401 lines
9.6 KiB
JSON
401 lines
9.6 KiB
JSON
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 3.0,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 519,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.057971014492753624,
|
||
|
|
"grad_norm": 29.952982357696865,
|
||
|
|
"learning_rate": 1.7307692307692308e-06,
|
||
|
|
"loss": 2.009,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11594202898550725,
|
||
|
|
"grad_norm": 11.512139435162283,
|
||
|
|
"learning_rate": 3.653846153846154e-06,
|
||
|
|
"loss": 1.8239,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17391304347826086,
|
||
|
|
"grad_norm": 6.626968877090522,
|
||
|
|
"learning_rate": 5.576923076923077e-06,
|
||
|
|
"loss": 1.1081,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2318840579710145,
|
||
|
|
"grad_norm": 5.700821840146368,
|
||
|
|
"learning_rate": 7.500000000000001e-06,
|
||
|
|
"loss": 0.4989,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2898550724637681,
|
||
|
|
"grad_norm": 3.4089761833344654,
|
||
|
|
"learning_rate": 9.423076923076923e-06,
|
||
|
|
"loss": 0.174,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34782608695652173,
|
||
|
|
"grad_norm": 0.5938704052704902,
|
||
|
|
"learning_rate": 9.994457294322858e-06,
|
||
|
|
"loss": 0.0671,
|
||
|
|
"step": 60
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4057971014492754,
|
||
|
|
"grad_norm": 0.3578679075693625,
|
||
|
|
"learning_rate": 9.967338926801066e-06,
|
||
|
|
"loss": 0.0486,
|
||
|
|
"step": 70
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.463768115942029,
|
||
|
|
"grad_norm": 0.27829454373593665,
|
||
|
|
"learning_rate": 9.917749373597506e-06,
|
||
|
|
"loss": 0.0374,
|
||
|
|
"step": 80
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5217391304347826,
|
||
|
|
"grad_norm": 0.41774241578480886,
|
||
|
|
"learning_rate": 9.84591296731245e-06,
|
||
|
|
"loss": 0.0304,
|
||
|
|
"step": 90
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5797101449275363,
|
||
|
|
"grad_norm": 0.7351453111684326,
|
||
|
|
"learning_rate": 9.752154680581783e-06,
|
||
|
|
"loss": 0.0313,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6376811594202898,
|
||
|
|
"grad_norm": 0.6835184704362615,
|
||
|
|
"learning_rate": 9.636898655969837e-06,
|
||
|
|
"loss": 0.0259,
|
||
|
|
"step": 110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6956521739130435,
|
||
|
|
"grad_norm": 0.7320159996448656,
|
||
|
|
"learning_rate": 9.500666287238573e-06,
|
||
|
|
"loss": 0.0267,
|
||
|
|
"step": 120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7536231884057971,
|
||
|
|
"grad_norm": 0.9706818562248596,
|
||
|
|
"learning_rate": 9.344073860673016e-06,
|
||
|
|
"loss": 0.022,
|
||
|
|
"step": 130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8115942028985508,
|
||
|
|
"grad_norm": 1.3747459132928828,
|
||
|
|
"learning_rate": 9.167829767133047e-06,
|
||
|
|
"loss": 0.0211,
|
||
|
|
"step": 140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8695652173913043,
|
||
|
|
"grad_norm": 0.3625030094557575,
|
||
|
|
"learning_rate": 8.972731297443722e-06,
|
||
|
|
"loss": 0.018,
|
||
|
|
"step": 150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.927536231884058,
|
||
|
|
"grad_norm": 0.32466805266951876,
|
||
|
|
"learning_rate": 8.759661035620992e-06,
|
||
|
|
"loss": 0.0151,
|
||
|
|
"step": 160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9855072463768116,
|
||
|
|
"grad_norm": 0.3393035460364341,
|
||
|
|
"learning_rate": 8.529582866249187e-06,
|
||
|
|
"loss": 0.0139,
|
||
|
|
"step": 170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0405797101449274,
|
||
|
|
"grad_norm": 0.3997122346830682,
|
||
|
|
"learning_rate": 8.283537614071987e-06,
|
||
|
|
"loss": 0.0113,
|
||
|
|
"step": 180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.098550724637681,
|
||
|
|
"grad_norm": 2.3217506351085917,
|
||
|
|
"learning_rate": 8.022638335522484e-06,
|
||
|
|
"loss": 0.0095,
|
||
|
|
"step": 190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.1565217391304348,
|
||
|
|
"grad_norm": 0.3656786470769303,
|
||
|
|
"learning_rate": 7.748065283492397e-06,
|
||
|
|
"loss": 0.0084,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.2144927536231884,
|
||
|
|
"grad_norm": 0.2974541884546707,
|
||
|
|
"learning_rate": 7.461060568118822e-06,
|
||
|
|
"loss": 0.0082,
|
||
|
|
"step": 210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.272463768115942,
|
||
|
|
"grad_norm": 0.3621483810363318,
|
||
|
|
"learning_rate": 7.162922537741937e-06,
|
||
|
|
"loss": 0.0079,
|
||
|
|
"step": 220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3304347826086955,
|
||
|
|
"grad_norm": 0.4604114786251555,
|
||
|
|
"learning_rate": 6.854999905453022e-06,
|
||
|
|
"loss": 0.0069,
|
||
|
|
"step": 230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.3884057971014494,
|
||
|
|
"grad_norm": 0.4826645672344626,
|
||
|
|
"learning_rate": 6.538685647803049e-06,
|
||
|
|
"loss": 0.0058,
|
||
|
|
"step": 240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.4463768115942028,
|
||
|
|
"grad_norm": 0.28726402847877164,
|
||
|
|
"learning_rate": 6.215410703272805e-06,
|
||
|
|
"loss": 0.0053,
|
||
|
|
"step": 250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5043478260869565,
|
||
|
|
"grad_norm": 0.1648274095907742,
|
||
|
|
"learning_rate": 5.8866374990112785e-06,
|
||
|
|
"loss": 0.0047,
|
||
|
|
"step": 260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.5623188405797102,
|
||
|
|
"grad_norm": 0.832116359446373,
|
||
|
|
"learning_rate": 5.5538533351260395e-06,
|
||
|
|
"loss": 0.0049,
|
||
|
|
"step": 270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6202898550724638,
|
||
|
|
"grad_norm": 0.35204832433816197,
|
||
|
|
"learning_rate": 5.218563656453609e-06,
|
||
|
|
"loss": 0.004,
|
||
|
|
"step": 280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.6782608695652175,
|
||
|
|
"grad_norm": 0.5085831024000373,
|
||
|
|
"learning_rate": 4.882285242246958e-06,
|
||
|
|
"loss": 0.0035,
|
||
|
|
"step": 290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.736231884057971,
|
||
|
|
"grad_norm": 0.5674575793970932,
|
||
|
|
"learning_rate": 4.546539344588486e-06,
|
||
|
|
"loss": 0.003,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.7942028985507248,
|
||
|
|
"grad_norm": 0.32551366582321456,
|
||
|
|
"learning_rate": 4.212844806568906e-06,
|
||
|
|
"loss": 0.003,
|
||
|
|
"step": 310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.8521739130434782,
|
||
|
|
"grad_norm": 0.18906388500680382,
|
||
|
|
"learning_rate": 3.88271119136386e-06,
|
||
|
|
"loss": 0.0022,
|
||
|
|
"step": 320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9101449275362319,
|
||
|
|
"grad_norm": 0.5113664546085201,
|
||
|
|
"learning_rate": 3.557631953290914e-06,
|
||
|
|
"loss": 0.0022,
|
||
|
|
"step": 330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.9681159420289855,
|
||
|
|
"grad_norm": 0.3861190420702106,
|
||
|
|
"learning_rate": 3.239077681739618e-06,
|
||
|
|
"loss": 0.0021,
|
||
|
|
"step": 340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.0231884057971015,
|
||
|
|
"grad_norm": 0.15897987765525454,
|
||
|
|
"learning_rate": 2.9284894485376057e-06,
|
||
|
|
"loss": 0.0017,
|
||
|
|
"step": 350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.081159420289855,
|
||
|
|
"grad_norm": 0.6453986830817099,
|
||
|
|
"learning_rate": 2.6272722888479152e-06,
|
||
|
|
"loss": 0.0011,
|
||
|
|
"step": 360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.139130434782609,
|
||
|
|
"grad_norm": 0.12469102836354488,
|
||
|
|
"learning_rate": 2.336788845088478e-06,
|
||
|
|
"loss": 0.0009,
|
||
|
|
"step": 370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.197101449275362,
|
||
|
|
"grad_norm": 0.5210826236391908,
|
||
|
|
"learning_rate": 2.058353202627417e-06,
|
||
|
|
"loss": 0.0009,
|
||
|
|
"step": 380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.255072463768116,
|
||
|
|
"grad_norm": 0.14107015806726528,
|
||
|
|
"learning_rate": 1.7932249451400863e-06,
|
||
|
|
"loss": 0.0007,
|
||
|
|
"step": 390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3130434782608695,
|
||
|
|
"grad_norm": 0.17094040993245266,
|
||
|
|
"learning_rate": 1.542603456520214e-06,
|
||
|
|
"loss": 0.0008,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.3710144927536234,
|
||
|
|
"grad_norm": 0.29173388058367744,
|
||
|
|
"learning_rate": 1.3076224951220413e-06,
|
||
|
|
"loss": 0.0008,
|
||
|
|
"step": 410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.428985507246377,
|
||
|
|
"grad_norm": 0.3029241786417147,
|
||
|
|
"learning_rate": 1.0893450648784736e-06,
|
||
|
|
"loss": 0.0007,
|
||
|
|
"step": 420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.4869565217391303,
|
||
|
|
"grad_norm": 0.17573523653808565,
|
||
|
|
"learning_rate": 8.887586064971859e-07,
|
||
|
|
"loss": 0.0007,
|
||
|
|
"step": 430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.544927536231884,
|
||
|
|
"grad_norm": 0.15567558268935872,
|
||
|
|
"learning_rate": 7.067705304887074e-07,
|
||
|
|
"loss": 0.0008,
|
||
|
|
"step": 440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.6028985507246376,
|
||
|
|
"grad_norm": 0.21209876832693972,
|
||
|
|
"learning_rate": 5.442041122341057e-07,
|
||
|
|
"loss": 0.0007,
|
||
|
|
"step": 450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.660869565217391,
|
||
|
|
"grad_norm": 0.09285510522354712,
|
||
|
|
"learning_rate": 4.0179476766211865e-07,
|
||
|
|
"loss": 0.0006,
|
||
|
|
"step": 460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.718840579710145,
|
||
|
|
"grad_norm": 0.13539077575014008,
|
||
|
|
"learning_rate": 2.8018672638378486e-07,
|
||
|
|
"loss": 0.0007,
|
||
|
|
"step": 470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.776811594202899,
|
||
|
|
"grad_norm": 0.9942412093476435,
|
||
|
|
"learning_rate": 1.7993011733458077e-07,
|
||
|
|
"loss": 0.0006,
|
||
|
|
"step": 480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8347826086956522,
|
||
|
|
"grad_norm": 0.13127636159227557,
|
||
|
|
"learning_rate": 1.0147848010803319e-07,
|
||
|
|
"loss": 0.0006,
|
||
|
|
"step": 490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.8927536231884057,
|
||
|
|
"grad_norm": 0.10614694682566984,
|
||
|
|
"learning_rate": 4.5186713238979385e-08,
|
||
|
|
"loss": 0.0005,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 2.9507246376811596,
|
||
|
|
"grad_norm": 0.13240069629057752,
|
||
|
|
"learning_rate": 1.1309468718013194e-08,
|
||
|
|
"loss": 0.0007,
|
||
|
|
"step": 510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 3.0,
|
||
|
|
"step": 519,
|
||
|
|
"total_flos": 457198286143488.0,
|
||
|
|
"train_loss": 0.11712381789758658,
|
||
|
|
"train_runtime": 11084.2911,
|
||
|
|
"train_samples_per_second": 2.988,
|
||
|
|
"train_steps_per_second": 0.047
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 10,
|
||
|
|
"max_steps": 519,
|
||
|
|
"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": 457198286143488.0,
|
||
|
|
"train_batch_size": 2,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|