GR00T_test / trainer_state.json
wikkard's picture
Upload folder using huggingface_hub
399015c verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 18.51851851851852,
"eval_steps": 500,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.018518518518518517,
"grad_norm": 8.40421199798584,
"learning_rate": 1.8e-06,
"loss": 0.7039,
"step": 10
},
{
"epoch": 0.037037037037037035,
"grad_norm": 2.8345916271209717,
"learning_rate": 3.8e-06,
"loss": 0.4905,
"step": 20
},
{
"epoch": 0.05555555555555555,
"grad_norm": 3.340608835220337,
"learning_rate": 5.8e-06,
"loss": 0.3101,
"step": 30
},
{
"epoch": 0.07407407407407407,
"grad_norm": 1.8772788047790527,
"learning_rate": 7.8e-06,
"loss": 0.212,
"step": 40
},
{
"epoch": 0.09259259259259259,
"grad_norm": 1.6856963634490967,
"learning_rate": 9.800000000000001e-06,
"loss": 0.1712,
"step": 50
},
{
"epoch": 0.1111111111111111,
"grad_norm": 0.9193410873413086,
"learning_rate": 1.18e-05,
"loss": 0.1594,
"step": 60
},
{
"epoch": 0.12962962962962962,
"grad_norm": 1.5567400455474854,
"learning_rate": 1.3800000000000002e-05,
"loss": 0.1423,
"step": 70
},
{
"epoch": 0.14814814814814814,
"grad_norm": 1.7364197969436646,
"learning_rate": 1.58e-05,
"loss": 0.1377,
"step": 80
},
{
"epoch": 0.16666666666666666,
"grad_norm": 0.7975390553474426,
"learning_rate": 1.78e-05,
"loss": 0.1255,
"step": 90
},
{
"epoch": 0.18518518518518517,
"grad_norm": 1.1305755376815796,
"learning_rate": 1.9800000000000004e-05,
"loss": 0.1161,
"step": 100
},
{
"epoch": 0.2037037037037037,
"grad_norm": 1.0273832082748413,
"learning_rate": 2.18e-05,
"loss": 0.1081,
"step": 110
},
{
"epoch": 0.2222222222222222,
"grad_norm": 0.6672540307044983,
"learning_rate": 2.38e-05,
"loss": 0.1037,
"step": 120
},
{
"epoch": 0.24074074074074073,
"grad_norm": 0.9112850427627563,
"learning_rate": 2.58e-05,
"loss": 0.0918,
"step": 130
},
{
"epoch": 0.25925925925925924,
"grad_norm": 0.7322865128517151,
"learning_rate": 2.7800000000000005e-05,
"loss": 0.0946,
"step": 140
},
{
"epoch": 0.2777777777777778,
"grad_norm": 1.086787462234497,
"learning_rate": 2.98e-05,
"loss": 0.0873,
"step": 150
},
{
"epoch": 0.2962962962962963,
"grad_norm": 0.8603954911231995,
"learning_rate": 3.18e-05,
"loss": 0.0816,
"step": 160
},
{
"epoch": 0.3148148148148148,
"grad_norm": 0.7118894457817078,
"learning_rate": 3.38e-05,
"loss": 0.0747,
"step": 170
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.9135509729385376,
"learning_rate": 3.58e-05,
"loss": 0.0672,
"step": 180
},
{
"epoch": 0.35185185185185186,
"grad_norm": 0.5829349756240845,
"learning_rate": 3.7800000000000004e-05,
"loss": 0.0632,
"step": 190
},
{
"epoch": 0.37037037037037035,
"grad_norm": 0.5715738534927368,
"learning_rate": 3.9800000000000005e-05,
"loss": 0.0604,
"step": 200
},
{
"epoch": 0.3888888888888889,
"grad_norm": 0.9450425505638123,
"learning_rate": 4.18e-05,
"loss": 0.0614,
"step": 210
},
{
"epoch": 0.4074074074074074,
"grad_norm": 0.5132505297660828,
"learning_rate": 4.38e-05,
"loss": 0.0581,
"step": 220
},
{
"epoch": 0.42592592592592593,
"grad_norm": 0.8497244119644165,
"learning_rate": 4.58e-05,
"loss": 0.0583,
"step": 230
},
{
"epoch": 0.4444444444444444,
"grad_norm": 1.0086556673049927,
"learning_rate": 4.78e-05,
"loss": 0.0563,
"step": 240
},
{
"epoch": 0.46296296296296297,
"grad_norm": 0.5299403667449951,
"learning_rate": 4.9800000000000004e-05,
"loss": 0.0539,
"step": 250
},
{
"epoch": 0.48148148148148145,
"grad_norm": 0.43011635541915894,
"learning_rate": 5.1800000000000005e-05,
"loss": 0.0513,
"step": 260
},
{
"epoch": 0.5,
"grad_norm": 0.6516759991645813,
"learning_rate": 5.380000000000001e-05,
"loss": 0.0517,
"step": 270
},
{
"epoch": 0.5185185185185185,
"grad_norm": 0.5121860504150391,
"learning_rate": 5.580000000000001e-05,
"loss": 0.0476,
"step": 280
},
{
"epoch": 0.5370370370370371,
"grad_norm": 0.6431804895401001,
"learning_rate": 5.7799999999999995e-05,
"loss": 0.0494,
"step": 290
},
{
"epoch": 0.5555555555555556,
"grad_norm": 0.5695860981941223,
"learning_rate": 5.9800000000000003e-05,
"loss": 0.0506,
"step": 300
},
{
"epoch": 0.5740740740740741,
"grad_norm": 0.5412283539772034,
"learning_rate": 6.18e-05,
"loss": 0.0523,
"step": 310
},
{
"epoch": 0.5925925925925926,
"grad_norm": 0.6154576539993286,
"learning_rate": 6.38e-05,
"loss": 0.0434,
"step": 320
},
{
"epoch": 0.6111111111111112,
"grad_norm": 0.5222446918487549,
"learning_rate": 6.58e-05,
"loss": 0.0525,
"step": 330
},
{
"epoch": 0.6296296296296297,
"grad_norm": 0.5334906578063965,
"learning_rate": 6.780000000000001e-05,
"loss": 0.0547,
"step": 340
},
{
"epoch": 0.6481481481481481,
"grad_norm": 0.49837547540664673,
"learning_rate": 6.98e-05,
"loss": 0.0452,
"step": 350
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.6329131722450256,
"learning_rate": 7.18e-05,
"loss": 0.051,
"step": 360
},
{
"epoch": 0.6851851851851852,
"grad_norm": 0.38624849915504456,
"learning_rate": 7.38e-05,
"loss": 0.0455,
"step": 370
},
{
"epoch": 0.7037037037037037,
"grad_norm": 0.6844066381454468,
"learning_rate": 7.58e-05,
"loss": 0.0428,
"step": 380
},
{
"epoch": 0.7222222222222222,
"grad_norm": 0.8023852109909058,
"learning_rate": 7.780000000000001e-05,
"loss": 0.0483,
"step": 390
},
{
"epoch": 0.7407407407407407,
"grad_norm": 0.7543495893478394,
"learning_rate": 7.98e-05,
"loss": 0.0478,
"step": 400
},
{
"epoch": 0.7592592592592593,
"grad_norm": 0.7769513130187988,
"learning_rate": 8.18e-05,
"loss": 0.0509,
"step": 410
},
{
"epoch": 0.7777777777777778,
"grad_norm": 0.5824040174484253,
"learning_rate": 8.38e-05,
"loss": 0.0461,
"step": 420
},
{
"epoch": 0.7962962962962963,
"grad_norm": 0.75037682056427,
"learning_rate": 8.58e-05,
"loss": 0.0483,
"step": 430
},
{
"epoch": 0.8148148148148148,
"grad_norm": 0.4793921113014221,
"learning_rate": 8.78e-05,
"loss": 0.0434,
"step": 440
},
{
"epoch": 0.8333333333333334,
"grad_norm": 0.6347702145576477,
"learning_rate": 8.98e-05,
"loss": 0.045,
"step": 450
},
{
"epoch": 0.8518518518518519,
"grad_norm": 0.5374931693077087,
"learning_rate": 9.180000000000001e-05,
"loss": 0.0425,
"step": 460
},
{
"epoch": 0.8703703703703703,
"grad_norm": 0.7056617140769958,
"learning_rate": 9.38e-05,
"loss": 0.0425,
"step": 470
},
{
"epoch": 0.8888888888888888,
"grad_norm": 0.7027549743652344,
"learning_rate": 9.58e-05,
"loss": 0.0459,
"step": 480
},
{
"epoch": 0.9074074074074074,
"grad_norm": 0.4468420743942261,
"learning_rate": 9.78e-05,
"loss": 0.0467,
"step": 490
},
{
"epoch": 0.9259259259259259,
"grad_norm": 0.4959528148174286,
"learning_rate": 9.98e-05,
"loss": 0.0457,
"step": 500
},
{
"epoch": 0.9444444444444444,
"grad_norm": 0.3801681101322174,
"learning_rate": 9.9999778549206e-05,
"loss": 0.0409,
"step": 510
},
{
"epoch": 0.9629629629629629,
"grad_norm": 0.5477695465087891,
"learning_rate": 9.999901304280685e-05,
"loss": 0.0445,
"step": 520
},
{
"epoch": 0.9814814814814815,
"grad_norm": 0.6408549547195435,
"learning_rate": 9.999770075521164e-05,
"loss": 0.047,
"step": 530
},
{
"epoch": 1.0,
"grad_norm": 1.3413923978805542,
"learning_rate": 9.99958417007713e-05,
"loss": 0.046,
"step": 540
},
{
"epoch": 1.0185185185185186,
"grad_norm": 0.4764229953289032,
"learning_rate": 9.999343589981615e-05,
"loss": 0.0398,
"step": 550
},
{
"epoch": 1.037037037037037,
"grad_norm": 0.4043806493282318,
"learning_rate": 9.999048337865568e-05,
"loss": 0.0415,
"step": 560
},
{
"epoch": 1.0555555555555556,
"grad_norm": 0.6044878959655762,
"learning_rate": 9.998698416957815e-05,
"loss": 0.0476,
"step": 570
},
{
"epoch": 1.074074074074074,
"grad_norm": 0.46906664967536926,
"learning_rate": 9.998293831085037e-05,
"loss": 0.041,
"step": 580
},
{
"epoch": 1.0925925925925926,
"grad_norm": 0.5724524259567261,
"learning_rate": 9.997834584671719e-05,
"loss": 0.0415,
"step": 590
},
{
"epoch": 1.1111111111111112,
"grad_norm": 0.6830921769142151,
"learning_rate": 9.997320682740107e-05,
"loss": 0.0431,
"step": 600
},
{
"epoch": 1.1296296296296295,
"grad_norm": 0.5186285972595215,
"learning_rate": 9.996752130910149e-05,
"loss": 0.0403,
"step": 610
},
{
"epoch": 1.1481481481481481,
"grad_norm": 0.4154529571533203,
"learning_rate": 9.99612893539944e-05,
"loss": 0.0439,
"step": 620
},
{
"epoch": 1.1666666666666667,
"grad_norm": 0.25487220287323,
"learning_rate": 9.995451103023144e-05,
"loss": 0.0416,
"step": 630
},
{
"epoch": 1.1851851851851851,
"grad_norm": 0.26666080951690674,
"learning_rate": 9.994718641193928e-05,
"loss": 0.0381,
"step": 640
},
{
"epoch": 1.2037037037037037,
"grad_norm": 0.43585100769996643,
"learning_rate": 9.993931557921874e-05,
"loss": 0.0427,
"step": 650
},
{
"epoch": 1.2222222222222223,
"grad_norm": 0.3880666196346283,
"learning_rate": 9.993089861814402e-05,
"loss": 0.0402,
"step": 660
},
{
"epoch": 1.2407407407407407,
"grad_norm": 0.4256173372268677,
"learning_rate": 9.992193562076166e-05,
"loss": 0.0393,
"step": 670
},
{
"epoch": 1.2592592592592593,
"grad_norm": 0.6624224185943604,
"learning_rate": 9.991242668508954e-05,
"loss": 0.0377,
"step": 680
},
{
"epoch": 1.2777777777777777,
"grad_norm": 0.44722142815589905,
"learning_rate": 9.990237191511587e-05,
"loss": 0.036,
"step": 690
},
{
"epoch": 1.2962962962962963,
"grad_norm": 0.4908144772052765,
"learning_rate": 9.989177142079802e-05,
"loss": 0.0358,
"step": 700
},
{
"epoch": 1.3148148148148149,
"grad_norm": 0.33555495738983154,
"learning_rate": 9.988062531806126e-05,
"loss": 0.0386,
"step": 710
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.21528975665569305,
"learning_rate": 9.986893372879762e-05,
"loss": 0.0382,
"step": 720
},
{
"epoch": 1.3518518518518519,
"grad_norm": 0.36565887928009033,
"learning_rate": 9.985669678086443e-05,
"loss": 0.0361,
"step": 730
},
{
"epoch": 1.3703703703703702,
"grad_norm": 0.3077153265476227,
"learning_rate": 9.984391460808298e-05,
"loss": 0.0428,
"step": 740
},
{
"epoch": 1.3888888888888888,
"grad_norm": 0.4870428740978241,
"learning_rate": 9.983058735023709e-05,
"loss": 0.0388,
"step": 750
},
{
"epoch": 1.4074074074074074,
"grad_norm": 0.4743514955043793,
"learning_rate": 9.98167151530715e-05,
"loss": 0.042,
"step": 760
},
{
"epoch": 1.425925925925926,
"grad_norm": 0.4561086893081665,
"learning_rate": 9.980229816829034e-05,
"loss": 0.0404,
"step": 770
},
{
"epoch": 1.4444444444444444,
"grad_norm": 0.37524229288101196,
"learning_rate": 9.978733655355544e-05,
"loss": 0.0379,
"step": 780
},
{
"epoch": 1.462962962962963,
"grad_norm": 0.4174795150756836,
"learning_rate": 9.977183047248464e-05,
"loss": 0.0365,
"step": 790
},
{
"epoch": 1.4814814814814814,
"grad_norm": 0.5232712626457214,
"learning_rate": 9.975578009464992e-05,
"loss": 0.0388,
"step": 800
},
{
"epoch": 1.5,
"grad_norm": 0.5445942282676697,
"learning_rate": 9.97391855955757e-05,
"loss": 0.0349,
"step": 810
},
{
"epoch": 1.5185185185185186,
"grad_norm": 0.4619694948196411,
"learning_rate": 9.972204715673669e-05,
"loss": 0.0345,
"step": 820
},
{
"epoch": 1.5370370370370372,
"grad_norm": 0.3875766396522522,
"learning_rate": 9.970436496555617e-05,
"loss": 0.0362,
"step": 830
},
{
"epoch": 1.5555555555555556,
"grad_norm": 0.371372789144516,
"learning_rate": 9.968613921540373e-05,
"loss": 0.0355,
"step": 840
},
{
"epoch": 1.574074074074074,
"grad_norm": 0.42488643527030945,
"learning_rate": 9.966737010559326e-05,
"loss": 0.0362,
"step": 850
},
{
"epoch": 1.5925925925925926,
"grad_norm": 0.4419741928577423,
"learning_rate": 9.964805784138072e-05,
"loss": 0.0355,
"step": 860
},
{
"epoch": 1.6111111111111112,
"grad_norm": 0.7232251167297363,
"learning_rate": 9.962820263396195e-05,
"loss": 0.0372,
"step": 870
},
{
"epoch": 1.6296296296296298,
"grad_norm": 0.42667055130004883,
"learning_rate": 9.960780470047033e-05,
"loss": 0.039,
"step": 880
},
{
"epoch": 1.6481481481481481,
"grad_norm": 0.5010427832603455,
"learning_rate": 9.958686426397437e-05,
"loss": 0.0365,
"step": 890
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.39860647916793823,
"learning_rate": 9.956538155347534e-05,
"loss": 0.0367,
"step": 900
},
{
"epoch": 1.6851851851851851,
"grad_norm": 0.4737634062767029,
"learning_rate": 9.95433568039047e-05,
"loss": 0.0349,
"step": 910
},
{
"epoch": 1.7037037037037037,
"grad_norm": 0.4173809587955475,
"learning_rate": 9.952079025612162e-05,
"loss": 0.0336,
"step": 920
},
{
"epoch": 1.7222222222222223,
"grad_norm": 0.43052947521209717,
"learning_rate": 9.949768215691022e-05,
"loss": 0.0372,
"step": 930
},
{
"epoch": 1.7407407407407407,
"grad_norm": 0.3775675296783447,
"learning_rate": 9.9474032758977e-05,
"loss": 0.0344,
"step": 940
},
{
"epoch": 1.7592592592592593,
"grad_norm": 0.4821918308734894,
"learning_rate": 9.944984232094794e-05,
"loss": 0.0332,
"step": 950
},
{
"epoch": 1.7777777777777777,
"grad_norm": 0.538884699344635,
"learning_rate": 9.942511110736584e-05,
"loss": 0.036,
"step": 960
},
{
"epoch": 1.7962962962962963,
"grad_norm": 0.3857181966304779,
"learning_rate": 9.939983938868726e-05,
"loss": 0.033,
"step": 970
},
{
"epoch": 1.8148148148148149,
"grad_norm": 0.3235240876674652,
"learning_rate": 9.93740274412797e-05,
"loss": 0.0333,
"step": 980
},
{
"epoch": 1.8333333333333335,
"grad_norm": 0.5306416153907776,
"learning_rate": 9.934767554741846e-05,
"loss": 0.0373,
"step": 990
},
{
"epoch": 1.8518518518518519,
"grad_norm": 0.37030676007270813,
"learning_rate": 9.932078399528361e-05,
"loss": 0.0327,
"step": 1000
},
{
"epoch": 1.8703703703703702,
"grad_norm": 0.3042660355567932,
"learning_rate": 9.929335307895689e-05,
"loss": 0.0324,
"step": 1010
},
{
"epoch": 1.8888888888888888,
"grad_norm": 0.3024146258831024,
"learning_rate": 9.926538309841839e-05,
"loss": 0.0388,
"step": 1020
},
{
"epoch": 1.9074074074074074,
"grad_norm": 0.4523896872997284,
"learning_rate": 9.923687435954334e-05,
"loss": 0.033,
"step": 1030
},
{
"epoch": 1.925925925925926,
"grad_norm": 0.41738566756248474,
"learning_rate": 9.920782717409873e-05,
"loss": 0.0321,
"step": 1040
},
{
"epoch": 1.9444444444444444,
"grad_norm": 0.45224565267562866,
"learning_rate": 9.917824185973994e-05,
"loss": 0.0315,
"step": 1050
},
{
"epoch": 1.9629629629629628,
"grad_norm": 0.5744406580924988,
"learning_rate": 9.914811874000723e-05,
"loss": 0.0336,
"step": 1060
},
{
"epoch": 1.9814814814814814,
"grad_norm": 0.3751019537448883,
"learning_rate": 9.911745814432218e-05,
"loss": 0.0338,
"step": 1070
},
{
"epoch": 2.0,
"grad_norm": 0.9458900094032288,
"learning_rate": 9.90862604079842e-05,
"loss": 0.0337,
"step": 1080
},
{
"epoch": 2.0185185185185186,
"grad_norm": 0.28884032368659973,
"learning_rate": 9.90545258721667e-05,
"loss": 0.032,
"step": 1090
},
{
"epoch": 2.037037037037037,
"grad_norm": 0.241240993142128,
"learning_rate": 9.90222548839135e-05,
"loss": 0.0333,
"step": 1100
},
{
"epoch": 2.0555555555555554,
"grad_norm": 0.24158261716365814,
"learning_rate": 9.898944779613495e-05,
"loss": 0.0288,
"step": 1110
},
{
"epoch": 2.074074074074074,
"grad_norm": 0.4657900631427765,
"learning_rate": 9.89561049676041e-05,
"loss": 0.0341,
"step": 1120
},
{
"epoch": 2.0925925925925926,
"grad_norm": 0.5393645763397217,
"learning_rate": 9.89222267629528e-05,
"loss": 0.0314,
"step": 1130
},
{
"epoch": 2.111111111111111,
"grad_norm": 0.4563658833503723,
"learning_rate": 9.888781355266763e-05,
"loss": 0.0314,
"step": 1140
},
{
"epoch": 2.1296296296296298,
"grad_norm": 0.3882819414138794,
"learning_rate": 9.885286571308598e-05,
"loss": 0.0288,
"step": 1150
},
{
"epoch": 2.148148148148148,
"grad_norm": 0.4034639000892639,
"learning_rate": 9.881738362639182e-05,
"loss": 0.0323,
"step": 1160
},
{
"epoch": 2.1666666666666665,
"grad_norm": 0.3633096516132355,
"learning_rate": 9.878136768061154e-05,
"loss": 0.0347,
"step": 1170
},
{
"epoch": 2.185185185185185,
"grad_norm": 0.3095862567424774,
"learning_rate": 9.874481826960979e-05,
"loss": 0.0327,
"step": 1180
},
{
"epoch": 2.2037037037037037,
"grad_norm": 0.4025212824344635,
"learning_rate": 9.870773579308503e-05,
"loss": 0.029,
"step": 1190
},
{
"epoch": 2.2222222222222223,
"grad_norm": 0.5413378477096558,
"learning_rate": 9.867012065656533e-05,
"loss": 0.0329,
"step": 1200
},
{
"epoch": 2.240740740740741,
"grad_norm": 0.23645605146884918,
"learning_rate": 9.863197327140376e-05,
"loss": 0.0307,
"step": 1210
},
{
"epoch": 2.259259259259259,
"grad_norm": 0.20576119422912598,
"learning_rate": 9.859329405477403e-05,
"loss": 0.028,
"step": 1220
},
{
"epoch": 2.2777777777777777,
"grad_norm": 0.3061058819293976,
"learning_rate": 9.855408342966585e-05,
"loss": 0.0297,
"step": 1230
},
{
"epoch": 2.2962962962962963,
"grad_norm": 0.3026726543903351,
"learning_rate": 9.851434182488033e-05,
"loss": 0.0295,
"step": 1240
},
{
"epoch": 2.314814814814815,
"grad_norm": 0.4563775658607483,
"learning_rate": 9.84740696750253e-05,
"loss": 0.0305,
"step": 1250
},
{
"epoch": 2.3333333333333335,
"grad_norm": 0.35343101620674133,
"learning_rate": 9.843326742051055e-05,
"loss": 0.0319,
"step": 1260
},
{
"epoch": 2.351851851851852,
"grad_norm": 0.2843523323535919,
"learning_rate": 9.839193550754297e-05,
"loss": 0.0266,
"step": 1270
},
{
"epoch": 2.3703703703703702,
"grad_norm": 0.25530385971069336,
"learning_rate": 9.835007438812177e-05,
"loss": 0.033,
"step": 1280
},
{
"epoch": 2.388888888888889,
"grad_norm": 0.3140629231929779,
"learning_rate": 9.830768452003341e-05,
"loss": 0.03,
"step": 1290
},
{
"epoch": 2.4074074074074074,
"grad_norm": 0.4510614573955536,
"learning_rate": 9.826476636684671e-05,
"loss": 0.0299,
"step": 1300
},
{
"epoch": 2.425925925925926,
"grad_norm": 0.4127548336982727,
"learning_rate": 9.822132039790773e-05,
"loss": 0.0318,
"step": 1310
},
{
"epoch": 2.4444444444444446,
"grad_norm": 0.31162577867507935,
"learning_rate": 9.817734708833461e-05,
"loss": 0.0269,
"step": 1320
},
{
"epoch": 2.462962962962963,
"grad_norm": 0.3712106943130493,
"learning_rate": 9.813284691901243e-05,
"loss": 0.032,
"step": 1330
},
{
"epoch": 2.4814814814814814,
"grad_norm": 0.4399408996105194,
"learning_rate": 9.808782037658792e-05,
"loss": 0.0327,
"step": 1340
},
{
"epoch": 2.5,
"grad_norm": 0.3978061378002167,
"learning_rate": 9.804226795346411e-05,
"loss": 0.0305,
"step": 1350
},
{
"epoch": 2.5185185185185186,
"grad_norm": 0.2510860860347748,
"learning_rate": 9.799619014779503e-05,
"loss": 0.0303,
"step": 1360
},
{
"epoch": 2.537037037037037,
"grad_norm": 0.44870445132255554,
"learning_rate": 9.794958746348013e-05,
"loss": 0.0318,
"step": 1370
},
{
"epoch": 2.5555555555555554,
"grad_norm": 0.6391816735267639,
"learning_rate": 9.790246041015896e-05,
"loss": 0.027,
"step": 1380
},
{
"epoch": 2.574074074074074,
"grad_norm": 0.5784100890159607,
"learning_rate": 9.785480950320538e-05,
"loss": 0.0259,
"step": 1390
},
{
"epoch": 2.5925925925925926,
"grad_norm": 0.39862194657325745,
"learning_rate": 9.78066352637221e-05,
"loss": 0.0284,
"step": 1400
},
{
"epoch": 2.611111111111111,
"grad_norm": 0.3439476191997528,
"learning_rate": 9.775793821853488e-05,
"loss": 0.026,
"step": 1410
},
{
"epoch": 2.6296296296296298,
"grad_norm": 0.3043521046638489,
"learning_rate": 9.77087189001868e-05,
"loss": 0.0292,
"step": 1420
},
{
"epoch": 2.648148148148148,
"grad_norm": 0.3189948797225952,
"learning_rate": 9.765897784693243e-05,
"loss": 0.0266,
"step": 1430
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.4445738196372986,
"learning_rate": 9.760871560273197e-05,
"loss": 0.0271,
"step": 1440
},
{
"epoch": 2.685185185185185,
"grad_norm": 0.32004910707473755,
"learning_rate": 9.755793271724526e-05,
"loss": 0.0275,
"step": 1450
},
{
"epoch": 2.7037037037037037,
"grad_norm": 0.24004176259040833,
"learning_rate": 9.750662974582584e-05,
"loss": 0.0333,
"step": 1460
},
{
"epoch": 2.7222222222222223,
"grad_norm": 0.3616419732570648,
"learning_rate": 9.745480724951473e-05,
"loss": 0.0267,
"step": 1470
},
{
"epoch": 2.7407407407407405,
"grad_norm": 0.409345418214798,
"learning_rate": 9.740246579503447e-05,
"loss": 0.0291,
"step": 1480
},
{
"epoch": 2.7592592592592595,
"grad_norm": 0.336698442697525,
"learning_rate": 9.734960595478284e-05,
"loss": 0.0285,
"step": 1490
},
{
"epoch": 2.7777777777777777,
"grad_norm": 0.3766871392726898,
"learning_rate": 9.729622830682657e-05,
"loss": 0.0251,
"step": 1500
},
{
"epoch": 2.7962962962962963,
"grad_norm": 0.25683119893074036,
"learning_rate": 9.724233343489504e-05,
"loss": 0.0282,
"step": 1510
},
{
"epoch": 2.814814814814815,
"grad_norm": 0.2245979607105255,
"learning_rate": 9.718792192837396e-05,
"loss": 0.027,
"step": 1520
},
{
"epoch": 2.8333333333333335,
"grad_norm": 0.3723769783973694,
"learning_rate": 9.713299438229886e-05,
"loss": 0.028,
"step": 1530
},
{
"epoch": 2.851851851851852,
"grad_norm": 0.20750750601291656,
"learning_rate": 9.707755139734855e-05,
"loss": 0.0282,
"step": 1540
},
{
"epoch": 2.8703703703703702,
"grad_norm": 0.25341686606407166,
"learning_rate": 9.702159357983866e-05,
"loss": 0.0269,
"step": 1550
},
{
"epoch": 2.888888888888889,
"grad_norm": 0.3434199392795563,
"learning_rate": 9.696512154171492e-05,
"loss": 0.0306,
"step": 1560
},
{
"epoch": 2.9074074074074074,
"grad_norm": 0.39540037512779236,
"learning_rate": 9.690813590054645e-05,
"loss": 0.0279,
"step": 1570
},
{
"epoch": 2.925925925925926,
"grad_norm": 0.2129739671945572,
"learning_rate": 9.685063727951914e-05,
"loss": 0.0243,
"step": 1580
},
{
"epoch": 2.9444444444444446,
"grad_norm": 0.2902113199234009,
"learning_rate": 9.679262630742865e-05,
"loss": 0.0282,
"step": 1590
},
{
"epoch": 2.962962962962963,
"grad_norm": 0.2857179641723633,
"learning_rate": 9.673410361867373e-05,
"loss": 0.0308,
"step": 1600
},
{
"epoch": 2.9814814814814814,
"grad_norm": 0.29263952374458313,
"learning_rate": 9.667506985324909e-05,
"loss": 0.0276,
"step": 1610
},
{
"epoch": 3.0,
"grad_norm": 0.6716474890708923,
"learning_rate": 9.661552565673855e-05,
"loss": 0.0237,
"step": 1620
},
{
"epoch": 3.0185185185185186,
"grad_norm": 0.35213494300842285,
"learning_rate": 9.655547168030789e-05,
"loss": 0.0276,
"step": 1630
},
{
"epoch": 3.037037037037037,
"grad_norm": 0.3571980595588684,
"learning_rate": 9.649490858069777e-05,
"loss": 0.0289,
"step": 1640
},
{
"epoch": 3.0555555555555554,
"grad_norm": 0.4329644739627838,
"learning_rate": 9.643383702021658e-05,
"loss": 0.0245,
"step": 1650
},
{
"epoch": 3.074074074074074,
"grad_norm": 0.33105847239494324,
"learning_rate": 9.637225766673307e-05,
"loss": 0.0245,
"step": 1660
},
{
"epoch": 3.0925925925925926,
"grad_norm": 0.399136483669281,
"learning_rate": 9.631017119366922e-05,
"loss": 0.0276,
"step": 1670
},
{
"epoch": 3.111111111111111,
"grad_norm": 0.4066803753376007,
"learning_rate": 9.624757827999273e-05,
"loss": 0.0268,
"step": 1680
},
{
"epoch": 3.1296296296296298,
"grad_norm": 0.2634310722351074,
"learning_rate": 9.618447961020971e-05,
"loss": 0.0289,
"step": 1690
},
{
"epoch": 3.148148148148148,
"grad_norm": 0.24467220902442932,
"learning_rate": 9.612087587435707e-05,
"loss": 0.0281,
"step": 1700
},
{
"epoch": 3.1666666666666665,
"grad_norm": 0.3698366582393646,
"learning_rate": 9.605676776799508e-05,
"loss": 0.0304,
"step": 1710
},
{
"epoch": 3.185185185185185,
"grad_norm": 0.2729560434818268,
"learning_rate": 9.599215599219973e-05,
"loss": 0.0234,
"step": 1720
},
{
"epoch": 3.2037037037037037,
"grad_norm": 0.4035135507583618,
"learning_rate": 9.592704125355505e-05,
"loss": 0.0282,
"step": 1730
},
{
"epoch": 3.2222222222222223,
"grad_norm": 0.28083720803260803,
"learning_rate": 9.586142426414538e-05,
"loss": 0.0248,
"step": 1740
},
{
"epoch": 3.240740740740741,
"grad_norm": 0.30893614888191223,
"learning_rate": 9.57953057415476e-05,
"loss": 0.0299,
"step": 1750
},
{
"epoch": 3.259259259259259,
"grad_norm": 0.29971230030059814,
"learning_rate": 9.572868640882328e-05,
"loss": 0.0267,
"step": 1760
},
{
"epoch": 3.2777777777777777,
"grad_norm": 0.42802125215530396,
"learning_rate": 9.56615669945108e-05,
"loss": 0.0263,
"step": 1770
},
{
"epoch": 3.2962962962962963,
"grad_norm": 0.49393147230148315,
"learning_rate": 9.55939482326173e-05,
"loss": 0.029,
"step": 1780
},
{
"epoch": 3.314814814814815,
"grad_norm": 0.2981662154197693,
"learning_rate": 9.552583086261069e-05,
"loss": 0.0288,
"step": 1790
},
{
"epoch": 3.3333333333333335,
"grad_norm": 0.352797269821167,
"learning_rate": 9.545721562941168e-05,
"loss": 0.0292,
"step": 1800
},
{
"epoch": 3.351851851851852,
"grad_norm": 0.3073903024196625,
"learning_rate": 9.538810328338543e-05,
"loss": 0.0277,
"step": 1810
},
{
"epoch": 3.3703703703703702,
"grad_norm": 0.311764657497406,
"learning_rate": 9.531849458033349e-05,
"loss": 0.0259,
"step": 1820
},
{
"epoch": 3.388888888888889,
"grad_norm": 0.3146636486053467,
"learning_rate": 9.524839028148547e-05,
"loss": 0.0255,
"step": 1830
},
{
"epoch": 3.4074074074074074,
"grad_norm": 0.3075493574142456,
"learning_rate": 9.517779115349077e-05,
"loss": 0.0229,
"step": 1840
},
{
"epoch": 3.425925925925926,
"grad_norm": 0.3865671157836914,
"learning_rate": 9.510669796841014e-05,
"loss": 0.0298,
"step": 1850
},
{
"epoch": 3.4444444444444446,
"grad_norm": 0.38654646277427673,
"learning_rate": 9.503511150370727e-05,
"loss": 0.0273,
"step": 1860
},
{
"epoch": 3.462962962962963,
"grad_norm": 0.3197333812713623,
"learning_rate": 9.496303254224024e-05,
"loss": 0.0246,
"step": 1870
},
{
"epoch": 3.4814814814814814,
"grad_norm": 0.3070426881313324,
"learning_rate": 9.489046187225306e-05,
"loss": 0.0274,
"step": 1880
},
{
"epoch": 3.5,
"grad_norm": 0.29184794425964355,
"learning_rate": 9.481740028736692e-05,
"loss": 0.0314,
"step": 1890
},
{
"epoch": 3.5185185185185186,
"grad_norm": 0.3075173795223236,
"learning_rate": 9.474384858657164e-05,
"loss": 0.0244,
"step": 1900
},
{
"epoch": 3.537037037037037,
"grad_norm": 0.31871992349624634,
"learning_rate": 9.466980757421679e-05,
"loss": 0.0258,
"step": 1910
},
{
"epoch": 3.5555555555555554,
"grad_norm": 0.40256521105766296,
"learning_rate": 9.459527806000305e-05,
"loss": 0.0289,
"step": 1920
},
{
"epoch": 3.574074074074074,
"grad_norm": 0.2957729995250702,
"learning_rate": 9.452026085897325e-05,
"loss": 0.0242,
"step": 1930
},
{
"epoch": 3.5925925925925926,
"grad_norm": 0.242269366979599,
"learning_rate": 9.444475679150348e-05,
"loss": 0.0277,
"step": 1940
},
{
"epoch": 3.611111111111111,
"grad_norm": 0.3702077567577362,
"learning_rate": 9.436876668329411e-05,
"loss": 0.0279,
"step": 1950
},
{
"epoch": 3.6296296296296298,
"grad_norm": 0.4864673614501953,
"learning_rate": 9.429229136536079e-05,
"loss": 0.0253,
"step": 1960
},
{
"epoch": 3.648148148148148,
"grad_norm": 0.3910199701786041,
"learning_rate": 9.421533167402534e-05,
"loss": 0.0265,
"step": 1970
},
{
"epoch": 3.6666666666666665,
"grad_norm": 0.28141796588897705,
"learning_rate": 9.413788845090666e-05,
"loss": 0.0281,
"step": 1980
},
{
"epoch": 3.685185185185185,
"grad_norm": 0.19294677674770355,
"learning_rate": 9.405996254291136e-05,
"loss": 0.0224,
"step": 1990
},
{
"epoch": 3.7037037037037037,
"grad_norm": 0.18972665071487427,
"learning_rate": 9.398155480222474e-05,
"loss": 0.0249,
"step": 2000
},
{
"epoch": 3.7222222222222223,
"grad_norm": 0.4179668128490448,
"learning_rate": 9.390266608630128e-05,
"loss": 0.0238,
"step": 2010
},
{
"epoch": 3.7407407407407405,
"grad_norm": 0.41081327199935913,
"learning_rate": 9.38232972578553e-05,
"loss": 0.0275,
"step": 2020
},
{
"epoch": 3.7592592592592595,
"grad_norm": 0.354586124420166,
"learning_rate": 9.374344918485164e-05,
"loss": 0.0275,
"step": 2030
},
{
"epoch": 3.7777777777777777,
"grad_norm": 0.27414700388908386,
"learning_rate": 9.366312274049602e-05,
"loss": 0.0255,
"step": 2040
},
{
"epoch": 3.7962962962962963,
"grad_norm": 0.5338584780693054,
"learning_rate": 9.358231880322554e-05,
"loss": 0.0271,
"step": 2050
},
{
"epoch": 3.814814814814815,
"grad_norm": 0.3043706715106964,
"learning_rate": 9.350103825669916e-05,
"loss": 0.0248,
"step": 2060
},
{
"epoch": 3.8333333333333335,
"grad_norm": 0.213580921292305,
"learning_rate": 9.341928198978787e-05,
"loss": 0.0247,
"step": 2070
},
{
"epoch": 3.851851851851852,
"grad_norm": 0.2466709166765213,
"learning_rate": 9.333705089656512e-05,
"loss": 0.0233,
"step": 2080
},
{
"epoch": 3.8703703703703702,
"grad_norm": 0.2591746151447296,
"learning_rate": 9.325434587629698e-05,
"loss": 0.0256,
"step": 2090
},
{
"epoch": 3.888888888888889,
"grad_norm": 0.2568264901638031,
"learning_rate": 9.31711678334323e-05,
"loss": 0.0256,
"step": 2100
},
{
"epoch": 3.9074074074074074,
"grad_norm": 0.2767021358013153,
"learning_rate": 9.308751767759282e-05,
"loss": 0.023,
"step": 2110
},
{
"epoch": 3.925925925925926,
"grad_norm": 0.21063950657844543,
"learning_rate": 9.300339632356325e-05,
"loss": 0.0233,
"step": 2120
},
{
"epoch": 3.9444444444444446,
"grad_norm": 0.2714349031448364,
"learning_rate": 9.291880469128124e-05,
"loss": 0.0294,
"step": 2130
},
{
"epoch": 3.962962962962963,
"grad_norm": 0.3519734740257263,
"learning_rate": 9.283374370582732e-05,
"loss": 0.023,
"step": 2140
},
{
"epoch": 3.9814814814814814,
"grad_norm": 0.2986918091773987,
"learning_rate": 9.274821429741482e-05,
"loss": 0.0259,
"step": 2150
},
{
"epoch": 4.0,
"grad_norm": 0.7932255864143372,
"learning_rate": 9.266221740137961e-05,
"loss": 0.0236,
"step": 2160
},
{
"epoch": 4.018518518518518,
"grad_norm": 0.2982627749443054,
"learning_rate": 9.257575395817001e-05,
"loss": 0.0239,
"step": 2170
},
{
"epoch": 4.037037037037037,
"grad_norm": 0.44402915239334106,
"learning_rate": 9.248882491333637e-05,
"loss": 0.0269,
"step": 2180
},
{
"epoch": 4.055555555555555,
"grad_norm": 0.3614986538887024,
"learning_rate": 9.240143121752076e-05,
"loss": 0.026,
"step": 2190
},
{
"epoch": 4.074074074074074,
"grad_norm": 0.2204546183347702,
"learning_rate": 9.23135738264467e-05,
"loss": 0.0275,
"step": 2200
},
{
"epoch": 4.092592592592593,
"grad_norm": 0.2746693789958954,
"learning_rate": 9.222525370090849e-05,
"loss": 0.026,
"step": 2210
},
{
"epoch": 4.111111111111111,
"grad_norm": 0.3058093190193176,
"learning_rate": 9.213647180676088e-05,
"loss": 0.0277,
"step": 2220
},
{
"epoch": 4.12962962962963,
"grad_norm": 0.2230917364358902,
"learning_rate": 9.204722911490846e-05,
"loss": 0.0268,
"step": 2230
},
{
"epoch": 4.148148148148148,
"grad_norm": 0.24511481821537018,
"learning_rate": 9.1957526601295e-05,
"loss": 0.0248,
"step": 2240
},
{
"epoch": 4.166666666666667,
"grad_norm": 0.3910863697528839,
"learning_rate": 9.186736524689281e-05,
"loss": 0.0233,
"step": 2250
},
{
"epoch": 4.185185185185185,
"grad_norm": 0.3914678692817688,
"learning_rate": 9.177674603769204e-05,
"loss": 0.0278,
"step": 2260
},
{
"epoch": 4.203703703703703,
"grad_norm": 0.21229635179042816,
"learning_rate": 9.168566996468983e-05,
"loss": 0.024,
"step": 2270
},
{
"epoch": 4.222222222222222,
"grad_norm": 0.25794699788093567,
"learning_rate": 9.159413802387951e-05,
"loss": 0.023,
"step": 2280
},
{
"epoch": 4.2407407407407405,
"grad_norm": 0.3773779273033142,
"learning_rate": 9.150215121623974e-05,
"loss": 0.0222,
"step": 2290
},
{
"epoch": 4.2592592592592595,
"grad_norm": 0.2078806757926941,
"learning_rate": 9.140971054772349e-05,
"loss": 0.025,
"step": 2300
},
{
"epoch": 4.277777777777778,
"grad_norm": 0.4091793894767761,
"learning_rate": 9.131681702924713e-05,
"loss": 0.025,
"step": 2310
},
{
"epoch": 4.296296296296296,
"grad_norm": 0.34888529777526855,
"learning_rate": 9.122347167667926e-05,
"loss": 0.0246,
"step": 2320
},
{
"epoch": 4.314814814814815,
"grad_norm": 0.4280141294002533,
"learning_rate": 9.112967551082973e-05,
"loss": 0.0255,
"step": 2330
},
{
"epoch": 4.333333333333333,
"grad_norm": 0.37729817628860474,
"learning_rate": 9.103542955743835e-05,
"loss": 0.0294,
"step": 2340
},
{
"epoch": 4.351851851851852,
"grad_norm": 0.3128068447113037,
"learning_rate": 9.094073484716381e-05,
"loss": 0.0247,
"step": 2350
},
{
"epoch": 4.37037037037037,
"grad_norm": 0.3114467263221741,
"learning_rate": 9.084559241557226e-05,
"loss": 0.0242,
"step": 2360
},
{
"epoch": 4.388888888888889,
"grad_norm": 0.19011156260967255,
"learning_rate": 9.075000330312608e-05,
"loss": 0.0261,
"step": 2370
},
{
"epoch": 4.407407407407407,
"grad_norm": 0.25602301955223083,
"learning_rate": 9.065396855517253e-05,
"loss": 0.0224,
"step": 2380
},
{
"epoch": 4.425925925925926,
"grad_norm": 0.2369946539402008,
"learning_rate": 9.055748922193219e-05,
"loss": 0.0233,
"step": 2390
},
{
"epoch": 4.444444444444445,
"grad_norm": 0.24912548065185547,
"learning_rate": 9.046056635848761e-05,
"loss": 0.0224,
"step": 2400
},
{
"epoch": 4.462962962962963,
"grad_norm": 0.3077091872692108,
"learning_rate": 9.036320102477169e-05,
"loss": 0.0242,
"step": 2410
},
{
"epoch": 4.481481481481482,
"grad_norm": 0.303082674741745,
"learning_rate": 9.02653942855561e-05,
"loss": 0.0242,
"step": 2420
},
{
"epoch": 4.5,
"grad_norm": 0.43740391731262207,
"learning_rate": 9.016714721043971e-05,
"loss": 0.0269,
"step": 2430
},
{
"epoch": 4.518518518518518,
"grad_norm": 0.18982768058776855,
"learning_rate": 9.006846087383675e-05,
"loss": 0.0257,
"step": 2440
},
{
"epoch": 4.537037037037037,
"grad_norm": 0.21268552541732788,
"learning_rate": 8.996933635496523e-05,
"loss": 0.0223,
"step": 2450
},
{
"epoch": 4.555555555555555,
"grad_norm": 0.26993295550346375,
"learning_rate": 8.986977473783498e-05,
"loss": 0.0227,
"step": 2460
},
{
"epoch": 4.574074074074074,
"grad_norm": 0.33540746569633484,
"learning_rate": 8.97697771112359e-05,
"loss": 0.0243,
"step": 2470
},
{
"epoch": 4.592592592592593,
"grad_norm": 0.2744498550891876,
"learning_rate": 8.966934456872602e-05,
"loss": 0.0249,
"step": 2480
},
{
"epoch": 4.611111111111111,
"grad_norm": 0.2517712414264679,
"learning_rate": 8.95684782086195e-05,
"loss": 0.0248,
"step": 2490
},
{
"epoch": 4.62962962962963,
"grad_norm": 0.24415573477745056,
"learning_rate": 8.946717913397476e-05,
"loss": 0.0258,
"step": 2500
},
{
"epoch": 4.648148148148148,
"grad_norm": 0.3292611539363861,
"learning_rate": 8.93654484525822e-05,
"loss": 0.027,
"step": 2510
},
{
"epoch": 4.666666666666667,
"grad_norm": 0.3423057794570923,
"learning_rate": 8.926328727695226e-05,
"loss": 0.0234,
"step": 2520
},
{
"epoch": 4.685185185185185,
"grad_norm": 0.3277914226055145,
"learning_rate": 8.916069672430319e-05,
"loss": 0.0228,
"step": 2530
},
{
"epoch": 4.703703703703704,
"grad_norm": 0.3065652549266815,
"learning_rate": 8.905767791654884e-05,
"loss": 0.0235,
"step": 2540
},
{
"epoch": 4.722222222222222,
"grad_norm": 0.2902333736419678,
"learning_rate": 8.895423198028638e-05,
"loss": 0.0246,
"step": 2550
},
{
"epoch": 4.7407407407407405,
"grad_norm": 0.31969141960144043,
"learning_rate": 8.885036004678402e-05,
"loss": 0.0235,
"step": 2560
},
{
"epoch": 4.7592592592592595,
"grad_norm": 0.3434152603149414,
"learning_rate": 8.874606325196857e-05,
"loss": 0.0257,
"step": 2570
},
{
"epoch": 4.777777777777778,
"grad_norm": 0.3704863488674164,
"learning_rate": 8.864134273641304e-05,
"loss": 0.0241,
"step": 2580
},
{
"epoch": 4.796296296296296,
"grad_norm": 0.3179256021976471,
"learning_rate": 8.853619964532427e-05,
"loss": 0.0217,
"step": 2590
},
{
"epoch": 4.814814814814815,
"grad_norm": 0.3592453598976135,
"learning_rate": 8.843063512853019e-05,
"loss": 0.0272,
"step": 2600
},
{
"epoch": 4.833333333333333,
"grad_norm": 0.2148142158985138,
"learning_rate": 8.832465034046749e-05,
"loss": 0.0241,
"step": 2610
},
{
"epoch": 4.851851851851852,
"grad_norm": 0.1970929205417633,
"learning_rate": 8.821824644016882e-05,
"loss": 0.023,
"step": 2620
},
{
"epoch": 4.87037037037037,
"grad_norm": 0.32399308681488037,
"learning_rate": 8.811142459125019e-05,
"loss": 0.022,
"step": 2630
},
{
"epoch": 4.888888888888889,
"grad_norm": 0.33707135915756226,
"learning_rate": 8.800418596189822e-05,
"loss": 0.0233,
"step": 2640
},
{
"epoch": 4.907407407407407,
"grad_norm": 0.19115638732910156,
"learning_rate": 8.789653172485737e-05,
"loss": 0.0248,
"step": 2650
},
{
"epoch": 4.925925925925926,
"grad_norm": 0.22406211495399475,
"learning_rate": 8.778846305741715e-05,
"loss": 0.0223,
"step": 2660
},
{
"epoch": 4.944444444444445,
"grad_norm": 0.2729659974575043,
"learning_rate": 8.767998114139918e-05,
"loss": 0.023,
"step": 2670
},
{
"epoch": 4.962962962962963,
"grad_norm": 0.22660554945468903,
"learning_rate": 8.757108716314429e-05,
"loss": 0.0238,
"step": 2680
},
{
"epoch": 4.981481481481482,
"grad_norm": 0.2774104177951813,
"learning_rate": 8.746178231349962e-05,
"loss": 0.0228,
"step": 2690
},
{
"epoch": 5.0,
"grad_norm": 0.8733599185943604,
"learning_rate": 8.735206778780549e-05,
"loss": 0.024,
"step": 2700
},
{
"epoch": 5.018518518518518,
"grad_norm": 0.45541656017303467,
"learning_rate": 8.724194478588234e-05,
"loss": 0.0286,
"step": 2710
},
{
"epoch": 5.037037037037037,
"grad_norm": 0.5369887948036194,
"learning_rate": 8.713141451201772e-05,
"loss": 0.0255,
"step": 2720
},
{
"epoch": 5.055555555555555,
"grad_norm": 0.3971169888973236,
"learning_rate": 8.702047817495295e-05,
"loss": 0.0262,
"step": 2730
},
{
"epoch": 5.074074074074074,
"grad_norm": 0.2992834746837616,
"learning_rate": 8.69091369878701e-05,
"loss": 0.024,
"step": 2740
},
{
"epoch": 5.092592592592593,
"grad_norm": 0.2394394725561142,
"learning_rate": 8.679739216837849e-05,
"loss": 0.022,
"step": 2750
},
{
"epoch": 5.111111111111111,
"grad_norm": 0.289729505777359,
"learning_rate": 8.66852449385016e-05,
"loss": 0.023,
"step": 2760
},
{
"epoch": 5.12962962962963,
"grad_norm": 0.27357134222984314,
"learning_rate": 8.657269652466356e-05,
"loss": 0.0236,
"step": 2770
},
{
"epoch": 5.148148148148148,
"grad_norm": 0.3606523871421814,
"learning_rate": 8.645974815767577e-05,
"loss": 0.0212,
"step": 2780
},
{
"epoch": 5.166666666666667,
"grad_norm": 0.36662402749061584,
"learning_rate": 8.634640107272351e-05,
"loss": 0.0211,
"step": 2790
},
{
"epoch": 5.185185185185185,
"grad_norm": 0.22495076060295105,
"learning_rate": 8.623265650935234e-05,
"loss": 0.0224,
"step": 2800
},
{
"epoch": 5.203703703703703,
"grad_norm": 0.28361451625823975,
"learning_rate": 8.611851571145456e-05,
"loss": 0.0243,
"step": 2810
},
{
"epoch": 5.222222222222222,
"grad_norm": 0.2966991066932678,
"learning_rate": 8.600397992725566e-05,
"loss": 0.0213,
"step": 2820
},
{
"epoch": 5.2407407407407405,
"grad_norm": 0.26136234402656555,
"learning_rate": 8.588905040930061e-05,
"loss": 0.0224,
"step": 2830
},
{
"epoch": 5.2592592592592595,
"grad_norm": 0.2862499952316284,
"learning_rate": 8.577372841444022e-05,
"loss": 0.0204,
"step": 2840
},
{
"epoch": 5.277777777777778,
"grad_norm": 0.33208534121513367,
"learning_rate": 8.565801520381736e-05,
"loss": 0.0229,
"step": 2850
},
{
"epoch": 5.296296296296296,
"grad_norm": 0.293374240398407,
"learning_rate": 8.554191204285313e-05,
"loss": 0.0208,
"step": 2860
},
{
"epoch": 5.314814814814815,
"grad_norm": 0.291591078042984,
"learning_rate": 8.542542020123315e-05,
"loss": 0.0251,
"step": 2870
},
{
"epoch": 5.333333333333333,
"grad_norm": 0.4706542193889618,
"learning_rate": 8.530854095289347e-05,
"loss": 0.0203,
"step": 2880
},
{
"epoch": 5.351851851851852,
"grad_norm": 0.34677860140800476,
"learning_rate": 8.519127557600688e-05,
"loss": 0.0211,
"step": 2890
},
{
"epoch": 5.37037037037037,
"grad_norm": 0.28802502155303955,
"learning_rate": 8.507362535296871e-05,
"loss": 0.0243,
"step": 2900
},
{
"epoch": 5.388888888888889,
"grad_norm": 0.26285848021507263,
"learning_rate": 8.495559157038299e-05,
"loss": 0.0206,
"step": 2910
},
{
"epoch": 5.407407407407407,
"grad_norm": 0.3403722643852234,
"learning_rate": 8.483717551904823e-05,
"loss": 0.0227,
"step": 2920
},
{
"epoch": 5.425925925925926,
"grad_norm": 0.27443116903305054,
"learning_rate": 8.47183784939434e-05,
"loss": 0.0258,
"step": 2930
},
{
"epoch": 5.444444444444445,
"grad_norm": 0.26819512248039246,
"learning_rate": 8.459920179421374e-05,
"loss": 0.0208,
"step": 2940
},
{
"epoch": 5.462962962962963,
"grad_norm": 0.30465978384017944,
"learning_rate": 8.447964672315656e-05,
"loss": 0.0238,
"step": 2950
},
{
"epoch": 5.481481481481482,
"grad_norm": 0.22769829630851746,
"learning_rate": 8.435971458820692e-05,
"loss": 0.0227,
"step": 2960
},
{
"epoch": 5.5,
"grad_norm": 0.27352967858314514,
"learning_rate": 8.423940670092345e-05,
"loss": 0.0218,
"step": 2970
},
{
"epoch": 5.518518518518518,
"grad_norm": 0.20965977013111115,
"learning_rate": 8.411872437697394e-05,
"loss": 0.0195,
"step": 2980
},
{
"epoch": 5.537037037037037,
"grad_norm": 0.1819644570350647,
"learning_rate": 8.399766893612096e-05,
"loss": 0.0194,
"step": 2990
},
{
"epoch": 5.555555555555555,
"grad_norm": 0.30472618341445923,
"learning_rate": 8.38762417022074e-05,
"loss": 0.0211,
"step": 3000
},
{
"epoch": 5.574074074074074,
"grad_norm": 0.2460194230079651,
"learning_rate": 8.375444400314204e-05,
"loss": 0.0207,
"step": 3010
},
{
"epoch": 5.592592592592593,
"grad_norm": 0.3365727663040161,
"learning_rate": 8.3632277170885e-05,
"loss": 0.0204,
"step": 3020
},
{
"epoch": 5.611111111111111,
"grad_norm": 0.3334239423274994,
"learning_rate": 8.350974254143318e-05,
"loss": 0.0195,
"step": 3030
},
{
"epoch": 5.62962962962963,
"grad_norm": 0.3364555835723877,
"learning_rate": 8.338684145480566e-05,
"loss": 0.0199,
"step": 3040
},
{
"epoch": 5.648148148148148,
"grad_norm": 0.2960301339626312,
"learning_rate": 8.326357525502904e-05,
"loss": 0.0222,
"step": 3050
},
{
"epoch": 5.666666666666667,
"grad_norm": 0.17312146723270416,
"learning_rate": 8.313994529012273e-05,
"loss": 0.018,
"step": 3060
},
{
"epoch": 5.685185185185185,
"grad_norm": 0.22202041745185852,
"learning_rate": 8.301595291208422e-05,
"loss": 0.0218,
"step": 3070
},
{
"epoch": 5.703703703703704,
"grad_norm": 0.2702822685241699,
"learning_rate": 8.289159947687427e-05,
"loss": 0.0178,
"step": 3080
},
{
"epoch": 5.722222222222222,
"grad_norm": 0.22096069157123566,
"learning_rate": 8.276688634440216e-05,
"loss": 0.0214,
"step": 3090
},
{
"epoch": 5.7407407407407405,
"grad_norm": 0.25706371665000916,
"learning_rate": 8.26418148785107e-05,
"loss": 0.0235,
"step": 3100
},
{
"epoch": 5.7592592592592595,
"grad_norm": 0.21485432982444763,
"learning_rate": 8.251638644696141e-05,
"loss": 0.0193,
"step": 3110
},
{
"epoch": 5.777777777777778,
"grad_norm": 0.2630963921546936,
"learning_rate": 8.23906024214195e-05,
"loss": 0.0192,
"step": 3120
},
{
"epoch": 5.796296296296296,
"grad_norm": 0.2886369526386261,
"learning_rate": 8.226446417743897e-05,
"loss": 0.0227,
"step": 3130
},
{
"epoch": 5.814814814814815,
"grad_norm": 0.4069041609764099,
"learning_rate": 8.213797309444742e-05,
"loss": 0.0217,
"step": 3140
},
{
"epoch": 5.833333333333333,
"grad_norm": 0.2805855870246887,
"learning_rate": 8.201113055573105e-05,
"loss": 0.0235,
"step": 3150
},
{
"epoch": 5.851851851851852,
"grad_norm": 0.2069445252418518,
"learning_rate": 8.188393794841958e-05,
"loss": 0.0227,
"step": 3160
},
{
"epoch": 5.87037037037037,
"grad_norm": 0.2260744422674179,
"learning_rate": 8.175639666347094e-05,
"loss": 0.0199,
"step": 3170
},
{
"epoch": 5.888888888888889,
"grad_norm": 0.3045082688331604,
"learning_rate": 8.162850809565623e-05,
"loss": 0.0186,
"step": 3180
},
{
"epoch": 5.907407407407407,
"grad_norm": 0.23879218101501465,
"learning_rate": 8.150027364354431e-05,
"loss": 0.0238,
"step": 3190
},
{
"epoch": 5.925925925925926,
"grad_norm": 0.2272648811340332,
"learning_rate": 8.137169470948662e-05,
"loss": 0.0208,
"step": 3200
},
{
"epoch": 5.944444444444445,
"grad_norm": 0.2475791573524475,
"learning_rate": 8.124277269960179e-05,
"loss": 0.0199,
"step": 3210
},
{
"epoch": 5.962962962962963,
"grad_norm": 0.27410802245140076,
"learning_rate": 8.111350902376023e-05,
"loss": 0.019,
"step": 3220
},
{
"epoch": 5.981481481481482,
"grad_norm": 0.23680570721626282,
"learning_rate": 8.098390509556883e-05,
"loss": 0.0201,
"step": 3230
},
{
"epoch": 6.0,
"grad_norm": 0.4440158009529114,
"learning_rate": 8.085396233235536e-05,
"loss": 0.0191,
"step": 3240
},
{
"epoch": 6.018518518518518,
"grad_norm": 0.4229031801223755,
"learning_rate": 8.072368215515306e-05,
"loss": 0.0203,
"step": 3250
},
{
"epoch": 6.037037037037037,
"grad_norm": 0.20239271223545074,
"learning_rate": 8.059306598868506e-05,
"loss": 0.02,
"step": 3260
},
{
"epoch": 6.055555555555555,
"grad_norm": 0.1755976676940918,
"learning_rate": 8.046211526134888e-05,
"loss": 0.021,
"step": 3270
},
{
"epoch": 6.074074074074074,
"grad_norm": 0.4123801589012146,
"learning_rate": 8.033083140520065e-05,
"loss": 0.0186,
"step": 3280
},
{
"epoch": 6.092592592592593,
"grad_norm": 0.24318896234035492,
"learning_rate": 8.019921585593962e-05,
"loss": 0.0244,
"step": 3290
},
{
"epoch": 6.111111111111111,
"grad_norm": 0.21515831351280212,
"learning_rate": 8.006727005289232e-05,
"loss": 0.022,
"step": 3300
},
{
"epoch": 6.12962962962963,
"grad_norm": 0.34318989515304565,
"learning_rate": 7.993499543899692e-05,
"loss": 0.0216,
"step": 3310
},
{
"epoch": 6.148148148148148,
"grad_norm": 0.16900299489498138,
"learning_rate": 7.980239346078742e-05,
"loss": 0.0217,
"step": 3320
},
{
"epoch": 6.166666666666667,
"grad_norm": 0.2132652848958969,
"learning_rate": 7.966946556837778e-05,
"loss": 0.0188,
"step": 3330
},
{
"epoch": 6.185185185185185,
"grad_norm": 0.24203772842884064,
"learning_rate": 7.953621321544616e-05,
"loss": 0.0185,
"step": 3340
},
{
"epoch": 6.203703703703703,
"grad_norm": 0.2927767336368561,
"learning_rate": 7.940263785921896e-05,
"loss": 0.0217,
"step": 3350
},
{
"epoch": 6.222222222222222,
"grad_norm": 0.2970094382762909,
"learning_rate": 7.926874096045482e-05,
"loss": 0.0199,
"step": 3360
},
{
"epoch": 6.2407407407407405,
"grad_norm": 0.1871601790189743,
"learning_rate": 7.913452398342881e-05,
"loss": 0.02,
"step": 3370
},
{
"epoch": 6.2592592592592595,
"grad_norm": 0.17269326746463776,
"learning_rate": 7.89999883959163e-05,
"loss": 0.0209,
"step": 3380
},
{
"epoch": 6.277777777777778,
"grad_norm": 0.2213200181722641,
"learning_rate": 7.886513566917687e-05,
"loss": 0.0173,
"step": 3390
},
{
"epoch": 6.296296296296296,
"grad_norm": 0.22219060361385345,
"learning_rate": 7.872996727793838e-05,
"loss": 0.0198,
"step": 3400
},
{
"epoch": 6.314814814814815,
"grad_norm": 0.276092529296875,
"learning_rate": 7.859448470038069e-05,
"loss": 0.0232,
"step": 3410
},
{
"epoch": 6.333333333333333,
"grad_norm": 0.25393709540367126,
"learning_rate": 7.845868941811956e-05,
"loss": 0.0169,
"step": 3420
},
{
"epoch": 6.351851851851852,
"grad_norm": 0.21830302476882935,
"learning_rate": 7.832258291619043e-05,
"loss": 0.0187,
"step": 3430
},
{
"epoch": 6.37037037037037,
"grad_norm": 0.22764980792999268,
"learning_rate": 7.81861666830322e-05,
"loss": 0.0209,
"step": 3440
},
{
"epoch": 6.388888888888889,
"grad_norm": 0.25005942583084106,
"learning_rate": 7.804944221047097e-05,
"loss": 0.0221,
"step": 3450
},
{
"epoch": 6.407407407407407,
"grad_norm": 0.22207368910312653,
"learning_rate": 7.791241099370364e-05,
"loss": 0.0211,
"step": 3460
},
{
"epoch": 6.425925925925926,
"grad_norm": 0.26478034257888794,
"learning_rate": 7.777507453128163e-05,
"loss": 0.0222,
"step": 3470
},
{
"epoch": 6.444444444444445,
"grad_norm": 0.2028956413269043,
"learning_rate": 7.763743432509451e-05,
"loss": 0.0185,
"step": 3480
},
{
"epoch": 6.462962962962963,
"grad_norm": 0.21192608773708344,
"learning_rate": 7.749949188035353e-05,
"loss": 0.0184,
"step": 3490
},
{
"epoch": 6.481481481481482,
"grad_norm": 0.21932436525821686,
"learning_rate": 7.736124870557516e-05,
"loss": 0.0203,
"step": 3500
},
{
"epoch": 6.5,
"grad_norm": 0.2874144911766052,
"learning_rate": 7.722270631256459e-05,
"loss": 0.0201,
"step": 3510
},
{
"epoch": 6.518518518518518,
"grad_norm": 0.2633146345615387,
"learning_rate": 7.708386621639925e-05,
"loss": 0.0179,
"step": 3520
},
{
"epoch": 6.537037037037037,
"grad_norm": 0.2622010111808777,
"learning_rate": 7.694472993541219e-05,
"loss": 0.0215,
"step": 3530
},
{
"epoch": 6.555555555555555,
"grad_norm": 0.23145030438899994,
"learning_rate": 7.680529899117547e-05,
"loss": 0.0199,
"step": 3540
},
{
"epoch": 6.574074074074074,
"grad_norm": 0.14719131588935852,
"learning_rate": 7.666557490848358e-05,
"loss": 0.0168,
"step": 3550
},
{
"epoch": 6.592592592592593,
"grad_norm": 0.3588024973869324,
"learning_rate": 7.65255592153367e-05,
"loss": 0.0192,
"step": 3560
},
{
"epoch": 6.611111111111111,
"grad_norm": 0.3507694602012634,
"learning_rate": 7.638525344292402e-05,
"loss": 0.0201,
"step": 3570
},
{
"epoch": 6.62962962962963,
"grad_norm": 0.2589399218559265,
"learning_rate": 7.624465912560697e-05,
"loss": 0.0217,
"step": 3580
},
{
"epoch": 6.648148148148148,
"grad_norm": 0.2474854290485382,
"learning_rate": 7.610377780090249e-05,
"loss": 0.0183,
"step": 3590
},
{
"epoch": 6.666666666666667,
"grad_norm": 0.25767403841018677,
"learning_rate": 7.596261100946618e-05,
"loss": 0.0204,
"step": 3600
},
{
"epoch": 6.685185185185185,
"grad_norm": 0.17771989107131958,
"learning_rate": 7.582116029507542e-05,
"loss": 0.0164,
"step": 3610
},
{
"epoch": 6.703703703703704,
"grad_norm": 0.17362728714942932,
"learning_rate": 7.56794272046126e-05,
"loss": 0.0194,
"step": 3620
},
{
"epoch": 6.722222222222222,
"grad_norm": 0.2660811245441437,
"learning_rate": 7.55374132880481e-05,
"loss": 0.0185,
"step": 3630
},
{
"epoch": 6.7407407407407405,
"grad_norm": 0.25032201409339905,
"learning_rate": 7.539512009842333e-05,
"loss": 0.0208,
"step": 3640
},
{
"epoch": 6.7592592592592595,
"grad_norm": 0.22758392989635468,
"learning_rate": 7.525254919183382e-05,
"loss": 0.0174,
"step": 3650
},
{
"epoch": 6.777777777777778,
"grad_norm": 0.2787107527256012,
"learning_rate": 7.510970212741215e-05,
"loss": 0.0221,
"step": 3660
},
{
"epoch": 6.796296296296296,
"grad_norm": 0.20214755833148956,
"learning_rate": 7.496658046731096e-05,
"loss": 0.0195,
"step": 3670
},
{
"epoch": 6.814814814814815,
"grad_norm": 0.1833910197019577,
"learning_rate": 7.482318577668578e-05,
"loss": 0.0198,
"step": 3680
},
{
"epoch": 6.833333333333333,
"grad_norm": 0.2200099229812622,
"learning_rate": 7.467951962367796e-05,
"loss": 0.0191,
"step": 3690
},
{
"epoch": 6.851851851851852,
"grad_norm": 0.22445610165596008,
"learning_rate": 7.453558357939755e-05,
"loss": 0.0211,
"step": 3700
},
{
"epoch": 6.87037037037037,
"grad_norm": 0.24373434484004974,
"learning_rate": 7.439137921790606e-05,
"loss": 0.0186,
"step": 3710
},
{
"epoch": 6.888888888888889,
"grad_norm": 0.23808437585830688,
"learning_rate": 7.42469081161993e-05,
"loss": 0.0188,
"step": 3720
},
{
"epoch": 6.907407407407407,
"grad_norm": 0.26300397515296936,
"learning_rate": 7.410217185419006e-05,
"loss": 0.0186,
"step": 3730
},
{
"epoch": 6.925925925925926,
"grad_norm": 0.19650381803512573,
"learning_rate": 7.395717201469095e-05,
"loss": 0.0171,
"step": 3740
},
{
"epoch": 6.944444444444445,
"grad_norm": 0.2303297519683838,
"learning_rate": 7.381191018339696e-05,
"loss": 0.02,
"step": 3750
},
{
"epoch": 6.962962962962963,
"grad_norm": 0.33165881037712097,
"learning_rate": 7.36663879488682e-05,
"loss": 0.0166,
"step": 3760
},
{
"epoch": 6.981481481481482,
"grad_norm": 0.307317316532135,
"learning_rate": 7.352060690251254e-05,
"loss": 0.0217,
"step": 3770
},
{
"epoch": 7.0,
"grad_norm": 0.2233235239982605,
"learning_rate": 7.337456863856811e-05,
"loss": 0.0149,
"step": 3780
},
{
"epoch": 7.018518518518518,
"grad_norm": 0.2888094484806061,
"learning_rate": 7.3228274754086e-05,
"loss": 0.0176,
"step": 3790
},
{
"epoch": 7.037037037037037,
"grad_norm": 0.18728870153427124,
"learning_rate": 7.308172684891267e-05,
"loss": 0.0198,
"step": 3800
},
{
"epoch": 7.055555555555555,
"grad_norm": 0.1812422126531601,
"learning_rate": 7.293492652567255e-05,
"loss": 0.0185,
"step": 3810
},
{
"epoch": 7.074074074074074,
"grad_norm": 0.2136453092098236,
"learning_rate": 7.278787538975043e-05,
"loss": 0.0172,
"step": 3820
},
{
"epoch": 7.092592592592593,
"grad_norm": 0.19265879690647125,
"learning_rate": 7.2640575049274e-05,
"loss": 0.0187,
"step": 3830
},
{
"epoch": 7.111111111111111,
"grad_norm": 0.19298294186592102,
"learning_rate": 7.249302711509616e-05,
"loss": 0.0178,
"step": 3840
},
{
"epoch": 7.12962962962963,
"grad_norm": 0.18972733616828918,
"learning_rate": 7.23452332007775e-05,
"loss": 0.0166,
"step": 3850
},
{
"epoch": 7.148148148148148,
"grad_norm": 0.3038862943649292,
"learning_rate": 7.219719492256858e-05,
"loss": 0.0162,
"step": 3860
},
{
"epoch": 7.166666666666667,
"grad_norm": 0.19389449059963226,
"learning_rate": 7.20489138993923e-05,
"loss": 0.0198,
"step": 3870
},
{
"epoch": 7.185185185185185,
"grad_norm": 0.3252969980239868,
"learning_rate": 7.190039175282614e-05,
"loss": 0.018,
"step": 3880
},
{
"epoch": 7.203703703703703,
"grad_norm": 0.29355478286743164,
"learning_rate": 7.175163010708455e-05,
"loss": 0.0169,
"step": 3890
},
{
"epoch": 7.222222222222222,
"grad_norm": 0.26375654339790344,
"learning_rate": 7.1602630589001e-05,
"loss": 0.0159,
"step": 3900
},
{
"epoch": 7.2407407407407405,
"grad_norm": 0.23320522904396057,
"learning_rate": 7.14533948280104e-05,
"loss": 0.0175,
"step": 3910
},
{
"epoch": 7.2592592592592595,
"grad_norm": 0.292441189289093,
"learning_rate": 7.130392445613109e-05,
"loss": 0.0245,
"step": 3920
},
{
"epoch": 7.277777777777778,
"grad_norm": 0.2864852249622345,
"learning_rate": 7.115422110794711e-05,
"loss": 0.0182,
"step": 3930
},
{
"epoch": 7.296296296296296,
"grad_norm": 0.24359479546546936,
"learning_rate": 7.100428642059033e-05,
"loss": 0.016,
"step": 3940
},
{
"epoch": 7.314814814814815,
"grad_norm": 0.2701646089553833,
"learning_rate": 7.08541220337224e-05,
"loss": 0.0177,
"step": 3950
},
{
"epoch": 7.333333333333333,
"grad_norm": 0.3457199037075043,
"learning_rate": 7.070372958951706e-05,
"loss": 0.0192,
"step": 3960
},
{
"epoch": 7.351851851851852,
"grad_norm": 0.25961625576019287,
"learning_rate": 7.055311073264194e-05,
"loss": 0.0197,
"step": 3970
},
{
"epoch": 7.37037037037037,
"grad_norm": 0.3077142536640167,
"learning_rate": 7.040226711024077e-05,
"loss": 0.022,
"step": 3980
},
{
"epoch": 7.388888888888889,
"grad_norm": 0.20970751345157623,
"learning_rate": 7.02512003719152e-05,
"loss": 0.0208,
"step": 3990
},
{
"epoch": 7.407407407407407,
"grad_norm": 0.20395323634147644,
"learning_rate": 7.00999121697069e-05,
"loss": 0.0227,
"step": 4000
},
{
"epoch": 7.425925925925926,
"grad_norm": 0.1897621750831604,
"learning_rate": 6.99484041580794e-05,
"loss": 0.0183,
"step": 4010
},
{
"epoch": 7.444444444444445,
"grad_norm": 0.286591500043869,
"learning_rate": 6.979667799390004e-05,
"loss": 0.0202,
"step": 4020
},
{
"epoch": 7.462962962962963,
"grad_norm": 0.281000018119812,
"learning_rate": 6.964473533642185e-05,
"loss": 0.0162,
"step": 4030
},
{
"epoch": 7.481481481481482,
"grad_norm": 0.3166241943836212,
"learning_rate": 6.949257784726539e-05,
"loss": 0.0193,
"step": 4040
},
{
"epoch": 7.5,
"grad_norm": 0.20640015602111816,
"learning_rate": 6.934020719040056e-05,
"loss": 0.0159,
"step": 4050
},
{
"epoch": 7.518518518518518,
"grad_norm": 0.1437007188796997,
"learning_rate": 6.918762503212848e-05,
"loss": 0.0162,
"step": 4060
},
{
"epoch": 7.537037037037037,
"grad_norm": 0.3471952974796295,
"learning_rate": 6.903483304106319e-05,
"loss": 0.0178,
"step": 4070
},
{
"epoch": 7.555555555555555,
"grad_norm": 0.3107878863811493,
"learning_rate": 6.888183288811341e-05,
"loss": 0.0189,
"step": 4080
},
{
"epoch": 7.574074074074074,
"grad_norm": 0.2234809547662735,
"learning_rate": 6.87286262464643e-05,
"loss": 0.0192,
"step": 4090
},
{
"epoch": 7.592592592592593,
"grad_norm": 0.1554151475429535,
"learning_rate": 6.857521479155915e-05,
"loss": 0.0202,
"step": 4100
},
{
"epoch": 7.611111111111111,
"grad_norm": 0.22985172271728516,
"learning_rate": 6.842160020108104e-05,
"loss": 0.0181,
"step": 4110
},
{
"epoch": 7.62962962962963,
"grad_norm": 0.15581990778446198,
"learning_rate": 6.826778415493455e-05,
"loss": 0.0197,
"step": 4120
},
{
"epoch": 7.648148148148148,
"grad_norm": 0.2925049662590027,
"learning_rate": 6.811376833522729e-05,
"loss": 0.0173,
"step": 4130
},
{
"epoch": 7.666666666666667,
"grad_norm": 0.2584574818611145,
"learning_rate": 6.795955442625159e-05,
"loss": 0.0188,
"step": 4140
},
{
"epoch": 7.685185185185185,
"grad_norm": 0.24499130249023438,
"learning_rate": 6.780514411446608e-05,
"loss": 0.0204,
"step": 4150
},
{
"epoch": 7.703703703703704,
"grad_norm": 0.3336038291454315,
"learning_rate": 6.765053908847716e-05,
"loss": 0.0188,
"step": 4160
},
{
"epoch": 7.722222222222222,
"grad_norm": 0.2917112708091736,
"learning_rate": 6.749574103902064e-05,
"loss": 0.0193,
"step": 4170
},
{
"epoch": 7.7407407407407405,
"grad_norm": 0.25376829504966736,
"learning_rate": 6.734075165894317e-05,
"loss": 0.0178,
"step": 4180
},
{
"epoch": 7.7592592592592595,
"grad_norm": 0.19700933992862701,
"learning_rate": 6.71855726431838e-05,
"loss": 0.0158,
"step": 4190
},
{
"epoch": 7.777777777777778,
"grad_norm": 0.2758132219314575,
"learning_rate": 6.703020568875538e-05,
"loss": 0.0164,
"step": 4200
},
{
"epoch": 7.796296296296296,
"grad_norm": 0.2460174709558487,
"learning_rate": 6.687465249472603e-05,
"loss": 0.0166,
"step": 4210
},
{
"epoch": 7.814814814814815,
"grad_norm": 0.2323077917098999,
"learning_rate": 6.671891476220055e-05,
"loss": 0.0171,
"step": 4220
},
{
"epoch": 7.833333333333333,
"grad_norm": 0.25105613470077515,
"learning_rate": 6.656299419430183e-05,
"loss": 0.0176,
"step": 4230
},
{
"epoch": 7.851851851851852,
"grad_norm": 0.2859521210193634,
"learning_rate": 6.640689249615223e-05,
"loss": 0.0161,
"step": 4240
},
{
"epoch": 7.87037037037037,
"grad_norm": 0.39157968759536743,
"learning_rate": 6.625061137485491e-05,
"loss": 0.0203,
"step": 4250
},
{
"epoch": 7.888888888888889,
"grad_norm": 0.2698723077774048,
"learning_rate": 6.609415253947517e-05,
"loss": 0.0167,
"step": 4260
},
{
"epoch": 7.907407407407407,
"grad_norm": 0.22730262577533722,
"learning_rate": 6.593751770102178e-05,
"loss": 0.0159,
"step": 4270
},
{
"epoch": 7.925925925925926,
"grad_norm": 0.2370423674583435,
"learning_rate": 6.578070857242823e-05,
"loss": 0.0162,
"step": 4280
},
{
"epoch": 7.944444444444445,
"grad_norm": 0.16310055553913116,
"learning_rate": 6.562372686853402e-05,
"loss": 0.016,
"step": 4290
},
{
"epoch": 7.962962962962963,
"grad_norm": 0.21592535078525543,
"learning_rate": 6.546657430606593e-05,
"loss": 0.0168,
"step": 4300
},
{
"epoch": 7.981481481481482,
"grad_norm": 0.20451724529266357,
"learning_rate": 6.530925260361918e-05,
"loss": 0.0182,
"step": 4310
},
{
"epoch": 8.0,
"grad_norm": 0.7038203477859497,
"learning_rate": 6.515176348163871e-05,
"loss": 0.0182,
"step": 4320
},
{
"epoch": 8.018518518518519,
"grad_norm": 0.24489392340183258,
"learning_rate": 6.499410866240032e-05,
"loss": 0.018,
"step": 4330
},
{
"epoch": 8.037037037037036,
"grad_norm": 0.3114599585533142,
"learning_rate": 6.48362898699919e-05,
"loss": 0.0172,
"step": 4340
},
{
"epoch": 8.055555555555555,
"grad_norm": 0.2280680537223816,
"learning_rate": 6.467830883029443e-05,
"loss": 0.0184,
"step": 4350
},
{
"epoch": 8.074074074074074,
"grad_norm": 0.23185963928699493,
"learning_rate": 6.452016727096326e-05,
"loss": 0.016,
"step": 4360
},
{
"epoch": 8.092592592592593,
"grad_norm": 0.22630156576633453,
"learning_rate": 6.436186692140916e-05,
"loss": 0.0172,
"step": 4370
},
{
"epoch": 8.11111111111111,
"grad_norm": 0.2376243621110916,
"learning_rate": 6.420340951277938e-05,
"loss": 0.0184,
"step": 4380
},
{
"epoch": 8.12962962962963,
"grad_norm": 0.17322981357574463,
"learning_rate": 6.404479677793874e-05,
"loss": 0.0162,
"step": 4390
},
{
"epoch": 8.148148148148149,
"grad_norm": 0.22323162853717804,
"learning_rate": 6.388603045145075e-05,
"loss": 0.0197,
"step": 4400
},
{
"epoch": 8.166666666666666,
"grad_norm": 0.19180379807949066,
"learning_rate": 6.372711226955843e-05,
"loss": 0.0183,
"step": 4410
},
{
"epoch": 8.185185185185185,
"grad_norm": 0.21892212331295013,
"learning_rate": 6.356804397016564e-05,
"loss": 0.0184,
"step": 4420
},
{
"epoch": 8.203703703703704,
"grad_norm": 0.2725140452384949,
"learning_rate": 6.340882729281779e-05,
"loss": 0.0157,
"step": 4430
},
{
"epoch": 8.222222222222221,
"grad_norm": 0.14921587705612183,
"learning_rate": 6.324946397868294e-05,
"loss": 0.0183,
"step": 4440
},
{
"epoch": 8.24074074074074,
"grad_norm": 0.14498989284038544,
"learning_rate": 6.308995577053276e-05,
"loss": 0.0172,
"step": 4450
},
{
"epoch": 8.25925925925926,
"grad_norm": 0.16282035410404205,
"learning_rate": 6.293030441272347e-05,
"loss": 0.0155,
"step": 4460
},
{
"epoch": 8.277777777777779,
"grad_norm": 0.19784556329250336,
"learning_rate": 6.277051165117677e-05,
"loss": 0.0157,
"step": 4470
},
{
"epoch": 8.296296296296296,
"grad_norm": 0.2002154141664505,
"learning_rate": 6.261057923336064e-05,
"loss": 0.0176,
"step": 4480
},
{
"epoch": 8.314814814814815,
"grad_norm": 0.2519257664680481,
"learning_rate": 6.245050890827042e-05,
"loss": 0.0186,
"step": 4490
},
{
"epoch": 8.333333333333334,
"grad_norm": 0.3008039593696594,
"learning_rate": 6.229030242640952e-05,
"loss": 0.0185,
"step": 4500
},
{
"epoch": 8.351851851851851,
"grad_norm": 0.16971473395824432,
"learning_rate": 6.212996153977037e-05,
"loss": 0.0168,
"step": 4510
},
{
"epoch": 8.37037037037037,
"grad_norm": 0.2413981854915619,
"learning_rate": 6.196948800181523e-05,
"loss": 0.0159,
"step": 4520
},
{
"epoch": 8.38888888888889,
"grad_norm": 0.2520849406719208,
"learning_rate": 6.180888356745695e-05,
"loss": 0.0144,
"step": 4530
},
{
"epoch": 8.407407407407407,
"grad_norm": 0.20414361357688904,
"learning_rate": 6.164814999303995e-05,
"loss": 0.0153,
"step": 4540
},
{
"epoch": 8.425925925925926,
"grad_norm": 0.16077838838100433,
"learning_rate": 6.148728903632081e-05,
"loss": 0.0152,
"step": 4550
},
{
"epoch": 8.444444444444445,
"grad_norm": 0.2709040343761444,
"learning_rate": 6.132630245644921e-05,
"loss": 0.0169,
"step": 4560
},
{
"epoch": 8.462962962962964,
"grad_norm": 0.26147571206092834,
"learning_rate": 6.116519201394857e-05,
"loss": 0.0161,
"step": 4570
},
{
"epoch": 8.481481481481481,
"grad_norm": 0.1886901706457138,
"learning_rate": 6.10039594706969e-05,
"loss": 0.0165,
"step": 4580
},
{
"epoch": 8.5,
"grad_norm": 0.3017023205757141,
"learning_rate": 6.084260658990744e-05,
"loss": 0.0166,
"step": 4590
},
{
"epoch": 8.518518518518519,
"grad_norm": 0.28350818157196045,
"learning_rate": 6.068113513610943e-05,
"loss": 0.0185,
"step": 4600
},
{
"epoch": 8.537037037037036,
"grad_norm": 0.1432543843984604,
"learning_rate": 6.0519546875128876e-05,
"loss": 0.0165,
"step": 4610
},
{
"epoch": 8.555555555555555,
"grad_norm": 0.22892121970653534,
"learning_rate": 6.035784357406906e-05,
"loss": 0.0173,
"step": 4620
},
{
"epoch": 8.574074074074074,
"grad_norm": 0.26011526584625244,
"learning_rate": 6.01960270012914e-05,
"loss": 0.0165,
"step": 4630
},
{
"epoch": 8.592592592592592,
"grad_norm": 0.1291012465953827,
"learning_rate": 6.003409892639599e-05,
"loss": 0.0153,
"step": 4640
},
{
"epoch": 8.61111111111111,
"grad_norm": 0.2323385626077652,
"learning_rate": 5.9872061120202336e-05,
"loss": 0.0139,
"step": 4650
},
{
"epoch": 8.62962962962963,
"grad_norm": 0.3262821137905121,
"learning_rate": 5.9709915354729914e-05,
"loss": 0.0167,
"step": 4660
},
{
"epoch": 8.648148148148149,
"grad_norm": 0.23939751088619232,
"learning_rate": 5.9547663403178824e-05,
"loss": 0.0166,
"step": 4670
},
{
"epoch": 8.666666666666666,
"grad_norm": 0.14883610606193542,
"learning_rate": 5.9385307039910445e-05,
"loss": 0.0171,
"step": 4680
},
{
"epoch": 8.685185185185185,
"grad_norm": 0.2196505218744278,
"learning_rate": 5.922284804042792e-05,
"loss": 0.0172,
"step": 4690
},
{
"epoch": 8.703703703703704,
"grad_norm": 0.21951188147068024,
"learning_rate": 5.906028818135687e-05,
"loss": 0.0191,
"step": 4700
},
{
"epoch": 8.722222222222221,
"grad_norm": 0.18664328753948212,
"learning_rate": 5.889762924042585e-05,
"loss": 0.0148,
"step": 4710
},
{
"epoch": 8.74074074074074,
"grad_norm": 0.1908455789089203,
"learning_rate": 5.873487299644699e-05,
"loss": 0.0154,
"step": 4720
},
{
"epoch": 8.75925925925926,
"grad_norm": 0.1796601414680481,
"learning_rate": 5.857202122929649e-05,
"loss": 0.0181,
"step": 4730
},
{
"epoch": 8.777777777777779,
"grad_norm": 0.19254940748214722,
"learning_rate": 5.840907571989518e-05,
"loss": 0.0169,
"step": 4740
},
{
"epoch": 8.796296296296296,
"grad_norm": 0.23145553469657898,
"learning_rate": 5.824603825018904e-05,
"loss": 0.0176,
"step": 4750
},
{
"epoch": 8.814814814814815,
"grad_norm": 0.2171379029750824,
"learning_rate": 5.808291060312975e-05,
"loss": 0.0188,
"step": 4760
},
{
"epoch": 8.833333333333334,
"grad_norm": 0.18328005075454712,
"learning_rate": 5.7919694562655083e-05,
"loss": 0.0164,
"step": 4770
},
{
"epoch": 8.851851851851851,
"grad_norm": 0.22032232582569122,
"learning_rate": 5.775639191366954e-05,
"loss": 0.0195,
"step": 4780
},
{
"epoch": 8.87037037037037,
"grad_norm": 0.1514636129140854,
"learning_rate": 5.75930044420247e-05,
"loss": 0.0158,
"step": 4790
},
{
"epoch": 8.88888888888889,
"grad_norm": 0.23274511098861694,
"learning_rate": 5.74295339344998e-05,
"loss": 0.0167,
"step": 4800
},
{
"epoch": 8.907407407407408,
"grad_norm": 0.18468771874904633,
"learning_rate": 5.726598217878211e-05,
"loss": 0.0155,
"step": 4810
},
{
"epoch": 8.925925925925926,
"grad_norm": 0.23459751904010773,
"learning_rate": 5.71023509634474e-05,
"loss": 0.0199,
"step": 4820
},
{
"epoch": 8.944444444444445,
"grad_norm": 0.2732996344566345,
"learning_rate": 5.693864207794049e-05,
"loss": 0.0155,
"step": 4830
},
{
"epoch": 8.962962962962964,
"grad_norm": 0.24473819136619568,
"learning_rate": 5.677485731255545e-05,
"loss": 0.0168,
"step": 4840
},
{
"epoch": 8.981481481481481,
"grad_norm": 0.17636768519878387,
"learning_rate": 5.6610998458416296e-05,
"loss": 0.0172,
"step": 4850
},
{
"epoch": 9.0,
"grad_norm": 0.14970089495182037,
"learning_rate": 5.644706730745716e-05,
"loss": 0.0148,
"step": 4860
},
{
"epoch": 9.018518518518519,
"grad_norm": 0.21171057224273682,
"learning_rate": 5.628306565240287e-05,
"loss": 0.0182,
"step": 4870
},
{
"epoch": 9.037037037037036,
"grad_norm": 0.2047504186630249,
"learning_rate": 5.611899528674923e-05,
"loss": 0.0162,
"step": 4880
},
{
"epoch": 9.055555555555555,
"grad_norm": 0.18343551456928253,
"learning_rate": 5.595485800474349e-05,
"loss": 0.0149,
"step": 4890
},
{
"epoch": 9.074074074074074,
"grad_norm": 0.27501538395881653,
"learning_rate": 5.579065560136467e-05,
"loss": 0.0138,
"step": 4900
},
{
"epoch": 9.092592592592593,
"grad_norm": 0.29593756794929504,
"learning_rate": 5.562638987230392e-05,
"loss": 0.016,
"step": 4910
},
{
"epoch": 9.11111111111111,
"grad_norm": 0.20254631340503693,
"learning_rate": 5.546206261394498e-05,
"loss": 0.016,
"step": 4920
},
{
"epoch": 9.12962962962963,
"grad_norm": 0.2504293620586395,
"learning_rate": 5.529767562334437e-05,
"loss": 0.0171,
"step": 4930
},
{
"epoch": 9.148148148148149,
"grad_norm": 0.32949069142341614,
"learning_rate": 5.5133230698211926e-05,
"loss": 0.0186,
"step": 4940
},
{
"epoch": 9.166666666666666,
"grad_norm": 0.11458481103181839,
"learning_rate": 5.496872963689096e-05,
"loss": 0.0149,
"step": 4950
},
{
"epoch": 9.185185185185185,
"grad_norm": 0.2192622870206833,
"learning_rate": 5.4804174238338756e-05,
"loss": 0.0171,
"step": 4960
},
{
"epoch": 9.203703703703704,
"grad_norm": 0.25588271021842957,
"learning_rate": 5.463956630210678e-05,
"loss": 0.0185,
"step": 4970
},
{
"epoch": 9.222222222222221,
"grad_norm": 0.16602252423763275,
"learning_rate": 5.4474907628321046e-05,
"loss": 0.0171,
"step": 4980
},
{
"epoch": 9.24074074074074,
"grad_norm": 0.24425970017910004,
"learning_rate": 5.431020001766244e-05,
"loss": 0.0172,
"step": 4990
},
{
"epoch": 9.25925925925926,
"grad_norm": 0.1667938232421875,
"learning_rate": 5.4145445271346986e-05,
"loss": 0.0158,
"step": 5000
},
{
"epoch": 9.277777777777779,
"grad_norm": 0.17069245874881744,
"learning_rate": 5.398064519110622e-05,
"loss": 0.0171,
"step": 5010
},
{
"epoch": 9.296296296296296,
"grad_norm": 0.18154054880142212,
"learning_rate": 5.3815801579167394e-05,
"loss": 0.0162,
"step": 5020
},
{
"epoch": 9.314814814814815,
"grad_norm": 0.22497673332691193,
"learning_rate": 5.365091623823382e-05,
"loss": 0.0159,
"step": 5030
},
{
"epoch": 9.333333333333334,
"grad_norm": 0.25690826773643494,
"learning_rate": 5.348599097146521e-05,
"loss": 0.0158,
"step": 5040
},
{
"epoch": 9.351851851851851,
"grad_norm": 0.20806878805160522,
"learning_rate": 5.3321027582457836e-05,
"loss": 0.0161,
"step": 5050
},
{
"epoch": 9.37037037037037,
"grad_norm": 0.2535109221935272,
"learning_rate": 5.315602787522491e-05,
"loss": 0.0137,
"step": 5060
},
{
"epoch": 9.38888888888889,
"grad_norm": 0.21145524084568024,
"learning_rate": 5.299099365417678e-05,
"loss": 0.0163,
"step": 5070
},
{
"epoch": 9.407407407407407,
"grad_norm": 0.15755097568035126,
"learning_rate": 5.2825926724101236e-05,
"loss": 0.0154,
"step": 5080
},
{
"epoch": 9.425925925925926,
"grad_norm": 0.23783256113529205,
"learning_rate": 5.26608288901438e-05,
"loss": 0.0148,
"step": 5090
},
{
"epoch": 9.444444444444445,
"grad_norm": 0.27243706583976746,
"learning_rate": 5.24957019577879e-05,
"loss": 0.0169,
"step": 5100
},
{
"epoch": 9.462962962962964,
"grad_norm": 0.25300362706184387,
"learning_rate": 5.2330547732835266e-05,
"loss": 0.0185,
"step": 5110
},
{
"epoch": 9.481481481481481,
"grad_norm": 0.33991149067878723,
"learning_rate": 5.2165368021385996e-05,
"loss": 0.0168,
"step": 5120
},
{
"epoch": 9.5,
"grad_norm": 0.25291067361831665,
"learning_rate": 5.200016462981897e-05,
"loss": 0.0159,
"step": 5130
},
{
"epoch": 9.518518518518519,
"grad_norm": 0.18026596307754517,
"learning_rate": 5.1834939364772015e-05,
"loss": 0.0161,
"step": 5140
},
{
"epoch": 9.537037037037036,
"grad_norm": 0.25530028343200684,
"learning_rate": 5.166969403312214e-05,
"loss": 0.0176,
"step": 5150
},
{
"epoch": 9.555555555555555,
"grad_norm": 0.25754010677337646,
"learning_rate": 5.1504430441965844e-05,
"loss": 0.0151,
"step": 5160
},
{
"epoch": 9.574074074074074,
"grad_norm": 0.16583363711833954,
"learning_rate": 5.133915039859923e-05,
"loss": 0.0175,
"step": 5170
},
{
"epoch": 9.592592592592592,
"grad_norm": 0.2954111099243164,
"learning_rate": 5.1173855710498444e-05,
"loss": 0.0138,
"step": 5180
},
{
"epoch": 9.61111111111111,
"grad_norm": 0.21461570262908936,
"learning_rate": 5.100854818529967e-05,
"loss": 0.0135,
"step": 5190
},
{
"epoch": 9.62962962962963,
"grad_norm": 0.26129066944122314,
"learning_rate": 5.084322963077951e-05,
"loss": 0.0162,
"step": 5200
},
{
"epoch": 9.648148148148149,
"grad_norm": 0.2057189792394638,
"learning_rate": 5.067790185483522e-05,
"loss": 0.0148,
"step": 5210
},
{
"epoch": 9.666666666666666,
"grad_norm": 0.20507746934890747,
"learning_rate": 5.0512566665464844e-05,
"loss": 0.015,
"step": 5220
},
{
"epoch": 9.685185185185185,
"grad_norm": 0.2096184641122818,
"learning_rate": 5.034722587074755e-05,
"loss": 0.0144,
"step": 5230
},
{
"epoch": 9.703703703703704,
"grad_norm": 0.16661977767944336,
"learning_rate": 5.018188127882375e-05,
"loss": 0.0148,
"step": 5240
},
{
"epoch": 9.722222222222221,
"grad_norm": 0.1946638822555542,
"learning_rate": 5.0016534697875417e-05,
"loss": 0.015,
"step": 5250
},
{
"epoch": 9.74074074074074,
"grad_norm": 0.18467889726161957,
"learning_rate": 4.9851187936106294e-05,
"loss": 0.0134,
"step": 5260
},
{
"epoch": 9.75925925925926,
"grad_norm": 0.16328614950180054,
"learning_rate": 4.968584280172206e-05,
"loss": 0.0176,
"step": 5270
},
{
"epoch": 9.777777777777779,
"grad_norm": 0.18084822595119476,
"learning_rate": 4.95205011029106e-05,
"loss": 0.0145,
"step": 5280
},
{
"epoch": 9.796296296296296,
"grad_norm": 0.1974593847990036,
"learning_rate": 4.935516464782227e-05,
"loss": 0.0132,
"step": 5290
},
{
"epoch": 9.814814814814815,
"grad_norm": 0.23900610208511353,
"learning_rate": 4.918983524455003e-05,
"loss": 0.0141,
"step": 5300
},
{
"epoch": 9.833333333333334,
"grad_norm": 0.15113961696624756,
"learning_rate": 4.9024514701109766e-05,
"loss": 0.0148,
"step": 5310
},
{
"epoch": 9.851851851851851,
"grad_norm": 0.2611234784126282,
"learning_rate": 4.885920482542043e-05,
"loss": 0.015,
"step": 5320
},
{
"epoch": 9.87037037037037,
"grad_norm": 0.23414680361747742,
"learning_rate": 4.869390742528438e-05,
"loss": 0.0137,
"step": 5330
},
{
"epoch": 9.88888888888889,
"grad_norm": 0.1417471170425415,
"learning_rate": 4.852862430836744e-05,
"loss": 0.0149,
"step": 5340
},
{
"epoch": 9.907407407407408,
"grad_norm": 0.1815817505121231,
"learning_rate": 4.836335728217933e-05,
"loss": 0.0144,
"step": 5350
},
{
"epoch": 9.925925925925926,
"grad_norm": 0.2732499837875366,
"learning_rate": 4.819810815405379e-05,
"loss": 0.0147,
"step": 5360
},
{
"epoch": 9.944444444444445,
"grad_norm": 0.2105555385351181,
"learning_rate": 4.803287873112877e-05,
"loss": 0.0153,
"step": 5370
},
{
"epoch": 9.962962962962964,
"grad_norm": 0.2245476096868515,
"learning_rate": 4.786767082032681e-05,
"loss": 0.0153,
"step": 5380
},
{
"epoch": 9.981481481481481,
"grad_norm": 0.18886566162109375,
"learning_rate": 4.77024862283351e-05,
"loss": 0.0138,
"step": 5390
},
{
"epoch": 10.0,
"grad_norm": 0.4246911406517029,
"learning_rate": 4.753732676158593e-05,
"loss": 0.0144,
"step": 5400
},
{
"epoch": 10.018518518518519,
"grad_norm": 0.19355298578739166,
"learning_rate": 4.737219422623672e-05,
"loss": 0.0156,
"step": 5410
},
{
"epoch": 10.037037037037036,
"grad_norm": 0.2155923992395401,
"learning_rate": 4.720709042815044e-05,
"loss": 0.0179,
"step": 5420
},
{
"epoch": 10.055555555555555,
"grad_norm": 0.18233537673950195,
"learning_rate": 4.704201717287578e-05,
"loss": 0.0174,
"step": 5430
},
{
"epoch": 10.074074074074074,
"grad_norm": 0.18409103155136108,
"learning_rate": 4.6876976265627404e-05,
"loss": 0.0134,
"step": 5440
},
{
"epoch": 10.092592592592593,
"grad_norm": 0.12578661739826202,
"learning_rate": 4.671196951126626e-05,
"loss": 0.0135,
"step": 5450
},
{
"epoch": 10.11111111111111,
"grad_norm": 0.20303009450435638,
"learning_rate": 4.654699871427971e-05,
"loss": 0.0145,
"step": 5460
},
{
"epoch": 10.12962962962963,
"grad_norm": 0.1388576477766037,
"learning_rate": 4.6382065678762034e-05,
"loss": 0.0135,
"step": 5470
},
{
"epoch": 10.148148148148149,
"grad_norm": 0.16757282614707947,
"learning_rate": 4.6217172208394424e-05,
"loss": 0.0139,
"step": 5480
},
{
"epoch": 10.166666666666666,
"grad_norm": 0.1706474870443344,
"learning_rate": 4.605232010642549e-05,
"loss": 0.0145,
"step": 5490
},
{
"epoch": 10.185185185185185,
"grad_norm": 0.1968596875667572,
"learning_rate": 4.588751117565142e-05,
"loss": 0.0145,
"step": 5500
},
{
"epoch": 10.203703703703704,
"grad_norm": 0.1136387437582016,
"learning_rate": 4.5722747218396214e-05,
"loss": 0.0171,
"step": 5510
},
{
"epoch": 10.222222222222221,
"grad_norm": 0.2556943893432617,
"learning_rate": 4.5558030036492194e-05,
"loss": 0.0131,
"step": 5520
},
{
"epoch": 10.24074074074074,
"grad_norm": 0.1556449681520462,
"learning_rate": 4.539336143125999e-05,
"loss": 0.0145,
"step": 5530
},
{
"epoch": 10.25925925925926,
"grad_norm": 0.13108809292316437,
"learning_rate": 4.522874320348916e-05,
"loss": 0.0144,
"step": 5540
},
{
"epoch": 10.277777777777779,
"grad_norm": 0.13954582810401917,
"learning_rate": 4.506417715341821e-05,
"loss": 0.0151,
"step": 5550
},
{
"epoch": 10.296296296296296,
"grad_norm": 0.22305041551589966,
"learning_rate": 4.489966508071511e-05,
"loss": 0.0145,
"step": 5560
},
{
"epoch": 10.314814814814815,
"grad_norm": 0.15616579353809357,
"learning_rate": 4.4735208784457575e-05,
"loss": 0.0149,
"step": 5570
},
{
"epoch": 10.333333333333334,
"grad_norm": 0.16979192197322845,
"learning_rate": 4.457081006311325e-05,
"loss": 0.0159,
"step": 5580
},
{
"epoch": 10.351851851851851,
"grad_norm": 0.2175489068031311,
"learning_rate": 4.440647071452027e-05,
"loss": 0.0191,
"step": 5590
},
{
"epoch": 10.37037037037037,
"grad_norm": 0.20710179209709167,
"learning_rate": 4.424219253586737e-05,
"loss": 0.0144,
"step": 5600
},
{
"epoch": 10.38888888888889,
"grad_norm": 0.1561076045036316,
"learning_rate": 4.407797732367443e-05,
"loss": 0.0174,
"step": 5610
},
{
"epoch": 10.407407407407407,
"grad_norm": 0.16007019579410553,
"learning_rate": 4.391382687377268e-05,
"loss": 0.0161,
"step": 5620
},
{
"epoch": 10.425925925925926,
"grad_norm": 0.19607481360435486,
"learning_rate": 4.374974298128512e-05,
"loss": 0.0152,
"step": 5630
},
{
"epoch": 10.444444444444445,
"grad_norm": 0.20307008922100067,
"learning_rate": 4.358572744060699e-05,
"loss": 0.0159,
"step": 5640
},
{
"epoch": 10.462962962962964,
"grad_norm": 0.13981832563877106,
"learning_rate": 4.342178204538588e-05,
"loss": 0.0137,
"step": 5650
},
{
"epoch": 10.481481481481481,
"grad_norm": 0.1687135249376297,
"learning_rate": 4.325790858850241e-05,
"loss": 0.0148,
"step": 5660
},
{
"epoch": 10.5,
"grad_norm": 0.2364518642425537,
"learning_rate": 4.309410886205043e-05,
"loss": 0.0138,
"step": 5670
},
{
"epoch": 10.518518518518519,
"grad_norm": 0.20697851479053497,
"learning_rate": 4.293038465731752e-05,
"loss": 0.0155,
"step": 5680
},
{
"epoch": 10.537037037037036,
"grad_norm": 0.1821158230304718,
"learning_rate": 4.276673776476533e-05,
"loss": 0.0131,
"step": 5690
},
{
"epoch": 10.555555555555555,
"grad_norm": 0.18130852282047272,
"learning_rate": 4.260316997401007e-05,
"loss": 0.0151,
"step": 5700
},
{
"epoch": 10.574074074074074,
"grad_norm": 0.22771954536437988,
"learning_rate": 4.243968307380293e-05,
"loss": 0.0123,
"step": 5710
},
{
"epoch": 10.592592592592592,
"grad_norm": 0.22695477306842804,
"learning_rate": 4.22762788520104e-05,
"loss": 0.0133,
"step": 5720
},
{
"epoch": 10.61111111111111,
"grad_norm": 0.1576288938522339,
"learning_rate": 4.211295909559491e-05,
"loss": 0.0138,
"step": 5730
},
{
"epoch": 10.62962962962963,
"grad_norm": 0.23055940866470337,
"learning_rate": 4.194972559059511e-05,
"loss": 0.0156,
"step": 5740
},
{
"epoch": 10.648148148148149,
"grad_norm": 0.15297625958919525,
"learning_rate": 4.178658012210651e-05,
"loss": 0.0146,
"step": 5750
},
{
"epoch": 10.666666666666666,
"grad_norm": 0.2038733959197998,
"learning_rate": 4.162352447426177e-05,
"loss": 0.0122,
"step": 5760
},
{
"epoch": 10.685185185185185,
"grad_norm": 0.18565721809864044,
"learning_rate": 4.146056043021135e-05,
"loss": 0.0135,
"step": 5770
},
{
"epoch": 10.703703703703704,
"grad_norm": 0.15847265720367432,
"learning_rate": 4.1297689772103944e-05,
"loss": 0.0142,
"step": 5780
},
{
"epoch": 10.722222222222221,
"grad_norm": 0.11047548800706863,
"learning_rate": 4.113491428106694e-05,
"loss": 0.0137,
"step": 5790
},
{
"epoch": 10.74074074074074,
"grad_norm": 0.19697855412960052,
"learning_rate": 4.0972235737187055e-05,
"loss": 0.0119,
"step": 5800
},
{
"epoch": 10.75925925925926,
"grad_norm": 0.1942683607339859,
"learning_rate": 4.080965591949076e-05,
"loss": 0.0135,
"step": 5810
},
{
"epoch": 10.777777777777779,
"grad_norm": 0.19134296476840973,
"learning_rate": 4.0647176605924924e-05,
"loss": 0.0152,
"step": 5820
},
{
"epoch": 10.796296296296296,
"grad_norm": 0.1203959733247757,
"learning_rate": 4.0484799573337255e-05,
"loss": 0.0122,
"step": 5830
},
{
"epoch": 10.814814814814815,
"grad_norm": 0.22707918286323547,
"learning_rate": 4.032252659745699e-05,
"loss": 0.0128,
"step": 5840
},
{
"epoch": 10.833333333333334,
"grad_norm": 0.16146455705165863,
"learning_rate": 4.016035945287539e-05,
"loss": 0.013,
"step": 5850
},
{
"epoch": 10.851851851851851,
"grad_norm": 0.19473141431808472,
"learning_rate": 3.999829991302635e-05,
"loss": 0.0157,
"step": 5860
},
{
"epoch": 10.87037037037037,
"grad_norm": 0.14040832221508026,
"learning_rate": 3.983634975016707e-05,
"loss": 0.0126,
"step": 5870
},
{
"epoch": 10.88888888888889,
"grad_norm": 0.15125393867492676,
"learning_rate": 3.967451073535854e-05,
"loss": 0.0125,
"step": 5880
},
{
"epoch": 10.907407407407408,
"grad_norm": 0.1352330893278122,
"learning_rate": 3.951278463844633e-05,
"loss": 0.0126,
"step": 5890
},
{
"epoch": 10.925925925925926,
"grad_norm": 0.12836302816867828,
"learning_rate": 3.935117322804111e-05,
"loss": 0.0107,
"step": 5900
},
{
"epoch": 10.944444444444445,
"grad_norm": 0.13717485964298248,
"learning_rate": 3.918967827149938e-05,
"loss": 0.0143,
"step": 5910
},
{
"epoch": 10.962962962962964,
"grad_norm": 0.17137043178081512,
"learning_rate": 3.9028301534904094e-05,
"loss": 0.0151,
"step": 5920
},
{
"epoch": 10.981481481481481,
"grad_norm": 0.15853627026081085,
"learning_rate": 3.88670447830454e-05,
"loss": 0.0123,
"step": 5930
},
{
"epoch": 11.0,
"grad_norm": 0.36139190196990967,
"learning_rate": 3.870590977940132e-05,
"loss": 0.0135,
"step": 5940
},
{
"epoch": 11.018518518518519,
"grad_norm": 0.21963530778884888,
"learning_rate": 3.8544898286118404e-05,
"loss": 0.0141,
"step": 5950
},
{
"epoch": 11.037037037037036,
"grad_norm": 0.21109621226787567,
"learning_rate": 3.838401206399257e-05,
"loss": 0.0162,
"step": 5960
},
{
"epoch": 11.055555555555555,
"grad_norm": 0.28180623054504395,
"learning_rate": 3.822325287244975e-05,
"loss": 0.0136,
"step": 5970
},
{
"epoch": 11.074074074074074,
"grad_norm": 0.22362850606441498,
"learning_rate": 3.8062622469526725e-05,
"loss": 0.0123,
"step": 5980
},
{
"epoch": 11.092592592592593,
"grad_norm": 0.2641434073448181,
"learning_rate": 3.790212261185183e-05,
"loss": 0.0124,
"step": 5990
},
{
"epoch": 11.11111111111111,
"grad_norm": 0.15765398740768433,
"learning_rate": 3.7741755054625794e-05,
"loss": 0.0143,
"step": 6000
},
{
"epoch": 11.12962962962963,
"grad_norm": 0.11513574421405792,
"learning_rate": 3.758152155160255e-05,
"loss": 0.0124,
"step": 6010
},
{
"epoch": 11.148148148148149,
"grad_norm": 0.164279043674469,
"learning_rate": 3.742142385506999e-05,
"loss": 0.0142,
"step": 6020
},
{
"epoch": 11.166666666666666,
"grad_norm": 0.1784529834985733,
"learning_rate": 3.72614637158309e-05,
"loss": 0.0139,
"step": 6030
},
{
"epoch": 11.185185185185185,
"grad_norm": 0.24391107261180878,
"learning_rate": 3.710164288318371e-05,
"loss": 0.0143,
"step": 6040
},
{
"epoch": 11.203703703703704,
"grad_norm": 0.11302986741065979,
"learning_rate": 3.694196310490345e-05,
"loss": 0.0115,
"step": 6050
},
{
"epoch": 11.222222222222221,
"grad_norm": 0.27688246965408325,
"learning_rate": 3.678242612722259e-05,
"loss": 0.0137,
"step": 6060
},
{
"epoch": 11.24074074074074,
"grad_norm": 0.14006178081035614,
"learning_rate": 3.6623033694811953e-05,
"loss": 0.0119,
"step": 6070
},
{
"epoch": 11.25925925925926,
"grad_norm": 0.1449032574892044,
"learning_rate": 3.6463787550761665e-05,
"loss": 0.0129,
"step": 6080
},
{
"epoch": 11.277777777777779,
"grad_norm": 0.15014226734638214,
"learning_rate": 3.630468943656202e-05,
"loss": 0.0166,
"step": 6090
},
{
"epoch": 11.296296296296296,
"grad_norm": 0.1299763023853302,
"learning_rate": 3.6145741092084523e-05,
"loss": 0.0129,
"step": 6100
},
{
"epoch": 11.314814814814815,
"grad_norm": 0.22849303483963013,
"learning_rate": 3.598694425556278e-05,
"loss": 0.0131,
"step": 6110
},
{
"epoch": 11.333333333333334,
"grad_norm": 0.15405701100826263,
"learning_rate": 3.58283006635736e-05,
"loss": 0.0143,
"step": 6120
},
{
"epoch": 11.351851851851851,
"grad_norm": 0.18705284595489502,
"learning_rate": 3.566981205101781e-05,
"loss": 0.0143,
"step": 6130
},
{
"epoch": 11.37037037037037,
"grad_norm": 0.15108929574489594,
"learning_rate": 3.5511480151101556e-05,
"loss": 0.0119,
"step": 6140
},
{
"epoch": 11.38888888888889,
"grad_norm": 0.1521301120519638,
"learning_rate": 3.5353306695317104e-05,
"loss": 0.0136,
"step": 6150
},
{
"epoch": 11.407407407407407,
"grad_norm": 0.08295134454965591,
"learning_rate": 3.519529341342402e-05,
"loss": 0.0124,
"step": 6160
},
{
"epoch": 11.425925925925926,
"grad_norm": 0.18803167343139648,
"learning_rate": 3.503744203343026e-05,
"loss": 0.012,
"step": 6170
},
{
"epoch": 11.444444444444445,
"grad_norm": 0.13088999688625336,
"learning_rate": 3.487975428157318e-05,
"loss": 0.0111,
"step": 6180
},
{
"epoch": 11.462962962962964,
"grad_norm": 0.24680213630199432,
"learning_rate": 3.472223188230083e-05,
"loss": 0.0115,
"step": 6190
},
{
"epoch": 11.481481481481481,
"grad_norm": 0.15881215035915375,
"learning_rate": 3.4564876558252866e-05,
"loss": 0.0137,
"step": 6200
},
{
"epoch": 11.5,
"grad_norm": 0.16456684470176697,
"learning_rate": 3.440769003024195e-05,
"loss": 0.0144,
"step": 6210
},
{
"epoch": 11.518518518518519,
"grad_norm": 0.25722959637641907,
"learning_rate": 3.425067401723477e-05,
"loss": 0.0147,
"step": 6220
},
{
"epoch": 11.537037037037036,
"grad_norm": 0.18458130955696106,
"learning_rate": 3.409383023633325e-05,
"loss": 0.0135,
"step": 6230
},
{
"epoch": 11.555555555555555,
"grad_norm": 0.10942313820123672,
"learning_rate": 3.3937160402755894e-05,
"loss": 0.0106,
"step": 6240
},
{
"epoch": 11.574074074074074,
"grad_norm": 0.16660438477993011,
"learning_rate": 3.378066622981885e-05,
"loss": 0.0129,
"step": 6250
},
{
"epoch": 11.592592592592592,
"grad_norm": 0.2543100118637085,
"learning_rate": 3.362434942891738e-05,
"loss": 0.0116,
"step": 6260
},
{
"epoch": 11.61111111111111,
"grad_norm": 0.13930945098400116,
"learning_rate": 3.346821170950693e-05,
"loss": 0.0119,
"step": 6270
},
{
"epoch": 11.62962962962963,
"grad_norm": 0.20048055052757263,
"learning_rate": 3.3312254779084585e-05,
"loss": 0.011,
"step": 6280
},
{
"epoch": 11.648148148148149,
"grad_norm": 0.1817791610956192,
"learning_rate": 3.315648034317039e-05,
"loss": 0.0118,
"step": 6290
},
{
"epoch": 11.666666666666666,
"grad_norm": 0.17183825373649597,
"learning_rate": 3.3000890105288564e-05,
"loss": 0.0138,
"step": 6300
},
{
"epoch": 11.685185185185185,
"grad_norm": 0.17662788927555084,
"learning_rate": 3.284548576694908e-05,
"loss": 0.0121,
"step": 6310
},
{
"epoch": 11.703703703703704,
"grad_norm": 0.16434216499328613,
"learning_rate": 3.2690269027628815e-05,
"loss": 0.0113,
"step": 6320
},
{
"epoch": 11.722222222222221,
"grad_norm": 0.2037888616323471,
"learning_rate": 3.253524158475324e-05,
"loss": 0.012,
"step": 6330
},
{
"epoch": 11.74074074074074,
"grad_norm": 0.17854318022727966,
"learning_rate": 3.238040513367757e-05,
"loss": 0.0116,
"step": 6340
},
{
"epoch": 11.75925925925926,
"grad_norm": 0.13727858662605286,
"learning_rate": 3.222576136766843e-05,
"loss": 0.0135,
"step": 6350
},
{
"epoch": 11.777777777777779,
"grad_norm": 0.1561526656150818,
"learning_rate": 3.2071311977885324e-05,
"loss": 0.0148,
"step": 6360
},
{
"epoch": 11.796296296296296,
"grad_norm": 0.1555299311876297,
"learning_rate": 3.191705865336197e-05,
"loss": 0.0119,
"step": 6370
},
{
"epoch": 11.814814814814815,
"grad_norm": 0.19848501682281494,
"learning_rate": 3.1763003080988075e-05,
"loss": 0.0126,
"step": 6380
},
{
"epoch": 11.833333333333334,
"grad_norm": 0.16384631395339966,
"learning_rate": 3.160914694549063e-05,
"loss": 0.0109,
"step": 6390
},
{
"epoch": 11.851851851851851,
"grad_norm": 0.14514987170696259,
"learning_rate": 3.145549192941573e-05,
"loss": 0.013,
"step": 6400
},
{
"epoch": 11.87037037037037,
"grad_norm": 0.16420520842075348,
"learning_rate": 3.130203971310999e-05,
"loss": 0.0142,
"step": 6410
},
{
"epoch": 11.88888888888889,
"grad_norm": 0.13385368883609772,
"learning_rate": 3.114879197470225e-05,
"loss": 0.0134,
"step": 6420
},
{
"epoch": 11.907407407407408,
"grad_norm": 0.1364380270242691,
"learning_rate": 3.0995750390085285e-05,
"loss": 0.0128,
"step": 6430
},
{
"epoch": 11.925925925925926,
"grad_norm": 0.13520628213882446,
"learning_rate": 3.084291663289728e-05,
"loss": 0.0117,
"step": 6440
},
{
"epoch": 11.944444444444445,
"grad_norm": 0.17970336973667145,
"learning_rate": 3.069029237450375e-05,
"loss": 0.012,
"step": 6450
},
{
"epoch": 11.962962962962964,
"grad_norm": 0.1322975605726242,
"learning_rate": 3.053787928397911e-05,
"loss": 0.0122,
"step": 6460
},
{
"epoch": 11.981481481481481,
"grad_norm": 0.189213827252388,
"learning_rate": 3.0385679028088526e-05,
"loss": 0.0125,
"step": 6470
},
{
"epoch": 12.0,
"grad_norm": 0.2747475504875183,
"learning_rate": 3.023369327126959e-05,
"loss": 0.0112,
"step": 6480
},
{
"epoch": 12.018518518518519,
"grad_norm": 0.13897903263568878,
"learning_rate": 3.0081923675614198e-05,
"loss": 0.0108,
"step": 6490
},
{
"epoch": 12.037037037037036,
"grad_norm": 0.16608110070228577,
"learning_rate": 2.993037190085034e-05,
"loss": 0.0137,
"step": 6500
},
{
"epoch": 12.055555555555555,
"grad_norm": 0.19581569731235504,
"learning_rate": 2.977903960432392e-05,
"loss": 0.0116,
"step": 6510
},
{
"epoch": 12.074074074074074,
"grad_norm": 0.12218277156352997,
"learning_rate": 2.9627928440980722e-05,
"loss": 0.0154,
"step": 6520
},
{
"epoch": 12.092592592592593,
"grad_norm": 0.16085056960582733,
"learning_rate": 2.9477040063348183e-05,
"loss": 0.0113,
"step": 6530
},
{
"epoch": 12.11111111111111,
"grad_norm": 0.16041666269302368,
"learning_rate": 2.9326376121517456e-05,
"loss": 0.0121,
"step": 6540
},
{
"epoch": 12.12962962962963,
"grad_norm": 0.14760176837444305,
"learning_rate": 2.9175938263125236e-05,
"loss": 0.0098,
"step": 6550
},
{
"epoch": 12.148148148148149,
"grad_norm": 0.1427803933620453,
"learning_rate": 2.9025728133335873e-05,
"loss": 0.0118,
"step": 6560
},
{
"epoch": 12.166666666666666,
"grad_norm": 0.16523391008377075,
"learning_rate": 2.8875747374823288e-05,
"loss": 0.0124,
"step": 6570
},
{
"epoch": 12.185185185185185,
"grad_norm": 0.09955291450023651,
"learning_rate": 2.872599762775298e-05,
"loss": 0.0131,
"step": 6580
},
{
"epoch": 12.203703703703704,
"grad_norm": 0.12456505745649338,
"learning_rate": 2.857648052976425e-05,
"loss": 0.0104,
"step": 6590
},
{
"epoch": 12.222222222222221,
"grad_norm": 0.22651228308677673,
"learning_rate": 2.8427197715952047e-05,
"loss": 0.012,
"step": 6600
},
{
"epoch": 12.24074074074074,
"grad_norm": 0.20232881605625153,
"learning_rate": 2.8278150818849393e-05,
"loss": 0.012,
"step": 6610
},
{
"epoch": 12.25925925925926,
"grad_norm": 0.21699605882167816,
"learning_rate": 2.812934146840922e-05,
"loss": 0.0125,
"step": 6620
},
{
"epoch": 12.277777777777779,
"grad_norm": 0.16731399297714233,
"learning_rate": 2.7980771291986764e-05,
"loss": 0.0148,
"step": 6630
},
{
"epoch": 12.296296296296296,
"grad_norm": 0.17831040918827057,
"learning_rate": 2.783244191432167e-05,
"loss": 0.013,
"step": 6640
},
{
"epoch": 12.314814814814815,
"grad_norm": 0.1302179992198944,
"learning_rate": 2.768435495752022e-05,
"loss": 0.0139,
"step": 6650
},
{
"epoch": 12.333333333333334,
"grad_norm": 0.17461994290351868,
"learning_rate": 2.753651204103771e-05,
"loss": 0.0144,
"step": 6660
},
{
"epoch": 12.351851851851851,
"grad_norm": 0.17328962683677673,
"learning_rate": 2.7388914781660523e-05,
"loss": 0.009,
"step": 6670
},
{
"epoch": 12.37037037037037,
"grad_norm": 0.07789794355630875,
"learning_rate": 2.7241564793488693e-05,
"loss": 0.0131,
"step": 6680
},
{
"epoch": 12.38888888888889,
"grad_norm": 0.14367705583572388,
"learning_rate": 2.7094463687918037e-05,
"loss": 0.0115,
"step": 6690
},
{
"epoch": 12.407407407407407,
"grad_norm": 0.16609890758991241,
"learning_rate": 2.694761307362268e-05,
"loss": 0.0113,
"step": 6700
},
{
"epoch": 12.425925925925926,
"grad_norm": 0.18608912825584412,
"learning_rate": 2.6801014556537467e-05,
"loss": 0.0105,
"step": 6710
},
{
"epoch": 12.444444444444445,
"grad_norm": 0.12139278650283813,
"learning_rate": 2.6654669739840243e-05,
"loss": 0.0117,
"step": 6720
},
{
"epoch": 12.462962962962964,
"grad_norm": 0.17434817552566528,
"learning_rate": 2.650858022393451e-05,
"loss": 0.0124,
"step": 6730
},
{
"epoch": 12.481481481481481,
"grad_norm": 0.18277306854724884,
"learning_rate": 2.6362747606431747e-05,
"loss": 0.0134,
"step": 6740
},
{
"epoch": 12.5,
"grad_norm": 0.14479221403598785,
"learning_rate": 2.6217173482134172e-05,
"loss": 0.0117,
"step": 6750
},
{
"epoch": 12.518518518518519,
"grad_norm": 0.1521540880203247,
"learning_rate": 2.6071859443017044e-05,
"loss": 0.0155,
"step": 6760
},
{
"epoch": 12.537037037037036,
"grad_norm": 0.13006556034088135,
"learning_rate": 2.5926807078211414e-05,
"loss": 0.0121,
"step": 6770
},
{
"epoch": 12.555555555555555,
"grad_norm": 0.14429277181625366,
"learning_rate": 2.5782017973986728e-05,
"loss": 0.0113,
"step": 6780
},
{
"epoch": 12.574074074074074,
"grad_norm": 0.16124217212200165,
"learning_rate": 2.5637493713733374e-05,
"loss": 0.0132,
"step": 6790
},
{
"epoch": 12.592592592592592,
"grad_norm": 0.18411949276924133,
"learning_rate": 2.549323587794559e-05,
"loss": 0.0119,
"step": 6800
},
{
"epoch": 12.61111111111111,
"grad_norm": 0.19239585101604462,
"learning_rate": 2.5349246044203895e-05,
"loss": 0.0119,
"step": 6810
},
{
"epoch": 12.62962962962963,
"grad_norm": 0.1233111321926117,
"learning_rate": 2.520552578715808e-05,
"loss": 0.0121,
"step": 6820
},
{
"epoch": 12.648148148148149,
"grad_norm": 0.20407912135124207,
"learning_rate": 2.506207667850981e-05,
"loss": 0.0145,
"step": 6830
},
{
"epoch": 12.666666666666666,
"grad_norm": 0.1662532389163971,
"learning_rate": 2.4918900286995555e-05,
"loss": 0.0115,
"step": 6840
},
{
"epoch": 12.685185185185185,
"grad_norm": 0.1678568720817566,
"learning_rate": 2.4775998178369458e-05,
"loss": 0.0107,
"step": 6850
},
{
"epoch": 12.703703703703704,
"grad_norm": 0.14796386659145355,
"learning_rate": 2.4633371915386017e-05,
"loss": 0.0112,
"step": 6860
},
{
"epoch": 12.722222222222221,
"grad_norm": 0.13624046742916107,
"learning_rate": 2.4491023057783235e-05,
"loss": 0.0106,
"step": 6870
},
{
"epoch": 12.74074074074074,
"grad_norm": 0.15286503732204437,
"learning_rate": 2.4348953162265375e-05,
"loss": 0.0124,
"step": 6880
},
{
"epoch": 12.75925925925926,
"grad_norm": 0.1121831014752388,
"learning_rate": 2.420716378248607e-05,
"loss": 0.0121,
"step": 6890
},
{
"epoch": 12.777777777777779,
"grad_norm": 0.1353166401386261,
"learning_rate": 2.4065656469031266e-05,
"loss": 0.0114,
"step": 6900
},
{
"epoch": 12.796296296296296,
"grad_norm": 0.19370369613170624,
"learning_rate": 2.3924432769402268e-05,
"loss": 0.0111,
"step": 6910
},
{
"epoch": 12.814814814814815,
"grad_norm": 0.18840144574642181,
"learning_rate": 2.3783494227998844e-05,
"loss": 0.0129,
"step": 6920
},
{
"epoch": 12.833333333333334,
"grad_norm": 0.16810353100299835,
"learning_rate": 2.3642842386102264e-05,
"loss": 0.0125,
"step": 6930
},
{
"epoch": 12.851851851851851,
"grad_norm": 0.13411790132522583,
"learning_rate": 2.3502478781858567e-05,
"loss": 0.0117,
"step": 6940
},
{
"epoch": 12.87037037037037,
"grad_norm": 0.1498802751302719,
"learning_rate": 2.3362404950261628e-05,
"loss": 0.0105,
"step": 6950
},
{
"epoch": 12.88888888888889,
"grad_norm": 0.1473735272884369,
"learning_rate": 2.3222622423136458e-05,
"loss": 0.0106,
"step": 6960
},
{
"epoch": 12.907407407407408,
"grad_norm": 0.15255799889564514,
"learning_rate": 2.3083132729122332e-05,
"loss": 0.0112,
"step": 6970
},
{
"epoch": 12.925925925925926,
"grad_norm": 0.10796765983104706,
"learning_rate": 2.294393739365621e-05,
"loss": 0.0123,
"step": 6980
},
{
"epoch": 12.944444444444445,
"grad_norm": 0.1290091574192047,
"learning_rate": 2.2805037938956e-05,
"loss": 0.012,
"step": 6990
},
{
"epoch": 12.962962962962964,
"grad_norm": 0.16411873698234558,
"learning_rate": 2.266643588400386e-05,
"loss": 0.0101,
"step": 7000
},
{
"epoch": 12.981481481481481,
"grad_norm": 0.23581312596797943,
"learning_rate": 2.252813274452969e-05,
"loss": 0.0105,
"step": 7010
},
{
"epoch": 13.0,
"grad_norm": 0.8078553676605225,
"learning_rate": 2.2390130032994427e-05,
"loss": 0.0109,
"step": 7020
},
{
"epoch": 13.018518518518519,
"grad_norm": 0.2610144317150116,
"learning_rate": 2.2252429258573633e-05,
"loss": 0.0125,
"step": 7030
},
{
"epoch": 13.037037037037036,
"grad_norm": 0.2771884799003601,
"learning_rate": 2.2115031927140904e-05,
"loss": 0.0117,
"step": 7040
},
{
"epoch": 13.055555555555555,
"grad_norm": 0.14431428909301758,
"learning_rate": 2.1977939541251463e-05,
"loss": 0.0126,
"step": 7050
},
{
"epoch": 13.074074074074074,
"grad_norm": 0.21846048533916473,
"learning_rate": 2.1841153600125684e-05,
"loss": 0.0105,
"step": 7060
},
{
"epoch": 13.092592592592593,
"grad_norm": 0.13519643247127533,
"learning_rate": 2.170467559963267e-05,
"loss": 0.0112,
"step": 7070
},
{
"epoch": 13.11111111111111,
"grad_norm": 0.15354624390602112,
"learning_rate": 2.1568507032273982e-05,
"loss": 0.0141,
"step": 7080
},
{
"epoch": 13.12962962962963,
"grad_norm": 0.10714532434940338,
"learning_rate": 2.1432649387167264e-05,
"loss": 0.0102,
"step": 7090
},
{
"epoch": 13.148148148148149,
"grad_norm": 0.18656988441944122,
"learning_rate": 2.1297104150029973e-05,
"loss": 0.0113,
"step": 7100
},
{
"epoch": 13.166666666666666,
"grad_norm": 0.15229035913944244,
"learning_rate": 2.116187280316307e-05,
"loss": 0.0104,
"step": 7110
},
{
"epoch": 13.185185185185185,
"grad_norm": 0.21051520109176636,
"learning_rate": 2.1026956825434908e-05,
"loss": 0.012,
"step": 7120
},
{
"epoch": 13.203703703703704,
"grad_norm": 0.17103920876979828,
"learning_rate": 2.0892357692265017e-05,
"loss": 0.0107,
"step": 7130
},
{
"epoch": 13.222222222222221,
"grad_norm": 0.15137511491775513,
"learning_rate": 2.0758076875607947e-05,
"loss": 0.0121,
"step": 7140
},
{
"epoch": 13.24074074074074,
"grad_norm": 0.15944904088974,
"learning_rate": 2.0624115843937207e-05,
"loss": 0.0094,
"step": 7150
},
{
"epoch": 13.25925925925926,
"grad_norm": 0.13502219319343567,
"learning_rate": 2.0490476062229157e-05,
"loss": 0.0089,
"step": 7160
},
{
"epoch": 13.277777777777779,
"grad_norm": 0.1867222785949707,
"learning_rate": 2.035715899194704e-05,
"loss": 0.0115,
"step": 7170
},
{
"epoch": 13.296296296296296,
"grad_norm": 0.1293974369764328,
"learning_rate": 2.022416609102499e-05,
"loss": 0.012,
"step": 7180
},
{
"epoch": 13.314814814814815,
"grad_norm": 0.13711346685886383,
"learning_rate": 2.009149881385205e-05,
"loss": 0.012,
"step": 7190
},
{
"epoch": 13.333333333333334,
"grad_norm": 0.2136504203081131,
"learning_rate": 1.995915861125634e-05,
"loss": 0.0122,
"step": 7200
},
{
"epoch": 13.351851851851851,
"grad_norm": 0.15976719558238983,
"learning_rate": 1.9827146930489065e-05,
"loss": 0.0109,
"step": 7210
},
{
"epoch": 13.37037037037037,
"grad_norm": 0.16626125574111938,
"learning_rate": 1.9695465215208848e-05,
"loss": 0.0114,
"step": 7220
},
{
"epoch": 13.38888888888889,
"grad_norm": 0.1644420176744461,
"learning_rate": 1.9564114905465813e-05,
"loss": 0.0094,
"step": 7230
},
{
"epoch": 13.407407407407407,
"grad_norm": 0.14022791385650635,
"learning_rate": 1.9433097437685936e-05,
"loss": 0.012,
"step": 7240
},
{
"epoch": 13.425925925925926,
"grad_norm": 0.1368105709552765,
"learning_rate": 1.930241424465521e-05,
"loss": 0.0116,
"step": 7250
},
{
"epoch": 13.444444444444445,
"grad_norm": 0.14134937524795532,
"learning_rate": 1.9172066755504115e-05,
"loss": 0.0121,
"step": 7260
},
{
"epoch": 13.462962962962964,
"grad_norm": 0.13752196729183197,
"learning_rate": 1.9042056395691914e-05,
"loss": 0.0088,
"step": 7270
},
{
"epoch": 13.481481481481481,
"grad_norm": 0.16244389116764069,
"learning_rate": 1.8912384586991066e-05,
"loss": 0.0123,
"step": 7280
},
{
"epoch": 13.5,
"grad_norm": 0.1296980381011963,
"learning_rate": 1.8783052747471717e-05,
"loss": 0.0118,
"step": 7290
},
{
"epoch": 13.518518518518519,
"grad_norm": 0.1727660745382309,
"learning_rate": 1.865406229148611e-05,
"loss": 0.0112,
"step": 7300
},
{
"epoch": 13.537037037037036,
"grad_norm": 0.17700189352035522,
"learning_rate": 1.8525414629653233e-05,
"loss": 0.0099,
"step": 7310
},
{
"epoch": 13.555555555555555,
"grad_norm": 0.0852062776684761,
"learning_rate": 1.8397111168843255e-05,
"loss": 0.0108,
"step": 7320
},
{
"epoch": 13.574074074074074,
"grad_norm": 0.1737292855978012,
"learning_rate": 1.8269153312162323e-05,
"loss": 0.0133,
"step": 7330
},
{
"epoch": 13.592592592592592,
"grad_norm": 0.15968166291713715,
"learning_rate": 1.8141542458937054e-05,
"loss": 0.0091,
"step": 7340
},
{
"epoch": 13.61111111111111,
"grad_norm": 0.16132958233356476,
"learning_rate": 1.8014280004699268e-05,
"loss": 0.0094,
"step": 7350
},
{
"epoch": 13.62962962962963,
"grad_norm": 0.1572614163160324,
"learning_rate": 1.788736734117078e-05,
"loss": 0.011,
"step": 7360
},
{
"epoch": 13.648148148148149,
"grad_norm": 0.19380661845207214,
"learning_rate": 1.7760805856248152e-05,
"loss": 0.0146,
"step": 7370
},
{
"epoch": 13.666666666666666,
"grad_norm": 0.14780157804489136,
"learning_rate": 1.7634596933987518e-05,
"loss": 0.0123,
"step": 7380
},
{
"epoch": 13.685185185185185,
"grad_norm": 0.12522783875465393,
"learning_rate": 1.7508741954589404e-05,
"loss": 0.0127,
"step": 7390
},
{
"epoch": 13.703703703703704,
"grad_norm": 0.15862339735031128,
"learning_rate": 1.7383242294383717e-05,
"loss": 0.0129,
"step": 7400
},
{
"epoch": 13.722222222222221,
"grad_norm": 0.19172026216983795,
"learning_rate": 1.7258099325814632e-05,
"loss": 0.01,
"step": 7410
},
{
"epoch": 13.74074074074074,
"grad_norm": 0.13690708577632904,
"learning_rate": 1.7133314417425594e-05,
"loss": 0.008,
"step": 7420
},
{
"epoch": 13.75925925925926,
"grad_norm": 0.12296158820390701,
"learning_rate": 1.7008888933844408e-05,
"loss": 0.0096,
"step": 7430
},
{
"epoch": 13.777777777777779,
"grad_norm": 0.16455790400505066,
"learning_rate": 1.6884824235768172e-05,
"loss": 0.0104,
"step": 7440
},
{
"epoch": 13.796296296296296,
"grad_norm": 0.12299106270074844,
"learning_rate": 1.6761121679948592e-05,
"loss": 0.0103,
"step": 7450
},
{
"epoch": 13.814814814814815,
"grad_norm": 0.10579638183116913,
"learning_rate": 1.663778261917695e-05,
"loss": 0.0098,
"step": 7460
},
{
"epoch": 13.833333333333334,
"grad_norm": 0.1649494767189026,
"learning_rate": 1.651480840226952e-05,
"loss": 0.0107,
"step": 7470
},
{
"epoch": 13.851851851851851,
"grad_norm": 0.1375342160463333,
"learning_rate": 1.639220037405258e-05,
"loss": 0.0106,
"step": 7480
},
{
"epoch": 13.87037037037037,
"grad_norm": 0.12565447390079498,
"learning_rate": 1.6269959875347906e-05,
"loss": 0.0091,
"step": 7490
},
{
"epoch": 13.88888888888889,
"grad_norm": 0.18669046461582184,
"learning_rate": 1.614808824295802e-05,
"loss": 0.0084,
"step": 7500
},
{
"epoch": 13.907407407407408,
"grad_norm": 0.1130140870809555,
"learning_rate": 1.602658680965152e-05,
"loss": 0.0093,
"step": 7510
},
{
"epoch": 13.925925925925926,
"grad_norm": 0.1748272180557251,
"learning_rate": 1.5905456904148686e-05,
"loss": 0.0097,
"step": 7520
},
{
"epoch": 13.944444444444445,
"grad_norm": 0.12322141975164413,
"learning_rate": 1.57846998511067e-05,
"loss": 0.0096,
"step": 7530
},
{
"epoch": 13.962962962962964,
"grad_norm": 0.11009227484464645,
"learning_rate": 1.566431697110538e-05,
"loss": 0.0101,
"step": 7540
},
{
"epoch": 13.981481481481481,
"grad_norm": 0.1575208306312561,
"learning_rate": 1.554430958063259e-05,
"loss": 0.0106,
"step": 7550
},
{
"epoch": 14.0,
"grad_norm": 0.11694421619176865,
"learning_rate": 1.5424678992069912e-05,
"loss": 0.0081,
"step": 7560
},
{
"epoch": 14.018518518518519,
"grad_norm": 0.15677650272846222,
"learning_rate": 1.5305426513678362e-05,
"loss": 0.0124,
"step": 7570
},
{
"epoch": 14.037037037037036,
"grad_norm": 0.14466744661331177,
"learning_rate": 1.518655344958388e-05,
"loss": 0.0098,
"step": 7580
},
{
"epoch": 14.055555555555555,
"grad_norm": 0.1616063117980957,
"learning_rate": 1.5068061099763275e-05,
"loss": 0.0087,
"step": 7590
},
{
"epoch": 14.074074074074074,
"grad_norm": 0.11272673308849335,
"learning_rate": 1.494995076002988e-05,
"loss": 0.0085,
"step": 7600
},
{
"epoch": 14.092592592592593,
"grad_norm": 0.18691617250442505,
"learning_rate": 1.4832223722019456e-05,
"loss": 0.0106,
"step": 7610
},
{
"epoch": 14.11111111111111,
"grad_norm": 0.10479917377233505,
"learning_rate": 1.4714881273176035e-05,
"loss": 0.0101,
"step": 7620
},
{
"epoch": 14.12962962962963,
"grad_norm": 0.12924514710903168,
"learning_rate": 1.4597924696737835e-05,
"loss": 0.0095,
"step": 7630
},
{
"epoch": 14.148148148148149,
"grad_norm": 0.07907586544752121,
"learning_rate": 1.4481355271723252e-05,
"loss": 0.0107,
"step": 7640
},
{
"epoch": 14.166666666666666,
"grad_norm": 0.1355125904083252,
"learning_rate": 1.4365174272916809e-05,
"loss": 0.0108,
"step": 7650
},
{
"epoch": 14.185185185185185,
"grad_norm": 0.08970734477043152,
"learning_rate": 1.4249382970855319e-05,
"loss": 0.0112,
"step": 7660
},
{
"epoch": 14.203703703703704,
"grad_norm": 0.12842874228954315,
"learning_rate": 1.4133982631813903e-05,
"loss": 0.0115,
"step": 7670
},
{
"epoch": 14.222222222222221,
"grad_norm": 0.10979616641998291,
"learning_rate": 1.4018974517792194e-05,
"loss": 0.0119,
"step": 7680
},
{
"epoch": 14.24074074074074,
"grad_norm": 0.10717067867517471,
"learning_rate": 1.390435988650048e-05,
"loss": 0.0082,
"step": 7690
},
{
"epoch": 14.25925925925926,
"grad_norm": 0.12461619824171066,
"learning_rate": 1.3790139991346006e-05,
"loss": 0.0089,
"step": 7700
},
{
"epoch": 14.277777777777779,
"grad_norm": 0.1722489297389984,
"learning_rate": 1.367631608141926e-05,
"loss": 0.0115,
"step": 7710
},
{
"epoch": 14.296296296296296,
"grad_norm": 0.12079145014286041,
"learning_rate": 1.3562889401480278e-05,
"loss": 0.0129,
"step": 7720
},
{
"epoch": 14.314814814814815,
"grad_norm": 0.15433591604232788,
"learning_rate": 1.3449861191945074e-05,
"loss": 0.0086,
"step": 7730
},
{
"epoch": 14.333333333333334,
"grad_norm": 0.150528222322464,
"learning_rate": 1.3337232688872009e-05,
"loss": 0.01,
"step": 7740
},
{
"epoch": 14.351851851851851,
"grad_norm": 0.10879624634981155,
"learning_rate": 1.3225005123948364e-05,
"loss": 0.0088,
"step": 7750
},
{
"epoch": 14.37037037037037,
"grad_norm": 0.1072290912270546,
"learning_rate": 1.311317972447681e-05,
"loss": 0.0093,
"step": 7760
},
{
"epoch": 14.38888888888889,
"grad_norm": 0.14874421060085297,
"learning_rate": 1.3001757713361996e-05,
"loss": 0.0096,
"step": 7770
},
{
"epoch": 14.407407407407407,
"grad_norm": 0.11189185082912445,
"learning_rate": 1.2890740309097204e-05,
"loss": 0.0131,
"step": 7780
},
{
"epoch": 14.425925925925926,
"grad_norm": 0.15256240963935852,
"learning_rate": 1.2780128725750944e-05,
"loss": 0.0107,
"step": 7790
},
{
"epoch": 14.444444444444445,
"grad_norm": 0.16773676872253418,
"learning_rate": 1.266992417295379e-05,
"loss": 0.0087,
"step": 7800
},
{
"epoch": 14.462962962962964,
"grad_norm": 0.15099745988845825,
"learning_rate": 1.2560127855885073e-05,
"loss": 0.01,
"step": 7810
},
{
"epoch": 14.481481481481481,
"grad_norm": 0.166538804769516,
"learning_rate": 1.2450740975259745e-05,
"loss": 0.0096,
"step": 7820
},
{
"epoch": 14.5,
"grad_norm": 0.14279592037200928,
"learning_rate": 1.234176472731517e-05,
"loss": 0.0074,
"step": 7830
},
{
"epoch": 14.518518518518519,
"grad_norm": 0.19192756712436676,
"learning_rate": 1.2233200303798158e-05,
"loss": 0.0115,
"step": 7840
},
{
"epoch": 14.537037037037036,
"grad_norm": 0.15167966485023499,
"learning_rate": 1.2125048891951846e-05,
"loss": 0.0106,
"step": 7850
},
{
"epoch": 14.555555555555555,
"grad_norm": 0.1554906815290451,
"learning_rate": 1.2017311674502745e-05,
"loss": 0.0098,
"step": 7860
},
{
"epoch": 14.574074074074074,
"grad_norm": 0.18407610058784485,
"learning_rate": 1.1909989829647822e-05,
"loss": 0.0102,
"step": 7870
},
{
"epoch": 14.592592592592592,
"grad_norm": 0.10603850334882736,
"learning_rate": 1.1803084531041553e-05,
"loss": 0.0096,
"step": 7880
},
{
"epoch": 14.61111111111111,
"grad_norm": 0.12585797905921936,
"learning_rate": 1.1696596947783162e-05,
"loss": 0.0102,
"step": 7890
},
{
"epoch": 14.62962962962963,
"grad_norm": 0.19426830112934113,
"learning_rate": 1.1590528244403803e-05,
"loss": 0.0101,
"step": 7900
},
{
"epoch": 14.648148148148149,
"grad_norm": 0.1230105459690094,
"learning_rate": 1.148487958085382e-05,
"loss": 0.0104,
"step": 7910
},
{
"epoch": 14.666666666666666,
"grad_norm": 0.14580340683460236,
"learning_rate": 1.1379652112490086e-05,
"loss": 0.0104,
"step": 7920
},
{
"epoch": 14.685185185185185,
"grad_norm": 0.13718611001968384,
"learning_rate": 1.1274846990063315e-05,
"loss": 0.0098,
"step": 7930
},
{
"epoch": 14.703703703703704,
"grad_norm": 0.15364445745944977,
"learning_rate": 1.117046535970554e-05,
"loss": 0.0096,
"step": 7940
},
{
"epoch": 14.722222222222221,
"grad_norm": 0.14727716147899628,
"learning_rate": 1.106650836291755e-05,
"loss": 0.0087,
"step": 7950
},
{
"epoch": 14.74074074074074,
"grad_norm": 0.14772866666316986,
"learning_rate": 1.0962977136556418e-05,
"loss": 0.0105,
"step": 7960
},
{
"epoch": 14.75925925925926,
"grad_norm": 0.16473093628883362,
"learning_rate": 1.0859872812823024e-05,
"loss": 0.0118,
"step": 7970
},
{
"epoch": 14.777777777777779,
"grad_norm": 0.10741712898015976,
"learning_rate": 1.0757196519249747e-05,
"loss": 0.0108,
"step": 7980
},
{
"epoch": 14.796296296296296,
"grad_norm": 0.17428821325302124,
"learning_rate": 1.0654949378688077e-05,
"loss": 0.0108,
"step": 7990
},
{
"epoch": 14.814814814814815,
"grad_norm": 0.17756251990795135,
"learning_rate": 1.0553132509296376e-05,
"loss": 0.0106,
"step": 8000
},
{
"epoch": 14.833333333333334,
"grad_norm": 0.11251235008239746,
"learning_rate": 1.0451747024527613e-05,
"loss": 0.0092,
"step": 8010
},
{
"epoch": 14.851851851851851,
"grad_norm": 0.07490465044975281,
"learning_rate": 1.0350794033117189e-05,
"loss": 0.0078,
"step": 8020
},
{
"epoch": 14.87037037037037,
"grad_norm": 0.09541039168834686,
"learning_rate": 1.0250274639070856e-05,
"loss": 0.0082,
"step": 8030
},
{
"epoch": 14.88888888888889,
"grad_norm": 0.15332746505737305,
"learning_rate": 1.0150189941652599e-05,
"loss": 0.0099,
"step": 8040
},
{
"epoch": 14.907407407407408,
"grad_norm": 0.148310586810112,
"learning_rate": 1.0050541035372635e-05,
"loss": 0.01,
"step": 8050
},
{
"epoch": 14.925925925925926,
"grad_norm": 0.11974785476922989,
"learning_rate": 9.951329009975458e-06,
"loss": 0.0094,
"step": 8060
},
{
"epoch": 14.944444444444445,
"grad_norm": 0.12376372516155243,
"learning_rate": 9.852554950427845e-06,
"loss": 0.0088,
"step": 8070
},
{
"epoch": 14.962962962962964,
"grad_norm": 0.1553984433412552,
"learning_rate": 9.754219936907105e-06,
"loss": 0.0093,
"step": 8080
},
{
"epoch": 14.981481481481481,
"grad_norm": 0.10422512888908386,
"learning_rate": 9.656325044789194e-06,
"loss": 0.0102,
"step": 8090
},
{
"epoch": 15.0,
"grad_norm": 0.3811832368373871,
"learning_rate": 9.55887134463697e-06,
"loss": 0.0092,
"step": 8100
},
{
"epoch": 15.018518518518519,
"grad_norm": 0.14635023474693298,
"learning_rate": 9.461859902188475e-06,
"loss": 0.011,
"step": 8110
},
{
"epoch": 15.037037037037036,
"grad_norm": 0.08813057094812393,
"learning_rate": 9.365291778345303e-06,
"loss": 0.0083,
"step": 8120
},
{
"epoch": 15.055555555555555,
"grad_norm": 0.14506255090236664,
"learning_rate": 9.269168029160991e-06,
"loss": 0.0096,
"step": 8130
},
{
"epoch": 15.074074074074074,
"grad_norm": 0.1673993617296219,
"learning_rate": 9.173489705829447e-06,
"loss": 0.0096,
"step": 8140
},
{
"epoch": 15.092592592592593,
"grad_norm": 0.10321272164583206,
"learning_rate": 9.078257854673516e-06,
"loss": 0.0085,
"step": 8150
},
{
"epoch": 15.11111111111111,
"grad_norm": 0.13575561344623566,
"learning_rate": 8.983473517133429e-06,
"loss": 0.0106,
"step": 8160
},
{
"epoch": 15.12962962962963,
"grad_norm": 0.12959037721157074,
"learning_rate": 8.889137729755537e-06,
"loss": 0.0094,
"step": 8170
},
{
"epoch": 15.148148148148149,
"grad_norm": 0.15763111412525177,
"learning_rate": 8.79525152418087e-06,
"loss": 0.0086,
"step": 8180
},
{
"epoch": 15.166666666666666,
"grad_norm": 0.11049918830394745,
"learning_rate": 8.701815927133961e-06,
"loss": 0.0083,
"step": 8190
},
{
"epoch": 15.185185185185185,
"grad_norm": 0.102789968252182,
"learning_rate": 8.608831960411534e-06,
"loss": 0.0096,
"step": 8200
},
{
"epoch": 15.203703703703704,
"grad_norm": 0.1005704402923584,
"learning_rate": 8.516300640871321e-06,
"loss": 0.0101,
"step": 8210
},
{
"epoch": 15.222222222222221,
"grad_norm": 0.1040162593126297,
"learning_rate": 8.424222980421038e-06,
"loss": 0.0095,
"step": 8220
},
{
"epoch": 15.24074074074074,
"grad_norm": 0.1338852196931839,
"learning_rate": 8.332599986007184e-06,
"loss": 0.0078,
"step": 8230
},
{
"epoch": 15.25925925925926,
"grad_norm": 0.1555245965719223,
"learning_rate": 8.241432659604203e-06,
"loss": 0.0074,
"step": 8240
},
{
"epoch": 15.277777777777779,
"grad_norm": 0.12137534469366074,
"learning_rate": 8.150721998203331e-06,
"loss": 0.0089,
"step": 8250
},
{
"epoch": 15.296296296296296,
"grad_norm": 0.19541344046592712,
"learning_rate": 8.06046899380184e-06,
"loss": 0.0108,
"step": 8260
},
{
"epoch": 15.314814814814815,
"grad_norm": 0.12632016837596893,
"learning_rate": 7.970674633392133e-06,
"loss": 0.0081,
"step": 8270
},
{
"epoch": 15.333333333333334,
"grad_norm": 0.11312908679246902,
"learning_rate": 7.881339898950924e-06,
"loss": 0.0071,
"step": 8280
},
{
"epoch": 15.351851851851851,
"grad_norm": 0.1144496351480484,
"learning_rate": 7.792465767428597e-06,
"loss": 0.0087,
"step": 8290
},
{
"epoch": 15.37037037037037,
"grad_norm": 0.12083644419908524,
"learning_rate": 7.704053210738376e-06,
"loss": 0.008,
"step": 8300
},
{
"epoch": 15.38888888888889,
"grad_norm": 0.16639256477355957,
"learning_rate": 7.6161031957458494e-06,
"loss": 0.0084,
"step": 8310
},
{
"epoch": 15.407407407407407,
"grad_norm": 0.12301436066627502,
"learning_rate": 7.5286166842582605e-06,
"loss": 0.0094,
"step": 8320
},
{
"epoch": 15.425925925925926,
"grad_norm": 0.10282429307699203,
"learning_rate": 7.4415946330140814e-06,
"loss": 0.0095,
"step": 8330
},
{
"epoch": 15.444444444444445,
"grad_norm": 0.15487389266490936,
"learning_rate": 7.3550379936725644e-06,
"loss": 0.0085,
"step": 8340
},
{
"epoch": 15.462962962962964,
"grad_norm": 0.11302638053894043,
"learning_rate": 7.2689477128032035e-06,
"loss": 0.0089,
"step": 8350
},
{
"epoch": 15.481481481481481,
"grad_norm": 0.1661285012960434,
"learning_rate": 7.183324731875551e-06,
"loss": 0.008,
"step": 8360
},
{
"epoch": 15.5,
"grad_norm": 0.10385741293430328,
"learning_rate": 7.098169987248782e-06,
"loss": 0.0093,
"step": 8370
},
{
"epoch": 15.518518518518519,
"grad_norm": 0.11806198209524155,
"learning_rate": 7.013484410161553e-06,
"loss": 0.0093,
"step": 8380
},
{
"epoch": 15.537037037037036,
"grad_norm": 0.1038653776049614,
"learning_rate": 6.92926892672176e-06,
"loss": 0.0077,
"step": 8390
},
{
"epoch": 15.555555555555555,
"grad_norm": 0.12973074615001678,
"learning_rate": 6.845524457896446e-06,
"loss": 0.0085,
"step": 8400
},
{
"epoch": 15.574074074074074,
"grad_norm": 0.10546618700027466,
"learning_rate": 6.7622519195017165e-06,
"loss": 0.0085,
"step": 8410
},
{
"epoch": 15.592592592592592,
"grad_norm": 0.12728683650493622,
"learning_rate": 6.679452222192684e-06,
"loss": 0.0107,
"step": 8420
},
{
"epoch": 15.61111111111111,
"grad_norm": 0.1038224920630455,
"learning_rate": 6.597126271453579e-06,
"loss": 0.0082,
"step": 8430
},
{
"epoch": 15.62962962962963,
"grad_norm": 0.11956731975078583,
"learning_rate": 6.51527496758782e-06,
"loss": 0.0085,
"step": 8440
},
{
"epoch": 15.648148148148149,
"grad_norm": 0.19235482811927795,
"learning_rate": 6.433899205708155e-06,
"loss": 0.008,
"step": 8450
},
{
"epoch": 15.666666666666666,
"grad_norm": 0.15918850898742676,
"learning_rate": 6.352999875726856e-06,
"loss": 0.012,
"step": 8460
},
{
"epoch": 15.685185185185185,
"grad_norm": 0.09061280637979507,
"learning_rate": 6.272577862346052e-06,
"loss": 0.0091,
"step": 8470
},
{
"epoch": 15.703703703703704,
"grad_norm": 0.15757571160793304,
"learning_rate": 6.192634045047996e-06,
"loss": 0.0087,
"step": 8480
},
{
"epoch": 15.722222222222221,
"grad_norm": 0.13536228239536285,
"learning_rate": 6.113169298085458e-06,
"loss": 0.0095,
"step": 8490
},
{
"epoch": 15.74074074074074,
"grad_norm": 0.12651894986629486,
"learning_rate": 6.034184490472195e-06,
"loss": 0.0106,
"step": 8500
},
{
"epoch": 15.75925925925926,
"grad_norm": 0.16352827847003937,
"learning_rate": 5.955680485973386e-06,
"loss": 0.0096,
"step": 8510
},
{
"epoch": 15.777777777777779,
"grad_norm": 0.12227349728345871,
"learning_rate": 5.877658143096265e-06,
"loss": 0.0086,
"step": 8520
},
{
"epoch": 15.796296296296296,
"grad_norm": 0.11193786561489105,
"learning_rate": 5.800118315080661e-06,
"loss": 0.0084,
"step": 8530
},
{
"epoch": 15.814814814814815,
"grad_norm": 0.08987289667129517,
"learning_rate": 5.723061849889716e-06,
"loss": 0.0077,
"step": 8540
},
{
"epoch": 15.833333333333334,
"grad_norm": 0.14479348063468933,
"learning_rate": 5.646489590200604e-06,
"loss": 0.0103,
"step": 8550
},
{
"epoch": 15.851851851851851,
"grad_norm": 0.12218350917100906,
"learning_rate": 5.570402373395256e-06,
"loss": 0.0073,
"step": 8560
},
{
"epoch": 15.87037037037037,
"grad_norm": 0.1457684338092804,
"learning_rate": 5.494801031551305e-06,
"loss": 0.0099,
"step": 8570
},
{
"epoch": 15.88888888888889,
"grad_norm": 0.12631197273731232,
"learning_rate": 5.41968639143291e-06,
"loss": 0.0104,
"step": 8580
},
{
"epoch": 15.907407407407408,
"grad_norm": 0.1562114655971527,
"learning_rate": 5.345059274481751e-06,
"loss": 0.0102,
"step": 8590
},
{
"epoch": 15.925925925925926,
"grad_norm": 0.10125651955604553,
"learning_rate": 5.270920496808002e-06,
"loss": 0.0075,
"step": 8600
},
{
"epoch": 15.944444444444445,
"grad_norm": 0.10967680811882019,
"learning_rate": 5.1972708691814695e-06,
"loss": 0.0079,
"step": 8610
},
{
"epoch": 15.962962962962964,
"grad_norm": 0.16501936316490173,
"learning_rate": 5.124111197022674e-06,
"loss": 0.0096,
"step": 8620
},
{
"epoch": 15.981481481481481,
"grad_norm": 0.08875839412212372,
"learning_rate": 5.051442280394081e-06,
"loss": 0.0082,
"step": 8630
},
{
"epoch": 16.0,
"grad_norm": 0.6014512181282043,
"learning_rate": 4.979264913991322e-06,
"loss": 0.0101,
"step": 8640
},
{
"epoch": 16.01851851851852,
"grad_norm": 0.09958849102258682,
"learning_rate": 4.907579887134489e-06,
"loss": 0.0085,
"step": 8650
},
{
"epoch": 16.037037037037038,
"grad_norm": 0.120966337621212,
"learning_rate": 4.836387983759572e-06,
"loss": 0.0087,
"step": 8660
},
{
"epoch": 16.055555555555557,
"grad_norm": 0.13525190949440002,
"learning_rate": 4.765689982409816e-06,
"loss": 0.0086,
"step": 8670
},
{
"epoch": 16.074074074074073,
"grad_norm": 0.1979486495256424,
"learning_rate": 4.695486656227233e-06,
"loss": 0.0089,
"step": 8680
},
{
"epoch": 16.09259259259259,
"grad_norm": 0.09045485407114029,
"learning_rate": 4.625778772944156e-06,
"loss": 0.0096,
"step": 8690
},
{
"epoch": 16.11111111111111,
"grad_norm": 0.11300152540206909,
"learning_rate": 4.556567094874825e-06,
"loss": 0.0083,
"step": 8700
},
{
"epoch": 16.12962962962963,
"grad_norm": 0.08737247437238693,
"learning_rate": 4.487852378907059e-06,
"loss": 0.0066,
"step": 8710
},
{
"epoch": 16.14814814814815,
"grad_norm": 0.18659187853336334,
"learning_rate": 4.419635376493986e-06,
"loss": 0.0076,
"step": 8720
},
{
"epoch": 16.166666666666668,
"grad_norm": 0.13316965103149414,
"learning_rate": 4.351916833645825e-06,
"loss": 0.0104,
"step": 8730
},
{
"epoch": 16.185185185185187,
"grad_norm": 0.11005474627017975,
"learning_rate": 4.284697490921691e-06,
"loss": 0.0072,
"step": 8740
},
{
"epoch": 16.203703703703702,
"grad_norm": 0.11568531394004822,
"learning_rate": 4.2179780834215585e-06,
"loss": 0.009,
"step": 8750
},
{
"epoch": 16.22222222222222,
"grad_norm": 0.12146589159965515,
"learning_rate": 4.151759340778178e-06,
"loss": 0.0082,
"step": 8760
},
{
"epoch": 16.24074074074074,
"grad_norm": 0.08866602182388306,
"learning_rate": 4.086041987149109e-06,
"loss": 0.0064,
"step": 8770
},
{
"epoch": 16.25925925925926,
"grad_norm": 0.17930375039577484,
"learning_rate": 4.020826741208811e-06,
"loss": 0.0093,
"step": 8780
},
{
"epoch": 16.27777777777778,
"grad_norm": 0.12084779888391495,
"learning_rate": 3.956114316140746e-06,
"loss": 0.0106,
"step": 8790
},
{
"epoch": 16.296296296296298,
"grad_norm": 0.10893933475017548,
"learning_rate": 3.891905419629643e-06,
"loss": 0.0085,
"step": 8800
},
{
"epoch": 16.314814814814813,
"grad_norm": 0.0690929964184761,
"learning_rate": 3.8282007538536946e-06,
"loss": 0.0085,
"step": 8810
},
{
"epoch": 16.333333333333332,
"grad_norm": 0.0713321641087532,
"learning_rate": 3.7650010154769265e-06,
"loss": 0.0071,
"step": 8820
},
{
"epoch": 16.35185185185185,
"grad_norm": 0.12631797790527344,
"learning_rate": 3.7023068956415608e-06,
"loss": 0.0088,
"step": 8830
},
{
"epoch": 16.37037037037037,
"grad_norm": 0.12863200902938843,
"learning_rate": 3.6401190799604303e-06,
"loss": 0.0092,
"step": 8840
},
{
"epoch": 16.38888888888889,
"grad_norm": 0.069121353328228,
"learning_rate": 3.578438248509536e-06,
"loss": 0.0105,
"step": 8850
},
{
"epoch": 16.40740740740741,
"grad_norm": 0.08070623874664307,
"learning_rate": 3.5172650758205583e-06,
"loss": 0.007,
"step": 8860
},
{
"epoch": 16.425925925925927,
"grad_norm": 0.13333022594451904,
"learning_rate": 3.45660023087353e-06,
"loss": 0.0102,
"step": 8870
},
{
"epoch": 16.444444444444443,
"grad_norm": 0.15298254787921906,
"learning_rate": 3.3964443770894528e-06,
"loss": 0.0094,
"step": 8880
},
{
"epoch": 16.462962962962962,
"grad_norm": 0.13243365287780762,
"learning_rate": 3.3367981723231245e-06,
"loss": 0.0093,
"step": 8890
},
{
"epoch": 16.48148148148148,
"grad_norm": 0.12942367792129517,
"learning_rate": 3.2776622688558746e-06,
"loss": 0.0089,
"step": 8900
},
{
"epoch": 16.5,
"grad_norm": 0.06228404864668846,
"learning_rate": 3.2190373133884677e-06,
"loss": 0.0077,
"step": 8910
},
{
"epoch": 16.51851851851852,
"grad_norm": 0.14292901754379272,
"learning_rate": 3.1609239470340446e-06,
"loss": 0.0108,
"step": 8920
},
{
"epoch": 16.537037037037038,
"grad_norm": 0.09717714786529541,
"learning_rate": 3.1033228053110373e-06,
"loss": 0.0069,
"step": 8930
},
{
"epoch": 16.555555555555557,
"grad_norm": 0.11619598418474197,
"learning_rate": 3.0462345181363314e-06,
"loss": 0.0132,
"step": 8940
},
{
"epoch": 16.574074074074073,
"grad_norm": 0.13606691360473633,
"learning_rate": 2.9896597098182654e-06,
"loss": 0.0074,
"step": 8950
},
{
"epoch": 16.59259259259259,
"grad_norm": 0.09636996686458588,
"learning_rate": 2.933598999049891e-06,
"loss": 0.0082,
"step": 8960
},
{
"epoch": 16.61111111111111,
"grad_norm": 0.10872022062540054,
"learning_rate": 2.8780529989021697e-06,
"loss": 0.0083,
"step": 8970
},
{
"epoch": 16.62962962962963,
"grad_norm": 0.1134805902838707,
"learning_rate": 2.823022316817242e-06,
"loss": 0.0099,
"step": 8980
},
{
"epoch": 16.64814814814815,
"grad_norm": 0.10475319623947144,
"learning_rate": 2.7685075546018456e-06,
"loss": 0.0082,
"step": 8990
},
{
"epoch": 16.666666666666668,
"grad_norm": 0.143001526594162,
"learning_rate": 2.7145093084206598e-06,
"loss": 0.0125,
"step": 9000
},
{
"epoch": 16.685185185185187,
"grad_norm": 0.11511417478322983,
"learning_rate": 2.661028168789892e-06,
"loss": 0.0088,
"step": 9010
},
{
"epoch": 16.703703703703702,
"grad_norm": 0.14228394627571106,
"learning_rate": 2.6080647205706855e-06,
"loss": 0.0095,
"step": 9020
},
{
"epoch": 16.72222222222222,
"grad_norm": 0.09285625070333481,
"learning_rate": 2.555619542962834e-06,
"loss": 0.0104,
"step": 9030
},
{
"epoch": 16.74074074074074,
"grad_norm": 0.08225535601377487,
"learning_rate": 2.503693209498409e-06,
"loss": 0.0111,
"step": 9040
},
{
"epoch": 16.75925925925926,
"grad_norm": 0.09045220911502838,
"learning_rate": 2.452286288035449e-06,
"loss": 0.0086,
"step": 9050
},
{
"epoch": 16.77777777777778,
"grad_norm": 0.10882551223039627,
"learning_rate": 2.4013993407518363e-06,
"loss": 0.0071,
"step": 9060
},
{
"epoch": 16.796296296296298,
"grad_norm": 0.1406363844871521,
"learning_rate": 2.351032924139063e-06,
"loss": 0.0077,
"step": 9070
},
{
"epoch": 16.814814814814813,
"grad_norm": 0.10963733494281769,
"learning_rate": 2.30118758899619e-06,
"loss": 0.0068,
"step": 9080
},
{
"epoch": 16.833333333333332,
"grad_norm": 0.08778999000787735,
"learning_rate": 2.2518638804238157e-06,
"loss": 0.0083,
"step": 9090
},
{
"epoch": 16.85185185185185,
"grad_norm": 0.12174137681722641,
"learning_rate": 2.203062337818118e-06,
"loss": 0.0098,
"step": 9100
},
{
"epoch": 16.87037037037037,
"grad_norm": 0.14932076632976532,
"learning_rate": 2.1547834948649483e-06,
"loss": 0.0098,
"step": 9110
},
{
"epoch": 16.88888888888889,
"grad_norm": 0.11017399281263351,
"learning_rate": 2.1070278795340017e-06,
"loss": 0.0062,
"step": 9120
},
{
"epoch": 16.90740740740741,
"grad_norm": 0.11405317485332489,
"learning_rate": 2.059796014073029e-06,
"loss": 0.0087,
"step": 9130
},
{
"epoch": 16.925925925925927,
"grad_norm": 0.17818747460842133,
"learning_rate": 2.01308841500214e-06,
"loss": 0.0092,
"step": 9140
},
{
"epoch": 16.944444444444443,
"grad_norm": 0.09499002248048782,
"learning_rate": 1.9669055931081704e-06,
"loss": 0.0074,
"step": 9150
},
{
"epoch": 16.962962962962962,
"grad_norm": 0.12164697796106339,
"learning_rate": 1.9212480534390507e-06,
"loss": 0.0078,
"step": 9160
},
{
"epoch": 16.98148148148148,
"grad_norm": 0.13032789528369904,
"learning_rate": 1.8761162952983246e-06,
"loss": 0.0095,
"step": 9170
},
{
"epoch": 17.0,
"grad_norm": 0.8052965998649597,
"learning_rate": 1.8315108122396618e-06,
"loss": 0.0066,
"step": 9180
},
{
"epoch": 17.01851851851852,
"grad_norm": 0.15812717378139496,
"learning_rate": 1.787432092061475e-06,
"loss": 0.0079,
"step": 9190
},
{
"epoch": 17.037037037037038,
"grad_norm": 0.17003348469734192,
"learning_rate": 1.743880616801602e-06,
"loss": 0.008,
"step": 9200
},
{
"epoch": 17.055555555555557,
"grad_norm": 0.13849136233329773,
"learning_rate": 1.7008568627319865e-06,
"loss": 0.0093,
"step": 9210
},
{
"epoch": 17.074074074074073,
"grad_norm": 0.20640458166599274,
"learning_rate": 1.6583613003535226e-06,
"loss": 0.0106,
"step": 9220
},
{
"epoch": 17.09259259259259,
"grad_norm": 0.1231725886464119,
"learning_rate": 1.6163943943908522e-06,
"loss": 0.0085,
"step": 9230
},
{
"epoch": 17.11111111111111,
"grad_norm": 0.1270870417356491,
"learning_rate": 1.5749566037873476e-06,
"loss": 0.0072,
"step": 9240
},
{
"epoch": 17.12962962962963,
"grad_norm": 0.11818239092826843,
"learning_rate": 1.5340483817000428e-06,
"loss": 0.0082,
"step": 9250
},
{
"epoch": 17.14814814814815,
"grad_norm": 0.11305985599756241,
"learning_rate": 1.4936701754947101e-06,
"loss": 0.0073,
"step": 9260
},
{
"epoch": 17.166666666666668,
"grad_norm": 0.09903740882873535,
"learning_rate": 1.4538224267409361e-06,
"loss": 0.0085,
"step": 9270
},
{
"epoch": 17.185185185185187,
"grad_norm": 0.11844265460968018,
"learning_rate": 1.414505571207314e-06,
"loss": 0.0099,
"step": 9280
},
{
"epoch": 17.203703703703702,
"grad_norm": 0.07644011080265045,
"learning_rate": 1.3757200388566816e-06,
"loss": 0.0113,
"step": 9290
},
{
"epoch": 17.22222222222222,
"grad_norm": 0.11540202796459198,
"learning_rate": 1.3374662538414074e-06,
"loss": 0.0098,
"step": 9300
},
{
"epoch": 17.24074074074074,
"grad_norm": 0.10887469351291656,
"learning_rate": 1.2997446344987617e-06,
"loss": 0.0078,
"step": 9310
},
{
"epoch": 17.25925925925926,
"grad_norm": 0.11233421415090561,
"learning_rate": 1.262555593346315e-06,
"loss": 0.0088,
"step": 9320
},
{
"epoch": 17.27777777777778,
"grad_norm": 0.1288515031337738,
"learning_rate": 1.2258995370774685e-06,
"loss": 0.0075,
"step": 9330
},
{
"epoch": 17.296296296296298,
"grad_norm": 0.0982125997543335,
"learning_rate": 1.1897768665569798e-06,
"loss": 0.0075,
"step": 9340
},
{
"epoch": 17.314814814814813,
"grad_norm": 0.13572707772254944,
"learning_rate": 1.1541879768165954e-06,
"loss": 0.0097,
"step": 9350
},
{
"epoch": 17.333333333333332,
"grad_norm": 0.14079776406288147,
"learning_rate": 1.1191332570507085e-06,
"loss": 0.0101,
"step": 9360
},
{
"epoch": 17.35185185185185,
"grad_norm": 0.12812009453773499,
"learning_rate": 1.0846130906121132e-06,
"loss": 0.0097,
"step": 9370
},
{
"epoch": 17.37037037037037,
"grad_norm": 0.12242446094751358,
"learning_rate": 1.0506278550078131e-06,
"loss": 0.0067,
"step": 9380
},
{
"epoch": 17.38888888888889,
"grad_norm": 0.1234826147556305,
"learning_rate": 1.0171779218949185e-06,
"loss": 0.0076,
"step": 9390
},
{
"epoch": 17.40740740740741,
"grad_norm": 0.10158117860555649,
"learning_rate": 9.842636570765174e-07,
"loss": 0.0066,
"step": 9400
},
{
"epoch": 17.425925925925927,
"grad_norm": 0.100815050303936,
"learning_rate": 9.518854204977612e-07,
"loss": 0.0084,
"step": 9410
},
{
"epoch": 17.444444444444443,
"grad_norm": 0.12093641608953476,
"learning_rate": 9.200435662418349e-07,
"loss": 0.0096,
"step": 9420
},
{
"epoch": 17.462962962962962,
"grad_norm": 0.08097100257873535,
"learning_rate": 8.887384425261658e-07,
"loss": 0.0074,
"step": 9430
},
{
"epoch": 17.48148148148148,
"grad_norm": 0.12034954130649567,
"learning_rate": 8.579703916985648e-07,
"loss": 0.0073,
"step": 9440
},
{
"epoch": 17.5,
"grad_norm": 0.1560267210006714,
"learning_rate": 8.277397502335194e-07,
"loss": 0.0072,
"step": 9450
},
{
"epoch": 17.51851851851852,
"grad_norm": 0.1436459869146347,
"learning_rate": 7.980468487284675e-07,
"loss": 0.0078,
"step": 9460
},
{
"epoch": 17.537037037037038,
"grad_norm": 0.07840181142091751,
"learning_rate": 7.688920119002297e-07,
"loss": 0.0074,
"step": 9470
},
{
"epoch": 17.555555555555557,
"grad_norm": 0.092781201004982,
"learning_rate": 7.402755585814269e-07,
"loss": 0.0104,
"step": 9480
},
{
"epoch": 17.574074074074073,
"grad_norm": 0.0861596092581749,
"learning_rate": 7.121978017170073e-07,
"loss": 0.0072,
"step": 9490
},
{
"epoch": 17.59259259259259,
"grad_norm": 0.1299380660057068,
"learning_rate": 6.846590483608306e-07,
"loss": 0.0076,
"step": 9500
},
{
"epoch": 17.61111111111111,
"grad_norm": 0.05626577138900757,
"learning_rate": 6.576595996722834e-07,
"loss": 0.0059,
"step": 9510
},
{
"epoch": 17.62962962962963,
"grad_norm": 0.09908762574195862,
"learning_rate": 6.311997509130141e-07,
"loss": 0.006,
"step": 9520
},
{
"epoch": 17.64814814814815,
"grad_norm": 0.12680689990520477,
"learning_rate": 6.052797914436803e-07,
"loss": 0.0092,
"step": 9530
},
{
"epoch": 17.666666666666668,
"grad_norm": 0.08880678564310074,
"learning_rate": 5.799000047208181e-07,
"loss": 0.0067,
"step": 9540
},
{
"epoch": 17.685185185185187,
"grad_norm": 0.08860208839178085,
"learning_rate": 5.550606682937054e-07,
"loss": 0.0058,
"step": 9550
},
{
"epoch": 17.703703703703702,
"grad_norm": 0.10537762939929962,
"learning_rate": 5.307620538013481e-07,
"loss": 0.0077,
"step": 9560
},
{
"epoch": 17.72222222222222,
"grad_norm": 0.09559914469718933,
"learning_rate": 5.070044269694874e-07,
"loss": 0.0075,
"step": 9570
},
{
"epoch": 17.74074074074074,
"grad_norm": 0.12621325254440308,
"learning_rate": 4.837880476077417e-07,
"loss": 0.0082,
"step": 9580
},
{
"epoch": 17.75925925925926,
"grad_norm": 0.14380834996700287,
"learning_rate": 4.6111316960670835e-07,
"loss": 0.0084,
"step": 9590
},
{
"epoch": 17.77777777777778,
"grad_norm": 0.09919670224189758,
"learning_rate": 4.389800409352218e-07,
"loss": 0.0099,
"step": 9600
},
{
"epoch": 17.796296296296298,
"grad_norm": 0.08830275386571884,
"learning_rate": 4.173889036376277e-07,
"loss": 0.0073,
"step": 9610
},
{
"epoch": 17.814814814814813,
"grad_norm": 0.1367911547422409,
"learning_rate": 3.963399938311463e-07,
"loss": 0.0092,
"step": 9620
},
{
"epoch": 17.833333333333332,
"grad_norm": 0.15709251165390015,
"learning_rate": 3.7583354170328545e-07,
"loss": 0.0075,
"step": 9630
},
{
"epoch": 17.85185185185185,
"grad_norm": 0.138651505112648,
"learning_rate": 3.558697715093207e-07,
"loss": 0.0081,
"step": 9640
},
{
"epoch": 17.87037037037037,
"grad_norm": 0.09270890057086945,
"learning_rate": 3.3644890156983576e-07,
"loss": 0.0073,
"step": 9650
},
{
"epoch": 17.88888888888889,
"grad_norm": 0.12990514934062958,
"learning_rate": 3.175711442683638e-07,
"loss": 0.0079,
"step": 9660
},
{
"epoch": 17.90740740740741,
"grad_norm": 0.13194292783737183,
"learning_rate": 2.9923670604902197e-07,
"loss": 0.0075,
"step": 9670
},
{
"epoch": 17.925925925925927,
"grad_norm": 0.0941421240568161,
"learning_rate": 2.814457874143028e-07,
"loss": 0.0076,
"step": 9680
},
{
"epoch": 17.944444444444443,
"grad_norm": 0.11929167062044144,
"learning_rate": 2.641985829228366e-07,
"loss": 0.0084,
"step": 9690
},
{
"epoch": 17.962962962962962,
"grad_norm": 0.11601762473583221,
"learning_rate": 2.474952811872877e-07,
"loss": 0.0076,
"step": 9700
},
{
"epoch": 17.98148148148148,
"grad_norm": 0.10628051310777664,
"learning_rate": 2.3133606487228397e-07,
"loss": 0.0083,
"step": 9710
},
{
"epoch": 18.0,
"grad_norm": 0.41246920824050903,
"learning_rate": 2.157211106924295e-07,
"loss": 0.0075,
"step": 9720
},
{
"epoch": 18.01851851851852,
"grad_norm": 0.16486990451812744,
"learning_rate": 2.006505894103672e-07,
"loss": 0.0083,
"step": 9730
},
{
"epoch": 18.037037037037038,
"grad_norm": 0.1112591102719307,
"learning_rate": 1.8612466583489696e-07,
"loss": 0.0085,
"step": 9740
},
{
"epoch": 18.055555555555557,
"grad_norm": 0.13400018215179443,
"learning_rate": 1.7214349881918834e-07,
"loss": 0.007,
"step": 9750
},
{
"epoch": 18.074074074074073,
"grad_norm": 0.13410645723342896,
"learning_rate": 1.5870724125904845e-07,
"loss": 0.0119,
"step": 9760
},
{
"epoch": 18.09259259259259,
"grad_norm": 0.06591040641069412,
"learning_rate": 1.4581604009124006e-07,
"loss": 0.0079,
"step": 9770
},
{
"epoch": 18.11111111111111,
"grad_norm": 0.14793242514133453,
"learning_rate": 1.334700362918717e-07,
"loss": 0.0087,
"step": 9780
},
{
"epoch": 18.12962962962963,
"grad_norm": 0.08889105916023254,
"learning_rate": 1.2166936487486015e-07,
"loss": 0.0067,
"step": 9790
},
{
"epoch": 18.14814814814815,
"grad_norm": 0.16043215990066528,
"learning_rate": 1.1041415489045914e-07,
"loss": 0.0113,
"step": 9800
},
{
"epoch": 18.166666666666668,
"grad_norm": 0.10628201812505722,
"learning_rate": 9.970452942384412e-08,
"loss": 0.0074,
"step": 9810
},
{
"epoch": 18.185185185185187,
"grad_norm": 0.07550280541181564,
"learning_rate": 8.954060559375754e-08,
"loss": 0.0073,
"step": 9820
},
{
"epoch": 18.203703703703702,
"grad_norm": 0.12803789973258972,
"learning_rate": 7.99224945512489e-08,
"loss": 0.008,
"step": 9830
},
{
"epoch": 18.22222222222222,
"grad_norm": 0.09466207027435303,
"learning_rate": 7.085030147843675e-08,
"loss": 0.0057,
"step": 9840
},
{
"epoch": 18.24074074074074,
"grad_norm": 0.17841790616512299,
"learning_rate": 6.232412558736523e-08,
"loss": 0.0093,
"step": 9850
},
{
"epoch": 18.25925925925926,
"grad_norm": 0.10296358913183212,
"learning_rate": 5.434406011893822e-08,
"loss": 0.0071,
"step": 9860
},
{
"epoch": 18.27777777777778,
"grad_norm": 0.16694875061511993,
"learning_rate": 4.6910192341864664e-08,
"loss": 0.0099,
"step": 9870
},
{
"epoch": 18.296296296296298,
"grad_norm": 0.20569169521331787,
"learning_rate": 4.0022603551737035e-08,
"loss": 0.0105,
"step": 9880
},
{
"epoch": 18.314814814814813,
"grad_norm": 0.18344692885875702,
"learning_rate": 3.3681369070120985e-08,
"loss": 0.0091,
"step": 9890
},
{
"epoch": 18.333333333333332,
"grad_norm": 0.11761084944009781,
"learning_rate": 2.7886558243744866e-08,
"loss": 0.0071,
"step": 9900
},
{
"epoch": 18.35185185185185,
"grad_norm": 0.0920497253537178,
"learning_rate": 2.2638234443722596e-08,
"loss": 0.0081,
"step": 9910
},
{
"epoch": 18.37037037037037,
"grad_norm": 0.0982198640704155,
"learning_rate": 1.7936455064887504e-08,
"loss": 0.01,
"step": 9920
},
{
"epoch": 18.38888888888889,
"grad_norm": 0.11440613120794296,
"learning_rate": 1.378127152514841e-08,
"loss": 0.0091,
"step": 9930
},
{
"epoch": 18.40740740740741,
"grad_norm": 0.12751302123069763,
"learning_rate": 1.0172729264917857e-08,
"loss": 0.0089,
"step": 9940
},
{
"epoch": 18.425925925925927,
"grad_norm": 0.1360410451889038,
"learning_rate": 7.1108677466458215e-09,
"loss": 0.0099,
"step": 9950
},
{
"epoch": 18.444444444444443,
"grad_norm": 0.11076436191797256,
"learning_rate": 4.595720454353414e-09,
"loss": 0.008,
"step": 9960
},
{
"epoch": 18.462962962962962,
"grad_norm": 0.10836225748062134,
"learning_rate": 2.627314893294264e-09,
"loss": 0.0081,
"step": 9970
},
{
"epoch": 18.48148148148148,
"grad_norm": 0.07448622584342957,
"learning_rate": 1.2056725896270048e-09,
"loss": 0.0063,
"step": 9980
},
{
"epoch": 18.5,
"grad_norm": 0.12873756885528564,
"learning_rate": 3.308090902098826e-10,
"loss": 0.0064,
"step": 9990
},
{
"epoch": 18.51851851851852,
"grad_norm": 0.11102540791034698,
"learning_rate": 2.7339624120159555e-12,
"loss": 0.0086,
"step": 10000
}
],
"logging_steps": 10,
"max_steps": 10000,
"num_input_tokens_seen": 0,
"num_train_epochs": 19,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 64,
"trial_name": null,
"trial_params": null
}