3509 lines
85 KiB
JSON
3509 lines
85 KiB
JSON
|
|
{
|
||
|
|
"best_global_step": null,
|
||
|
|
"best_metric": null,
|
||
|
|
"best_model_checkpoint": null,
|
||
|
|
"epoch": 1.0,
|
||
|
|
"eval_steps": 500,
|
||
|
|
"global_step": 4954,
|
||
|
|
"is_hyper_param_search": false,
|
||
|
|
"is_local_process_zero": true,
|
||
|
|
"is_world_process_zero": true,
|
||
|
|
"log_history": [
|
||
|
|
{
|
||
|
|
"epoch": 0.002018927444794953,
|
||
|
|
"grad_norm": 2.3389563305266345,
|
||
|
|
"learning_rate": 1.8145161290322583e-07,
|
||
|
|
"loss": 0.7881,
|
||
|
|
"step": 10
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.004037854889589906,
|
||
|
|
"grad_norm": 2.024167058240825,
|
||
|
|
"learning_rate": 3.830645161290323e-07,
|
||
|
|
"loss": 0.8121,
|
||
|
|
"step": 20
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.006056782334384858,
|
||
|
|
"grad_norm": 1.4351269819866417,
|
||
|
|
"learning_rate": 5.846774193548388e-07,
|
||
|
|
"loss": 0.7797,
|
||
|
|
"step": 30
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.008075709779179811,
|
||
|
|
"grad_norm": 0.9310252821124614,
|
||
|
|
"learning_rate": 7.862903225806453e-07,
|
||
|
|
"loss": 0.7318,
|
||
|
|
"step": 40
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.010094637223974764,
|
||
|
|
"grad_norm": 0.9557517206875442,
|
||
|
|
"learning_rate": 9.879032258064516e-07,
|
||
|
|
"loss": 0.7018,
|
||
|
|
"step": 50
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.012113564668769715,
|
||
|
|
"grad_norm": 0.5829217778568054,
|
||
|
|
"learning_rate": 1.1895161290322582e-06,
|
||
|
|
"loss": 0.6708,
|
||
|
|
"step": 60
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.014132492113564668,
|
||
|
|
"grad_norm": 0.5112881851342524,
|
||
|
|
"learning_rate": 1.3911290322580644e-06,
|
||
|
|
"loss": 0.6417,
|
||
|
|
"step": 70
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.016151419558359623,
|
||
|
|
"grad_norm": 0.5140187941233466,
|
||
|
|
"learning_rate": 1.5927419354838712e-06,
|
||
|
|
"loss": 0.6227,
|
||
|
|
"step": 80
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.018170347003154574,
|
||
|
|
"grad_norm": 0.4872446144280347,
|
||
|
|
"learning_rate": 1.7943548387096777e-06,
|
||
|
|
"loss": 0.615,
|
||
|
|
"step": 90
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.02018927444794953,
|
||
|
|
"grad_norm": 0.5191969395080858,
|
||
|
|
"learning_rate": 1.9959677419354837e-06,
|
||
|
|
"loss": 0.6105,
|
||
|
|
"step": 100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.02220820189274448,
|
||
|
|
"grad_norm": 0.451802382492337,
|
||
|
|
"learning_rate": 2.1975806451612907e-06,
|
||
|
|
"loss": 0.5911,
|
||
|
|
"step": 110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.02422712933753943,
|
||
|
|
"grad_norm": 0.42163928465445194,
|
||
|
|
"learning_rate": 2.399193548387097e-06,
|
||
|
|
"loss": 0.5921,
|
||
|
|
"step": 120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.026246056782334385,
|
||
|
|
"grad_norm": 0.4552027864136731,
|
||
|
|
"learning_rate": 2.6008064516129032e-06,
|
||
|
|
"loss": 0.5818,
|
||
|
|
"step": 130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.028264984227129336,
|
||
|
|
"grad_norm": 0.44985963500335385,
|
||
|
|
"learning_rate": 2.8024193548387097e-06,
|
||
|
|
"loss": 0.5876,
|
||
|
|
"step": 140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03028391167192429,
|
||
|
|
"grad_norm": 0.5111630527751778,
|
||
|
|
"learning_rate": 3.0040322580645167e-06,
|
||
|
|
"loss": 0.5728,
|
||
|
|
"step": 150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.032302839116719245,
|
||
|
|
"grad_norm": 0.44240607593167414,
|
||
|
|
"learning_rate": 3.2056451612903228e-06,
|
||
|
|
"loss": 0.5643,
|
||
|
|
"step": 160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03432176656151419,
|
||
|
|
"grad_norm": 0.4865981695157576,
|
||
|
|
"learning_rate": 3.4072580645161293e-06,
|
||
|
|
"loss": 0.5617,
|
||
|
|
"step": 170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.03634069400630915,
|
||
|
|
"grad_norm": 0.6085288533687223,
|
||
|
|
"learning_rate": 3.6088709677419358e-06,
|
||
|
|
"loss": 0.5501,
|
||
|
|
"step": 180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0383596214511041,
|
||
|
|
"grad_norm": 0.47289717909858964,
|
||
|
|
"learning_rate": 3.8104838709677423e-06,
|
||
|
|
"loss": 0.559,
|
||
|
|
"step": 190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04037854889589906,
|
||
|
|
"grad_norm": 0.5389683685989423,
|
||
|
|
"learning_rate": 4.012096774193548e-06,
|
||
|
|
"loss": 0.5601,
|
||
|
|
"step": 200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.042397476340694004,
|
||
|
|
"grad_norm": 0.44905068010794785,
|
||
|
|
"learning_rate": 4.213709677419355e-06,
|
||
|
|
"loss": 0.546,
|
||
|
|
"step": 210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04441640378548896,
|
||
|
|
"grad_norm": 0.4738265544351665,
|
||
|
|
"learning_rate": 4.415322580645161e-06,
|
||
|
|
"loss": 0.5527,
|
||
|
|
"step": 220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04643533123028391,
|
||
|
|
"grad_norm": 0.48116551901846,
|
||
|
|
"learning_rate": 4.616935483870968e-06,
|
||
|
|
"loss": 0.5428,
|
||
|
|
"step": 230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.04845425867507886,
|
||
|
|
"grad_norm": 0.4514773717137442,
|
||
|
|
"learning_rate": 4.818548387096774e-06,
|
||
|
|
"loss": 0.5338,
|
||
|
|
"step": 240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.050473186119873815,
|
||
|
|
"grad_norm": 0.5092040454559039,
|
||
|
|
"learning_rate": 5.020161290322581e-06,
|
||
|
|
"loss": 0.5546,
|
||
|
|
"step": 250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05249211356466877,
|
||
|
|
"grad_norm": 0.45858657596184155,
|
||
|
|
"learning_rate": 5.221774193548388e-06,
|
||
|
|
"loss": 0.5408,
|
||
|
|
"step": 260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.054511041009463725,
|
||
|
|
"grad_norm": 0.47766211942842146,
|
||
|
|
"learning_rate": 5.423387096774194e-06,
|
||
|
|
"loss": 0.5427,
|
||
|
|
"step": 270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05652996845425867,
|
||
|
|
"grad_norm": 0.5108229913280311,
|
||
|
|
"learning_rate": 5.625e-06,
|
||
|
|
"loss": 0.5357,
|
||
|
|
"step": 280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.05854889589905363,
|
||
|
|
"grad_norm": 0.5232937676969737,
|
||
|
|
"learning_rate": 5.8266129032258064e-06,
|
||
|
|
"loss": 0.5316,
|
||
|
|
"step": 290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06056782334384858,
|
||
|
|
"grad_norm": 0.5241114539289579,
|
||
|
|
"learning_rate": 6.028225806451613e-06,
|
||
|
|
"loss": 0.5444,
|
||
|
|
"step": 300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06258675078864354,
|
||
|
|
"grad_norm": 0.9566252867340465,
|
||
|
|
"learning_rate": 6.2298387096774194e-06,
|
||
|
|
"loss": 0.5256,
|
||
|
|
"step": 310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06460567823343849,
|
||
|
|
"grad_norm": 0.4700069264487592,
|
||
|
|
"learning_rate": 6.431451612903226e-06,
|
||
|
|
"loss": 0.535,
|
||
|
|
"step": 320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06662460567823345,
|
||
|
|
"grad_norm": 0.46343674018411934,
|
||
|
|
"learning_rate": 6.633064516129033e-06,
|
||
|
|
"loss": 0.5281,
|
||
|
|
"step": 330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.06864353312302839,
|
||
|
|
"grad_norm": 0.46610206235853135,
|
||
|
|
"learning_rate": 6.834677419354839e-06,
|
||
|
|
"loss": 0.5324,
|
||
|
|
"step": 340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07066246056782334,
|
||
|
|
"grad_norm": 0.5217091961421014,
|
||
|
|
"learning_rate": 7.0362903225806454e-06,
|
||
|
|
"loss": 0.5321,
|
||
|
|
"step": 350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0726813880126183,
|
||
|
|
"grad_norm": 0.4532872443875366,
|
||
|
|
"learning_rate": 7.2379032258064515e-06,
|
||
|
|
"loss": 0.5235,
|
||
|
|
"step": 360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07470031545741325,
|
||
|
|
"grad_norm": 0.5481191485070994,
|
||
|
|
"learning_rate": 7.4395161290322585e-06,
|
||
|
|
"loss": 0.5203,
|
||
|
|
"step": 370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.0767192429022082,
|
||
|
|
"grad_norm": 0.4817884817896886,
|
||
|
|
"learning_rate": 7.641129032258065e-06,
|
||
|
|
"loss": 0.5289,
|
||
|
|
"step": 380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.07873817034700316,
|
||
|
|
"grad_norm": 0.463062633118143,
|
||
|
|
"learning_rate": 7.842741935483872e-06,
|
||
|
|
"loss": 0.5199,
|
||
|
|
"step": 390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08075709779179811,
|
||
|
|
"grad_norm": 0.5267129255916911,
|
||
|
|
"learning_rate": 8.044354838709678e-06,
|
||
|
|
"loss": 0.5239,
|
||
|
|
"step": 400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08277602523659305,
|
||
|
|
"grad_norm": 0.4703183395271576,
|
||
|
|
"learning_rate": 8.245967741935484e-06,
|
||
|
|
"loss": 0.5217,
|
||
|
|
"step": 410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08479495268138801,
|
||
|
|
"grad_norm": 0.4559407675951848,
|
||
|
|
"learning_rate": 8.447580645161291e-06,
|
||
|
|
"loss": 0.5187,
|
||
|
|
"step": 420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08681388012618296,
|
||
|
|
"grad_norm": 0.4638358181971901,
|
||
|
|
"learning_rate": 8.649193548387097e-06,
|
||
|
|
"loss": 0.5194,
|
||
|
|
"step": 430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.08883280757097792,
|
||
|
|
"grad_norm": 0.47768712570505284,
|
||
|
|
"learning_rate": 8.850806451612905e-06,
|
||
|
|
"loss": 0.513,
|
||
|
|
"step": 440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09085173501577287,
|
||
|
|
"grad_norm": 0.5140723294282747,
|
||
|
|
"learning_rate": 9.05241935483871e-06,
|
||
|
|
"loss": 0.5161,
|
||
|
|
"step": 450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09287066246056783,
|
||
|
|
"grad_norm": 0.4719590441343577,
|
||
|
|
"learning_rate": 9.254032258064517e-06,
|
||
|
|
"loss": 0.5033,
|
||
|
|
"step": 460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09488958990536278,
|
||
|
|
"grad_norm": 0.5031363320062582,
|
||
|
|
"learning_rate": 9.455645161290323e-06,
|
||
|
|
"loss": 0.5145,
|
||
|
|
"step": 470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09690851735015772,
|
||
|
|
"grad_norm": 0.5027908604602157,
|
||
|
|
"learning_rate": 9.65725806451613e-06,
|
||
|
|
"loss": 0.5178,
|
||
|
|
"step": 480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.09892744479495268,
|
||
|
|
"grad_norm": 0.5046628095943158,
|
||
|
|
"learning_rate": 9.858870967741936e-06,
|
||
|
|
"loss": 0.5069,
|
||
|
|
"step": 490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10094637223974763,
|
||
|
|
"grad_norm": 0.5522780839548007,
|
||
|
|
"learning_rate": 9.99998882617222e-06,
|
||
|
|
"loss": 0.5169,
|
||
|
|
"step": 500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10296529968454259,
|
||
|
|
"grad_norm": 0.48260445699138166,
|
||
|
|
"learning_rate": 9.999790181734308e-06,
|
||
|
|
"loss": 0.5067,
|
||
|
|
"step": 510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10498422712933754,
|
||
|
|
"grad_norm": 0.4875096921972213,
|
||
|
|
"learning_rate": 9.999343241367322e-06,
|
||
|
|
"loss": 0.5152,
|
||
|
|
"step": 520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1070031545741325,
|
||
|
|
"grad_norm": 0.5080882476390746,
|
||
|
|
"learning_rate": 9.998648027266884e-06,
|
||
|
|
"loss": 0.5131,
|
||
|
|
"step": 530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.10902208201892745,
|
||
|
|
"grad_norm": 0.5107719615388011,
|
||
|
|
"learning_rate": 9.997704573958209e-06,
|
||
|
|
"loss": 0.5053,
|
||
|
|
"step": 540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1110410094637224,
|
||
|
|
"grad_norm": 0.5361602322829081,
|
||
|
|
"learning_rate": 9.996512928294389e-06,
|
||
|
|
"loss": 0.5083,
|
||
|
|
"step": 550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11305993690851734,
|
||
|
|
"grad_norm": 0.4851941991561818,
|
||
|
|
"learning_rate": 9.995073149454057e-06,
|
||
|
|
"loss": 0.4997,
|
||
|
|
"step": 560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1150788643533123,
|
||
|
|
"grad_norm": 0.49258962377904003,
|
||
|
|
"learning_rate": 9.99338530893846e-06,
|
||
|
|
"loss": 0.516,
|
||
|
|
"step": 570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11709779179810725,
|
||
|
|
"grad_norm": 0.4835880405802275,
|
||
|
|
"learning_rate": 9.991449490567901e-06,
|
||
|
|
"loss": 0.5073,
|
||
|
|
"step": 580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.11911671924290221,
|
||
|
|
"grad_norm": 0.507245673357914,
|
||
|
|
"learning_rate": 9.989265790477581e-06,
|
||
|
|
"loss": 0.5014,
|
||
|
|
"step": 590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12113564668769716,
|
||
|
|
"grad_norm": 0.5146575464998254,
|
||
|
|
"learning_rate": 9.986834317112817e-06,
|
||
|
|
"loss": 0.5177,
|
||
|
|
"step": 600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12315457413249212,
|
||
|
|
"grad_norm": 0.4662328996304102,
|
||
|
|
"learning_rate": 9.984155191223663e-06,
|
||
|
|
"loss": 0.5054,
|
||
|
|
"step": 610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12517350157728707,
|
||
|
|
"grad_norm": 0.4571893765964666,
|
||
|
|
"learning_rate": 9.981228545858913e-06,
|
||
|
|
"loss": 0.5086,
|
||
|
|
"step": 620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12719242902208203,
|
||
|
|
"grad_norm": 0.4697827268486345,
|
||
|
|
"learning_rate": 9.978054526359493e-06,
|
||
|
|
"loss": 0.5029,
|
||
|
|
"step": 630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.12921135646687698,
|
||
|
|
"grad_norm": 0.5031062337107917,
|
||
|
|
"learning_rate": 9.97463329035124e-06,
|
||
|
|
"loss": 0.5021,
|
||
|
|
"step": 640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13123028391167194,
|
||
|
|
"grad_norm": 0.571029993720401,
|
||
|
|
"learning_rate": 9.970965007737081e-06,
|
||
|
|
"loss": 0.5077,
|
||
|
|
"step": 650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1332492113564669,
|
||
|
|
"grad_norm": 0.45753840087765046,
|
||
|
|
"learning_rate": 9.967049860688587e-06,
|
||
|
|
"loss": 0.4961,
|
||
|
|
"step": 660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13526813880126182,
|
||
|
|
"grad_norm": 0.5189371422618692,
|
||
|
|
"learning_rate": 9.962888043636931e-06,
|
||
|
|
"loss": 0.5036,
|
||
|
|
"step": 670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13728706624605677,
|
||
|
|
"grad_norm": 0.47066047930695143,
|
||
|
|
"learning_rate": 9.958479763263234e-06,
|
||
|
|
"loss": 0.4972,
|
||
|
|
"step": 680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.13930599369085173,
|
||
|
|
"grad_norm": 0.49380951082560154,
|
||
|
|
"learning_rate": 9.953825238488296e-06,
|
||
|
|
"loss": 0.4955,
|
||
|
|
"step": 690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14132492113564668,
|
||
|
|
"grad_norm": 0.49487805704511495,
|
||
|
|
"learning_rate": 9.948924700461727e-06,
|
||
|
|
"loss": 0.5013,
|
||
|
|
"step": 700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14334384858044164,
|
||
|
|
"grad_norm": 0.4768179474157311,
|
||
|
|
"learning_rate": 9.94377839255047e-06,
|
||
|
|
"loss": 0.4945,
|
||
|
|
"step": 710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1453627760252366,
|
||
|
|
"grad_norm": 0.5086022999656893,
|
||
|
|
"learning_rate": 9.938386570326707e-06,
|
||
|
|
"loss": 0.5033,
|
||
|
|
"step": 720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.14738170347003154,
|
||
|
|
"grad_norm": 0.4947936431118982,
|
||
|
|
"learning_rate": 9.93274950155518e-06,
|
||
|
|
"loss": 0.5013,
|
||
|
|
"step": 730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1494006309148265,
|
||
|
|
"grad_norm": 0.49370518434763616,
|
||
|
|
"learning_rate": 9.926867466179883e-06,
|
||
|
|
"loss": 0.4923,
|
||
|
|
"step": 740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15141955835962145,
|
||
|
|
"grad_norm": 0.49665569470371385,
|
||
|
|
"learning_rate": 9.920740756310156e-06,
|
||
|
|
"loss": 0.5038,
|
||
|
|
"step": 750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1534384858044164,
|
||
|
|
"grad_norm": 0.46813454785234254,
|
||
|
|
"learning_rate": 9.914369676206195e-06,
|
||
|
|
"loss": 0.4875,
|
||
|
|
"step": 760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15545741324921136,
|
||
|
|
"grad_norm": 0.49814972796278695,
|
||
|
|
"learning_rate": 9.907754542263926e-06,
|
||
|
|
"loss": 0.4939,
|
||
|
|
"step": 770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15747634069400632,
|
||
|
|
"grad_norm": 0.5218798292518447,
|
||
|
|
"learning_rate": 9.900895682999301e-06,
|
||
|
|
"loss": 0.501,
|
||
|
|
"step": 780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.15949526813880127,
|
||
|
|
"grad_norm": 0.5913103279447773,
|
||
|
|
"learning_rate": 9.893793439031982e-06,
|
||
|
|
"loss": 0.4927,
|
||
|
|
"step": 790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16151419558359623,
|
||
|
|
"grad_norm": 0.5125446596453367,
|
||
|
|
"learning_rate": 9.88644816306842e-06,
|
||
|
|
"loss": 0.4971,
|
||
|
|
"step": 800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16353312302839118,
|
||
|
|
"grad_norm": 0.4483762668624327,
|
||
|
|
"learning_rate": 9.878860219884347e-06,
|
||
|
|
"loss": 0.4981,
|
||
|
|
"step": 810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1655520504731861,
|
||
|
|
"grad_norm": 0.4787860866647759,
|
||
|
|
"learning_rate": 9.871029986306658e-06,
|
||
|
|
"loss": 0.4964,
|
||
|
|
"step": 820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16757097791798106,
|
||
|
|
"grad_norm": 0.5030394247546482,
|
||
|
|
"learning_rate": 9.862957851194694e-06,
|
||
|
|
"loss": 0.4891,
|
||
|
|
"step": 830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.16958990536277602,
|
||
|
|
"grad_norm": 0.5239985323894606,
|
||
|
|
"learning_rate": 9.854644215420933e-06,
|
||
|
|
"loss": 0.501,
|
||
|
|
"step": 840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17160883280757097,
|
||
|
|
"grad_norm": 0.4934736601735223,
|
||
|
|
"learning_rate": 9.846089491851089e-06,
|
||
|
|
"loss": 0.5012,
|
||
|
|
"step": 850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17362776025236593,
|
||
|
|
"grad_norm": 0.5649851363661309,
|
||
|
|
"learning_rate": 9.837294105323592e-06,
|
||
|
|
"loss": 0.501,
|
||
|
|
"step": 860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17564668769716088,
|
||
|
|
"grad_norm": 0.5032183289408247,
|
||
|
|
"learning_rate": 9.828258492628506e-06,
|
||
|
|
"loss": 0.5019,
|
||
|
|
"step": 870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.17766561514195583,
|
||
|
|
"grad_norm": 0.5408394773542722,
|
||
|
|
"learning_rate": 9.818983102485832e-06,
|
||
|
|
"loss": 0.4914,
|
||
|
|
"step": 880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1796845425867508,
|
||
|
|
"grad_norm": 0.47080262308715565,
|
||
|
|
"learning_rate": 9.809468395523225e-06,
|
||
|
|
"loss": 0.4874,
|
||
|
|
"step": 890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18170347003154574,
|
||
|
|
"grad_norm": 0.5707749159791521,
|
||
|
|
"learning_rate": 9.799714844253109e-06,
|
||
|
|
"loss": 0.4995,
|
||
|
|
"step": 900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1837223974763407,
|
||
|
|
"grad_norm": 0.48122327046673424,
|
||
|
|
"learning_rate": 9.789722933049227e-06,
|
||
|
|
"loss": 0.4892,
|
||
|
|
"step": 910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18574132492113565,
|
||
|
|
"grad_norm": 0.45082444705324104,
|
||
|
|
"learning_rate": 9.779493158122578e-06,
|
||
|
|
"loss": 0.4909,
|
||
|
|
"step": 920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1877602523659306,
|
||
|
|
"grad_norm": 0.5068314998947344,
|
||
|
|
"learning_rate": 9.769026027496773e-06,
|
||
|
|
"loss": 0.4881,
|
||
|
|
"step": 930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.18977917981072556,
|
||
|
|
"grad_norm": 0.44089908003969003,
|
||
|
|
"learning_rate": 9.758322060982813e-06,
|
||
|
|
"loss": 0.4923,
|
||
|
|
"step": 940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19179810725552052,
|
||
|
|
"grad_norm": 0.4786006068183123,
|
||
|
|
"learning_rate": 9.747381790153262e-06,
|
||
|
|
"loss": 0.4912,
|
||
|
|
"step": 950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19381703470031544,
|
||
|
|
"grad_norm": 0.4531421936277408,
|
||
|
|
"learning_rate": 9.736205758315864e-06,
|
||
|
|
"loss": 0.4744,
|
||
|
|
"step": 960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1958359621451104,
|
||
|
|
"grad_norm": 0.4492650045467347,
|
||
|
|
"learning_rate": 9.724794520486551e-06,
|
||
|
|
"loss": 0.485,
|
||
|
|
"step": 970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.19785488958990535,
|
||
|
|
"grad_norm": 0.4616235423638888,
|
||
|
|
"learning_rate": 9.713148643361884e-06,
|
||
|
|
"loss": 0.488,
|
||
|
|
"step": 980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.1998738170347003,
|
||
|
|
"grad_norm": 0.5078516721505932,
|
||
|
|
"learning_rate": 9.701268705290907e-06,
|
||
|
|
"loss": 0.4886,
|
||
|
|
"step": 990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20189274447949526,
|
||
|
|
"grad_norm": 0.5292871585531458,
|
||
|
|
"learning_rate": 9.689155296246431e-06,
|
||
|
|
"loss": 0.4897,
|
||
|
|
"step": 1000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20391167192429022,
|
||
|
|
"grad_norm": 0.4614297555636878,
|
||
|
|
"learning_rate": 9.67680901779573e-06,
|
||
|
|
"loss": 0.4809,
|
||
|
|
"step": 1010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20593059936908517,
|
||
|
|
"grad_norm": 0.4853770956696164,
|
||
|
|
"learning_rate": 9.664230483070669e-06,
|
||
|
|
"loss": 0.4848,
|
||
|
|
"step": 1020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20794952681388013,
|
||
|
|
"grad_norm": 0.45954875393494693,
|
||
|
|
"learning_rate": 9.651420316737254e-06,
|
||
|
|
"loss": 0.4817,
|
||
|
|
"step": 1030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.20996845425867508,
|
||
|
|
"grad_norm": 0.4539790244117171,
|
||
|
|
"learning_rate": 9.638379154964615e-06,
|
||
|
|
"loss": 0.4927,
|
||
|
|
"step": 1040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21198738170347003,
|
||
|
|
"grad_norm": 0.4999117191134038,
|
||
|
|
"learning_rate": 9.625107645393398e-06,
|
||
|
|
"loss": 0.4906,
|
||
|
|
"step": 1050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.214006309148265,
|
||
|
|
"grad_norm": 0.5315614747354604,
|
||
|
|
"learning_rate": 9.611606447103625e-06,
|
||
|
|
"loss": 0.4701,
|
||
|
|
"step": 1060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.21602523659305994,
|
||
|
|
"grad_norm": 0.4465543761029164,
|
||
|
|
"learning_rate": 9.597876230581946e-06,
|
||
|
|
"loss": 0.4844,
|
||
|
|
"step": 1070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2180441640378549,
|
||
|
|
"grad_norm": 0.5240713468337838,
|
||
|
|
"learning_rate": 9.583917677688348e-06,
|
||
|
|
"loss": 0.4802,
|
||
|
|
"step": 1080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22006309148264985,
|
||
|
|
"grad_norm": 0.4797396408083518,
|
||
|
|
"learning_rate": 9.56973148162229e-06,
|
||
|
|
"loss": 0.4822,
|
||
|
|
"step": 1090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2220820189274448,
|
||
|
|
"grad_norm": 0.5124904557581254,
|
||
|
|
"learning_rate": 9.555318346888287e-06,
|
||
|
|
"loss": 0.4887,
|
||
|
|
"step": 1100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22410094637223973,
|
||
|
|
"grad_norm": 0.448720333545146,
|
||
|
|
"learning_rate": 9.540678989260907e-06,
|
||
|
|
"loss": 0.4804,
|
||
|
|
"step": 1110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2261198738170347,
|
||
|
|
"grad_norm": 0.4479157257263002,
|
||
|
|
"learning_rate": 9.525814135749238e-06,
|
||
|
|
"loss": 0.4776,
|
||
|
|
"step": 1120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.22813880126182964,
|
||
|
|
"grad_norm": 0.4630993156519356,
|
||
|
|
"learning_rate": 9.510724524560783e-06,
|
||
|
|
"loss": 0.4753,
|
||
|
|
"step": 1130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2301577287066246,
|
||
|
|
"grad_norm": 0.46574160333254905,
|
||
|
|
"learning_rate": 9.495410905064795e-06,
|
||
|
|
"loss": 0.4802,
|
||
|
|
"step": 1140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23217665615141955,
|
||
|
|
"grad_norm": 0.532351729710861,
|
||
|
|
"learning_rate": 9.479874037755061e-06,
|
||
|
|
"loss": 0.4826,
|
||
|
|
"step": 1150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2341955835962145,
|
||
|
|
"grad_norm": 0.4629220502691385,
|
||
|
|
"learning_rate": 9.46411469421214e-06,
|
||
|
|
"loss": 0.4802,
|
||
|
|
"step": 1160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23621451104100946,
|
||
|
|
"grad_norm": 0.459458980682449,
|
||
|
|
"learning_rate": 9.448133657065045e-06,
|
||
|
|
"loss": 0.4863,
|
||
|
|
"step": 1170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.23823343848580442,
|
||
|
|
"grad_norm": 0.4652002127480942,
|
||
|
|
"learning_rate": 9.431931719952372e-06,
|
||
|
|
"loss": 0.4785,
|
||
|
|
"step": 1180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24025236593059937,
|
||
|
|
"grad_norm": 0.466540922054408,
|
||
|
|
"learning_rate": 9.415509687482892e-06,
|
||
|
|
"loss": 0.4719,
|
||
|
|
"step": 1190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24227129337539433,
|
||
|
|
"grad_norm": 0.49341259273484933,
|
||
|
|
"learning_rate": 9.398868375195587e-06,
|
||
|
|
"loss": 0.4926,
|
||
|
|
"step": 1200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24429022082018928,
|
||
|
|
"grad_norm": 0.4728385366149605,
|
||
|
|
"learning_rate": 9.382008609519159e-06,
|
||
|
|
"loss": 0.4741,
|
||
|
|
"step": 1210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.24630914826498423,
|
||
|
|
"grad_norm": 0.5027959432042377,
|
||
|
|
"learning_rate": 9.364931227730978e-06,
|
||
|
|
"loss": 0.4826,
|
||
|
|
"step": 1220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2483280757097792,
|
||
|
|
"grad_norm": 0.4883346568815589,
|
||
|
|
"learning_rate": 9.34763707791551e-06,
|
||
|
|
"loss": 0.4837,
|
||
|
|
"step": 1230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25034700315457414,
|
||
|
|
"grad_norm": 0.4689919758806841,
|
||
|
|
"learning_rate": 9.330127018922195e-06,
|
||
|
|
"loss": 0.4834,
|
||
|
|
"step": 1240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25236593059936907,
|
||
|
|
"grad_norm": 0.5054719137784303,
|
||
|
|
"learning_rate": 9.312401920322793e-06,
|
||
|
|
"loss": 0.4858,
|
||
|
|
"step": 1250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25438485804416405,
|
||
|
|
"grad_norm": 0.4829012063372003,
|
||
|
|
"learning_rate": 9.294462662368211e-06,
|
||
|
|
"loss": 0.4757,
|
||
|
|
"step": 1260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.256403785488959,
|
||
|
|
"grad_norm": 0.4627206332424085,
|
||
|
|
"learning_rate": 9.276310135944777e-06,
|
||
|
|
"loss": 0.4783,
|
||
|
|
"step": 1270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.25842271293375396,
|
||
|
|
"grad_norm": 0.4497460790448332,
|
||
|
|
"learning_rate": 9.257945242530002e-06,
|
||
|
|
"loss": 0.4785,
|
||
|
|
"step": 1280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2604416403785489,
|
||
|
|
"grad_norm": 0.44985051561838785,
|
||
|
|
"learning_rate": 9.239368894147812e-06,
|
||
|
|
"loss": 0.4786,
|
||
|
|
"step": 1290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.26246056782334387,
|
||
|
|
"grad_norm": 0.48240095905903635,
|
||
|
|
"learning_rate": 9.220582013323258e-06,
|
||
|
|
"loss": 0.4807,
|
||
|
|
"step": 1300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2644794952681388,
|
||
|
|
"grad_norm": 0.48132105969556493,
|
||
|
|
"learning_rate": 9.201585533036696e-06,
|
||
|
|
"loss": 0.4706,
|
||
|
|
"step": 1310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2664984227129338,
|
||
|
|
"grad_norm": 0.46675857398369297,
|
||
|
|
"learning_rate": 9.18238039667746e-06,
|
||
|
|
"loss": 0.4879,
|
||
|
|
"step": 1320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2685173501577287,
|
||
|
|
"grad_norm": 0.4665634722250575,
|
||
|
|
"learning_rate": 9.162967557997003e-06,
|
||
|
|
"loss": 0.4799,
|
||
|
|
"step": 1330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27053627760252363,
|
||
|
|
"grad_norm": 0.4432315628923224,
|
||
|
|
"learning_rate": 9.14334798106155e-06,
|
||
|
|
"loss": 0.4757,
|
||
|
|
"step": 1340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2725552050473186,
|
||
|
|
"grad_norm": 0.5000831797296084,
|
||
|
|
"learning_rate": 9.123522640204198e-06,
|
||
|
|
"loss": 0.4805,
|
||
|
|
"step": 1350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27457413249211354,
|
||
|
|
"grad_norm": 0.4441933490286196,
|
||
|
|
"learning_rate": 9.103492519976555e-06,
|
||
|
|
"loss": 0.4769,
|
||
|
|
"step": 1360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2765930599369085,
|
||
|
|
"grad_norm": 0.4805005293661477,
|
||
|
|
"learning_rate": 9.083258615099819e-06,
|
||
|
|
"loss": 0.4777,
|
||
|
|
"step": 1370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.27861198738170345,
|
||
|
|
"grad_norm": 0.4445460559749252,
|
||
|
|
"learning_rate": 9.062821930415404e-06,
|
||
|
|
"loss": 0.4728,
|
||
|
|
"step": 1380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28063091482649843,
|
||
|
|
"grad_norm": 0.4333540273890526,
|
||
|
|
"learning_rate": 9.042183480835018e-06,
|
||
|
|
"loss": 0.4705,
|
||
|
|
"step": 1390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28264984227129336,
|
||
|
|
"grad_norm": 0.48259999792627817,
|
||
|
|
"learning_rate": 9.021344291290275e-06,
|
||
|
|
"loss": 0.4782,
|
||
|
|
"step": 1400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28466876971608834,
|
||
|
|
"grad_norm": 0.439829603884435,
|
||
|
|
"learning_rate": 9.000305396681788e-06,
|
||
|
|
"loss": 0.4705,
|
||
|
|
"step": 1410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28668769716088327,
|
||
|
|
"grad_norm": 0.4776753948778827,
|
||
|
|
"learning_rate": 8.97906784182778e-06,
|
||
|
|
"loss": 0.4694,
|
||
|
|
"step": 1420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.28870662460567825,
|
||
|
|
"grad_norm": 0.49821375314109334,
|
||
|
|
"learning_rate": 8.95763268141219e-06,
|
||
|
|
"loss": 0.4698,
|
||
|
|
"step": 1430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2907255520504732,
|
||
|
|
"grad_norm": 0.4626475218122892,
|
||
|
|
"learning_rate": 8.936000979932304e-06,
|
||
|
|
"loss": 0.4766,
|
||
|
|
"step": 1440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29274447949526816,
|
||
|
|
"grad_norm": 0.4797061046529278,
|
||
|
|
"learning_rate": 8.914173811645884e-06,
|
||
|
|
"loss": 0.4773,
|
||
|
|
"step": 1450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.2947634069400631,
|
||
|
|
"grad_norm": 0.5067089149884695,
|
||
|
|
"learning_rate": 8.892152260517824e-06,
|
||
|
|
"loss": 0.4652,
|
||
|
|
"step": 1460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.29678233438485807,
|
||
|
|
"grad_norm": 0.5216910139465747,
|
||
|
|
"learning_rate": 8.869937420166312e-06,
|
||
|
|
"loss": 0.475,
|
||
|
|
"step": 1470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.298801261829653,
|
||
|
|
"grad_norm": 0.4655884354714463,
|
||
|
|
"learning_rate": 8.847530393808534e-06,
|
||
|
|
"loss": 0.4811,
|
||
|
|
"step": 1480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3008201892744479,
|
||
|
|
"grad_norm": 0.47533704121654896,
|
||
|
|
"learning_rate": 8.824932294205868e-06,
|
||
|
|
"loss": 0.4723,
|
||
|
|
"step": 1490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3028391167192429,
|
||
|
|
"grad_norm": 0.479585079565894,
|
||
|
|
"learning_rate": 8.802144243608633e-06,
|
||
|
|
"loss": 0.4728,
|
||
|
|
"step": 1500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30485804416403783,
|
||
|
|
"grad_norm": 0.4338897481002412,
|
||
|
|
"learning_rate": 8.779167373700359e-06,
|
||
|
|
"loss": 0.4687,
|
||
|
|
"step": 1510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3068769716088328,
|
||
|
|
"grad_norm": 0.51284609503485,
|
||
|
|
"learning_rate": 8.756002825541583e-06,
|
||
|
|
"loss": 0.4795,
|
||
|
|
"step": 1520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.30889589905362774,
|
||
|
|
"grad_norm": 0.5019515223957067,
|
||
|
|
"learning_rate": 8.73265174951318e-06,
|
||
|
|
"loss": 0.4782,
|
||
|
|
"step": 1530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3109148264984227,
|
||
|
|
"grad_norm": 0.5764514697725889,
|
||
|
|
"learning_rate": 8.709115305259233e-06,
|
||
|
|
"loss": 0.4722,
|
||
|
|
"step": 1540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31293375394321765,
|
||
|
|
"grad_norm": 0.48602580093856246,
|
||
|
|
"learning_rate": 8.685394661629453e-06,
|
||
|
|
"loss": 0.4764,
|
||
|
|
"step": 1550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31495268138801263,
|
||
|
|
"grad_norm": 0.43106495742778833,
|
||
|
|
"learning_rate": 8.661490996621125e-06,
|
||
|
|
"loss": 0.4749,
|
||
|
|
"step": 1560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31697160883280756,
|
||
|
|
"grad_norm": 0.4879400234119546,
|
||
|
|
"learning_rate": 8.637405497320599e-06,
|
||
|
|
"loss": 0.4685,
|
||
|
|
"step": 1570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.31899053627760254,
|
||
|
|
"grad_norm": 0.48144632266967335,
|
||
|
|
"learning_rate": 8.61313935984436e-06,
|
||
|
|
"loss": 0.473,
|
||
|
|
"step": 1580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32100946372239747,
|
||
|
|
"grad_norm": 0.4824945838927638,
|
||
|
|
"learning_rate": 8.588693789279605e-06,
|
||
|
|
"loss": 0.467,
|
||
|
|
"step": 1590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32302839116719245,
|
||
|
|
"grad_norm": 0.48854234491790616,
|
||
|
|
"learning_rate": 8.56406999962441e-06,
|
||
|
|
"loss": 0.4715,
|
||
|
|
"step": 1600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3250473186119874,
|
||
|
|
"grad_norm": 0.43356661154496373,
|
||
|
|
"learning_rate": 8.539269213727434e-06,
|
||
|
|
"loss": 0.4643,
|
||
|
|
"step": 1610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.32706624605678236,
|
||
|
|
"grad_norm": 0.45038524364573934,
|
||
|
|
"learning_rate": 8.514292663227196e-06,
|
||
|
|
"loss": 0.4753,
|
||
|
|
"step": 1620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3290851735015773,
|
||
|
|
"grad_norm": 0.5390076730569227,
|
||
|
|
"learning_rate": 8.489141588490914e-06,
|
||
|
|
"loss": 0.4692,
|
||
|
|
"step": 1630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3311041009463722,
|
||
|
|
"grad_norm": 0.5166665541948204,
|
||
|
|
"learning_rate": 8.46381723855289e-06,
|
||
|
|
"loss": 0.4626,
|
||
|
|
"step": 1640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3331230283911672,
|
||
|
|
"grad_norm": 0.4973742178832253,
|
||
|
|
"learning_rate": 8.438320871052501e-06,
|
||
|
|
"loss": 0.4722,
|
||
|
|
"step": 1650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3351419558359621,
|
||
|
|
"grad_norm": 0.4625137579015689,
|
||
|
|
"learning_rate": 8.412653752171732e-06,
|
||
|
|
"loss": 0.4591,
|
||
|
|
"step": 1660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3371608832807571,
|
||
|
|
"grad_norm": 0.48424933592035946,
|
||
|
|
"learning_rate": 8.386817156572298e-06,
|
||
|
|
"loss": 0.48,
|
||
|
|
"step": 1670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.33917981072555203,
|
||
|
|
"grad_norm": 0.4532582849486311,
|
||
|
|
"learning_rate": 8.360812367332342e-06,
|
||
|
|
"loss": 0.469,
|
||
|
|
"step": 1680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.341198738170347,
|
||
|
|
"grad_norm": 0.4456128521040015,
|
||
|
|
"learning_rate": 8.33464067588272e-06,
|
||
|
|
"loss": 0.4701,
|
||
|
|
"step": 1690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34321766561514194,
|
||
|
|
"grad_norm": 0.47975070454468693,
|
||
|
|
"learning_rate": 8.30830338194286e-06,
|
||
|
|
"loss": 0.4742,
|
||
|
|
"step": 1700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3452365930599369,
|
||
|
|
"grad_norm": 0.44758135515111036,
|
||
|
|
"learning_rate": 8.281801793456223e-06,
|
||
|
|
"loss": 0.4563,
|
||
|
|
"step": 1710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34725552050473185,
|
||
|
|
"grad_norm": 0.539530429343652,
|
||
|
|
"learning_rate": 8.255137226525346e-06,
|
||
|
|
"loss": 0.4678,
|
||
|
|
"step": 1720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.34927444794952683,
|
||
|
|
"grad_norm": 0.44634229380095736,
|
||
|
|
"learning_rate": 8.22831100534648e-06,
|
||
|
|
"loss": 0.4688,
|
||
|
|
"step": 1730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35129337539432176,
|
||
|
|
"grad_norm": 0.41960408937669463,
|
||
|
|
"learning_rate": 8.201324462143827e-06,
|
||
|
|
"loss": 0.4542,
|
||
|
|
"step": 1740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35331230283911674,
|
||
|
|
"grad_norm": 0.4791531544406239,
|
||
|
|
"learning_rate": 8.174178937103397e-06,
|
||
|
|
"loss": 0.4743,
|
||
|
|
"step": 1750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35533123028391167,
|
||
|
|
"grad_norm": 0.4352818312435405,
|
||
|
|
"learning_rate": 8.14687577830643e-06,
|
||
|
|
"loss": 0.4604,
|
||
|
|
"step": 1760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.35735015772870665,
|
||
|
|
"grad_norm": 0.4672874879206997,
|
||
|
|
"learning_rate": 8.11941634166246e-06,
|
||
|
|
"loss": 0.4712,
|
||
|
|
"step": 1770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3593690851735016,
|
||
|
|
"grad_norm": 0.4637056308756207,
|
||
|
|
"learning_rate": 8.09180199084198e-06,
|
||
|
|
"loss": 0.4668,
|
||
|
|
"step": 1780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3613880126182965,
|
||
|
|
"grad_norm": 0.44264459219210456,
|
||
|
|
"learning_rate": 8.064034097208712e-06,
|
||
|
|
"loss": 0.4596,
|
||
|
|
"step": 1790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3634069400630915,
|
||
|
|
"grad_norm": 0.49462230789807315,
|
||
|
|
"learning_rate": 8.036114039751516e-06,
|
||
|
|
"loss": 0.4667,
|
||
|
|
"step": 1800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3654258675078864,
|
||
|
|
"grad_norm": 0.4634309345247295,
|
||
|
|
"learning_rate": 8.0080432050159e-06,
|
||
|
|
"loss": 0.4604,
|
||
|
|
"step": 1810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3674447949526814,
|
||
|
|
"grad_norm": 0.4999816402185204,
|
||
|
|
"learning_rate": 7.979822987035157e-06,
|
||
|
|
"loss": 0.4644,
|
||
|
|
"step": 1820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3694637223974763,
|
||
|
|
"grad_norm": 0.47642327436037435,
|
||
|
|
"learning_rate": 7.951454787261148e-06,
|
||
|
|
"loss": 0.4677,
|
||
|
|
"step": 1830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3714826498422713,
|
||
|
|
"grad_norm": 0.4953986824691425,
|
||
|
|
"learning_rate": 7.922940014494693e-06,
|
||
|
|
"loss": 0.4593,
|
||
|
|
"step": 1840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37350157728706623,
|
||
|
|
"grad_norm": 0.4817430868786167,
|
||
|
|
"learning_rate": 7.894280084815618e-06,
|
||
|
|
"loss": 0.4689,
|
||
|
|
"step": 1850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3755205047318612,
|
||
|
|
"grad_norm": 0.4356746961237441,
|
||
|
|
"learning_rate": 7.86547642151242e-06,
|
||
|
|
"loss": 0.4611,
|
||
|
|
"step": 1860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.37753943217665614,
|
||
|
|
"grad_norm": 0.4905696785040258,
|
||
|
|
"learning_rate": 7.836530455011594e-06,
|
||
|
|
"loss": 0.4647,
|
||
|
|
"step": 1870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3795583596214511,
|
||
|
|
"grad_norm": 0.4730705035791906,
|
||
|
|
"learning_rate": 7.807443622806592e-06,
|
||
|
|
"loss": 0.4643,
|
||
|
|
"step": 1880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38157728706624605,
|
||
|
|
"grad_norm": 0.4626721574097504,
|
||
|
|
"learning_rate": 7.778217369386434e-06,
|
||
|
|
"loss": 0.4621,
|
||
|
|
"step": 1890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38359621451104103,
|
||
|
|
"grad_norm": 0.5095748907639542,
|
||
|
|
"learning_rate": 7.748853146163978e-06,
|
||
|
|
"loss": 0.4702,
|
||
|
|
"step": 1900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38561514195583596,
|
||
|
|
"grad_norm": 0.45666908318697785,
|
||
|
|
"learning_rate": 7.719352411403836e-06,
|
||
|
|
"loss": 0.4584,
|
||
|
|
"step": 1910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3876340694006309,
|
||
|
|
"grad_norm": 0.5831960484166556,
|
||
|
|
"learning_rate": 7.68971663014995e-06,
|
||
|
|
"loss": 0.4581,
|
||
|
|
"step": 1920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.38965299684542587,
|
||
|
|
"grad_norm": 0.43083136772614855,
|
||
|
|
"learning_rate": 7.659947274152857e-06,
|
||
|
|
"loss": 0.4595,
|
||
|
|
"step": 1930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3916719242902208,
|
||
|
|
"grad_norm": 0.44527597409724967,
|
||
|
|
"learning_rate": 7.630045821796567e-06,
|
||
|
|
"loss": 0.4603,
|
||
|
|
"step": 1940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3936908517350158,
|
||
|
|
"grad_norm": 0.5085813030698595,
|
||
|
|
"learning_rate": 7.60001375802518e-06,
|
||
|
|
"loss": 0.4726,
|
||
|
|
"step": 1950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3957097791798107,
|
||
|
|
"grad_norm": 0.5516780680340057,
|
||
|
|
"learning_rate": 7.5698525742691145e-06,
|
||
|
|
"loss": 0.4553,
|
||
|
|
"step": 1960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3977287066246057,
|
||
|
|
"grad_norm": 0.4473799858576117,
|
||
|
|
"learning_rate": 7.5395637683710534e-06,
|
||
|
|
"loss": 0.4652,
|
||
|
|
"step": 1970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.3997476340694006,
|
||
|
|
"grad_norm": 0.4594068744960154,
|
||
|
|
"learning_rate": 7.509148844511556e-06,
|
||
|
|
"loss": 0.4569,
|
||
|
|
"step": 1980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4017665615141956,
|
||
|
|
"grad_norm": 0.4195303708920141,
|
||
|
|
"learning_rate": 7.478609313134367e-06,
|
||
|
|
"loss": 0.4608,
|
||
|
|
"step": 1990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4037854889589905,
|
||
|
|
"grad_norm": 0.5224611450690858,
|
||
|
|
"learning_rate": 7.447946690871387e-06,
|
||
|
|
"loss": 0.4618,
|
||
|
|
"step": 2000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4058044164037855,
|
||
|
|
"grad_norm": 0.5037868351130359,
|
||
|
|
"learning_rate": 7.4171625004673765e-06,
|
||
|
|
"loss": 0.4549,
|
||
|
|
"step": 2010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.40782334384858043,
|
||
|
|
"grad_norm": 0.6617292048565804,
|
||
|
|
"learning_rate": 7.386258270704318e-06,
|
||
|
|
"loss": 0.4648,
|
||
|
|
"step": 2020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4098422712933754,
|
||
|
|
"grad_norm": 0.46778481767707863,
|
||
|
|
"learning_rate": 7.355235536325507e-06,
|
||
|
|
"loss": 0.4618,
|
||
|
|
"step": 2030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.41186119873817034,
|
||
|
|
"grad_norm": 0.4467025867029599,
|
||
|
|
"learning_rate": 7.32409583795932e-06,
|
||
|
|
"loss": 0.4556,
|
||
|
|
"step": 2040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4138801261829653,
|
||
|
|
"grad_norm": 0.4864826536631601,
|
||
|
|
"learning_rate": 7.292840722042721e-06,
|
||
|
|
"loss": 0.4649,
|
||
|
|
"step": 2050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.41589905362776025,
|
||
|
|
"grad_norm": 0.44674303304698915,
|
||
|
|
"learning_rate": 7.261471740744453e-06,
|
||
|
|
"loss": 0.4546,
|
||
|
|
"step": 2060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4179179810725552,
|
||
|
|
"grad_norm": 0.4638264920705765,
|
||
|
|
"learning_rate": 7.229990451887961e-06,
|
||
|
|
"loss": 0.459,
|
||
|
|
"step": 2070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.41993690851735016,
|
||
|
|
"grad_norm": 0.44062835097290626,
|
||
|
|
"learning_rate": 7.1983984188740215e-06,
|
||
|
|
"loss": 0.4627,
|
||
|
|
"step": 2080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4219558359621451,
|
||
|
|
"grad_norm": 0.4596889143174662,
|
||
|
|
"learning_rate": 7.1666972106031105e-06,
|
||
|
|
"loss": 0.4584,
|
||
|
|
"step": 2090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.42397476340694007,
|
||
|
|
"grad_norm": 0.4866244769293159,
|
||
|
|
"learning_rate": 7.134888401397484e-06,
|
||
|
|
"loss": 0.4629,
|
||
|
|
"step": 2100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.425993690851735,
|
||
|
|
"grad_norm": 0.5201676761862537,
|
||
|
|
"learning_rate": 7.102973570922997e-06,
|
||
|
|
"loss": 0.4549,
|
||
|
|
"step": 2110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.42801261829653,
|
||
|
|
"grad_norm": 0.4790871394995968,
|
||
|
|
"learning_rate": 7.070954304110656e-06,
|
||
|
|
"loss": 0.4632,
|
||
|
|
"step": 2120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4300315457413249,
|
||
|
|
"grad_norm": 0.4659786396553266,
|
||
|
|
"learning_rate": 7.03883219107791e-06,
|
||
|
|
"loss": 0.4581,
|
||
|
|
"step": 2130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4320504731861199,
|
||
|
|
"grad_norm": 0.4104121258537053,
|
||
|
|
"learning_rate": 7.0066088270496775e-06,
|
||
|
|
"loss": 0.4596,
|
||
|
|
"step": 2140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4340694006309148,
|
||
|
|
"grad_norm": 0.46988504036618683,
|
||
|
|
"learning_rate": 6.9742858122791355e-06,
|
||
|
|
"loss": 0.4648,
|
||
|
|
"step": 2150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4360883280757098,
|
||
|
|
"grad_norm": 0.6098242782979109,
|
||
|
|
"learning_rate": 6.941864751968239e-06,
|
||
|
|
"loss": 0.4548,
|
||
|
|
"step": 2160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4381072555205047,
|
||
|
|
"grad_norm": 0.7557069029520974,
|
||
|
|
"learning_rate": 6.90934725618801e-06,
|
||
|
|
"loss": 0.461,
|
||
|
|
"step": 2170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4401261829652997,
|
||
|
|
"grad_norm": 0.46196984209426056,
|
||
|
|
"learning_rate": 6.876734939798584e-06,
|
||
|
|
"loss": 0.4631,
|
||
|
|
"step": 2180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.44214511041009463,
|
||
|
|
"grad_norm": 0.45367052105694244,
|
||
|
|
"learning_rate": 6.844029422368998e-06,
|
||
|
|
"loss": 0.4567,
|
||
|
|
"step": 2190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4441640378548896,
|
||
|
|
"grad_norm": 0.4923825492028847,
|
||
|
|
"learning_rate": 6.811232328096778e-06,
|
||
|
|
"loss": 0.4619,
|
||
|
|
"step": 2200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.44618296529968454,
|
||
|
|
"grad_norm": 0.4412882669817311,
|
||
|
|
"learning_rate": 6.7783452857272725e-06,
|
||
|
|
"loss": 0.4472,
|
||
|
|
"step": 2210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.44820189274447947,
|
||
|
|
"grad_norm": 0.4804805684578094,
|
||
|
|
"learning_rate": 6.745369928472764e-06,
|
||
|
|
"loss": 0.4634,
|
||
|
|
"step": 2220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.45022082018927445,
|
||
|
|
"grad_norm": 0.46765384977250785,
|
||
|
|
"learning_rate": 6.712307893931364e-06,
|
||
|
|
"loss": 0.4595,
|
||
|
|
"step": 2230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4522397476340694,
|
||
|
|
"grad_norm": 1.0068029764875739,
|
||
|
|
"learning_rate": 6.67916082400569e-06,
|
||
|
|
"loss": 0.457,
|
||
|
|
"step": 2240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.45425867507886436,
|
||
|
|
"grad_norm": 0.4828558935559972,
|
||
|
|
"learning_rate": 6.645930364821327e-06,
|
||
|
|
"loss": 0.465,
|
||
|
|
"step": 2250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4562776025236593,
|
||
|
|
"grad_norm": 0.4571716607551568,
|
||
|
|
"learning_rate": 6.612618166645068e-06,
|
||
|
|
"loss": 0.4521,
|
||
|
|
"step": 2260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.45829652996845427,
|
||
|
|
"grad_norm": 0.49891810921169605,
|
||
|
|
"learning_rate": 6.579225883802981e-06,
|
||
|
|
"loss": 0.4621,
|
||
|
|
"step": 2270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4603154574132492,
|
||
|
|
"grad_norm": 0.6812817990255348,
|
||
|
|
"learning_rate": 6.5457551745982275e-06,
|
||
|
|
"loss": 0.4553,
|
||
|
|
"step": 2280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4623343848580442,
|
||
|
|
"grad_norm": 0.466296009966539,
|
||
|
|
"learning_rate": 6.512207701228727e-06,
|
||
|
|
"loss": 0.4585,
|
||
|
|
"step": 2290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4643533123028391,
|
||
|
|
"grad_norm": 0.48144166164862184,
|
||
|
|
"learning_rate": 6.478585129704608e-06,
|
||
|
|
"loss": 0.4579,
|
||
|
|
"step": 2300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4663722397476341,
|
||
|
|
"grad_norm": 0.4575454070862058,
|
||
|
|
"learning_rate": 6.444889129765465e-06,
|
||
|
|
"loss": 0.453,
|
||
|
|
"step": 2310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.468391167192429,
|
||
|
|
"grad_norm": 0.5049212463992705,
|
||
|
|
"learning_rate": 6.4111213747974385e-06,
|
||
|
|
"loss": 0.4606,
|
||
|
|
"step": 2320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.470410094637224,
|
||
|
|
"grad_norm": 0.44196403700173553,
|
||
|
|
"learning_rate": 6.377283541750123e-06,
|
||
|
|
"loss": 0.4528,
|
||
|
|
"step": 2330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4724290220820189,
|
||
|
|
"grad_norm": 0.44557725863878417,
|
||
|
|
"learning_rate": 6.343377311053271e-06,
|
||
|
|
"loss": 0.4614,
|
||
|
|
"step": 2340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4744479495268139,
|
||
|
|
"grad_norm": 0.485138333868983,
|
||
|
|
"learning_rate": 6.309404366533355e-06,
|
||
|
|
"loss": 0.4662,
|
||
|
|
"step": 2350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.47646687697160883,
|
||
|
|
"grad_norm": 0.44703181568848005,
|
||
|
|
"learning_rate": 6.275366395329934e-06,
|
||
|
|
"loss": 0.4484,
|
||
|
|
"step": 2360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.47848580441640376,
|
||
|
|
"grad_norm": 0.46684239360571517,
|
||
|
|
"learning_rate": 6.241265087811883e-06,
|
||
|
|
"loss": 0.461,
|
||
|
|
"step": 2370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.48050473186119874,
|
||
|
|
"grad_norm": 0.4308686375763305,
|
||
|
|
"learning_rate": 6.207102137493433e-06,
|
||
|
|
"loss": 0.4572,
|
||
|
|
"step": 2380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.48252365930599367,
|
||
|
|
"grad_norm": 0.4349332573681559,
|
||
|
|
"learning_rate": 6.172879240950075e-06,
|
||
|
|
"loss": 0.4586,
|
||
|
|
"step": 2390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.48454258675078865,
|
||
|
|
"grad_norm": 0.4946893193107334,
|
||
|
|
"learning_rate": 6.138598097734308e-06,
|
||
|
|
"loss": 0.4613,
|
||
|
|
"step": 2400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4865615141955836,
|
||
|
|
"grad_norm": 0.4278600145969151,
|
||
|
|
"learning_rate": 6.104260410291237e-06,
|
||
|
|
"loss": 0.4524,
|
||
|
|
"step": 2410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.48858044164037856,
|
||
|
|
"grad_norm": 0.438261930357501,
|
||
|
|
"learning_rate": 6.069867883874017e-06,
|
||
|
|
"loss": 0.4546,
|
||
|
|
"step": 2420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4905993690851735,
|
||
|
|
"grad_norm": 0.4834328335496326,
|
||
|
|
"learning_rate": 6.035422226459183e-06,
|
||
|
|
"loss": 0.4601,
|
||
|
|
"step": 2430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.49261829652996847,
|
||
|
|
"grad_norm": 0.4560202224259225,
|
||
|
|
"learning_rate": 6.000925148661822e-06,
|
||
|
|
"loss": 0.4552,
|
||
|
|
"step": 2440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4946372239747634,
|
||
|
|
"grad_norm": 0.46242134458299394,
|
||
|
|
"learning_rate": 5.966378363650618e-06,
|
||
|
|
"loss": 0.4534,
|
||
|
|
"step": 2450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4966561514195584,
|
||
|
|
"grad_norm": 0.509798379798209,
|
||
|
|
"learning_rate": 5.931783587062785e-06,
|
||
|
|
"loss": 0.4478,
|
||
|
|
"step": 2460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.4986750788643533,
|
||
|
|
"grad_norm": 0.4916270732208669,
|
||
|
|
"learning_rate": 5.897142536918854e-06,
|
||
|
|
"loss": 0.4606,
|
||
|
|
"step": 2470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5006940063091483,
|
||
|
|
"grad_norm": 0.4493237090311729,
|
||
|
|
"learning_rate": 5.8624569335373625e-06,
|
||
|
|
"loss": 0.4512,
|
||
|
|
"step": 2480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5027129337539432,
|
||
|
|
"grad_norm": 0.4570244860395198,
|
||
|
|
"learning_rate": 5.827728499449417e-06,
|
||
|
|
"loss": 0.4503,
|
||
|
|
"step": 2490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5047318611987381,
|
||
|
|
"grad_norm": 0.5044698787582333,
|
||
|
|
"learning_rate": 5.792958959313155e-06,
|
||
|
|
"loss": 0.4586,
|
||
|
|
"step": 2500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5067507886435331,
|
||
|
|
"grad_norm": 0.4567238497826546,
|
||
|
|
"learning_rate": 5.758150039828084e-06,
|
||
|
|
"loss": 0.442,
|
||
|
|
"step": 2510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5087697160883281,
|
||
|
|
"grad_norm": 0.49456514738525703,
|
||
|
|
"learning_rate": 5.723303469649352e-06,
|
||
|
|
"loss": 0.4524,
|
||
|
|
"step": 2520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.510788643533123,
|
||
|
|
"grad_norm": 0.46129063291343286,
|
||
|
|
"learning_rate": 5.688420979301876e-06,
|
||
|
|
"loss": 0.4533,
|
||
|
|
"step": 2530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.512807570977918,
|
||
|
|
"grad_norm": 0.46780830492612324,
|
||
|
|
"learning_rate": 5.6535043010944226e-06,
|
||
|
|
"loss": 0.4537,
|
||
|
|
"step": 2540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5148264984227129,
|
||
|
|
"grad_norm": 0.49663726459668367,
|
||
|
|
"learning_rate": 5.618555169033567e-06,
|
||
|
|
"loss": 0.4624,
|
||
|
|
"step": 2550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5168454258675079,
|
||
|
|
"grad_norm": 0.4323698197626643,
|
||
|
|
"learning_rate": 5.583575318737588e-06,
|
||
|
|
"loss": 0.4394,
|
||
|
|
"step": 2560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5188643533123029,
|
||
|
|
"grad_norm": 0.433967581020834,
|
||
|
|
"learning_rate": 5.548566487350264e-06,
|
||
|
|
"loss": 0.4482,
|
||
|
|
"step": 2570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5208832807570978,
|
||
|
|
"grad_norm": 0.5093937708469708,
|
||
|
|
"learning_rate": 5.513530413454618e-06,
|
||
|
|
"loss": 0.4488,
|
||
|
|
"step": 2580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5229022082018927,
|
||
|
|
"grad_norm": 0.4479313579903872,
|
||
|
|
"learning_rate": 5.478468836986568e-06,
|
||
|
|
"loss": 0.4547,
|
||
|
|
"step": 2590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5249211356466877,
|
||
|
|
"grad_norm": 0.4784953987086218,
|
||
|
|
"learning_rate": 5.443383499148519e-06,
|
||
|
|
"loss": 0.4536,
|
||
|
|
"step": 2600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5269400630914827,
|
||
|
|
"grad_norm": 0.4311672354723456,
|
||
|
|
"learning_rate": 5.408276142322904e-06,
|
||
|
|
"loss": 0.4489,
|
||
|
|
"step": 2610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5289589905362776,
|
||
|
|
"grad_norm": 0.4320377418884208,
|
||
|
|
"learning_rate": 5.373148509985636e-06,
|
||
|
|
"loss": 0.457,
|
||
|
|
"step": 2620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5309779179810725,
|
||
|
|
"grad_norm": 0.4475608921551904,
|
||
|
|
"learning_rate": 5.3380023466195455e-06,
|
||
|
|
"loss": 0.4491,
|
||
|
|
"step": 2630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5329968454258676,
|
||
|
|
"grad_norm": 0.4150312768857391,
|
||
|
|
"learning_rate": 5.302839397627732e-06,
|
||
|
|
"loss": 0.4509,
|
||
|
|
"step": 2640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5350157728706625,
|
||
|
|
"grad_norm": 0.4842663915075593,
|
||
|
|
"learning_rate": 5.267661409246893e-06,
|
||
|
|
"loss": 0.4497,
|
||
|
|
"step": 2650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5370347003154574,
|
||
|
|
"grad_norm": 0.48020608359904937,
|
||
|
|
"learning_rate": 5.232470128460597e-06,
|
||
|
|
"loss": 0.4455,
|
||
|
|
"step": 2660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5390536277602523,
|
||
|
|
"grad_norm": 0.460861374110654,
|
||
|
|
"learning_rate": 5.197267302912535e-06,
|
||
|
|
"loss": 0.4633,
|
||
|
|
"step": 2670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5410725552050473,
|
||
|
|
"grad_norm": 0.4861656655295904,
|
||
|
|
"learning_rate": 5.162054680819724e-06,
|
||
|
|
"loss": 0.455,
|
||
|
|
"step": 2680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5430914826498423,
|
||
|
|
"grad_norm": 0.4280331150720691,
|
||
|
|
"learning_rate": 5.12683401088569e-06,
|
||
|
|
"loss": 0.4527,
|
||
|
|
"step": 2690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5451104100946372,
|
||
|
|
"grad_norm": 0.4803934844859794,
|
||
|
|
"learning_rate": 5.091607042213621e-06,
|
||
|
|
"loss": 0.4507,
|
||
|
|
"step": 2700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5471293375394322,
|
||
|
|
"grad_norm": 0.44679887041219724,
|
||
|
|
"learning_rate": 5.056375524219516e-06,
|
||
|
|
"loss": 0.4467,
|
||
|
|
"step": 2710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5491482649842271,
|
||
|
|
"grad_norm": 0.4562693718725788,
|
||
|
|
"learning_rate": 5.021141206545293e-06,
|
||
|
|
"loss": 0.4481,
|
||
|
|
"step": 2720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5511671924290221,
|
||
|
|
"grad_norm": 0.4568391909726876,
|
||
|
|
"learning_rate": 4.9859058389719095e-06,
|
||
|
|
"loss": 0.4467,
|
||
|
|
"step": 2730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.553186119873817,
|
||
|
|
"grad_norm": 0.4402113900197009,
|
||
|
|
"learning_rate": 4.950671171332458e-06,
|
||
|
|
"loss": 0.4438,
|
||
|
|
"step": 2740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.555205047318612,
|
||
|
|
"grad_norm": 0.4694901686129437,
|
||
|
|
"learning_rate": 4.915438953425277e-06,
|
||
|
|
"loss": 0.455,
|
||
|
|
"step": 2750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5572239747634069,
|
||
|
|
"grad_norm": 0.5092377388071306,
|
||
|
|
"learning_rate": 4.8802109349270415e-06,
|
||
|
|
"loss": 0.4431,
|
||
|
|
"step": 2760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5592429022082019,
|
||
|
|
"grad_norm": 0.5334478668234227,
|
||
|
|
"learning_rate": 4.844988865305883e-06,
|
||
|
|
"loss": 0.4457,
|
||
|
|
"step": 2770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5612618296529969,
|
||
|
|
"grad_norm": 0.450190134489615,
|
||
|
|
"learning_rate": 4.809774493734507e-06,
|
||
|
|
"loss": 0.4468,
|
||
|
|
"step": 2780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5632807570977918,
|
||
|
|
"grad_norm": 0.6136460189487958,
|
||
|
|
"learning_rate": 4.774569569003314e-06,
|
||
|
|
"loss": 0.4531,
|
||
|
|
"step": 2790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5652996845425867,
|
||
|
|
"grad_norm": 0.47669716943494894,
|
||
|
|
"learning_rate": 4.739375839433575e-06,
|
||
|
|
"loss": 0.4559,
|
||
|
|
"step": 2800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5673186119873816,
|
||
|
|
"grad_norm": 0.47396268716032475,
|
||
|
|
"learning_rate": 4.704195052790584e-06,
|
||
|
|
"loss": 0.4463,
|
||
|
|
"step": 2810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5693375394321767,
|
||
|
|
"grad_norm": 0.5115114860936024,
|
||
|
|
"learning_rate": 4.66902895619688e-06,
|
||
|
|
"loss": 0.4494,
|
||
|
|
"step": 2820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5713564668769716,
|
||
|
|
"grad_norm": 0.4561358589231192,
|
||
|
|
"learning_rate": 4.633879296045474e-06,
|
||
|
|
"loss": 0.4477,
|
||
|
|
"step": 2830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5733753943217665,
|
||
|
|
"grad_norm": 0.4547946639578122,
|
||
|
|
"learning_rate": 4.59874781791312e-06,
|
||
|
|
"loss": 0.4502,
|
||
|
|
"step": 2840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5753943217665615,
|
||
|
|
"grad_norm": 0.5036899795994354,
|
||
|
|
"learning_rate": 4.56363626647363e-06,
|
||
|
|
"loss": 0.4553,
|
||
|
|
"step": 2850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5774132492113565,
|
||
|
|
"grad_norm": 0.46515739613875146,
|
||
|
|
"learning_rate": 4.528546385411239e-06,
|
||
|
|
"loss": 0.4373,
|
||
|
|
"step": 2860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5794321766561514,
|
||
|
|
"grad_norm": 0.4834989187992637,
|
||
|
|
"learning_rate": 4.493479917333989e-06,
|
||
|
|
"loss": 0.453,
|
||
|
|
"step": 2870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5814511041009464,
|
||
|
|
"grad_norm": 0.45922181466381384,
|
||
|
|
"learning_rate": 4.4584386036872155e-06,
|
||
|
|
"loss": 0.4527,
|
||
|
|
"step": 2880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5834700315457413,
|
||
|
|
"grad_norm": 0.46803871925708684,
|
||
|
|
"learning_rate": 4.42342418466705e-06,
|
||
|
|
"loss": 0.4426,
|
||
|
|
"step": 2890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5854889589905363,
|
||
|
|
"grad_norm": 0.47319916436776355,
|
||
|
|
"learning_rate": 4.388438399134e-06,
|
||
|
|
"loss": 0.4547,
|
||
|
|
"step": 2900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5875078864353313,
|
||
|
|
"grad_norm": 0.4598569796278221,
|
||
|
|
"learning_rate": 4.3534829845265996e-06,
|
||
|
|
"loss": 0.4441,
|
||
|
|
"step": 2910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5895268138801262,
|
||
|
|
"grad_norm": 0.47100191118178447,
|
||
|
|
"learning_rate": 4.318559676775126e-06,
|
||
|
|
"loss": 0.4472,
|
||
|
|
"step": 2920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5915457413249211,
|
||
|
|
"grad_norm": 0.495923691841689,
|
||
|
|
"learning_rate": 4.283670210215384e-06,
|
||
|
|
"loss": 0.4516,
|
||
|
|
"step": 2930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5935646687697161,
|
||
|
|
"grad_norm": 0.4391764043906036,
|
||
|
|
"learning_rate": 4.248816317502585e-06,
|
||
|
|
"loss": 0.4468,
|
||
|
|
"step": 2940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5955835962145111,
|
||
|
|
"grad_norm": 0.47900304182889214,
|
||
|
|
"learning_rate": 4.213999729525304e-06,
|
||
|
|
"loss": 0.452,
|
||
|
|
"step": 2950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.597602523659306,
|
||
|
|
"grad_norm": 0.4894376668991494,
|
||
|
|
"learning_rate": 4.179222175319502e-06,
|
||
|
|
"loss": 0.4418,
|
||
|
|
"step": 2960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.5996214511041009,
|
||
|
|
"grad_norm": 0.4698266233405637,
|
||
|
|
"learning_rate": 4.1444853819826845e-06,
|
||
|
|
"loss": 0.4503,
|
||
|
|
"step": 2970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6016403785488958,
|
||
|
|
"grad_norm": 0.4601994675920038,
|
||
|
|
"learning_rate": 4.1097910745881175e-06,
|
||
|
|
"loss": 0.4421,
|
||
|
|
"step": 2980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6036593059936909,
|
||
|
|
"grad_norm": 0.6167906848346414,
|
||
|
|
"learning_rate": 4.0751409760991585e-06,
|
||
|
|
"loss": 0.4469,
|
||
|
|
"step": 2990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6056782334384858,
|
||
|
|
"grad_norm": 0.5051790895352023,
|
||
|
|
"learning_rate": 4.040536807283696e-06,
|
||
|
|
"loss": 0.4502,
|
||
|
|
"step": 3000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6076971608832807,
|
||
|
|
"grad_norm": 0.45322450050363433,
|
||
|
|
"learning_rate": 4.005980286628698e-06,
|
||
|
|
"loss": 0.4505,
|
||
|
|
"step": 3010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6097160883280757,
|
||
|
|
"grad_norm": 0.4910869337898713,
|
||
|
|
"learning_rate": 3.9714731302548525e-06,
|
||
|
|
"loss": 0.4486,
|
||
|
|
"step": 3020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6117350157728707,
|
||
|
|
"grad_norm": 0.447631013542978,
|
||
|
|
"learning_rate": 3.937017051831367e-06,
|
||
|
|
"loss": 0.4478,
|
||
|
|
"step": 3030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6137539432176656,
|
||
|
|
"grad_norm": 0.7018069369756337,
|
||
|
|
"learning_rate": 3.9026137624908466e-06,
|
||
|
|
"loss": 0.4492,
|
||
|
|
"step": 3040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6157728706624606,
|
||
|
|
"grad_norm": 0.4762076148803905,
|
||
|
|
"learning_rate": 3.868264970744324e-06,
|
||
|
|
"loss": 0.4513,
|
||
|
|
"step": 3050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6177917981072555,
|
||
|
|
"grad_norm": 0.5020509623933802,
|
||
|
|
"learning_rate": 3.833972382396412e-06,
|
||
|
|
"loss": 0.4482,
|
||
|
|
"step": 3060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6198107255520505,
|
||
|
|
"grad_norm": 0.44347011131653674,
|
||
|
|
"learning_rate": 3.799737700460594e-06,
|
||
|
|
"loss": 0.4441,
|
||
|
|
"step": 3070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6218296529968455,
|
||
|
|
"grad_norm": 0.45353690385677753,
|
||
|
|
"learning_rate": 3.7655626250746446e-06,
|
||
|
|
"loss": 0.4443,
|
||
|
|
"step": 3080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6238485804416404,
|
||
|
|
"grad_norm": 0.44404552147051596,
|
||
|
|
"learning_rate": 3.7314488534162007e-06,
|
||
|
|
"loss": 0.4509,
|
||
|
|
"step": 3090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6258675078864353,
|
||
|
|
"grad_norm": 0.4759693579595236,
|
||
|
|
"learning_rate": 3.6973980796184843e-06,
|
||
|
|
"loss": 0.4573,
|
||
|
|
"step": 3100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6278864353312302,
|
||
|
|
"grad_norm": 0.45248182104327295,
|
||
|
|
"learning_rate": 3.6634119946861556e-06,
|
||
|
|
"loss": 0.4442,
|
||
|
|
"step": 3110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6299053627760253,
|
||
|
|
"grad_norm": 0.44605725901650006,
|
||
|
|
"learning_rate": 3.629492286411349e-06,
|
||
|
|
"loss": 0.4515,
|
||
|
|
"step": 3120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6319242902208202,
|
||
|
|
"grad_norm": 0.4770842161429305,
|
||
|
|
"learning_rate": 3.5956406392898484e-06,
|
||
|
|
"loss": 0.4481,
|
||
|
|
"step": 3130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6339432176656151,
|
||
|
|
"grad_norm": 0.4611511904942963,
|
||
|
|
"learning_rate": 3.561858734437432e-06,
|
||
|
|
"loss": 0.4553,
|
||
|
|
"step": 3140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.63596214511041,
|
||
|
|
"grad_norm": 0.4886490480648605,
|
||
|
|
"learning_rate": 3.528148249506391e-06,
|
||
|
|
"loss": 0.4541,
|
||
|
|
"step": 3150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6379810725552051,
|
||
|
|
"grad_norm": 0.5063011038652176,
|
||
|
|
"learning_rate": 3.494510858602217e-06,
|
||
|
|
"loss": 0.448,
|
||
|
|
"step": 3160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.64,
|
||
|
|
"grad_norm": 0.5297665998286369,
|
||
|
|
"learning_rate": 3.4609482322004463e-06,
|
||
|
|
"loss": 0.45,
|
||
|
|
"step": 3170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6420189274447949,
|
||
|
|
"grad_norm": 0.789602090268308,
|
||
|
|
"learning_rate": 3.427462037063729e-06,
|
||
|
|
"loss": 0.4499,
|
||
|
|
"step": 3180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6440378548895899,
|
||
|
|
"grad_norm": 0.4595511607863831,
|
||
|
|
"learning_rate": 3.394053936159035e-06,
|
||
|
|
"loss": 0.4451,
|
||
|
|
"step": 3190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6460567823343849,
|
||
|
|
"grad_norm": 0.5053416927018038,
|
||
|
|
"learning_rate": 3.3607255885750756e-06,
|
||
|
|
"loss": 0.4488,
|
||
|
|
"step": 3200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6480757097791798,
|
||
|
|
"grad_norm": 0.47050871567377006,
|
||
|
|
"learning_rate": 3.3274786494399102e-06,
|
||
|
|
"loss": 0.4461,
|
||
|
|
"step": 3210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6500946372239748,
|
||
|
|
"grad_norm": 0.46433780403203007,
|
||
|
|
"learning_rate": 3.2943147698387562e-06,
|
||
|
|
"loss": 0.4497,
|
||
|
|
"step": 3220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6521135646687697,
|
||
|
|
"grad_norm": 0.45409078162209787,
|
||
|
|
"learning_rate": 3.261235596731982e-06,
|
||
|
|
"loss": 0.4457,
|
||
|
|
"step": 3230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6541324921135647,
|
||
|
|
"grad_norm": 0.4218245687222197,
|
||
|
|
"learning_rate": 3.228242772873328e-06,
|
||
|
|
"loss": 0.4437,
|
||
|
|
"step": 3240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6561514195583596,
|
||
|
|
"grad_norm": 0.47127385494827945,
|
||
|
|
"learning_rate": 3.195337936728323e-06,
|
||
|
|
"loss": 0.4488,
|
||
|
|
"step": 3250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6581703470031546,
|
||
|
|
"grad_norm": 0.43568607166498186,
|
||
|
|
"learning_rate": 3.1625227223929105e-06,
|
||
|
|
"loss": 0.4415,
|
||
|
|
"step": 3260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6601892744479495,
|
||
|
|
"grad_norm": 0.4200623699882891,
|
||
|
|
"learning_rate": 3.1297987595123024e-06,
|
||
|
|
"loss": 0.4423,
|
||
|
|
"step": 3270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6622082018927444,
|
||
|
|
"grad_norm": 0.45902924831849934,
|
||
|
|
"learning_rate": 3.0971676732000494e-06,
|
||
|
|
"loss": 0.4488,
|
||
|
|
"step": 3280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6642271293375395,
|
||
|
|
"grad_norm": 0.48610535709106045,
|
||
|
|
"learning_rate": 3.0646310839573302e-06,
|
||
|
|
"loss": 0.441,
|
||
|
|
"step": 3290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6662460567823344,
|
||
|
|
"grad_norm": 0.5237594196219926,
|
||
|
|
"learning_rate": 3.03219060759248e-06,
|
||
|
|
"loss": 0.454,
|
||
|
|
"step": 3300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6682649842271293,
|
||
|
|
"grad_norm": 0.4228282553579445,
|
||
|
|
"learning_rate": 2.9998478551407524e-06,
|
||
|
|
"loss": 0.4378,
|
||
|
|
"step": 3310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6702839116719242,
|
||
|
|
"grad_norm": 0.46926328208349977,
|
||
|
|
"learning_rate": 2.9676044327842963e-06,
|
||
|
|
"loss": 0.4475,
|
||
|
|
"step": 3320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6723028391167193,
|
||
|
|
"grad_norm": 0.47029450291352376,
|
||
|
|
"learning_rate": 2.935461941772412e-06,
|
||
|
|
"loss": 0.4418,
|
||
|
|
"step": 3330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6743217665615142,
|
||
|
|
"grad_norm": 0.461936788764074,
|
||
|
|
"learning_rate": 2.9034219783420183e-06,
|
||
|
|
"loss": 0.438,
|
||
|
|
"step": 3340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6763406940063091,
|
||
|
|
"grad_norm": 0.4879384251763818,
|
||
|
|
"learning_rate": 2.871486133638378e-06,
|
||
|
|
"loss": 0.4498,
|
||
|
|
"step": 3350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6783596214511041,
|
||
|
|
"grad_norm": 0.47390435742643133,
|
||
|
|
"learning_rate": 2.8396559936360934e-06,
|
||
|
|
"loss": 0.4444,
|
||
|
|
"step": 3360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6803785488958991,
|
||
|
|
"grad_norm": 0.43781719559966065,
|
||
|
|
"learning_rate": 2.807933139060337e-06,
|
||
|
|
"loss": 0.4481,
|
||
|
|
"step": 3370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.682397476340694,
|
||
|
|
"grad_norm": 0.452166890113836,
|
||
|
|
"learning_rate": 2.776319145308345e-06,
|
||
|
|
"loss": 0.4397,
|
||
|
|
"step": 3380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.684416403785489,
|
||
|
|
"grad_norm": 0.4480696892989974,
|
||
|
|
"learning_rate": 2.7448155823711897e-06,
|
||
|
|
"loss": 0.4415,
|
||
|
|
"step": 3390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6864353312302839,
|
||
|
|
"grad_norm": 0.482631983733039,
|
||
|
|
"learning_rate": 2.7134240147558124e-06,
|
||
|
|
"loss": 0.4494,
|
||
|
|
"step": 3400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6884542586750788,
|
||
|
|
"grad_norm": 0.4215351521880855,
|
||
|
|
"learning_rate": 2.6821460014073187e-06,
|
||
|
|
"loss": 0.4359,
|
||
|
|
"step": 3410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6904731861198738,
|
||
|
|
"grad_norm": 0.44170545621316426,
|
||
|
|
"learning_rate": 2.6509830956315684e-06,
|
||
|
|
"loss": 0.4516,
|
||
|
|
"step": 3420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6924921135646688,
|
||
|
|
"grad_norm": 0.43846533524679027,
|
||
|
|
"learning_rate": 2.619936845018031e-06,
|
||
|
|
"loss": 0.4357,
|
||
|
|
"step": 3430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6945110410094637,
|
||
|
|
"grad_norm": 0.5008934216933979,
|
||
|
|
"learning_rate": 2.5890087913629358e-06,
|
||
|
|
"loss": 0.4467,
|
||
|
|
"step": 3440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6965299684542586,
|
||
|
|
"grad_norm": 0.5250213600953301,
|
||
|
|
"learning_rate": 2.558200470592701e-06,
|
||
|
|
"loss": 0.449,
|
||
|
|
"step": 3450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.6985488958990537,
|
||
|
|
"grad_norm": 0.5362972001626155,
|
||
|
|
"learning_rate": 2.527513412687654e-06,
|
||
|
|
"loss": 0.436,
|
||
|
|
"step": 3460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7005678233438486,
|
||
|
|
"grad_norm": 0.4447368579357635,
|
||
|
|
"learning_rate": 2.4969491416060542e-06,
|
||
|
|
"loss": 0.4393,
|
||
|
|
"step": 3470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7025867507886435,
|
||
|
|
"grad_norm": 0.42761346324330285,
|
||
|
|
"learning_rate": 2.4665091752084175e-06,
|
||
|
|
"loss": 0.4423,
|
||
|
|
"step": 3480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7046056782334384,
|
||
|
|
"grad_norm": 0.45180644732668995,
|
||
|
|
"learning_rate": 2.4361950251821253e-06,
|
||
|
|
"loss": 0.4351,
|
||
|
|
"step": 3490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7066246056782335,
|
||
|
|
"grad_norm": 0.4655597774869552,
|
||
|
|
"learning_rate": 2.406008196966359e-06,
|
||
|
|
"loss": 0.445,
|
||
|
|
"step": 3500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7086435331230284,
|
||
|
|
"grad_norm": 0.44607134328593223,
|
||
|
|
"learning_rate": 2.375950189677342e-06,
|
||
|
|
"loss": 0.4322,
|
||
|
|
"step": 3510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7106624605678233,
|
||
|
|
"grad_norm": 0.45886523824853853,
|
||
|
|
"learning_rate": 2.346022496033879e-06,
|
||
|
|
"loss": 0.439,
|
||
|
|
"step": 3520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7126813880126183,
|
||
|
|
"grad_norm": 0.439206301308698,
|
||
|
|
"learning_rate": 2.316226602283243e-06,
|
||
|
|
"loss": 0.445,
|
||
|
|
"step": 3530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7147003154574133,
|
||
|
|
"grad_norm": 0.42444173885080955,
|
||
|
|
"learning_rate": 2.2865639881273494e-06,
|
||
|
|
"loss": 0.4402,
|
||
|
|
"step": 3540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7167192429022082,
|
||
|
|
"grad_norm": 0.48281990057391616,
|
||
|
|
"learning_rate": 2.25703612664928e-06,
|
||
|
|
"loss": 0.445,
|
||
|
|
"step": 3550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7187381703470032,
|
||
|
|
"grad_norm": 0.46127829547820787,
|
||
|
|
"learning_rate": 2.227644484240135e-06,
|
||
|
|
"loss": 0.4356,
|
||
|
|
"step": 3560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7207570977917981,
|
||
|
|
"grad_norm": 0.4610423355631118,
|
||
|
|
"learning_rate": 2.198390520526194e-06,
|
||
|
|
"loss": 0.4438,
|
||
|
|
"step": 3570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.722776025236593,
|
||
|
|
"grad_norm": 0.4711885745663891,
|
||
|
|
"learning_rate": 2.1692756882964404e-06,
|
||
|
|
"loss": 0.4464,
|
||
|
|
"step": 3580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.724794952681388,
|
||
|
|
"grad_norm": 0.5420237747030388,
|
||
|
|
"learning_rate": 2.140301433430414e-06,
|
||
|
|
"loss": 0.4436,
|
||
|
|
"step": 3590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.726813880126183,
|
||
|
|
"grad_norm": 0.47745493287406926,
|
||
|
|
"learning_rate": 2.1114691948264038e-06,
|
||
|
|
"loss": 0.4433,
|
||
|
|
"step": 3600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7288328075709779,
|
||
|
|
"grad_norm": 0.43683824200923127,
|
||
|
|
"learning_rate": 2.082780404329988e-06,
|
||
|
|
"loss": 0.4429,
|
||
|
|
"step": 3610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7308517350157728,
|
||
|
|
"grad_norm": 0.42141576682539833,
|
||
|
|
"learning_rate": 2.0542364866629292e-06,
|
||
|
|
"loss": 0.4402,
|
||
|
|
"step": 3620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7328706624605679,
|
||
|
|
"grad_norm": 0.4304526228372951,
|
||
|
|
"learning_rate": 2.025838859352428e-06,
|
||
|
|
"loss": 0.4462,
|
||
|
|
"step": 3630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7348895899053628,
|
||
|
|
"grad_norm": 0.4960972537231228,
|
||
|
|
"learning_rate": 1.997588932660713e-06,
|
||
|
|
"loss": 0.4298,
|
||
|
|
"step": 3640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7369085173501577,
|
||
|
|
"grad_norm": 0.5055917828329974,
|
||
|
|
"learning_rate": 1.9694881095150125e-06,
|
||
|
|
"loss": 0.4429,
|
||
|
|
"step": 3650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7389274447949526,
|
||
|
|
"grad_norm": 0.4756073151541865,
|
||
|
|
"learning_rate": 1.9415377854378903e-06,
|
||
|
|
"loss": 0.4423,
|
||
|
|
"step": 3660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7409463722397477,
|
||
|
|
"grad_norm": 0.4293238953341605,
|
||
|
|
"learning_rate": 1.913739348477927e-06,
|
||
|
|
"loss": 0.444,
|
||
|
|
"step": 3670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7429652996845426,
|
||
|
|
"grad_norm": 0.4476184732471472,
|
||
|
|
"learning_rate": 1.8860941791408038e-06,
|
||
|
|
"loss": 0.4432,
|
||
|
|
"step": 3680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7449842271293375,
|
||
|
|
"grad_norm": 0.4453341449071723,
|
||
|
|
"learning_rate": 1.8586036503207316e-06,
|
||
|
|
"loss": 0.4351,
|
||
|
|
"step": 3690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7470031545741325,
|
||
|
|
"grad_norm": 0.4741404781356218,
|
||
|
|
"learning_rate": 1.8312691272322768e-06,
|
||
|
|
"loss": 0.4445,
|
||
|
|
"step": 3700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7490220820189274,
|
||
|
|
"grad_norm": 0.4314595245056323,
|
||
|
|
"learning_rate": 1.804091967342569e-06,
|
||
|
|
"loss": 0.4392,
|
||
|
|
"step": 3710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7510410094637224,
|
||
|
|
"grad_norm": 0.4960471983973586,
|
||
|
|
"learning_rate": 1.777073520303874e-06,
|
||
|
|
"loss": 0.4406,
|
||
|
|
"step": 3720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7530599369085174,
|
||
|
|
"grad_norm": 0.47724657720097824,
|
||
|
|
"learning_rate": 1.7502151278865798e-06,
|
||
|
|
"loss": 0.4391,
|
||
|
|
"step": 3730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7550788643533123,
|
||
|
|
"grad_norm": 0.45566327509907684,
|
||
|
|
"learning_rate": 1.7235181239125587e-06,
|
||
|
|
"loss": 0.4411,
|
||
|
|
"step": 3740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7570977917981072,
|
||
|
|
"grad_norm": 0.47859031492685,
|
||
|
|
"learning_rate": 1.6969838341889288e-06,
|
||
|
|
"loss": 0.4412,
|
||
|
|
"step": 3750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7591167192429022,
|
||
|
|
"grad_norm": 0.4239970850626806,
|
||
|
|
"learning_rate": 1.67061357644221e-06,
|
||
|
|
"loss": 0.4327,
|
||
|
|
"step": 3760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7611356466876972,
|
||
|
|
"grad_norm": 0.41439508221708543,
|
||
|
|
"learning_rate": 1.6444086602528835e-06,
|
||
|
|
"loss": 0.4403,
|
||
|
|
"step": 3770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7631545741324921,
|
||
|
|
"grad_norm": 0.4724942386548484,
|
||
|
|
"learning_rate": 1.6183703869903666e-06,
|
||
|
|
"loss": 0.4399,
|
||
|
|
"step": 3780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.765173501577287,
|
||
|
|
"grad_norm": 0.5068083200686946,
|
||
|
|
"learning_rate": 1.59250004974837e-06,
|
||
|
|
"loss": 0.4484,
|
||
|
|
"step": 3790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7671924290220821,
|
||
|
|
"grad_norm": 0.47623913375530136,
|
||
|
|
"learning_rate": 1.5667989332806894e-06,
|
||
|
|
"loss": 0.442,
|
||
|
|
"step": 3800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.769211356466877,
|
||
|
|
"grad_norm": 0.5448383844625657,
|
||
|
|
"learning_rate": 1.5412683139374068e-06,
|
||
|
|
"loss": 0.4339,
|
||
|
|
"step": 3810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7712302839116719,
|
||
|
|
"grad_norm": 0.4819612793724808,
|
||
|
|
"learning_rate": 1.5159094596014934e-06,
|
||
|
|
"loss": 0.4478,
|
||
|
|
"step": 3820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7732492113564668,
|
||
|
|
"grad_norm": 0.4456007417706316,
|
||
|
|
"learning_rate": 1.4907236296258598e-06,
|
||
|
|
"loss": 0.4423,
|
||
|
|
"step": 3830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7752681388012618,
|
||
|
|
"grad_norm": 0.43979298393219013,
|
||
|
|
"learning_rate": 1.4657120747708014e-06,
|
||
|
|
"loss": 0.4417,
|
||
|
|
"step": 3840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7772870662460568,
|
||
|
|
"grad_norm": 0.5066920888924865,
|
||
|
|
"learning_rate": 1.4408760371418895e-06,
|
||
|
|
"loss": 0.4397,
|
||
|
|
"step": 3850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7793059936908517,
|
||
|
|
"grad_norm": 0.47046124228822955,
|
||
|
|
"learning_rate": 1.4162167501282925e-06,
|
||
|
|
"loss": 0.4416,
|
||
|
|
"step": 3860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7813249211356467,
|
||
|
|
"grad_norm": 0.4558898604297212,
|
||
|
|
"learning_rate": 1.391735438341514e-06,
|
||
|
|
"loss": 0.437,
|
||
|
|
"step": 3870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7833438485804416,
|
||
|
|
"grad_norm": 0.4635342780752439,
|
||
|
|
"learning_rate": 1.3674333175545818e-06,
|
||
|
|
"loss": 0.4425,
|
||
|
|
"step": 3880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7853627760252366,
|
||
|
|
"grad_norm": 0.450134125379297,
|
||
|
|
"learning_rate": 1.343311594641674e-06,
|
||
|
|
"loss": 0.4371,
|
||
|
|
"step": 3890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7873817034700316,
|
||
|
|
"grad_norm": 0.48501901287598154,
|
||
|
|
"learning_rate": 1.3193714675181813e-06,
|
||
|
|
"loss": 0.4428,
|
||
|
|
"step": 3900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7894006309148265,
|
||
|
|
"grad_norm": 0.41955405495588977,
|
||
|
|
"learning_rate": 1.2956141250812144e-06,
|
||
|
|
"loss": 0.4307,
|
||
|
|
"step": 3910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7914195583596214,
|
||
|
|
"grad_norm": 0.41659445986128424,
|
||
|
|
"learning_rate": 1.272040747150563e-06,
|
||
|
|
"loss": 0.447,
|
||
|
|
"step": 3920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7934384858044164,
|
||
|
|
"grad_norm": 0.44707673152149113,
|
||
|
|
"learning_rate": 1.2486525044101123e-06,
|
||
|
|
"loss": 0.4333,
|
||
|
|
"step": 3930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7954574132492114,
|
||
|
|
"grad_norm": 0.4698766268506151,
|
||
|
|
"learning_rate": 1.2254505583496922e-06,
|
||
|
|
"loss": 0.4403,
|
||
|
|
"step": 3940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7974763406940063,
|
||
|
|
"grad_norm": 0.47512471796296707,
|
||
|
|
"learning_rate": 1.2024360612074055e-06,
|
||
|
|
"loss": 0.4454,
|
||
|
|
"step": 3950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.7994952681388012,
|
||
|
|
"grad_norm": 0.4585910400457452,
|
||
|
|
"learning_rate": 1.179610155912405e-06,
|
||
|
|
"loss": 0.436,
|
||
|
|
"step": 3960
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8015141955835963,
|
||
|
|
"grad_norm": 0.4557389746705491,
|
||
|
|
"learning_rate": 1.1569739760281295e-06,
|
||
|
|
"loss": 0.4393,
|
||
|
|
"step": 3970
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8035331230283912,
|
||
|
|
"grad_norm": 0.4553106102818403,
|
||
|
|
"learning_rate": 1.1345286456960186e-06,
|
||
|
|
"loss": 0.4439,
|
||
|
|
"step": 3980
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8055520504731861,
|
||
|
|
"grad_norm": 0.4443887135897382,
|
||
|
|
"learning_rate": 1.1122752795796748e-06,
|
||
|
|
"loss": 0.4413,
|
||
|
|
"step": 3990
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.807570977917981,
|
||
|
|
"grad_norm": 0.47272003456758904,
|
||
|
|
"learning_rate": 1.0902149828095154e-06,
|
||
|
|
"loss": 0.4422,
|
||
|
|
"step": 4000
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.809589905362776,
|
||
|
|
"grad_norm": 0.43703108025040205,
|
||
|
|
"learning_rate": 1.0683488509278928e-06,
|
||
|
|
"loss": 0.4386,
|
||
|
|
"step": 4010
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.811608832807571,
|
||
|
|
"grad_norm": 0.4441202945224273,
|
||
|
|
"learning_rate": 1.0466779698346797e-06,
|
||
|
|
"loss": 0.4363,
|
||
|
|
"step": 4020
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8136277602523659,
|
||
|
|
"grad_norm": 0.42622028909468485,
|
||
|
|
"learning_rate": 1.025203415733348e-06,
|
||
|
|
"loss": 0.4365,
|
||
|
|
"step": 4030
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8156466876971609,
|
||
|
|
"grad_norm": 0.440821020950604,
|
||
|
|
"learning_rate": 1.0039262550775235e-06,
|
||
|
|
"loss": 0.4345,
|
||
|
|
"step": 4040
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8176656151419558,
|
||
|
|
"grad_norm": 0.4670601139166515,
|
||
|
|
"learning_rate": 9.828475445180212e-07,
|
||
|
|
"loss": 0.4404,
|
||
|
|
"step": 4050
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8196845425867508,
|
||
|
|
"grad_norm": 0.4363041969561276,
|
||
|
|
"learning_rate": 9.61968330850374e-07,
|
||
|
|
"loss": 0.4369,
|
||
|
|
"step": 4060
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8217034700315458,
|
||
|
|
"grad_norm": 0.4866069797752605,
|
||
|
|
"learning_rate": 9.412896509628439e-07,
|
||
|
|
"loss": 0.4434,
|
||
|
|
"step": 4070
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8237223974763407,
|
||
|
|
"grad_norm": 0.4360948594973755,
|
||
|
|
"learning_rate": 9.208125317849304e-07,
|
||
|
|
"loss": 0.4395,
|
||
|
|
"step": 4080
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8257413249211356,
|
||
|
|
"grad_norm": 0.4490621547393745,
|
||
|
|
"learning_rate": 9.005379902363753e-07,
|
||
|
|
"loss": 0.4405,
|
||
|
|
"step": 4090
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8277602523659306,
|
||
|
|
"grad_norm": 0.4809678412101839,
|
||
|
|
"learning_rate": 8.804670331766557e-07,
|
||
|
|
"loss": 0.4449,
|
||
|
|
"step": 4100
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8297791798107256,
|
||
|
|
"grad_norm": 0.4626290445825763,
|
||
|
|
"learning_rate": 8.606006573549836e-07,
|
||
|
|
"loss": 0.4359,
|
||
|
|
"step": 4110
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8317981072555205,
|
||
|
|
"grad_norm": 0.5187638972410576,
|
||
|
|
"learning_rate": 8.409398493608089e-07,
|
||
|
|
"loss": 0.4413,
|
||
|
|
"step": 4120
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8338170347003154,
|
||
|
|
"grad_norm": 0.46587792642977804,
|
||
|
|
"learning_rate": 8.214855855748249e-07,
|
||
|
|
"loss": 0.4409,
|
||
|
|
"step": 4130
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8358359621451104,
|
||
|
|
"grad_norm": 0.4706901825283824,
|
||
|
|
"learning_rate": 8.022388321204688e-07,
|
||
|
|
"loss": 0.4354,
|
||
|
|
"step": 4140
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8378548895899054,
|
||
|
|
"grad_norm": 0.562791948402191,
|
||
|
|
"learning_rate": 7.832005448159585e-07,
|
||
|
|
"loss": 0.4445,
|
||
|
|
"step": 4150
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8398738170347003,
|
||
|
|
"grad_norm": 0.4543738469403552,
|
||
|
|
"learning_rate": 7.643716691268171e-07,
|
||
|
|
"loss": 0.4374,
|
||
|
|
"step": 4160
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8418927444794952,
|
||
|
|
"grad_norm": 1.6174532184269235,
|
||
|
|
"learning_rate": 7.457531401189188e-07,
|
||
|
|
"loss": 0.4408,
|
||
|
|
"step": 4170
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8439116719242902,
|
||
|
|
"grad_norm": 0.46922997370194347,
|
||
|
|
"learning_rate": 7.273458824120555e-07,
|
||
|
|
"loss": 0.4376,
|
||
|
|
"step": 4180
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8459305993690852,
|
||
|
|
"grad_norm": 0.44210687899012985,
|
||
|
|
"learning_rate": 7.091508101340189e-07,
|
||
|
|
"loss": 0.4407,
|
||
|
|
"step": 4190
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8479495268138801,
|
||
|
|
"grad_norm": 0.488809217492704,
|
||
|
|
"learning_rate": 6.91168826875201e-07,
|
||
|
|
"loss": 0.4489,
|
||
|
|
"step": 4200
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8499684542586751,
|
||
|
|
"grad_norm": 0.4335482219101724,
|
||
|
|
"learning_rate": 6.734008256437253e-07,
|
||
|
|
"loss": 0.435,
|
||
|
|
"step": 4210
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.85198738170347,
|
||
|
|
"grad_norm": 0.4549494452467777,
|
||
|
|
"learning_rate": 6.558476888210935e-07,
|
||
|
|
"loss": 0.4375,
|
||
|
|
"step": 4220
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.854006309148265,
|
||
|
|
"grad_norm": 0.44896647048001603,
|
||
|
|
"learning_rate": 6.38510288118368e-07,
|
||
|
|
"loss": 0.4409,
|
||
|
|
"step": 4230
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.85602523659306,
|
||
|
|
"grad_norm": 0.43685823117883865,
|
||
|
|
"learning_rate": 6.213894845328844e-07,
|
||
|
|
"loss": 0.4418,
|
||
|
|
"step": 4240
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8580441640378549,
|
||
|
|
"grad_norm": 0.47792243578520655,
|
||
|
|
"learning_rate": 6.044861283054876e-07,
|
||
|
|
"loss": 0.4452,
|
||
|
|
"step": 4250
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8600630914826498,
|
||
|
|
"grad_norm": 0.43885235740502504,
|
||
|
|
"learning_rate": 5.878010588783101e-07,
|
||
|
|
"loss": 0.4423,
|
||
|
|
"step": 4260
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8620820189274448,
|
||
|
|
"grad_norm": 0.5121668333108534,
|
||
|
|
"learning_rate": 5.713351048530874e-07,
|
||
|
|
"loss": 0.4429,
|
||
|
|
"step": 4270
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8641009463722398,
|
||
|
|
"grad_norm": 0.44403508398564756,
|
||
|
|
"learning_rate": 5.550890839500067e-07,
|
||
|
|
"loss": 0.4511,
|
||
|
|
"step": 4280
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8661198738170347,
|
||
|
|
"grad_norm": 0.42769598938785636,
|
||
|
|
"learning_rate": 5.390638029670908e-07,
|
||
|
|
"loss": 0.4322,
|
||
|
|
"step": 4290
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8681388012618296,
|
||
|
|
"grad_norm": 0.4949663561795469,
|
||
|
|
"learning_rate": 5.232600577401436e-07,
|
||
|
|
"loss": 0.4457,
|
||
|
|
"step": 4300
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8701577287066246,
|
||
|
|
"grad_norm": 0.4489314407451146,
|
||
|
|
"learning_rate": 5.076786331032224e-07,
|
||
|
|
"loss": 0.4327,
|
||
|
|
"step": 4310
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8721766561514196,
|
||
|
|
"grad_norm": 0.48025463446625855,
|
||
|
|
"learning_rate": 4.923203028496576e-07,
|
||
|
|
"loss": 0.4447,
|
||
|
|
"step": 4320
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8741955835962145,
|
||
|
|
"grad_norm": 0.5743773868003562,
|
||
|
|
"learning_rate": 4.771858296936311e-07,
|
||
|
|
"loss": 0.4402,
|
||
|
|
"step": 4330
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8762145110410094,
|
||
|
|
"grad_norm": 0.4329672101913256,
|
||
|
|
"learning_rate": 4.622759652322989e-07,
|
||
|
|
"loss": 0.4455,
|
||
|
|
"step": 4340
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8782334384858044,
|
||
|
|
"grad_norm": 0.49048963795844497,
|
||
|
|
"learning_rate": 4.475914499084605e-07,
|
||
|
|
"loss": 0.4467,
|
||
|
|
"step": 4350
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8802523659305994,
|
||
|
|
"grad_norm": 0.4829611101804599,
|
||
|
|
"learning_rate": 4.331330129737954e-07,
|
||
|
|
"loss": 0.4287,
|
||
|
|
"step": 4360
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8822712933753943,
|
||
|
|
"grad_norm": 0.44140910883095635,
|
||
|
|
"learning_rate": 4.189013724526392e-07,
|
||
|
|
"loss": 0.4446,
|
||
|
|
"step": 4370
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8842902208201893,
|
||
|
|
"grad_norm": 0.45272062627566334,
|
||
|
|
"learning_rate": 4.048972351063313e-07,
|
||
|
|
"loss": 0.4354,
|
||
|
|
"step": 4380
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8863091482649842,
|
||
|
|
"grad_norm": 0.7049959167129851,
|
||
|
|
"learning_rate": 3.9112129639811535e-07,
|
||
|
|
"loss": 0.4346,
|
||
|
|
"step": 4390
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8883280757097792,
|
||
|
|
"grad_norm": 0.4811238411002169,
|
||
|
|
"learning_rate": 3.7757424045860025e-07,
|
||
|
|
"loss": 0.4372,
|
||
|
|
"step": 4400
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8903470031545742,
|
||
|
|
"grad_norm": 0.4540483251684103,
|
||
|
|
"learning_rate": 3.642567400517838e-07,
|
||
|
|
"loss": 0.4314,
|
||
|
|
"step": 4410
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8923659305993691,
|
||
|
|
"grad_norm": 0.47603601308214405,
|
||
|
|
"learning_rate": 3.511694565416485e-07,
|
||
|
|
"loss": 0.4382,
|
||
|
|
"step": 4420
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.894384858044164,
|
||
|
|
"grad_norm": 0.4286142580967842,
|
||
|
|
"learning_rate": 3.383130398593121e-07,
|
||
|
|
"loss": 0.4399,
|
||
|
|
"step": 4430
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.8964037854889589,
|
||
|
|
"grad_norm": 0.4337017987281163,
|
||
|
|
"learning_rate": 3.2568812847074913e-07,
|
||
|
|
"loss": 0.4437,
|
||
|
|
"step": 4440
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.898422712933754,
|
||
|
|
"grad_norm": 0.48386529314250487,
|
||
|
|
"learning_rate": 3.1329534934509186e-07,
|
||
|
|
"loss": 0.4415,
|
||
|
|
"step": 4450
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9004416403785489,
|
||
|
|
"grad_norm": 0.46990087367597777,
|
||
|
|
"learning_rate": 3.011353179234888e-07,
|
||
|
|
"loss": 0.4317,
|
||
|
|
"step": 4460
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9024605678233438,
|
||
|
|
"grad_norm": 0.42506034975479673,
|
||
|
|
"learning_rate": 2.8920863808853947e-07,
|
||
|
|
"loss": 0.444,
|
||
|
|
"step": 4470
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9044794952681388,
|
||
|
|
"grad_norm": 0.6216117310744336,
|
||
|
|
"learning_rate": 2.775159021343082e-07,
|
||
|
|
"loss": 0.4367,
|
||
|
|
"step": 4480
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9064984227129338,
|
||
|
|
"grad_norm": 0.4583995655944248,
|
||
|
|
"learning_rate": 2.660576907369111e-07,
|
||
|
|
"loss": 0.4412,
|
||
|
|
"step": 4490
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9085173501577287,
|
||
|
|
"grad_norm": 0.4889160958013681,
|
||
|
|
"learning_rate": 2.5483457292567517e-07,
|
||
|
|
"loss": 0.4403,
|
||
|
|
"step": 4500
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9105362776025236,
|
||
|
|
"grad_norm": 0.4267838945833425,
|
||
|
|
"learning_rate": 2.4384710605488195e-07,
|
||
|
|
"loss": 0.4402,
|
||
|
|
"step": 4510
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9125552050473186,
|
||
|
|
"grad_norm": 0.4986776395611631,
|
||
|
|
"learning_rate": 2.330958357760882e-07,
|
||
|
|
"loss": 0.4399,
|
||
|
|
"step": 4520
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9145741324921136,
|
||
|
|
"grad_norm": 0.459784020601065,
|
||
|
|
"learning_rate": 2.225812960110263e-07,
|
||
|
|
"loss": 0.4316,
|
||
|
|
"step": 4530
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9165930599369085,
|
||
|
|
"grad_norm": 0.44426443349474987,
|
||
|
|
"learning_rate": 2.123040089250944e-07,
|
||
|
|
"loss": 0.4402,
|
||
|
|
"step": 4540
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9186119873817035,
|
||
|
|
"grad_norm": 0.49235316049109906,
|
||
|
|
"learning_rate": 2.0226448490141715e-07,
|
||
|
|
"loss": 0.4481,
|
||
|
|
"step": 4550
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9206309148264984,
|
||
|
|
"grad_norm": 0.43058790479448,
|
||
|
|
"learning_rate": 1.9246322251550653e-07,
|
||
|
|
"loss": 0.4317,
|
||
|
|
"step": 4560
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9226498422712934,
|
||
|
|
"grad_norm": 0.4966838385984267,
|
||
|
|
"learning_rate": 1.829007085104978e-07,
|
||
|
|
"loss": 0.4427,
|
||
|
|
"step": 4570
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9246687697160884,
|
||
|
|
"grad_norm": 0.4666275183152453,
|
||
|
|
"learning_rate": 1.7357741777297987e-07,
|
||
|
|
"loss": 0.4485,
|
||
|
|
"step": 4580
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9266876971608833,
|
||
|
|
"grad_norm": 0.4412830191687912,
|
||
|
|
"learning_rate": 1.644938133094076e-07,
|
||
|
|
"loss": 0.4312,
|
||
|
|
"step": 4590
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9287066246056782,
|
||
|
|
"grad_norm": 0.4886548137136921,
|
||
|
|
"learning_rate": 1.5565034622311349e-07,
|
||
|
|
"loss": 0.4461,
|
||
|
|
"step": 4600
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9307255520504731,
|
||
|
|
"grad_norm": 0.4652701957357695,
|
||
|
|
"learning_rate": 1.4704745569190327e-07,
|
||
|
|
"loss": 0.4375,
|
||
|
|
"step": 4610
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9327444794952682,
|
||
|
|
"grad_norm": 0.5298674695548953,
|
||
|
|
"learning_rate": 1.3868556894624353e-07,
|
||
|
|
"loss": 0.4399,
|
||
|
|
"step": 4620
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9347634069400631,
|
||
|
|
"grad_norm": 0.4562802443840814,
|
||
|
|
"learning_rate": 1.30565101248048e-07,
|
||
|
|
"loss": 0.4385,
|
||
|
|
"step": 4630
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.936782334384858,
|
||
|
|
"grad_norm": 0.4578530629768795,
|
||
|
|
"learning_rate": 1.226864558700541e-07,
|
||
|
|
"loss": 0.4416,
|
||
|
|
"step": 4640
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.938801261829653,
|
||
|
|
"grad_norm": 0.4951660779551572,
|
||
|
|
"learning_rate": 1.1505002407579512e-07,
|
||
|
|
"loss": 0.4424,
|
||
|
|
"step": 4650
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.940820189274448,
|
||
|
|
"grad_norm": 0.44757841438265406,
|
||
|
|
"learning_rate": 1.0765618510017118e-07,
|
||
|
|
"loss": 0.4368,
|
||
|
|
"step": 4660
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9428391167192429,
|
||
|
|
"grad_norm": 1.5035499500846865,
|
||
|
|
"learning_rate": 1.0050530613061327e-07,
|
||
|
|
"loss": 0.4353,
|
||
|
|
"step": 4670
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9448580441640378,
|
||
|
|
"grad_norm": 0.4938389329390479,
|
||
|
|
"learning_rate": 9.359774228885177e-08,
|
||
|
|
"loss": 0.4391,
|
||
|
|
"step": 4680
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9468769716088328,
|
||
|
|
"grad_norm": 0.4711850612472287,
|
||
|
|
"learning_rate": 8.693383661327882e-08,
|
||
|
|
"loss": 0.4388,
|
||
|
|
"step": 4690
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9488958990536278,
|
||
|
|
"grad_norm": 0.48652374519922575,
|
||
|
|
"learning_rate": 8.051392004191139e-08,
|
||
|
|
"loss": 0.4486,
|
||
|
|
"step": 4700
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9509148264984227,
|
||
|
|
"grad_norm": 0.4473704324045684,
|
||
|
|
"learning_rate": 7.433831139595727e-08,
|
||
|
|
"loss": 0.4385,
|
||
|
|
"step": 4710
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9529337539432177,
|
||
|
|
"grad_norm": 0.509944701555344,
|
||
|
|
"learning_rate": 6.840731736398542e-08,
|
||
|
|
"loss": 0.4362,
|
||
|
|
"step": 4720
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9549526813880126,
|
||
|
|
"grad_norm": 0.4468123798191573,
|
||
|
|
"learning_rate": 6.272123248668993e-08,
|
||
|
|
"loss": 0.4356,
|
||
|
|
"step": 4730
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9569716088328075,
|
||
|
|
"grad_norm": 0.4432856047100861,
|
||
|
|
"learning_rate": 5.728033914226494e-08,
|
||
|
|
"loss": 0.4406,
|
||
|
|
"step": 4740
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9589905362776026,
|
||
|
|
"grad_norm": 0.8312641118207356,
|
||
|
|
"learning_rate": 5.2084907532383156e-08,
|
||
|
|
"loss": 0.4457,
|
||
|
|
"step": 4750
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9610094637223975,
|
||
|
|
"grad_norm": 0.4308428180533281,
|
||
|
|
"learning_rate": 4.7135195668776e-08,
|
||
|
|
"loss": 0.4337,
|
||
|
|
"step": 4760
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9630283911671924,
|
||
|
|
"grad_norm": 0.47706748470749133,
|
||
|
|
"learning_rate": 4.243144936041943e-08,
|
||
|
|
"loss": 0.4413,
|
||
|
|
"step": 4770
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9650473186119873,
|
||
|
|
"grad_norm": 0.46426542825118344,
|
||
|
|
"learning_rate": 3.797390220132757e-08,
|
||
|
|
"loss": 0.4325,
|
||
|
|
"step": 4780
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9670662460567824,
|
||
|
|
"grad_norm": 0.4324982289321672,
|
||
|
|
"learning_rate": 3.3762775558952576e-08,
|
||
|
|
"loss": 0.4377,
|
||
|
|
"step": 4790
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9690851735015773,
|
||
|
|
"grad_norm": 0.47719223462129745,
|
||
|
|
"learning_rate": 2.9798278563190618e-08,
|
||
|
|
"loss": 0.4439,
|
||
|
|
"step": 4800
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9711041009463722,
|
||
|
|
"grad_norm": 0.4540530080947163,
|
||
|
|
"learning_rate": 2.6080608095997438e-08,
|
||
|
|
"loss": 0.4377,
|
||
|
|
"step": 4810
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9731230283911672,
|
||
|
|
"grad_norm": 0.4413909184803219,
|
||
|
|
"learning_rate": 2.2609948781607827e-08,
|
||
|
|
"loss": 0.4424,
|
||
|
|
"step": 4820
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9751419558359622,
|
||
|
|
"grad_norm": 0.4712177319951425,
|
||
|
|
"learning_rate": 1.9386472977370728e-08,
|
||
|
|
"loss": 0.4412,
|
||
|
|
"step": 4830
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9771608832807571,
|
||
|
|
"grad_norm": 0.443140708468747,
|
||
|
|
"learning_rate": 1.6410340765187772e-08,
|
||
|
|
"loss": 0.4382,
|
||
|
|
"step": 4840
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.979179810725552,
|
||
|
|
"grad_norm": 0.4782858905945508,
|
||
|
|
"learning_rate": 1.3681699943564053e-08,
|
||
|
|
"loss": 0.4387,
|
||
|
|
"step": 4850
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.981198738170347,
|
||
|
|
"grad_norm": 0.6206662222549734,
|
||
|
|
"learning_rate": 1.1200686020266805e-08,
|
||
|
|
"loss": 0.4431,
|
||
|
|
"step": 4860
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.983217665615142,
|
||
|
|
"grad_norm": 2.8460570799996985,
|
||
|
|
"learning_rate": 8.967422205598541e-09,
|
||
|
|
"loss": 0.4393,
|
||
|
|
"step": 4870
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9852365930599369,
|
||
|
|
"grad_norm": 0.4435087118229933,
|
||
|
|
"learning_rate": 6.982019406275853e-09,
|
||
|
|
"loss": 0.436,
|
||
|
|
"step": 4880
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9872555205047319,
|
||
|
|
"grad_norm": 0.4300508595922509,
|
||
|
|
"learning_rate": 5.244576219922693e-09,
|
||
|
|
"loss": 0.4316,
|
||
|
|
"step": 4890
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9892744479495268,
|
||
|
|
"grad_norm": 0.49712300151356087,
|
||
|
|
"learning_rate": 3.755178930173742e-09,
|
||
|
|
"loss": 0.4463,
|
||
|
|
"step": 4900
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9912933753943217,
|
||
|
|
"grad_norm": 0.4798533947795922,
|
||
|
|
"learning_rate": 2.513901502389504e-09,
|
||
|
|
"loss": 0.4327,
|
||
|
|
"step": 4910
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9933123028391168,
|
||
|
|
"grad_norm": 0.4936501161499046,
|
||
|
|
"learning_rate": 1.520805579983131e-09,
|
||
|
|
"loss": 0.4403,
|
||
|
|
"step": 4920
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9953312302839117,
|
||
|
|
"grad_norm": 0.4638813307919199,
|
||
|
|
"learning_rate": 7.759404813600935e-10,
|
||
|
|
"loss": 0.4385,
|
||
|
|
"step": 4930
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9973501577287066,
|
||
|
|
"grad_norm": 0.43789683220104403,
|
||
|
|
"learning_rate": 2.7934319746625306e-10,
|
||
|
|
"loss": 0.4387,
|
||
|
|
"step": 4940
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 0.9993690851735015,
|
||
|
|
"grad_norm": 0.4980004114526748,
|
||
|
|
"learning_rate": 3.103838995321962e-11,
|
||
|
|
"loss": 0.4391,
|
||
|
|
"step": 4950
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"epoch": 1.0,
|
||
|
|
"step": 4954,
|
||
|
|
"total_flos": 1.1092292323919266e+19,
|
||
|
|
"train_loss": 0.46906891030845593,
|
||
|
|
"train_runtime": 81648.0087,
|
||
|
|
"train_samples_per_second": 7.765,
|
||
|
|
"train_steps_per_second": 0.061
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"logging_steps": 10,
|
||
|
|
"max_steps": 4954,
|
||
|
|
"num_input_tokens_seen": 0,
|
||
|
|
"num_train_epochs": 1,
|
||
|
|
"save_steps": 50000,
|
||
|
|
"stateful_callbacks": {
|
||
|
|
"TrainerControl": {
|
||
|
|
"args": {
|
||
|
|
"should_epoch_stop": false,
|
||
|
|
"should_evaluate": false,
|
||
|
|
"should_log": false,
|
||
|
|
"should_save": true,
|
||
|
|
"should_training_stop": true
|
||
|
|
},
|
||
|
|
"attributes": {}
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"total_flos": 1.1092292323919266e+19,
|
||
|
|
"train_batch_size": 1,
|
||
|
|
"trial_name": null,
|
||
|
|
"trial_params": null
|
||
|
|
}
|