.3943196078924247e-07,
|
||
|
|
"loss": 0.6829,
|
||
|
|
"step": 2000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.428720808644127,
|
||
|
|
"grad_norm": 2.5430848598480225,
|
||
|
|
"learning_rate": 4.155184436196669e-08,
|
||
|
|
"loss": 0.6795,
|
||
|
|
"step": 2050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4635761589403973,
|
||
|
|
"grad_norm": 2.450352430343628,
|
||
|
|
"learning_rate": 1.154995882924892e-09,
|
||
|
|
"loss": 0.6782,
|
||
|
|
"step": 2100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.469850121993726,
|
||
|
|
"step": 2109,
|
||
|
|
"total_flos": 7.41829675403182e+16,
|
||
|
|
"train_loss": 0.8431025520214372,
|
||
|
|
"train_runtime": 2810.3208,
|
||
|
|
"train_samples_per_second": 6.004,
|
||
|
|
"train_steps_per_second": 0.75
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 50,
|
||
|
|
"max_steps": 2109,
|
||
|
|
"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": true
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 7.41829675403182e+16,
|
||
|
|
"train_batch_size": 1,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|