434 lines
10 KiB
JSON
434 lines
10 KiB
JSON
{
|
|
"best_global_step": null,
|
|
"best_metric": null,
|
|
"best_model_checkpoint": null,
|
|
"epoch": 1.9965277777777777,
|
|
"eval_steps": 500,
|
|
"global_step": 1150,
|
|
"is_hyper_param_search": false,
|
|
"is_local_process_zero": true,
|
|
"is_world_process_zero": true,
|
|
"log_history": [
|
|
{
|
|
"epoch": 0.034722222222222224,
|
|
"grad_norm": 2.975640058517456,
|
|
"learning_rate": 1.6379310344827587e-06,
|
|
"loss": 2.6527,
|
|
"step": 20
|
|
},
|
|
{
|
|
"epoch": 0.06944444444444445,
|
|
"grad_norm": 2.329663038253784,
|
|
"learning_rate": 3.362068965517242e-06,
|
|
"loss": 2.4375,
|
|
"step": 40
|
|
},
|
|
{
|
|
"epoch": 0.10416666666666667,
|
|
"grad_norm": 1.8956992626190186,
|
|
"learning_rate": 5.086206896551724e-06,
|
|
"loss": 2.2566,
|
|
"step": 60
|
|
},
|
|
{
|
|
"epoch": 0.1388888888888889,
|
|
"grad_norm": 1.8109517097473145,
|
|
"learning_rate": 6.810344827586207e-06,
|
|
"loss": 2.0567,
|
|
"step": 80
|
|
},
|
|
{
|
|
"epoch": 0.1736111111111111,
|
|
"grad_norm": 2.382589817047119,
|
|
"learning_rate": 8.53448275862069e-06,
|
|
"loss": 1.9338,
|
|
"step": 100
|
|
},
|
|
{
|
|
"epoch": 0.20833333333333334,
|
|
"grad_norm": 2.378267526626587,
|
|
"learning_rate": 9.999793100349294e-06,
|
|
"loss": 1.9218,
|
|
"step": 120
|
|
},
|
|
{
|
|
"epoch": 0.24305555555555555,
|
|
"grad_norm": 2.6260263919830322,
|
|
"learning_rate": 9.987843743451796e-06,
|
|
"loss": 1.7895,
|
|
"step": 140
|
|
},
|
|
{
|
|
"epoch": 0.2777777777777778,
|
|
"grad_norm": 2.2361156940460205,
|
|
"learning_rate": 9.957553516178782e-06,
|
|
"loss": 1.7166,
|
|
"step": 160
|
|
},
|
|
{
|
|
"epoch": 0.3125,
|
|
"grad_norm": 2.345552682876587,
|
|
"learning_rate": 9.909033799150947e-06,
|
|
"loss": 1.7573,
|
|
"step": 180
|
|
},
|
|
{
|
|
"epoch": 0.3472222222222222,
|
|
"grad_norm": 2.4918200969696045,
|
|
"learning_rate": 9.842463004902127e-06,
|
|
"loss": 1.6668,
|
|
"step": 200
|
|
},
|
|
{
|
|
"epoch": 0.3819444444444444,
|
|
"grad_norm": 2.5107343196868896,
|
|
"learning_rate": 9.758085921836076e-06,
|
|
"loss": 1.6556,
|
|
"step": 220
|
|
},
|
|
{
|
|
"epoch": 0.4166666666666667,
|
|
"grad_norm": 2.69317626953125,
|
|
"learning_rate": 9.656212814111567e-06,
|
|
"loss": 1.6606,
|
|
"step": 240
|
|
},
|
|
{
|
|
"epoch": 0.4513888888888889,
|
|
"grad_norm": 2.759850025177002,
|
|
"learning_rate": 9.53721828076571e-06,
|
|
"loss": 1.6359,
|
|
"step": 260
|
|
},
|
|
{
|
|
"epoch": 0.4861111111111111,
|
|
"grad_norm": 2.751281976699829,
|
|
"learning_rate": 9.401539878270545e-06,
|
|
"loss": 1.635,
|
|
"step": 280
|
|
},
|
|
{
|
|
"epoch": 0.5208333333333334,
|
|
"grad_norm": 3.0062153339385986,
|
|
"learning_rate": 9.249676511588e-06,
|
|
"loss": 1.6278,
|
|
"step": 300
|
|
},
|
|
{
|
|
"epoch": 0.5555555555555556,
|
|
"grad_norm": 2.9671173095703125,
|
|
"learning_rate": 9.082186599639429e-06,
|
|
"loss": 1.6051,
|
|
"step": 320
|
|
},
|
|
{
|
|
"epoch": 0.5902777777777778,
|
|
"grad_norm": 2.6128034591674805,
|
|
"learning_rate": 8.899686021935554e-06,
|
|
"loss": 1.5921,
|
|
"step": 340
|
|
},
|
|
{
|
|
"epoch": 0.625,
|
|
"grad_norm": 2.750213384628296,
|
|
"learning_rate": 8.702845853917242e-06,
|
|
"loss": 1.5715,
|
|
"step": 360
|
|
},
|
|
{
|
|
"epoch": 0.6597222222222222,
|
|
"grad_norm": 2.4882044792175293,
|
|
"learning_rate": 8.492389899334572e-06,
|
|
"loss": 1.5607,
|
|
"step": 380
|
|
},
|
|
{
|
|
"epoch": 0.6944444444444444,
|
|
"grad_norm": 2.949469566345215,
|
|
"learning_rate": 8.269092028737885e-06,
|
|
"loss": 1.5498,
|
|
"step": 400
|
|
},
|
|
{
|
|
"epoch": 0.7291666666666666,
|
|
"grad_norm": 3.360396146774292,
|
|
"learning_rate": 8.033773333867498e-06,
|
|
"loss": 1.5246,
|
|
"step": 420
|
|
},
|
|
{
|
|
"epoch": 0.7638888888888888,
|
|
"grad_norm": 2.726001739501953,
|
|
"learning_rate": 7.78729910840572e-06,
|
|
"loss": 1.5172,
|
|
"step": 440
|
|
},
|
|
{
|
|
"epoch": 0.7986111111111112,
|
|
"grad_norm": 3.169776678085327,
|
|
"learning_rate": 7.530575666193283e-06,
|
|
"loss": 1.5212,
|
|
"step": 460
|
|
},
|
|
{
|
|
"epoch": 0.8333333333333334,
|
|
"grad_norm": 3.2836718559265137,
|
|
"learning_rate": 7.26454700860997e-06,
|
|
"loss": 1.5022,
|
|
"step": 480
|
|
},
|
|
{
|
|
"epoch": 0.8680555555555556,
|
|
"grad_norm": 3.3847920894622803,
|
|
"learning_rate": 6.990191353373876e-06,
|
|
"loss": 1.4968,
|
|
"step": 500
|
|
},
|
|
{
|
|
"epoch": 0.9027777777777778,
|
|
"grad_norm": 3.5661308765411377,
|
|
"learning_rate": 6.708517537523264e-06,
|
|
"loss": 1.477,
|
|
"step": 520
|
|
},
|
|
{
|
|
"epoch": 0.9375,
|
|
"grad_norm": 3.243346929550171,
|
|
"learning_rate": 6.420561307807713e-06,
|
|
"loss": 1.483,
|
|
"step": 540
|
|
},
|
|
{
|
|
"epoch": 0.9722222222222222,
|
|
"grad_norm": 2.8526790142059326,
|
|
"learning_rate": 6.12738151212918e-06,
|
|
"loss": 1.5018,
|
|
"step": 560
|
|
},
|
|
{
|
|
"epoch": 1.0069444444444444,
|
|
"grad_norm": 3.678152322769165,
|
|
"learning_rate": 5.830056206037482e-06,
|
|
"loss": 1.4604,
|
|
"step": 580
|
|
},
|
|
{
|
|
"epoch": 1.0416666666666667,
|
|
"grad_norm": 2.920789957046509,
|
|
"learning_rate": 5.529678688597081e-06,
|
|
"loss": 1.3984,
|
|
"step": 600
|
|
},
|
|
{
|
|
"epoch": 1.0763888888888888,
|
|
"grad_norm": 3.198030710220337,
|
|
"learning_rate": 5.2273534822017105e-06,
|
|
"loss": 1.4028,
|
|
"step": 620
|
|
},
|
|
{
|
|
"epoch": 1.1111111111111112,
|
|
"grad_norm": 3.737910270690918,
|
|
"learning_rate": 4.924192271119554e-06,
|
|
"loss": 1.3788,
|
|
"step": 640
|
|
},
|
|
{
|
|
"epoch": 1.1458333333333333,
|
|
"grad_norm": 2.9169039726257324,
|
|
"learning_rate": 4.621309813703385e-06,
|
|
"loss": 1.3787,
|
|
"step": 660
|
|
},
|
|
{
|
|
"epoch": 1.1805555555555556,
|
|
"grad_norm": 4.108018398284912,
|
|
"learning_rate": 4.319819843296952e-06,
|
|
"loss": 1.3854,
|
|
"step": 680
|
|
},
|
|
{
|
|
"epoch": 1.2152777777777777,
|
|
"grad_norm": 3.489570140838623,
|
|
"learning_rate": 4.020830972910433e-06,
|
|
"loss": 1.394,
|
|
"step": 700
|
|
},
|
|
{
|
|
"epoch": 1.25,
|
|
"grad_norm": 3.5733399391174316,
|
|
"learning_rate": 3.7254426187239567e-06,
|
|
"loss": 1.347,
|
|
"step": 720
|
|
},
|
|
{
|
|
"epoch": 1.2847222222222223,
|
|
"grad_norm": 3.479602098464966,
|
|
"learning_rate": 3.4347409574088896e-06,
|
|
"loss": 1.3647,
|
|
"step": 740
|
|
},
|
|
{
|
|
"epoch": 1.3194444444444444,
|
|
"grad_norm": 4.031429767608643,
|
|
"learning_rate": 3.149794932132331e-06,
|
|
"loss": 1.3879,
|
|
"step": 760
|
|
},
|
|
{
|
|
"epoch": 1.3541666666666667,
|
|
"grad_norm": 4.149319171905518,
|
|
"learning_rate": 2.871652321931161e-06,
|
|
"loss": 1.3618,
|
|
"step": 780
|
|
},
|
|
{
|
|
"epoch": 1.3888888888888888,
|
|
"grad_norm": 2.849364757537842,
|
|
"learning_rate": 2.601335888909005e-06,
|
|
"loss": 1.3687,
|
|
"step": 800
|
|
},
|
|
{
|
|
"epoch": 1.4236111111111112,
|
|
"grad_norm": 3.924811601638794,
|
|
"learning_rate": 2.339839617423318e-06,
|
|
"loss": 1.3329,
|
|
"step": 820
|
|
},
|
|
{
|
|
"epoch": 1.4583333333333333,
|
|
"grad_norm": 3.762462615966797,
|
|
"learning_rate": 2.0881250590915316e-06,
|
|
"loss": 1.3513,
|
|
"step": 840
|
|
},
|
|
{
|
|
"epoch": 1.4930555555555556,
|
|
"grad_norm": 3.951873302459717,
|
|
"learning_rate": 1.8471177970560712e-06,
|
|
"loss": 1.3655,
|
|
"step": 860
|
|
},
|
|
{
|
|
"epoch": 1.5277777777777777,
|
|
"grad_norm": 3.1432993412017822,
|
|
"learning_rate": 1.6177040425095664e-06,
|
|
"loss": 1.3276,
|
|
"step": 880
|
|
},
|
|
{
|
|
"epoch": 1.5625,
|
|
"grad_norm": 3.6409666538238525,
|
|
"learning_rate": 1.40072737599522e-06,
|
|
"loss": 1.3631,
|
|
"step": 900
|
|
},
|
|
{
|
|
"epoch": 1.5972222222222223,
|
|
"grad_norm": 3.062958002090454,
|
|
"learning_rate": 1.196985645464921e-06,
|
|
"loss": 1.3374,
|
|
"step": 920
|
|
},
|
|
{
|
|
"epoch": 1.6319444444444444,
|
|
"grad_norm": 4.230245590209961,
|
|
"learning_rate": 1.0072280325013185e-06,
|
|
"loss": 1.3438,
|
|
"step": 940
|
|
},
|
|
{
|
|
"epoch": 1.6666666666666665,
|
|
"grad_norm": 3.834761381149292,
|
|
"learning_rate": 8.321522974916968e-07,
|
|
"loss": 1.3611,
|
|
"step": 960
|
|
},
|
|
{
|
|
"epoch": 1.7013888888888888,
|
|
"grad_norm": 4.029264450073242,
|
|
"learning_rate": 6.724022138834341e-07,
|
|
"loss": 1.3323,
|
|
"step": 980
|
|
},
|
|
{
|
|
"epoch": 1.7361111111111112,
|
|
"grad_norm": 3.260033369064331,
|
|
"learning_rate": 5.285652009556075e-07,
|
|
"loss": 1.3273,
|
|
"step": 1000
|
|
},
|
|
{
|
|
"epoch": 1.7708333333333335,
|
|
"grad_norm": 4.1092071533203125,
|
|
"learning_rate": 4.0117016381130636e-07,
|
|
"loss": 1.3372,
|
|
"step": 1020
|
|
},
|
|
{
|
|
"epoch": 1.8055555555555556,
|
|
"grad_norm": 3.8922550678253174,
|
|
"learning_rate": 2.906855485332305e-07,
|
|
"loss": 1.3055,
|
|
"step": 1040
|
|
},
|
|
{
|
|
"epoch": 1.8402777777777777,
|
|
"grad_norm": 3.679527759552002,
|
|
"learning_rate": 1.975176196540185e-07,
|
|
"loss": 1.3342,
|
|
"step": 1060
|
|
},
|
|
{
|
|
"epoch": 1.875,
|
|
"grad_norm": 3.9715416431427,
|
|
"learning_rate": 1.2200896627521718e-07,
|
|
"loss": 1.3447,
|
|
"step": 1080
|
|
},
|
|
{
|
|
"epoch": 1.9097222222222223,
|
|
"grad_norm": 3.626037120819092,
|
|
"learning_rate": 6.443724232808146e-08,
|
|
"loss": 1.3219,
|
|
"step": 1100
|
|
},
|
|
{
|
|
"epoch": 1.9444444444444444,
|
|
"grad_norm": 4.019833564758301,
|
|
"learning_rate": 2.501414560839577e-08,
|
|
"loss": 1.2977,
|
|
"step": 1120
|
|
},
|
|
{
|
|
"epoch": 1.9791666666666665,
|
|
"grad_norm": 3.5683982372283936,
|
|
"learning_rate": 3.884639339534202e-09,
|
|
"loss": 1.3165,
|
|
"step": 1140
|
|
}
|
|
],
|
|
"logging_steps": 20,
|
|
"max_steps": 1152,
|
|
"num_input_tokens_seen": 0,
|
|
"num_train_epochs": 2,
|
|
"save_steps": 50,
|
|
"stateful_callbacks": {
|
|
"TrainerControl": {
|
|
"args": {
|
|
"should_epoch_stop": false,
|
|
"should_evaluate": false,
|
|
"should_log": false,
|
|
"should_save": true,
|
|
"should_training_stop": false
|
|
},
|
|
"attributes": {}
|
|
}
|
|
},
|
|
"total_flos": 2.3405557230187315e+17,
|
|
"train_batch_size": 4,
|
|
"trial_name": null,
|
|
"trial_params": null
|
|
}
|