gr00t-3b-sl-4-sa / trainer_state.json
nextbig's picture
Upload folder using huggingface_hub
48f59e7 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 59.880239520958085,
"eval_steps": 500,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.059880239520958084,
"grad_norm": 10.409473419189453,
"learning_rate": 1.8e-06,
"loss": 1.0926,
"step": 10
},
{
"epoch": 0.11976047904191617,
"grad_norm": 2.721712589263916,
"learning_rate": 3.8e-06,
"loss": 0.8079,
"step": 20
},
{
"epoch": 0.17964071856287425,
"grad_norm": 2.668743133544922,
"learning_rate": 5.8e-06,
"loss": 0.3786,
"step": 30
},
{
"epoch": 0.23952095808383234,
"grad_norm": 1.9073970317840576,
"learning_rate": 7.8e-06,
"loss": 0.2267,
"step": 40
},
{
"epoch": 0.2994011976047904,
"grad_norm": 1.314000129699707,
"learning_rate": 9.800000000000001e-06,
"loss": 0.1605,
"step": 50
},
{
"epoch": 0.3592814371257485,
"grad_norm": 0.8861719369888306,
"learning_rate": 1.18e-05,
"loss": 0.1262,
"step": 60
},
{
"epoch": 0.41916167664670656,
"grad_norm": 0.767314612865448,
"learning_rate": 1.3800000000000002e-05,
"loss": 0.1167,
"step": 70
},
{
"epoch": 0.47904191616766467,
"grad_norm": 0.5740051865577698,
"learning_rate": 1.58e-05,
"loss": 0.0986,
"step": 80
},
{
"epoch": 0.5389221556886228,
"grad_norm": 0.731820285320282,
"learning_rate": 1.78e-05,
"loss": 0.0935,
"step": 90
},
{
"epoch": 0.5988023952095808,
"grad_norm": 0.7142053246498108,
"learning_rate": 1.9800000000000004e-05,
"loss": 0.0852,
"step": 100
},
{
"epoch": 0.6586826347305389,
"grad_norm": 0.3792620301246643,
"learning_rate": 2.18e-05,
"loss": 0.0769,
"step": 110
},
{
"epoch": 0.718562874251497,
"grad_norm": 0.4527987241744995,
"learning_rate": 2.38e-05,
"loss": 0.0703,
"step": 120
},
{
"epoch": 0.7784431137724551,
"grad_norm": 0.5975872278213501,
"learning_rate": 2.58e-05,
"loss": 0.0631,
"step": 130
},
{
"epoch": 0.8383233532934131,
"grad_norm": 0.7328669428825378,
"learning_rate": 2.7800000000000005e-05,
"loss": 0.06,
"step": 140
},
{
"epoch": 0.8982035928143712,
"grad_norm": 0.5794113874435425,
"learning_rate": 2.98e-05,
"loss": 0.0546,
"step": 150
},
{
"epoch": 0.9580838323353293,
"grad_norm": 0.6779555678367615,
"learning_rate": 3.18e-05,
"loss": 0.0481,
"step": 160
},
{
"epoch": 1.0179640718562875,
"grad_norm": 0.6833993792533875,
"learning_rate": 3.38e-05,
"loss": 0.0419,
"step": 170
},
{
"epoch": 1.0778443113772456,
"grad_norm": 0.4933662414550781,
"learning_rate": 3.58e-05,
"loss": 0.0437,
"step": 180
},
{
"epoch": 1.1377245508982037,
"grad_norm": 0.5869358777999878,
"learning_rate": 3.7800000000000004e-05,
"loss": 0.0439,
"step": 190
},
{
"epoch": 1.1976047904191618,
"grad_norm": 0.4607003927230835,
"learning_rate": 3.9800000000000005e-05,
"loss": 0.0406,
"step": 200
},
{
"epoch": 1.2574850299401197,
"grad_norm": 0.5321051478385925,
"learning_rate": 4.18e-05,
"loss": 0.0349,
"step": 210
},
{
"epoch": 1.3173652694610778,
"grad_norm": 0.4957868158817291,
"learning_rate": 4.38e-05,
"loss": 0.0345,
"step": 220
},
{
"epoch": 1.377245508982036,
"grad_norm": 0.8227276802062988,
"learning_rate": 4.58e-05,
"loss": 0.0361,
"step": 230
},
{
"epoch": 1.437125748502994,
"grad_norm": 0.5242947340011597,
"learning_rate": 4.78e-05,
"loss": 0.0352,
"step": 240
},
{
"epoch": 1.4970059880239521,
"grad_norm": 0.3747622072696686,
"learning_rate": 4.9800000000000004e-05,
"loss": 0.0334,
"step": 250
},
{
"epoch": 1.55688622754491,
"grad_norm": 0.443041056394577,
"learning_rate": 5.1800000000000005e-05,
"loss": 0.03,
"step": 260
},
{
"epoch": 1.6167664670658684,
"grad_norm": 0.5506426095962524,
"learning_rate": 5.380000000000001e-05,
"loss": 0.0322,
"step": 270
},
{
"epoch": 1.6766467065868262,
"grad_norm": 0.5093207955360413,
"learning_rate": 5.580000000000001e-05,
"loss": 0.0298,
"step": 280
},
{
"epoch": 1.7365269461077846,
"grad_norm": 0.6352999210357666,
"learning_rate": 5.7799999999999995e-05,
"loss": 0.0298,
"step": 290
},
{
"epoch": 1.7964071856287425,
"grad_norm": 0.5009264349937439,
"learning_rate": 5.9800000000000003e-05,
"loss": 0.0345,
"step": 300
},
{
"epoch": 1.8562874251497006,
"grad_norm": 0.4747627377510071,
"learning_rate": 6.18e-05,
"loss": 0.0279,
"step": 310
},
{
"epoch": 1.9161676646706587,
"grad_norm": 0.5248665809631348,
"learning_rate": 6.38e-05,
"loss": 0.0287,
"step": 320
},
{
"epoch": 1.9760479041916168,
"grad_norm": 0.43971648812294006,
"learning_rate": 6.58e-05,
"loss": 0.0301,
"step": 330
},
{
"epoch": 2.035928143712575,
"grad_norm": 0.39686694741249084,
"learning_rate": 6.780000000000001e-05,
"loss": 0.0287,
"step": 340
},
{
"epoch": 2.095808383233533,
"grad_norm": 0.37652140855789185,
"learning_rate": 6.98e-05,
"loss": 0.0262,
"step": 350
},
{
"epoch": 2.155688622754491,
"grad_norm": 0.263740211725235,
"learning_rate": 7.18e-05,
"loss": 0.0251,
"step": 360
},
{
"epoch": 2.215568862275449,
"grad_norm": 0.44441261887550354,
"learning_rate": 7.38e-05,
"loss": 0.0245,
"step": 370
},
{
"epoch": 2.2754491017964074,
"grad_norm": 0.3769250810146332,
"learning_rate": 7.58e-05,
"loss": 0.0263,
"step": 380
},
{
"epoch": 2.3353293413173652,
"grad_norm": 0.5479228496551514,
"learning_rate": 7.780000000000001e-05,
"loss": 0.0294,
"step": 390
},
{
"epoch": 2.3952095808383236,
"grad_norm": 0.3823162615299225,
"learning_rate": 7.98e-05,
"loss": 0.0257,
"step": 400
},
{
"epoch": 2.4550898203592815,
"grad_norm": 0.36714932322502136,
"learning_rate": 8.18e-05,
"loss": 0.0231,
"step": 410
},
{
"epoch": 2.5149700598802394,
"grad_norm": 0.411713182926178,
"learning_rate": 8.38e-05,
"loss": 0.0248,
"step": 420
},
{
"epoch": 2.5748502994011977,
"grad_norm": 0.5007436275482178,
"learning_rate": 8.58e-05,
"loss": 0.0294,
"step": 430
},
{
"epoch": 2.6347305389221556,
"grad_norm": 0.6436091661453247,
"learning_rate": 8.78e-05,
"loss": 0.0274,
"step": 440
},
{
"epoch": 2.694610778443114,
"grad_norm": 0.5559032559394836,
"learning_rate": 8.98e-05,
"loss": 0.0272,
"step": 450
},
{
"epoch": 2.754491017964072,
"grad_norm": 0.5250544548034668,
"learning_rate": 9.180000000000001e-05,
"loss": 0.0257,
"step": 460
},
{
"epoch": 2.81437125748503,
"grad_norm": 0.3654075562953949,
"learning_rate": 9.38e-05,
"loss": 0.0226,
"step": 470
},
{
"epoch": 2.874251497005988,
"grad_norm": 0.2955826222896576,
"learning_rate": 9.58e-05,
"loss": 0.0225,
"step": 480
},
{
"epoch": 2.934131736526946,
"grad_norm": 0.40357694029808044,
"learning_rate": 9.78e-05,
"loss": 0.02,
"step": 490
},
{
"epoch": 2.9940119760479043,
"grad_norm": 0.4705279767513275,
"learning_rate": 9.98e-05,
"loss": 0.0204,
"step": 500
},
{
"epoch": 3.053892215568862,
"grad_norm": 0.35780131816864014,
"learning_rate": 9.9999778549206e-05,
"loss": 0.022,
"step": 510
},
{
"epoch": 3.1137724550898205,
"grad_norm": 0.3284674882888794,
"learning_rate": 9.999901304280685e-05,
"loss": 0.0249,
"step": 520
},
{
"epoch": 3.1736526946107784,
"grad_norm": 0.3672216236591339,
"learning_rate": 9.999770075521164e-05,
"loss": 0.0217,
"step": 530
},
{
"epoch": 3.2335329341317367,
"grad_norm": 0.42824992537498474,
"learning_rate": 9.99958417007713e-05,
"loss": 0.021,
"step": 540
},
{
"epoch": 3.2934131736526946,
"grad_norm": 0.5583121180534363,
"learning_rate": 9.999343589981615e-05,
"loss": 0.0239,
"step": 550
},
{
"epoch": 3.3532934131736525,
"grad_norm": 0.4128214716911316,
"learning_rate": 9.999048337865568e-05,
"loss": 0.0196,
"step": 560
},
{
"epoch": 3.413173652694611,
"grad_norm": 0.43404772877693176,
"learning_rate": 9.998698416957815e-05,
"loss": 0.0194,
"step": 570
},
{
"epoch": 3.4730538922155687,
"grad_norm": 0.383382111787796,
"learning_rate": 9.998293831085037e-05,
"loss": 0.0226,
"step": 580
},
{
"epoch": 3.532934131736527,
"grad_norm": 0.6052247285842896,
"learning_rate": 9.997834584671719e-05,
"loss": 0.0208,
"step": 590
},
{
"epoch": 3.592814371257485,
"grad_norm": 0.4105810523033142,
"learning_rate": 9.997320682740107e-05,
"loss": 0.0205,
"step": 600
},
{
"epoch": 3.6526946107784433,
"grad_norm": 0.32404133677482605,
"learning_rate": 9.996752130910149e-05,
"loss": 0.0219,
"step": 610
},
{
"epoch": 3.712574850299401,
"grad_norm": 0.33681678771972656,
"learning_rate": 9.99612893539944e-05,
"loss": 0.0176,
"step": 620
},
{
"epoch": 3.772455089820359,
"grad_norm": 0.3996165692806244,
"learning_rate": 9.995451103023144e-05,
"loss": 0.0202,
"step": 630
},
{
"epoch": 3.8323353293413174,
"grad_norm": 0.37920981645584106,
"learning_rate": 9.994718641193928e-05,
"loss": 0.02,
"step": 640
},
{
"epoch": 3.8922155688622757,
"grad_norm": 0.4280896782875061,
"learning_rate": 9.993931557921874e-05,
"loss": 0.0166,
"step": 650
},
{
"epoch": 3.9520958083832336,
"grad_norm": 0.6001282334327698,
"learning_rate": 9.993089861814402e-05,
"loss": 0.0186,
"step": 660
},
{
"epoch": 4.0119760479041915,
"grad_norm": 0.3995504081249237,
"learning_rate": 9.992193562076166e-05,
"loss": 0.0177,
"step": 670
},
{
"epoch": 4.07185628742515,
"grad_norm": 0.4000985622406006,
"learning_rate": 9.991242668508954e-05,
"loss": 0.0212,
"step": 680
},
{
"epoch": 4.131736526946108,
"grad_norm": 0.5066369771957397,
"learning_rate": 9.990237191511587e-05,
"loss": 0.0187,
"step": 690
},
{
"epoch": 4.191616766467066,
"grad_norm": 0.25399208068847656,
"learning_rate": 9.989177142079802e-05,
"loss": 0.0157,
"step": 700
},
{
"epoch": 4.251497005988024,
"grad_norm": 0.38600268959999084,
"learning_rate": 9.988062531806126e-05,
"loss": 0.0179,
"step": 710
},
{
"epoch": 4.311377245508982,
"grad_norm": 0.305318146944046,
"learning_rate": 9.986893372879762e-05,
"loss": 0.018,
"step": 720
},
{
"epoch": 4.37125748502994,
"grad_norm": 0.2971837818622589,
"learning_rate": 9.985669678086443e-05,
"loss": 0.0204,
"step": 730
},
{
"epoch": 4.431137724550898,
"grad_norm": 0.23951543867588043,
"learning_rate": 9.984391460808298e-05,
"loss": 0.0158,
"step": 740
},
{
"epoch": 4.491017964071856,
"grad_norm": 0.26355302333831787,
"learning_rate": 9.983058735023709e-05,
"loss": 0.0174,
"step": 750
},
{
"epoch": 4.550898203592815,
"grad_norm": 0.3793099522590637,
"learning_rate": 9.98167151530715e-05,
"loss": 0.0167,
"step": 760
},
{
"epoch": 4.610778443113772,
"grad_norm": 0.2782505750656128,
"learning_rate": 9.980229816829034e-05,
"loss": 0.017,
"step": 770
},
{
"epoch": 4.6706586826347305,
"grad_norm": 0.3714260756969452,
"learning_rate": 9.978733655355544e-05,
"loss": 0.0166,
"step": 780
},
{
"epoch": 4.730538922155689,
"grad_norm": 0.3202928900718689,
"learning_rate": 9.977183047248464e-05,
"loss": 0.018,
"step": 790
},
{
"epoch": 4.790419161676647,
"grad_norm": 0.31857386231422424,
"learning_rate": 9.975578009464992e-05,
"loss": 0.0189,
"step": 800
},
{
"epoch": 4.850299401197605,
"grad_norm": 0.3495095372200012,
"learning_rate": 9.97391855955757e-05,
"loss": 0.0162,
"step": 810
},
{
"epoch": 4.910179640718563,
"grad_norm": 0.34920454025268555,
"learning_rate": 9.972204715673669e-05,
"loss": 0.0181,
"step": 820
},
{
"epoch": 4.970059880239521,
"grad_norm": 0.5020573735237122,
"learning_rate": 9.970436496555617e-05,
"loss": 0.0179,
"step": 830
},
{
"epoch": 5.029940119760479,
"grad_norm": 0.500695526599884,
"learning_rate": 9.968613921540373e-05,
"loss": 0.0168,
"step": 840
},
{
"epoch": 5.089820359281437,
"grad_norm": 0.36557909846305847,
"learning_rate": 9.966737010559326e-05,
"loss": 0.015,
"step": 850
},
{
"epoch": 5.149700598802395,
"grad_norm": 0.40890976786613464,
"learning_rate": 9.964805784138072e-05,
"loss": 0.0155,
"step": 860
},
{
"epoch": 5.209580838323353,
"grad_norm": 0.32353684306144714,
"learning_rate": 9.962820263396195e-05,
"loss": 0.0149,
"step": 870
},
{
"epoch": 5.269461077844311,
"grad_norm": 0.5726319551467896,
"learning_rate": 9.960780470047033e-05,
"loss": 0.0144,
"step": 880
},
{
"epoch": 5.3293413173652695,
"grad_norm": 0.5006370544433594,
"learning_rate": 9.958686426397437e-05,
"loss": 0.014,
"step": 890
},
{
"epoch": 5.389221556886228,
"grad_norm": 0.4396548569202423,
"learning_rate": 9.956538155347534e-05,
"loss": 0.0155,
"step": 900
},
{
"epoch": 5.449101796407185,
"grad_norm": 0.41286391019821167,
"learning_rate": 9.95433568039047e-05,
"loss": 0.0169,
"step": 910
},
{
"epoch": 5.508982035928144,
"grad_norm": 0.30442172288894653,
"learning_rate": 9.952079025612162e-05,
"loss": 0.0161,
"step": 920
},
{
"epoch": 5.568862275449102,
"grad_norm": 0.36612051725387573,
"learning_rate": 9.949768215691022e-05,
"loss": 0.017,
"step": 930
},
{
"epoch": 5.62874251497006,
"grad_norm": 0.45451632142066956,
"learning_rate": 9.9474032758977e-05,
"loss": 0.0169,
"step": 940
},
{
"epoch": 5.688622754491018,
"grad_norm": 0.4567358195781708,
"learning_rate": 9.944984232094794e-05,
"loss": 0.0184,
"step": 950
},
{
"epoch": 5.748502994011976,
"grad_norm": 0.3491564393043518,
"learning_rate": 9.942511110736584e-05,
"loss": 0.0186,
"step": 960
},
{
"epoch": 5.808383233532934,
"grad_norm": 0.4113578796386719,
"learning_rate": 9.939983938868726e-05,
"loss": 0.0166,
"step": 970
},
{
"epoch": 5.868263473053892,
"grad_norm": 0.30258816480636597,
"learning_rate": 9.93740274412797e-05,
"loss": 0.0164,
"step": 980
},
{
"epoch": 5.92814371257485,
"grad_norm": 0.34874433279037476,
"learning_rate": 9.934767554741846e-05,
"loss": 0.0182,
"step": 990
},
{
"epoch": 5.9880239520958085,
"grad_norm": 0.19492897391319275,
"learning_rate": 9.932078399528361e-05,
"loss": 0.015,
"step": 1000
},
{
"epoch": 6.047904191616767,
"grad_norm": 0.2990895211696625,
"learning_rate": 9.929335307895689e-05,
"loss": 0.0136,
"step": 1010
},
{
"epoch": 6.107784431137724,
"grad_norm": 0.24870064854621887,
"learning_rate": 9.926538309841839e-05,
"loss": 0.0138,
"step": 1020
},
{
"epoch": 6.167664670658683,
"grad_norm": 0.3146226108074188,
"learning_rate": 9.923687435954334e-05,
"loss": 0.0143,
"step": 1030
},
{
"epoch": 6.227544910179641,
"grad_norm": 0.39984560012817383,
"learning_rate": 9.920782717409873e-05,
"loss": 0.0121,
"step": 1040
},
{
"epoch": 6.287425149700598,
"grad_norm": 0.4670575261116028,
"learning_rate": 9.917824185973994e-05,
"loss": 0.0135,
"step": 1050
},
{
"epoch": 6.347305389221557,
"grad_norm": 0.4787517189979553,
"learning_rate": 9.914811874000723e-05,
"loss": 0.0148,
"step": 1060
},
{
"epoch": 6.407185628742515,
"grad_norm": 0.5922192931175232,
"learning_rate": 9.911745814432218e-05,
"loss": 0.014,
"step": 1070
},
{
"epoch": 6.467065868263473,
"grad_norm": 0.4581575393676758,
"learning_rate": 9.90862604079842e-05,
"loss": 0.0163,
"step": 1080
},
{
"epoch": 6.526946107784431,
"grad_norm": 0.35336896777153015,
"learning_rate": 9.90545258721667e-05,
"loss": 0.0132,
"step": 1090
},
{
"epoch": 6.586826347305389,
"grad_norm": 0.18546943366527557,
"learning_rate": 9.90222548839135e-05,
"loss": 0.0153,
"step": 1100
},
{
"epoch": 6.6467065868263475,
"grad_norm": 0.3376789391040802,
"learning_rate": 9.898944779613495e-05,
"loss": 0.0152,
"step": 1110
},
{
"epoch": 6.706586826347305,
"grad_norm": 0.3100164830684662,
"learning_rate": 9.89561049676041e-05,
"loss": 0.0156,
"step": 1120
},
{
"epoch": 6.766467065868263,
"grad_norm": 0.2680951952934265,
"learning_rate": 9.89222267629528e-05,
"loss": 0.0149,
"step": 1130
},
{
"epoch": 6.826347305389222,
"grad_norm": 0.30346718430519104,
"learning_rate": 9.888781355266763e-05,
"loss": 0.0148,
"step": 1140
},
{
"epoch": 6.88622754491018,
"grad_norm": 0.260969340801239,
"learning_rate": 9.885286571308598e-05,
"loss": 0.0163,
"step": 1150
},
{
"epoch": 6.946107784431137,
"grad_norm": 0.4359760284423828,
"learning_rate": 9.881738362639182e-05,
"loss": 0.0135,
"step": 1160
},
{
"epoch": 7.005988023952096,
"grad_norm": 0.21016572415828705,
"learning_rate": 9.878136768061154e-05,
"loss": 0.0124,
"step": 1170
},
{
"epoch": 7.065868263473054,
"grad_norm": 0.32056793570518494,
"learning_rate": 9.874481826960979e-05,
"loss": 0.0122,
"step": 1180
},
{
"epoch": 7.125748502994012,
"grad_norm": 0.2880637049674988,
"learning_rate": 9.870773579308503e-05,
"loss": 0.0139,
"step": 1190
},
{
"epoch": 7.18562874251497,
"grad_norm": 0.2694314420223236,
"learning_rate": 9.867012065656533e-05,
"loss": 0.0143,
"step": 1200
},
{
"epoch": 7.245508982035928,
"grad_norm": 0.25571173429489136,
"learning_rate": 9.863197327140376e-05,
"loss": 0.0112,
"step": 1210
},
{
"epoch": 7.3053892215568865,
"grad_norm": 0.3050115406513214,
"learning_rate": 9.859329405477403e-05,
"loss": 0.0161,
"step": 1220
},
{
"epoch": 7.365269461077844,
"grad_norm": 0.21913251280784607,
"learning_rate": 9.855408342966585e-05,
"loss": 0.0126,
"step": 1230
},
{
"epoch": 7.425149700598802,
"grad_norm": 0.2909991443157196,
"learning_rate": 9.851434182488033e-05,
"loss": 0.013,
"step": 1240
},
{
"epoch": 7.485029940119761,
"grad_norm": 0.3046460449695587,
"learning_rate": 9.84740696750253e-05,
"loss": 0.0129,
"step": 1250
},
{
"epoch": 7.544910179640718,
"grad_norm": 0.2511163353919983,
"learning_rate": 9.843326742051055e-05,
"loss": 0.013,
"step": 1260
},
{
"epoch": 7.604790419161676,
"grad_norm": 0.22432568669319153,
"learning_rate": 9.839193550754297e-05,
"loss": 0.0155,
"step": 1270
},
{
"epoch": 7.664670658682635,
"grad_norm": 0.338826984167099,
"learning_rate": 9.835007438812177e-05,
"loss": 0.0125,
"step": 1280
},
{
"epoch": 7.724550898203593,
"grad_norm": 0.20700474083423615,
"learning_rate": 9.830768452003341e-05,
"loss": 0.0113,
"step": 1290
},
{
"epoch": 7.7844311377245505,
"grad_norm": 0.2013084590435028,
"learning_rate": 9.826476636684671e-05,
"loss": 0.0146,
"step": 1300
},
{
"epoch": 7.844311377245509,
"grad_norm": 0.24529442191123962,
"learning_rate": 9.822132039790773e-05,
"loss": 0.0139,
"step": 1310
},
{
"epoch": 7.904191616766467,
"grad_norm": 0.2629704773426056,
"learning_rate": 9.817734708833461e-05,
"loss": 0.0122,
"step": 1320
},
{
"epoch": 7.9640718562874255,
"grad_norm": 0.28070858120918274,
"learning_rate": 9.813284691901243e-05,
"loss": 0.0121,
"step": 1330
},
{
"epoch": 8.023952095808383,
"grad_norm": 0.25485098361968994,
"learning_rate": 9.808782037658792e-05,
"loss": 0.0116,
"step": 1340
},
{
"epoch": 8.08383233532934,
"grad_norm": 0.21787486970424652,
"learning_rate": 9.804226795346411e-05,
"loss": 0.0113,
"step": 1350
},
{
"epoch": 8.1437125748503,
"grad_norm": 0.17861317098140717,
"learning_rate": 9.799619014779503e-05,
"loss": 0.0116,
"step": 1360
},
{
"epoch": 8.203592814371257,
"grad_norm": 0.3499056100845337,
"learning_rate": 9.794958746348013e-05,
"loss": 0.0134,
"step": 1370
},
{
"epoch": 8.263473053892216,
"grad_norm": 0.27046534419059753,
"learning_rate": 9.790246041015896e-05,
"loss": 0.0122,
"step": 1380
},
{
"epoch": 8.323353293413174,
"grad_norm": 0.20849302411079407,
"learning_rate": 9.785480950320538e-05,
"loss": 0.0112,
"step": 1390
},
{
"epoch": 8.383233532934131,
"grad_norm": 0.45044535398483276,
"learning_rate": 9.78066352637221e-05,
"loss": 0.013,
"step": 1400
},
{
"epoch": 8.44311377245509,
"grad_norm": 0.18034030497074127,
"learning_rate": 9.775793821853488e-05,
"loss": 0.0117,
"step": 1410
},
{
"epoch": 8.502994011976048,
"grad_norm": 0.3212395906448364,
"learning_rate": 9.77087189001868e-05,
"loss": 0.0104,
"step": 1420
},
{
"epoch": 8.562874251497005,
"grad_norm": 0.2533542513847351,
"learning_rate": 9.765897784693243e-05,
"loss": 0.0113,
"step": 1430
},
{
"epoch": 8.622754491017965,
"grad_norm": 0.17141425609588623,
"learning_rate": 9.760871560273197e-05,
"loss": 0.0119,
"step": 1440
},
{
"epoch": 8.682634730538922,
"grad_norm": 0.3235777020454407,
"learning_rate": 9.755793271724526e-05,
"loss": 0.0117,
"step": 1450
},
{
"epoch": 8.74251497005988,
"grad_norm": 0.2590118944644928,
"learning_rate": 9.750662974582584e-05,
"loss": 0.0126,
"step": 1460
},
{
"epoch": 8.802395209580839,
"grad_norm": 0.20043741166591644,
"learning_rate": 9.745480724951473e-05,
"loss": 0.0095,
"step": 1470
},
{
"epoch": 8.862275449101796,
"grad_norm": 0.22708261013031006,
"learning_rate": 9.740246579503447e-05,
"loss": 0.0124,
"step": 1480
},
{
"epoch": 8.922155688622755,
"grad_norm": 0.2595358192920685,
"learning_rate": 9.734960595478284e-05,
"loss": 0.0132,
"step": 1490
},
{
"epoch": 8.982035928143713,
"grad_norm": 0.21153053641319275,
"learning_rate": 9.729622830682657e-05,
"loss": 0.0147,
"step": 1500
},
{
"epoch": 9.04191616766467,
"grad_norm": 0.34915104508399963,
"learning_rate": 9.724233343489504e-05,
"loss": 0.0112,
"step": 1510
},
{
"epoch": 9.10179640718563,
"grad_norm": 0.23510225117206573,
"learning_rate": 9.718792192837396e-05,
"loss": 0.0103,
"step": 1520
},
{
"epoch": 9.161676646706587,
"grad_norm": 0.2854577600955963,
"learning_rate": 9.713299438229886e-05,
"loss": 0.0125,
"step": 1530
},
{
"epoch": 9.221556886227544,
"grad_norm": 0.21470534801483154,
"learning_rate": 9.707755139734855e-05,
"loss": 0.012,
"step": 1540
},
{
"epoch": 9.281437125748504,
"grad_norm": 0.31662946939468384,
"learning_rate": 9.702159357983866e-05,
"loss": 0.0105,
"step": 1550
},
{
"epoch": 9.341317365269461,
"grad_norm": 0.29741784930229187,
"learning_rate": 9.696512154171492e-05,
"loss": 0.0114,
"step": 1560
},
{
"epoch": 9.401197604790418,
"grad_norm": 0.31248733401298523,
"learning_rate": 9.690813590054645e-05,
"loss": 0.0116,
"step": 1570
},
{
"epoch": 9.461077844311378,
"grad_norm": 0.2677183747291565,
"learning_rate": 9.685063727951914e-05,
"loss": 0.0114,
"step": 1580
},
{
"epoch": 9.520958083832335,
"grad_norm": 0.15428173542022705,
"learning_rate": 9.679262630742865e-05,
"loss": 0.0111,
"step": 1590
},
{
"epoch": 9.580838323353294,
"grad_norm": 0.22496920824050903,
"learning_rate": 9.673410361867373e-05,
"loss": 0.0102,
"step": 1600
},
{
"epoch": 9.640718562874252,
"grad_norm": 0.33756619691848755,
"learning_rate": 9.667506985324909e-05,
"loss": 0.0116,
"step": 1610
},
{
"epoch": 9.70059880239521,
"grad_norm": 0.3057170510292053,
"learning_rate": 9.661552565673855e-05,
"loss": 0.0098,
"step": 1620
},
{
"epoch": 9.760479041916168,
"grad_norm": 0.292972207069397,
"learning_rate": 9.655547168030789e-05,
"loss": 0.013,
"step": 1630
},
{
"epoch": 9.820359281437126,
"grad_norm": 0.3326488733291626,
"learning_rate": 9.649490858069777e-05,
"loss": 0.0122,
"step": 1640
},
{
"epoch": 9.880239520958083,
"grad_norm": 0.22227708995342255,
"learning_rate": 9.643383702021658e-05,
"loss": 0.0114,
"step": 1650
},
{
"epoch": 9.940119760479043,
"grad_norm": 0.22172564268112183,
"learning_rate": 9.637225766673307e-05,
"loss": 0.0113,
"step": 1660
},
{
"epoch": 10.0,
"grad_norm": 0.25080209970474243,
"learning_rate": 9.631017119366922e-05,
"loss": 0.0108,
"step": 1670
},
{
"epoch": 10.059880239520957,
"grad_norm": 0.2742719352245331,
"learning_rate": 9.624757827999273e-05,
"loss": 0.0113,
"step": 1680
},
{
"epoch": 10.119760479041917,
"grad_norm": 0.32316118478775024,
"learning_rate": 9.618447961020971e-05,
"loss": 0.0101,
"step": 1690
},
{
"epoch": 10.179640718562874,
"grad_norm": 0.3154122531414032,
"learning_rate": 9.612087587435707e-05,
"loss": 0.0145,
"step": 1700
},
{
"epoch": 10.239520958083832,
"grad_norm": 0.32044708728790283,
"learning_rate": 9.605676776799508e-05,
"loss": 0.0121,
"step": 1710
},
{
"epoch": 10.29940119760479,
"grad_norm": 0.23354724049568176,
"learning_rate": 9.599215599219973e-05,
"loss": 0.0121,
"step": 1720
},
{
"epoch": 10.359281437125748,
"grad_norm": 0.2523927092552185,
"learning_rate": 9.592704125355505e-05,
"loss": 0.0126,
"step": 1730
},
{
"epoch": 10.419161676646706,
"grad_norm": 0.29253828525543213,
"learning_rate": 9.586142426414538e-05,
"loss": 0.0112,
"step": 1740
},
{
"epoch": 10.479041916167665,
"grad_norm": 0.30786383152008057,
"learning_rate": 9.57953057415476e-05,
"loss": 0.0112,
"step": 1750
},
{
"epoch": 10.538922155688622,
"grad_norm": 0.2749510407447815,
"learning_rate": 9.572868640882328e-05,
"loss": 0.009,
"step": 1760
},
{
"epoch": 10.598802395209582,
"grad_norm": 0.2448800653219223,
"learning_rate": 9.56615669945108e-05,
"loss": 0.0097,
"step": 1770
},
{
"epoch": 10.658682634730539,
"grad_norm": 0.22953806817531586,
"learning_rate": 9.55939482326173e-05,
"loss": 0.011,
"step": 1780
},
{
"epoch": 10.718562874251496,
"grad_norm": 0.30279576778411865,
"learning_rate": 9.552583086261069e-05,
"loss": 0.0093,
"step": 1790
},
{
"epoch": 10.778443113772456,
"grad_norm": 0.2829456925392151,
"learning_rate": 9.545721562941168e-05,
"loss": 0.0103,
"step": 1800
},
{
"epoch": 10.838323353293413,
"grad_norm": 0.3553027808666229,
"learning_rate": 9.538810328338543e-05,
"loss": 0.0103,
"step": 1810
},
{
"epoch": 10.89820359281437,
"grad_norm": 0.32986587285995483,
"learning_rate": 9.531849458033349e-05,
"loss": 0.0099,
"step": 1820
},
{
"epoch": 10.95808383233533,
"grad_norm": 0.30219295620918274,
"learning_rate": 9.524839028148547e-05,
"loss": 0.0112,
"step": 1830
},
{
"epoch": 11.017964071856287,
"grad_norm": 0.35268446803092957,
"learning_rate": 9.517779115349077e-05,
"loss": 0.0099,
"step": 1840
},
{
"epoch": 11.077844311377245,
"grad_norm": 0.25568851828575134,
"learning_rate": 9.510669796841014e-05,
"loss": 0.0095,
"step": 1850
},
{
"epoch": 11.137724550898204,
"grad_norm": 0.26727548241615295,
"learning_rate": 9.503511150370727e-05,
"loss": 0.0127,
"step": 1860
},
{
"epoch": 11.197604790419161,
"grad_norm": 0.24547088146209717,
"learning_rate": 9.496303254224024e-05,
"loss": 0.011,
"step": 1870
},
{
"epoch": 11.25748502994012,
"grad_norm": 0.2939532399177551,
"learning_rate": 9.489046187225306e-05,
"loss": 0.0114,
"step": 1880
},
{
"epoch": 11.317365269461078,
"grad_norm": 0.2843455672264099,
"learning_rate": 9.481740028736692e-05,
"loss": 0.0111,
"step": 1890
},
{
"epoch": 11.377245508982035,
"grad_norm": 0.376820832490921,
"learning_rate": 9.474384858657164e-05,
"loss": 0.0133,
"step": 1900
},
{
"epoch": 11.437125748502995,
"grad_norm": 0.39941832423210144,
"learning_rate": 9.466980757421679e-05,
"loss": 0.0104,
"step": 1910
},
{
"epoch": 11.497005988023952,
"grad_norm": 0.3612916171550751,
"learning_rate": 9.459527806000305e-05,
"loss": 0.0115,
"step": 1920
},
{
"epoch": 11.55688622754491,
"grad_norm": 0.2397584319114685,
"learning_rate": 9.452026085897325e-05,
"loss": 0.0116,
"step": 1930
},
{
"epoch": 11.616766467065869,
"grad_norm": 0.3787127137184143,
"learning_rate": 9.444475679150348e-05,
"loss": 0.0117,
"step": 1940
},
{
"epoch": 11.676646706586826,
"grad_norm": 0.2517344653606415,
"learning_rate": 9.436876668329411e-05,
"loss": 0.0099,
"step": 1950
},
{
"epoch": 11.736526946107784,
"grad_norm": 0.17566126585006714,
"learning_rate": 9.429229136536079e-05,
"loss": 0.0122,
"step": 1960
},
{
"epoch": 11.796407185628743,
"grad_norm": 0.18655864894390106,
"learning_rate": 9.421533167402534e-05,
"loss": 0.0104,
"step": 1970
},
{
"epoch": 11.8562874251497,
"grad_norm": 0.15407340228557587,
"learning_rate": 9.413788845090666e-05,
"loss": 0.0103,
"step": 1980
},
{
"epoch": 11.91616766467066,
"grad_norm": 0.18353429436683655,
"learning_rate": 9.405996254291136e-05,
"loss": 0.0126,
"step": 1990
},
{
"epoch": 11.976047904191617,
"grad_norm": 0.17884893715381622,
"learning_rate": 9.398155480222474e-05,
"loss": 0.0095,
"step": 2000
},
{
"epoch": 12.035928143712574,
"grad_norm": 0.22959740459918976,
"learning_rate": 9.390266608630128e-05,
"loss": 0.0113,
"step": 2010
},
{
"epoch": 12.095808383233534,
"grad_norm": 0.17174960672855377,
"learning_rate": 9.38232972578553e-05,
"loss": 0.0089,
"step": 2020
},
{
"epoch": 12.155688622754491,
"grad_norm": 0.2418939769268036,
"learning_rate": 9.374344918485164e-05,
"loss": 0.0125,
"step": 2030
},
{
"epoch": 12.215568862275449,
"grad_norm": 0.1747925877571106,
"learning_rate": 9.366312274049602e-05,
"loss": 0.0098,
"step": 2040
},
{
"epoch": 12.275449101796408,
"grad_norm": 0.37728214263916016,
"learning_rate": 9.358231880322554e-05,
"loss": 0.0137,
"step": 2050
},
{
"epoch": 12.335329341317365,
"grad_norm": 0.31272563338279724,
"learning_rate": 9.350103825669916e-05,
"loss": 0.0101,
"step": 2060
},
{
"epoch": 12.395209580838323,
"grad_norm": 0.40574055910110474,
"learning_rate": 9.341928198978787e-05,
"loss": 0.0104,
"step": 2070
},
{
"epoch": 12.455089820359282,
"grad_norm": 0.2597385346889496,
"learning_rate": 9.333705089656512e-05,
"loss": 0.0087,
"step": 2080
},
{
"epoch": 12.51497005988024,
"grad_norm": 0.31778234243392944,
"learning_rate": 9.325434587629698e-05,
"loss": 0.0131,
"step": 2090
},
{
"epoch": 12.574850299401197,
"grad_norm": 0.27370357513427734,
"learning_rate": 9.31711678334323e-05,
"loss": 0.0121,
"step": 2100
},
{
"epoch": 12.634730538922156,
"grad_norm": 0.2294006645679474,
"learning_rate": 9.308751767759282e-05,
"loss": 0.0094,
"step": 2110
},
{
"epoch": 12.694610778443113,
"grad_norm": 0.27382636070251465,
"learning_rate": 9.300339632356325e-05,
"loss": 0.0091,
"step": 2120
},
{
"epoch": 12.754491017964071,
"grad_norm": 0.21049852669239044,
"learning_rate": 9.291880469128124e-05,
"loss": 0.009,
"step": 2130
},
{
"epoch": 12.81437125748503,
"grad_norm": 0.2991500198841095,
"learning_rate": 9.283374370582732e-05,
"loss": 0.0102,
"step": 2140
},
{
"epoch": 12.874251497005988,
"grad_norm": 0.27590519189834595,
"learning_rate": 9.274821429741482e-05,
"loss": 0.0102,
"step": 2150
},
{
"epoch": 12.934131736526947,
"grad_norm": 0.2655685544013977,
"learning_rate": 9.266221740137961e-05,
"loss": 0.0126,
"step": 2160
},
{
"epoch": 12.994011976047904,
"grad_norm": 0.3225456178188324,
"learning_rate": 9.257575395817001e-05,
"loss": 0.0126,
"step": 2170
},
{
"epoch": 13.053892215568862,
"grad_norm": 0.21614615619182587,
"learning_rate": 9.248882491333637e-05,
"loss": 0.012,
"step": 2180
},
{
"epoch": 13.113772455089821,
"grad_norm": 0.1930590271949768,
"learning_rate": 9.240143121752076e-05,
"loss": 0.0098,
"step": 2190
},
{
"epoch": 13.173652694610778,
"grad_norm": 0.1839047521352768,
"learning_rate": 9.23135738264467e-05,
"loss": 0.0088,
"step": 2200
},
{
"epoch": 13.233532934131736,
"grad_norm": 0.32084015011787415,
"learning_rate": 9.222525370090849e-05,
"loss": 0.0114,
"step": 2210
},
{
"epoch": 13.293413173652695,
"grad_norm": 0.2828623950481415,
"learning_rate": 9.213647180676088e-05,
"loss": 0.0101,
"step": 2220
},
{
"epoch": 13.353293413173652,
"grad_norm": 0.2671723961830139,
"learning_rate": 9.204722911490846e-05,
"loss": 0.0098,
"step": 2230
},
{
"epoch": 13.41317365269461,
"grad_norm": 0.211421936750412,
"learning_rate": 9.1957526601295e-05,
"loss": 0.0118,
"step": 2240
},
{
"epoch": 13.47305389221557,
"grad_norm": 0.22886018455028534,
"learning_rate": 9.186736524689281e-05,
"loss": 0.01,
"step": 2250
},
{
"epoch": 13.532934131736527,
"grad_norm": 0.2203945815563202,
"learning_rate": 9.177674603769204e-05,
"loss": 0.0088,
"step": 2260
},
{
"epoch": 13.592814371257486,
"grad_norm": 0.16183653473854065,
"learning_rate": 9.168566996468983e-05,
"loss": 0.012,
"step": 2270
},
{
"epoch": 13.652694610778443,
"grad_norm": 0.2912003993988037,
"learning_rate": 9.159413802387951e-05,
"loss": 0.0104,
"step": 2280
},
{
"epoch": 13.7125748502994,
"grad_norm": 0.27864718437194824,
"learning_rate": 9.150215121623974e-05,
"loss": 0.0108,
"step": 2290
},
{
"epoch": 13.77245508982036,
"grad_norm": 0.32082399725914,
"learning_rate": 9.140971054772349e-05,
"loss": 0.0116,
"step": 2300
},
{
"epoch": 13.832335329341317,
"grad_norm": 0.22473551332950592,
"learning_rate": 9.131681702924713e-05,
"loss": 0.0108,
"step": 2310
},
{
"epoch": 13.892215568862275,
"grad_norm": 0.2645914554595947,
"learning_rate": 9.122347167667926e-05,
"loss": 0.0104,
"step": 2320
},
{
"epoch": 13.952095808383234,
"grad_norm": 0.1774655282497406,
"learning_rate": 9.112967551082973e-05,
"loss": 0.0106,
"step": 2330
},
{
"epoch": 14.011976047904191,
"grad_norm": 0.3298044204711914,
"learning_rate": 9.103542955743835e-05,
"loss": 0.0105,
"step": 2340
},
{
"epoch": 14.071856287425149,
"grad_norm": 0.18368478119373322,
"learning_rate": 9.094073484716381e-05,
"loss": 0.0098,
"step": 2350
},
{
"epoch": 14.131736526946108,
"grad_norm": 0.3527606427669525,
"learning_rate": 9.084559241557226e-05,
"loss": 0.011,
"step": 2360
},
{
"epoch": 14.191616766467066,
"grad_norm": 0.28607675433158875,
"learning_rate": 9.075000330312608e-05,
"loss": 0.009,
"step": 2370
},
{
"epoch": 14.251497005988025,
"grad_norm": 0.16145572066307068,
"learning_rate": 9.065396855517253e-05,
"loss": 0.0117,
"step": 2380
},
{
"epoch": 14.311377245508982,
"grad_norm": 0.33116358518600464,
"learning_rate": 9.055748922193219e-05,
"loss": 0.0092,
"step": 2390
},
{
"epoch": 14.37125748502994,
"grad_norm": 0.3597061038017273,
"learning_rate": 9.046056635848761e-05,
"loss": 0.011,
"step": 2400
},
{
"epoch": 14.431137724550899,
"grad_norm": 0.11118652671575546,
"learning_rate": 9.036320102477169e-05,
"loss": 0.0116,
"step": 2410
},
{
"epoch": 14.491017964071856,
"grad_norm": 0.21635514497756958,
"learning_rate": 9.02653942855561e-05,
"loss": 0.0099,
"step": 2420
},
{
"epoch": 14.550898203592814,
"grad_norm": 0.2644486725330353,
"learning_rate": 9.016714721043971e-05,
"loss": 0.0091,
"step": 2430
},
{
"epoch": 14.610778443113773,
"grad_norm": 0.3248916566371918,
"learning_rate": 9.006846087383675e-05,
"loss": 0.011,
"step": 2440
},
{
"epoch": 14.67065868263473,
"grad_norm": 0.12523305416107178,
"learning_rate": 8.996933635496523e-05,
"loss": 0.011,
"step": 2450
},
{
"epoch": 14.730538922155688,
"grad_norm": 0.3054371476173401,
"learning_rate": 8.986977473783498e-05,
"loss": 0.01,
"step": 2460
},
{
"epoch": 14.790419161676647,
"grad_norm": 0.28510555624961853,
"learning_rate": 8.97697771112359e-05,
"loss": 0.0113,
"step": 2470
},
{
"epoch": 14.850299401197605,
"grad_norm": 0.39528530836105347,
"learning_rate": 8.966934456872602e-05,
"loss": 0.0108,
"step": 2480
},
{
"epoch": 14.910179640718562,
"grad_norm": 0.2956768870353699,
"learning_rate": 8.95684782086195e-05,
"loss": 0.0087,
"step": 2490
},
{
"epoch": 14.970059880239521,
"grad_norm": 0.187069833278656,
"learning_rate": 8.946717913397476e-05,
"loss": 0.0104,
"step": 2500
},
{
"epoch": 15.029940119760479,
"grad_norm": 0.2029697149991989,
"learning_rate": 8.93654484525822e-05,
"loss": 0.0089,
"step": 2510
},
{
"epoch": 15.089820359281438,
"grad_norm": 0.38674280047416687,
"learning_rate": 8.926328727695226e-05,
"loss": 0.0103,
"step": 2520
},
{
"epoch": 15.149700598802395,
"grad_norm": 0.15843753516674042,
"learning_rate": 8.916069672430319e-05,
"loss": 0.0107,
"step": 2530
},
{
"epoch": 15.209580838323353,
"grad_norm": 0.2515184283256531,
"learning_rate": 8.905767791654884e-05,
"loss": 0.0101,
"step": 2540
},
{
"epoch": 15.269461077844312,
"grad_norm": 0.26208603382110596,
"learning_rate": 8.895423198028638e-05,
"loss": 0.0093,
"step": 2550
},
{
"epoch": 15.32934131736527,
"grad_norm": 0.23202282190322876,
"learning_rate": 8.885036004678402e-05,
"loss": 0.0096,
"step": 2560
},
{
"epoch": 15.389221556886227,
"grad_norm": 0.18799848854541779,
"learning_rate": 8.874606325196857e-05,
"loss": 0.009,
"step": 2570
},
{
"epoch": 15.449101796407186,
"grad_norm": 0.13833986222743988,
"learning_rate": 8.864134273641304e-05,
"loss": 0.0095,
"step": 2580
},
{
"epoch": 15.508982035928144,
"grad_norm": 0.19574953615665436,
"learning_rate": 8.853619964532427e-05,
"loss": 0.0085,
"step": 2590
},
{
"epoch": 15.568862275449101,
"grad_norm": 0.22622154653072357,
"learning_rate": 8.843063512853019e-05,
"loss": 0.0109,
"step": 2600
},
{
"epoch": 15.62874251497006,
"grad_norm": 0.22898878157138824,
"learning_rate": 8.832465034046749e-05,
"loss": 0.0089,
"step": 2610
},
{
"epoch": 15.688622754491018,
"grad_norm": 0.201274573802948,
"learning_rate": 8.821824644016882e-05,
"loss": 0.0097,
"step": 2620
},
{
"epoch": 15.748502994011975,
"grad_norm": 0.16331319510936737,
"learning_rate": 8.811142459125019e-05,
"loss": 0.01,
"step": 2630
},
{
"epoch": 15.808383233532934,
"grad_norm": 0.2613867521286011,
"learning_rate": 8.800418596189822e-05,
"loss": 0.0096,
"step": 2640
},
{
"epoch": 15.868263473053892,
"grad_norm": 0.21544219553470612,
"learning_rate": 8.789653172485737e-05,
"loss": 0.0102,
"step": 2650
},
{
"epoch": 15.928143712574851,
"grad_norm": 0.20314428210258484,
"learning_rate": 8.778846305741715e-05,
"loss": 0.0071,
"step": 2660
},
{
"epoch": 15.988023952095809,
"grad_norm": 0.21594394743442535,
"learning_rate": 8.767998114139918e-05,
"loss": 0.011,
"step": 2670
},
{
"epoch": 16.047904191616766,
"grad_norm": 0.24611404538154602,
"learning_rate": 8.757108716314429e-05,
"loss": 0.0119,
"step": 2680
},
{
"epoch": 16.107784431137723,
"grad_norm": 0.4281372129917145,
"learning_rate": 8.746178231349962e-05,
"loss": 0.0082,
"step": 2690
},
{
"epoch": 16.16766467065868,
"grad_norm": 0.2246774137020111,
"learning_rate": 8.735206778780549e-05,
"loss": 0.0076,
"step": 2700
},
{
"epoch": 16.227544910179642,
"grad_norm": 0.2802632749080658,
"learning_rate": 8.724194478588234e-05,
"loss": 0.0079,
"step": 2710
},
{
"epoch": 16.2874251497006,
"grad_norm": 0.28266945481300354,
"learning_rate": 8.713141451201772e-05,
"loss": 0.0104,
"step": 2720
},
{
"epoch": 16.347305389221557,
"grad_norm": 0.23547963798046112,
"learning_rate": 8.702047817495295e-05,
"loss": 0.011,
"step": 2730
},
{
"epoch": 16.407185628742514,
"grad_norm": 0.28189995884895325,
"learning_rate": 8.69091369878701e-05,
"loss": 0.0078,
"step": 2740
},
{
"epoch": 16.46706586826347,
"grad_norm": 0.14232125878334045,
"learning_rate": 8.679739216837849e-05,
"loss": 0.0087,
"step": 2750
},
{
"epoch": 16.526946107784433,
"grad_norm": 0.2087014615535736,
"learning_rate": 8.66852449385016e-05,
"loss": 0.0095,
"step": 2760
},
{
"epoch": 16.58682634730539,
"grad_norm": 0.19207759201526642,
"learning_rate": 8.657269652466356e-05,
"loss": 0.0074,
"step": 2770
},
{
"epoch": 16.646706586826348,
"grad_norm": 0.3309922516345978,
"learning_rate": 8.645974815767577e-05,
"loss": 0.008,
"step": 2780
},
{
"epoch": 16.706586826347305,
"grad_norm": 0.36144140362739563,
"learning_rate": 8.634640107272351e-05,
"loss": 0.0082,
"step": 2790
},
{
"epoch": 16.766467065868262,
"grad_norm": 0.2286171019077301,
"learning_rate": 8.623265650935234e-05,
"loss": 0.0085,
"step": 2800
},
{
"epoch": 16.82634730538922,
"grad_norm": 0.36014845967292786,
"learning_rate": 8.611851571145456e-05,
"loss": 0.0113,
"step": 2810
},
{
"epoch": 16.88622754491018,
"grad_norm": 0.19778671860694885,
"learning_rate": 8.600397992725566e-05,
"loss": 0.011,
"step": 2820
},
{
"epoch": 16.94610778443114,
"grad_norm": 0.20682662725448608,
"learning_rate": 8.588905040930061e-05,
"loss": 0.0099,
"step": 2830
},
{
"epoch": 17.005988023952096,
"grad_norm": 0.1719847321510315,
"learning_rate": 8.577372841444022e-05,
"loss": 0.0117,
"step": 2840
},
{
"epoch": 17.065868263473053,
"grad_norm": 0.27442604303359985,
"learning_rate": 8.565801520381736e-05,
"loss": 0.0096,
"step": 2850
},
{
"epoch": 17.12574850299401,
"grad_norm": 0.22688908874988556,
"learning_rate": 8.554191204285313e-05,
"loss": 0.0087,
"step": 2860
},
{
"epoch": 17.18562874251497,
"grad_norm": 0.19836221635341644,
"learning_rate": 8.542542020123315e-05,
"loss": 0.009,
"step": 2870
},
{
"epoch": 17.24550898203593,
"grad_norm": 0.18135212361812592,
"learning_rate": 8.530854095289347e-05,
"loss": 0.0098,
"step": 2880
},
{
"epoch": 17.305389221556887,
"grad_norm": 0.2687807083129883,
"learning_rate": 8.519127557600688e-05,
"loss": 0.01,
"step": 2890
},
{
"epoch": 17.365269461077844,
"grad_norm": 0.28633272647857666,
"learning_rate": 8.507362535296871e-05,
"loss": 0.0084,
"step": 2900
},
{
"epoch": 17.4251497005988,
"grad_norm": 0.24726998805999756,
"learning_rate": 8.495559157038299e-05,
"loss": 0.0086,
"step": 2910
},
{
"epoch": 17.48502994011976,
"grad_norm": 0.2627945840358734,
"learning_rate": 8.483717551904823e-05,
"loss": 0.0075,
"step": 2920
},
{
"epoch": 17.54491017964072,
"grad_norm": 0.26374438405036926,
"learning_rate": 8.47183784939434e-05,
"loss": 0.0086,
"step": 2930
},
{
"epoch": 17.604790419161677,
"grad_norm": 0.24662402272224426,
"learning_rate": 8.459920179421374e-05,
"loss": 0.0084,
"step": 2940
},
{
"epoch": 17.664670658682635,
"grad_norm": 0.27646318078041077,
"learning_rate": 8.447964672315656e-05,
"loss": 0.0088,
"step": 2950
},
{
"epoch": 17.724550898203592,
"grad_norm": 0.19818875193595886,
"learning_rate": 8.435971458820692e-05,
"loss": 0.0093,
"step": 2960
},
{
"epoch": 17.78443113772455,
"grad_norm": 0.30022165179252625,
"learning_rate": 8.423940670092345e-05,
"loss": 0.0087,
"step": 2970
},
{
"epoch": 17.84431137724551,
"grad_norm": 0.23861311376094818,
"learning_rate": 8.411872437697394e-05,
"loss": 0.012,
"step": 2980
},
{
"epoch": 17.904191616766468,
"grad_norm": 0.3647650182247162,
"learning_rate": 8.399766893612096e-05,
"loss": 0.01,
"step": 2990
},
{
"epoch": 17.964071856287426,
"grad_norm": 0.31965184211730957,
"learning_rate": 8.38762417022074e-05,
"loss": 0.0088,
"step": 3000
},
{
"epoch": 18.023952095808383,
"grad_norm": 0.17509804666042328,
"learning_rate": 8.375444400314204e-05,
"loss": 0.0106,
"step": 3010
},
{
"epoch": 18.08383233532934,
"grad_norm": 0.28793132305145264,
"learning_rate": 8.3632277170885e-05,
"loss": 0.0097,
"step": 3020
},
{
"epoch": 18.143712574850298,
"grad_norm": 0.27870792150497437,
"learning_rate": 8.350974254143318e-05,
"loss": 0.0098,
"step": 3030
},
{
"epoch": 18.20359281437126,
"grad_norm": 0.20702967047691345,
"learning_rate": 8.338684145480566e-05,
"loss": 0.0084,
"step": 3040
},
{
"epoch": 18.263473053892216,
"grad_norm": 0.2859809994697571,
"learning_rate": 8.326357525502904e-05,
"loss": 0.0088,
"step": 3050
},
{
"epoch": 18.323353293413174,
"grad_norm": 0.25322043895721436,
"learning_rate": 8.313994529012273e-05,
"loss": 0.0095,
"step": 3060
},
{
"epoch": 18.38323353293413,
"grad_norm": 0.2492699921131134,
"learning_rate": 8.301595291208422e-05,
"loss": 0.0097,
"step": 3070
},
{
"epoch": 18.44311377245509,
"grad_norm": 0.2027384340763092,
"learning_rate": 8.289159947687427e-05,
"loss": 0.0079,
"step": 3080
},
{
"epoch": 18.50299401197605,
"grad_norm": 0.2573034465312958,
"learning_rate": 8.276688634440216e-05,
"loss": 0.008,
"step": 3090
},
{
"epoch": 18.562874251497007,
"grad_norm": 0.24480800330638885,
"learning_rate": 8.26418148785107e-05,
"loss": 0.0138,
"step": 3100
},
{
"epoch": 18.622754491017965,
"grad_norm": 0.2102058082818985,
"learning_rate": 8.251638644696141e-05,
"loss": 0.0107,
"step": 3110
},
{
"epoch": 18.682634730538922,
"grad_norm": 0.21686388552188873,
"learning_rate": 8.23906024214195e-05,
"loss": 0.0096,
"step": 3120
},
{
"epoch": 18.74251497005988,
"grad_norm": 0.17702405154705048,
"learning_rate": 8.226446417743897e-05,
"loss": 0.0093,
"step": 3130
},
{
"epoch": 18.802395209580837,
"grad_norm": 0.21044769883155823,
"learning_rate": 8.213797309444742e-05,
"loss": 0.0098,
"step": 3140
},
{
"epoch": 18.862275449101798,
"grad_norm": 0.27455055713653564,
"learning_rate": 8.201113055573105e-05,
"loss": 0.0079,
"step": 3150
},
{
"epoch": 18.922155688622755,
"grad_norm": 0.2763652205467224,
"learning_rate": 8.188393794841958e-05,
"loss": 0.0078,
"step": 3160
},
{
"epoch": 18.982035928143713,
"grad_norm": 0.17499370872974396,
"learning_rate": 8.175639666347094e-05,
"loss": 0.0096,
"step": 3170
},
{
"epoch": 19.04191616766467,
"grad_norm": 0.21008409559726715,
"learning_rate": 8.162850809565623e-05,
"loss": 0.01,
"step": 3180
},
{
"epoch": 19.101796407185628,
"grad_norm": 0.243890643119812,
"learning_rate": 8.150027364354431e-05,
"loss": 0.0106,
"step": 3190
},
{
"epoch": 19.161676646706585,
"grad_norm": 0.384591281414032,
"learning_rate": 8.137169470948662e-05,
"loss": 0.009,
"step": 3200
},
{
"epoch": 19.221556886227546,
"grad_norm": 0.22937975823879242,
"learning_rate": 8.124277269960179e-05,
"loss": 0.01,
"step": 3210
},
{
"epoch": 19.281437125748504,
"grad_norm": 0.3041241466999054,
"learning_rate": 8.111350902376023e-05,
"loss": 0.0104,
"step": 3220
},
{
"epoch": 19.34131736526946,
"grad_norm": 0.2071114033460617,
"learning_rate": 8.098390509556883e-05,
"loss": 0.0097,
"step": 3230
},
{
"epoch": 19.40119760479042,
"grad_norm": 0.2389390617609024,
"learning_rate": 8.085396233235536e-05,
"loss": 0.0086,
"step": 3240
},
{
"epoch": 19.461077844311376,
"grad_norm": 0.20335648953914642,
"learning_rate": 8.072368215515306e-05,
"loss": 0.0088,
"step": 3250
},
{
"epoch": 19.520958083832337,
"grad_norm": 0.21788887679576874,
"learning_rate": 8.059306598868506e-05,
"loss": 0.0097,
"step": 3260
},
{
"epoch": 19.580838323353294,
"grad_norm": 0.2601717412471771,
"learning_rate": 8.046211526134888e-05,
"loss": 0.0098,
"step": 3270
},
{
"epoch": 19.64071856287425,
"grad_norm": 0.3103547692298889,
"learning_rate": 8.033083140520065e-05,
"loss": 0.0115,
"step": 3280
},
{
"epoch": 19.70059880239521,
"grad_norm": 0.35989537835121155,
"learning_rate": 8.019921585593962e-05,
"loss": 0.0091,
"step": 3290
},
{
"epoch": 19.760479041916167,
"grad_norm": 0.24372735619544983,
"learning_rate": 8.006727005289232e-05,
"loss": 0.0095,
"step": 3300
},
{
"epoch": 19.820359281437124,
"grad_norm": 0.10749700665473938,
"learning_rate": 7.993499543899692e-05,
"loss": 0.007,
"step": 3310
},
{
"epoch": 19.880239520958085,
"grad_norm": 0.19475200772285461,
"learning_rate": 7.980239346078742e-05,
"loss": 0.0105,
"step": 3320
},
{
"epoch": 19.940119760479043,
"grad_norm": 0.13693422079086304,
"learning_rate": 7.966946556837778e-05,
"loss": 0.0094,
"step": 3330
},
{
"epoch": 20.0,
"grad_norm": 0.33291926980018616,
"learning_rate": 7.953621321544616e-05,
"loss": 0.0102,
"step": 3340
},
{
"epoch": 20.059880239520957,
"grad_norm": 0.32077088952064514,
"learning_rate": 7.940263785921896e-05,
"loss": 0.0095,
"step": 3350
},
{
"epoch": 20.119760479041915,
"grad_norm": 0.25905555486679077,
"learning_rate": 7.926874096045482e-05,
"loss": 0.008,
"step": 3360
},
{
"epoch": 20.179640718562876,
"grad_norm": 0.23526518046855927,
"learning_rate": 7.913452398342881e-05,
"loss": 0.0104,
"step": 3370
},
{
"epoch": 20.239520958083833,
"grad_norm": 0.2624652087688446,
"learning_rate": 7.89999883959163e-05,
"loss": 0.0091,
"step": 3380
},
{
"epoch": 20.29940119760479,
"grad_norm": 0.2630319595336914,
"learning_rate": 7.886513566917687e-05,
"loss": 0.0095,
"step": 3390
},
{
"epoch": 20.35928143712575,
"grad_norm": 0.17402897775173187,
"learning_rate": 7.872996727793838e-05,
"loss": 0.0084,
"step": 3400
},
{
"epoch": 20.419161676646706,
"grad_norm": 0.1966816931962967,
"learning_rate": 7.859448470038069e-05,
"loss": 0.0081,
"step": 3410
},
{
"epoch": 20.479041916167663,
"grad_norm": 0.11593285948038101,
"learning_rate": 7.845868941811956e-05,
"loss": 0.0087,
"step": 3420
},
{
"epoch": 20.538922155688624,
"grad_norm": 0.1766042560338974,
"learning_rate": 7.832258291619043e-05,
"loss": 0.0086,
"step": 3430
},
{
"epoch": 20.59880239520958,
"grad_norm": 0.17047084867954254,
"learning_rate": 7.81861666830322e-05,
"loss": 0.0092,
"step": 3440
},
{
"epoch": 20.65868263473054,
"grad_norm": 0.20493540167808533,
"learning_rate": 7.804944221047097e-05,
"loss": 0.0079,
"step": 3450
},
{
"epoch": 20.718562874251496,
"grad_norm": 0.28203222155570984,
"learning_rate": 7.791241099370364e-05,
"loss": 0.0095,
"step": 3460
},
{
"epoch": 20.778443113772454,
"grad_norm": 0.23536105453968048,
"learning_rate": 7.777507453128163e-05,
"loss": 0.0072,
"step": 3470
},
{
"epoch": 20.83832335329341,
"grad_norm": 0.19733834266662598,
"learning_rate": 7.763743432509451e-05,
"loss": 0.0077,
"step": 3480
},
{
"epoch": 20.898203592814372,
"grad_norm": 0.18416664004325867,
"learning_rate": 7.749949188035353e-05,
"loss": 0.01,
"step": 3490
},
{
"epoch": 20.95808383233533,
"grad_norm": 0.24035850167274475,
"learning_rate": 7.736124870557516e-05,
"loss": 0.0081,
"step": 3500
},
{
"epoch": 21.017964071856287,
"grad_norm": 0.31664758920669556,
"learning_rate": 7.722270631256459e-05,
"loss": 0.0075,
"step": 3510
},
{
"epoch": 21.077844311377245,
"grad_norm": 0.25902247428894043,
"learning_rate": 7.708386621639925e-05,
"loss": 0.0072,
"step": 3520
},
{
"epoch": 21.137724550898202,
"grad_norm": 0.18188126385211945,
"learning_rate": 7.694472993541219e-05,
"loss": 0.008,
"step": 3530
},
{
"epoch": 21.197604790419163,
"grad_norm": 0.18373800814151764,
"learning_rate": 7.680529899117547e-05,
"loss": 0.011,
"step": 3540
},
{
"epoch": 21.25748502994012,
"grad_norm": 0.28492408990859985,
"learning_rate": 7.666557490848358e-05,
"loss": 0.0071,
"step": 3550
},
{
"epoch": 21.317365269461078,
"grad_norm": 0.20545673370361328,
"learning_rate": 7.65255592153367e-05,
"loss": 0.0077,
"step": 3560
},
{
"epoch": 21.377245508982035,
"grad_norm": 0.21985077857971191,
"learning_rate": 7.638525344292402e-05,
"loss": 0.0083,
"step": 3570
},
{
"epoch": 21.437125748502993,
"grad_norm": 0.2656082808971405,
"learning_rate": 7.624465912560697e-05,
"loss": 0.0102,
"step": 3580
},
{
"epoch": 21.49700598802395,
"grad_norm": 0.18971195816993713,
"learning_rate": 7.610377780090249e-05,
"loss": 0.0095,
"step": 3590
},
{
"epoch": 21.55688622754491,
"grad_norm": 0.19257457554340363,
"learning_rate": 7.596261100946618e-05,
"loss": 0.0078,
"step": 3600
},
{
"epoch": 21.61676646706587,
"grad_norm": 0.1880970150232315,
"learning_rate": 7.582116029507542e-05,
"loss": 0.01,
"step": 3610
},
{
"epoch": 21.676646706586826,
"grad_norm": 0.14620891213417053,
"learning_rate": 7.56794272046126e-05,
"loss": 0.006,
"step": 3620
},
{
"epoch": 21.736526946107784,
"grad_norm": 0.18656040728092194,
"learning_rate": 7.55374132880481e-05,
"loss": 0.0084,
"step": 3630
},
{
"epoch": 21.79640718562874,
"grad_norm": 0.2397589236497879,
"learning_rate": 7.539512009842333e-05,
"loss": 0.009,
"step": 3640
},
{
"epoch": 21.856287425149702,
"grad_norm": 0.2860172688961029,
"learning_rate": 7.525254919183382e-05,
"loss": 0.0079,
"step": 3650
},
{
"epoch": 21.91616766467066,
"grad_norm": 0.26243826746940613,
"learning_rate": 7.510970212741215e-05,
"loss": 0.0115,
"step": 3660
},
{
"epoch": 21.976047904191617,
"grad_norm": 0.26309916377067566,
"learning_rate": 7.496658046731096e-05,
"loss": 0.01,
"step": 3670
},
{
"epoch": 22.035928143712574,
"grad_norm": 0.21241137385368347,
"learning_rate": 7.482318577668578e-05,
"loss": 0.0087,
"step": 3680
},
{
"epoch": 22.095808383233532,
"grad_norm": 0.24924255907535553,
"learning_rate": 7.467951962367796e-05,
"loss": 0.0091,
"step": 3690
},
{
"epoch": 22.15568862275449,
"grad_norm": 0.27013659477233887,
"learning_rate": 7.453558357939755e-05,
"loss": 0.0088,
"step": 3700
},
{
"epoch": 22.21556886227545,
"grad_norm": 0.17430929839611053,
"learning_rate": 7.439137921790606e-05,
"loss": 0.0086,
"step": 3710
},
{
"epoch": 22.275449101796408,
"grad_norm": 0.15088723599910736,
"learning_rate": 7.42469081161993e-05,
"loss": 0.0068,
"step": 3720
},
{
"epoch": 22.335329341317365,
"grad_norm": 0.25066283345222473,
"learning_rate": 7.410217185419006e-05,
"loss": 0.0096,
"step": 3730
},
{
"epoch": 22.395209580838323,
"grad_norm": 0.24439945816993713,
"learning_rate": 7.395717201469095e-05,
"loss": 0.009,
"step": 3740
},
{
"epoch": 22.45508982035928,
"grad_norm": 0.14095449447631836,
"learning_rate": 7.381191018339696e-05,
"loss": 0.0102,
"step": 3750
},
{
"epoch": 22.51497005988024,
"grad_norm": 0.23285959661006927,
"learning_rate": 7.36663879488682e-05,
"loss": 0.0089,
"step": 3760
},
{
"epoch": 22.5748502994012,
"grad_norm": 0.12560361623764038,
"learning_rate": 7.352060690251254e-05,
"loss": 0.0092,
"step": 3770
},
{
"epoch": 22.634730538922156,
"grad_norm": 0.22383177280426025,
"learning_rate": 7.337456863856811e-05,
"loss": 0.0087,
"step": 3780
},
{
"epoch": 22.694610778443113,
"grad_norm": 0.30464059114456177,
"learning_rate": 7.3228274754086e-05,
"loss": 0.0094,
"step": 3790
},
{
"epoch": 22.75449101796407,
"grad_norm": 0.2332906723022461,
"learning_rate": 7.308172684891267e-05,
"loss": 0.008,
"step": 3800
},
{
"epoch": 22.81437125748503,
"grad_norm": 0.20523525774478912,
"learning_rate": 7.293492652567255e-05,
"loss": 0.007,
"step": 3810
},
{
"epoch": 22.87425149700599,
"grad_norm": 0.12237464636564255,
"learning_rate": 7.278787538975043e-05,
"loss": 0.0067,
"step": 3820
},
{
"epoch": 22.934131736526947,
"grad_norm": 0.10689786076545715,
"learning_rate": 7.2640575049274e-05,
"loss": 0.0069,
"step": 3830
},
{
"epoch": 22.994011976047904,
"grad_norm": 0.1963309943675995,
"learning_rate": 7.249302711509616e-05,
"loss": 0.009,
"step": 3840
},
{
"epoch": 23.05389221556886,
"grad_norm": 0.15283331274986267,
"learning_rate": 7.23452332007775e-05,
"loss": 0.0068,
"step": 3850
},
{
"epoch": 23.11377245508982,
"grad_norm": 0.2632969319820404,
"learning_rate": 7.219719492256858e-05,
"loss": 0.0075,
"step": 3860
},
{
"epoch": 23.17365269461078,
"grad_norm": 0.17407459020614624,
"learning_rate": 7.20489138993923e-05,
"loss": 0.0072,
"step": 3870
},
{
"epoch": 23.233532934131738,
"grad_norm": 0.18360528349876404,
"learning_rate": 7.190039175282614e-05,
"loss": 0.0072,
"step": 3880
},
{
"epoch": 23.293413173652695,
"grad_norm": 0.19253474473953247,
"learning_rate": 7.175163010708455e-05,
"loss": 0.0087,
"step": 3890
},
{
"epoch": 23.353293413173652,
"grad_norm": 0.17147378623485565,
"learning_rate": 7.1602630589001e-05,
"loss": 0.0069,
"step": 3900
},
{
"epoch": 23.41317365269461,
"grad_norm": 0.1267763376235962,
"learning_rate": 7.14533948280104e-05,
"loss": 0.0057,
"step": 3910
},
{
"epoch": 23.473053892215567,
"grad_norm": 0.2177012413740158,
"learning_rate": 7.130392445613109e-05,
"loss": 0.008,
"step": 3920
},
{
"epoch": 23.53293413173653,
"grad_norm": 0.19764775037765503,
"learning_rate": 7.115422110794711e-05,
"loss": 0.007,
"step": 3930
},
{
"epoch": 23.592814371257486,
"grad_norm": 0.1799926459789276,
"learning_rate": 7.100428642059033e-05,
"loss": 0.0076,
"step": 3940
},
{
"epoch": 23.652694610778443,
"grad_norm": 0.24496091902256012,
"learning_rate": 7.08541220337224e-05,
"loss": 0.0083,
"step": 3950
},
{
"epoch": 23.7125748502994,
"grad_norm": 0.23630625009536743,
"learning_rate": 7.070372958951706e-05,
"loss": 0.0085,
"step": 3960
},
{
"epoch": 23.772455089820358,
"grad_norm": 0.2908124029636383,
"learning_rate": 7.055311073264194e-05,
"loss": 0.0087,
"step": 3970
},
{
"epoch": 23.83233532934132,
"grad_norm": 0.3030369281768799,
"learning_rate": 7.040226711024077e-05,
"loss": 0.0078,
"step": 3980
},
{
"epoch": 23.892215568862277,
"grad_norm": 0.26541629433631897,
"learning_rate": 7.02512003719152e-05,
"loss": 0.0071,
"step": 3990
},
{
"epoch": 23.952095808383234,
"grad_norm": 0.2075839638710022,
"learning_rate": 7.00999121697069e-05,
"loss": 0.0095,
"step": 4000
},
{
"epoch": 24.01197604790419,
"grad_norm": 0.20263533294200897,
"learning_rate": 6.99484041580794e-05,
"loss": 0.0089,
"step": 4010
},
{
"epoch": 24.07185628742515,
"grad_norm": 0.19597840309143066,
"learning_rate": 6.979667799390004e-05,
"loss": 0.01,
"step": 4020
},
{
"epoch": 24.131736526946106,
"grad_norm": 0.14223524928092957,
"learning_rate": 6.964473533642185e-05,
"loss": 0.0074,
"step": 4030
},
{
"epoch": 24.191616766467067,
"grad_norm": 0.13311758637428284,
"learning_rate": 6.949257784726539e-05,
"loss": 0.0067,
"step": 4040
},
{
"epoch": 24.251497005988025,
"grad_norm": 0.20629455149173737,
"learning_rate": 6.934020719040056e-05,
"loss": 0.0073,
"step": 4050
},
{
"epoch": 24.311377245508982,
"grad_norm": 0.32997483015060425,
"learning_rate": 6.918762503212848e-05,
"loss": 0.0076,
"step": 4060
},
{
"epoch": 24.37125748502994,
"grad_norm": 0.18799160420894623,
"learning_rate": 6.903483304106319e-05,
"loss": 0.0072,
"step": 4070
},
{
"epoch": 24.431137724550897,
"grad_norm": 0.1267097294330597,
"learning_rate": 6.888183288811341e-05,
"loss": 0.0076,
"step": 4080
},
{
"epoch": 24.491017964071855,
"grad_norm": 0.20079046487808228,
"learning_rate": 6.87286262464643e-05,
"loss": 0.0075,
"step": 4090
},
{
"epoch": 24.550898203592816,
"grad_norm": 0.1365060806274414,
"learning_rate": 6.857521479155915e-05,
"loss": 0.0072,
"step": 4100
},
{
"epoch": 24.610778443113773,
"grad_norm": 0.15212292969226837,
"learning_rate": 6.842160020108104e-05,
"loss": 0.0069,
"step": 4110
},
{
"epoch": 24.67065868263473,
"grad_norm": 0.1241065114736557,
"learning_rate": 6.826778415493455e-05,
"loss": 0.0063,
"step": 4120
},
{
"epoch": 24.730538922155688,
"grad_norm": 0.15440422296524048,
"learning_rate": 6.811376833522729e-05,
"loss": 0.0056,
"step": 4130
},
{
"epoch": 24.790419161676645,
"grad_norm": 0.31122419238090515,
"learning_rate": 6.795955442625159e-05,
"loss": 0.0079,
"step": 4140
},
{
"epoch": 24.850299401197606,
"grad_norm": 0.18489384651184082,
"learning_rate": 6.780514411446608e-05,
"loss": 0.0066,
"step": 4150
},
{
"epoch": 24.910179640718564,
"grad_norm": 0.17013992369174957,
"learning_rate": 6.765053908847716e-05,
"loss": 0.0069,
"step": 4160
},
{
"epoch": 24.97005988023952,
"grad_norm": 0.20781290531158447,
"learning_rate": 6.749574103902064e-05,
"loss": 0.008,
"step": 4170
},
{
"epoch": 25.02994011976048,
"grad_norm": 0.20181341469287872,
"learning_rate": 6.734075165894317e-05,
"loss": 0.0089,
"step": 4180
},
{
"epoch": 25.089820359281436,
"grad_norm": 0.1467220038175583,
"learning_rate": 6.71855726431838e-05,
"loss": 0.0067,
"step": 4190
},
{
"epoch": 25.149700598802394,
"grad_norm": 0.158147931098938,
"learning_rate": 6.703020568875538e-05,
"loss": 0.0078,
"step": 4200
},
{
"epoch": 25.209580838323355,
"grad_norm": 0.3235675096511841,
"learning_rate": 6.687465249472603e-05,
"loss": 0.012,
"step": 4210
},
{
"epoch": 25.269461077844312,
"grad_norm": 0.15191662311553955,
"learning_rate": 6.671891476220055e-05,
"loss": 0.0072,
"step": 4220
},
{
"epoch": 25.32934131736527,
"grad_norm": 0.12914660573005676,
"learning_rate": 6.656299419430183e-05,
"loss": 0.0077,
"step": 4230
},
{
"epoch": 25.389221556886227,
"grad_norm": 0.19954808056354523,
"learning_rate": 6.640689249615223e-05,
"loss": 0.0076,
"step": 4240
},
{
"epoch": 25.449101796407184,
"grad_norm": 0.12728682160377502,
"learning_rate": 6.625061137485491e-05,
"loss": 0.0079,
"step": 4250
},
{
"epoch": 25.508982035928145,
"grad_norm": 0.19139660894870758,
"learning_rate": 6.609415253947517e-05,
"loss": 0.0071,
"step": 4260
},
{
"epoch": 25.568862275449103,
"grad_norm": 0.18319684267044067,
"learning_rate": 6.593751770102178e-05,
"loss": 0.0066,
"step": 4270
},
{
"epoch": 25.62874251497006,
"grad_norm": 0.349337100982666,
"learning_rate": 6.578070857242823e-05,
"loss": 0.0088,
"step": 4280
},
{
"epoch": 25.688622754491018,
"grad_norm": 0.19814909994602203,
"learning_rate": 6.562372686853402e-05,
"loss": 0.0068,
"step": 4290
},
{
"epoch": 25.748502994011975,
"grad_norm": 0.291513592004776,
"learning_rate": 6.546657430606593e-05,
"loss": 0.008,
"step": 4300
},
{
"epoch": 25.808383233532933,
"grad_norm": 0.1787031590938568,
"learning_rate": 6.530925260361918e-05,
"loss": 0.0085,
"step": 4310
},
{
"epoch": 25.868263473053894,
"grad_norm": 0.16379982233047485,
"learning_rate": 6.515176348163871e-05,
"loss": 0.0076,
"step": 4320
},
{
"epoch": 25.92814371257485,
"grad_norm": 0.13094156980514526,
"learning_rate": 6.499410866240032e-05,
"loss": 0.0059,
"step": 4330
},
{
"epoch": 25.98802395209581,
"grad_norm": 0.11630963534116745,
"learning_rate": 6.48362898699919e-05,
"loss": 0.0062,
"step": 4340
},
{
"epoch": 26.047904191616766,
"grad_norm": 0.2187400609254837,
"learning_rate": 6.467830883029443e-05,
"loss": 0.0082,
"step": 4350
},
{
"epoch": 26.107784431137723,
"grad_norm": 0.16844452917575836,
"learning_rate": 6.452016727096326e-05,
"loss": 0.0076,
"step": 4360
},
{
"epoch": 26.16766467065868,
"grad_norm": 0.1773519665002823,
"learning_rate": 6.436186692140916e-05,
"loss": 0.0055,
"step": 4370
},
{
"epoch": 26.227544910179642,
"grad_norm": 0.17812564969062805,
"learning_rate": 6.420340951277938e-05,
"loss": 0.0062,
"step": 4380
},
{
"epoch": 26.2874251497006,
"grad_norm": 0.15265731513500214,
"learning_rate": 6.404479677793874e-05,
"loss": 0.0066,
"step": 4390
},
{
"epoch": 26.347305389221557,
"grad_norm": 0.30377325415611267,
"learning_rate": 6.388603045145075e-05,
"loss": 0.0095,
"step": 4400
},
{
"epoch": 26.407185628742514,
"grad_norm": 0.25060153007507324,
"learning_rate": 6.372711226955843e-05,
"loss": 0.0086,
"step": 4410
},
{
"epoch": 26.46706586826347,
"grad_norm": 0.23492346704006195,
"learning_rate": 6.356804397016564e-05,
"loss": 0.0076,
"step": 4420
},
{
"epoch": 26.526946107784433,
"grad_norm": 0.299795001745224,
"learning_rate": 6.340882729281779e-05,
"loss": 0.0108,
"step": 4430
},
{
"epoch": 26.58682634730539,
"grad_norm": 0.23072198033332825,
"learning_rate": 6.324946397868294e-05,
"loss": 0.0069,
"step": 4440
},
{
"epoch": 26.646706586826348,
"grad_norm": 0.34057268500328064,
"learning_rate": 6.308995577053276e-05,
"loss": 0.0085,
"step": 4450
},
{
"epoch": 26.706586826347305,
"grad_norm": 0.19974052906036377,
"learning_rate": 6.293030441272347e-05,
"loss": 0.0062,
"step": 4460
},
{
"epoch": 26.766467065868262,
"grad_norm": 0.150442436337471,
"learning_rate": 6.277051165117677e-05,
"loss": 0.0085,
"step": 4470
},
{
"epoch": 26.82634730538922,
"grad_norm": 0.159685418009758,
"learning_rate": 6.261057923336064e-05,
"loss": 0.0068,
"step": 4480
},
{
"epoch": 26.88622754491018,
"grad_norm": 0.13874687254428864,
"learning_rate": 6.245050890827042e-05,
"loss": 0.0067,
"step": 4490
},
{
"epoch": 26.94610778443114,
"grad_norm": 0.14795225858688354,
"learning_rate": 6.229030242640952e-05,
"loss": 0.0063,
"step": 4500
},
{
"epoch": 27.005988023952096,
"grad_norm": 0.15675286948680878,
"learning_rate": 6.212996153977037e-05,
"loss": 0.0096,
"step": 4510
},
{
"epoch": 27.065868263473053,
"grad_norm": 0.2097497284412384,
"learning_rate": 6.196948800181523e-05,
"loss": 0.0073,
"step": 4520
},
{
"epoch": 27.12574850299401,
"grad_norm": 0.22438617050647736,
"learning_rate": 6.180888356745695e-05,
"loss": 0.0064,
"step": 4530
},
{
"epoch": 27.18562874251497,
"grad_norm": 0.14291080832481384,
"learning_rate": 6.164814999303995e-05,
"loss": 0.0058,
"step": 4540
},
{
"epoch": 27.24550898203593,
"grad_norm": 0.18571148812770844,
"learning_rate": 6.148728903632081e-05,
"loss": 0.0069,
"step": 4550
},
{
"epoch": 27.305389221556887,
"grad_norm": 0.25261014699935913,
"learning_rate": 6.132630245644921e-05,
"loss": 0.0071,
"step": 4560
},
{
"epoch": 27.365269461077844,
"grad_norm": 0.3090629577636719,
"learning_rate": 6.116519201394857e-05,
"loss": 0.0079,
"step": 4570
},
{
"epoch": 27.4251497005988,
"grad_norm": 0.19101376831531525,
"learning_rate": 6.10039594706969e-05,
"loss": 0.0093,
"step": 4580
},
{
"epoch": 27.48502994011976,
"grad_norm": 0.18836107850074768,
"learning_rate": 6.084260658990744e-05,
"loss": 0.0068,
"step": 4590
},
{
"epoch": 27.54491017964072,
"grad_norm": 0.26658037304878235,
"learning_rate": 6.068113513610943e-05,
"loss": 0.0077,
"step": 4600
},
{
"epoch": 27.604790419161677,
"grad_norm": 0.3041040003299713,
"learning_rate": 6.0519546875128876e-05,
"loss": 0.0075,
"step": 4610
},
{
"epoch": 27.664670658682635,
"grad_norm": 0.10149551182985306,
"learning_rate": 6.035784357406906e-05,
"loss": 0.0078,
"step": 4620
},
{
"epoch": 27.724550898203592,
"grad_norm": 0.12548454105854034,
"learning_rate": 6.01960270012914e-05,
"loss": 0.0062,
"step": 4630
},
{
"epoch": 27.78443113772455,
"grad_norm": 0.12445248663425446,
"learning_rate": 6.003409892639599e-05,
"loss": 0.0066,
"step": 4640
},
{
"epoch": 27.84431137724551,
"grad_norm": 0.17534667253494263,
"learning_rate": 5.9872061120202336e-05,
"loss": 0.0069,
"step": 4650
},
{
"epoch": 27.904191616766468,
"grad_norm": 0.12828294932842255,
"learning_rate": 5.9709915354729914e-05,
"loss": 0.006,
"step": 4660
},
{
"epoch": 27.964071856287426,
"grad_norm": 0.216370090842247,
"learning_rate": 5.9547663403178824e-05,
"loss": 0.0079,
"step": 4670
},
{
"epoch": 28.023952095808383,
"grad_norm": 0.12458262592554092,
"learning_rate": 5.9385307039910445e-05,
"loss": 0.0079,
"step": 4680
},
{
"epoch": 28.08383233532934,
"grad_norm": 0.10598327219486237,
"learning_rate": 5.922284804042792e-05,
"loss": 0.0074,
"step": 4690
},
{
"epoch": 28.143712574850298,
"grad_norm": 0.16289222240447998,
"learning_rate": 5.906028818135687e-05,
"loss": 0.0068,
"step": 4700
},
{
"epoch": 28.20359281437126,
"grad_norm": 0.19962510466575623,
"learning_rate": 5.889762924042585e-05,
"loss": 0.006,
"step": 4710
},
{
"epoch": 28.263473053892216,
"grad_norm": 0.22746624052524567,
"learning_rate": 5.873487299644699e-05,
"loss": 0.0076,
"step": 4720
},
{
"epoch": 28.323353293413174,
"grad_norm": 0.26980334520339966,
"learning_rate": 5.857202122929649e-05,
"loss": 0.0083,
"step": 4730
},
{
"epoch": 28.38323353293413,
"grad_norm": 0.22785213589668274,
"learning_rate": 5.840907571989518e-05,
"loss": 0.0067,
"step": 4740
},
{
"epoch": 28.44311377245509,
"grad_norm": 0.21585190296173096,
"learning_rate": 5.824603825018904e-05,
"loss": 0.0068,
"step": 4750
},
{
"epoch": 28.50299401197605,
"grad_norm": 0.20298276841640472,
"learning_rate": 5.808291060312975e-05,
"loss": 0.0078,
"step": 4760
},
{
"epoch": 28.562874251497007,
"grad_norm": 0.10802645236253738,
"learning_rate": 5.7919694562655083e-05,
"loss": 0.0056,
"step": 4770
},
{
"epoch": 28.622754491017965,
"grad_norm": 0.11957808583974838,
"learning_rate": 5.775639191366954e-05,
"loss": 0.0078,
"step": 4780
},
{
"epoch": 28.682634730538922,
"grad_norm": 0.2116246074438095,
"learning_rate": 5.75930044420247e-05,
"loss": 0.0078,
"step": 4790
},
{
"epoch": 28.74251497005988,
"grad_norm": 0.18922735750675201,
"learning_rate": 5.74295339344998e-05,
"loss": 0.0065,
"step": 4800
},
{
"epoch": 28.802395209580837,
"grad_norm": 0.22901888191699982,
"learning_rate": 5.726598217878211e-05,
"loss": 0.0068,
"step": 4810
},
{
"epoch": 28.862275449101798,
"grad_norm": 0.22344359755516052,
"learning_rate": 5.71023509634474e-05,
"loss": 0.0075,
"step": 4820
},
{
"epoch": 28.922155688622755,
"grad_norm": 0.12992268800735474,
"learning_rate": 5.693864207794049e-05,
"loss": 0.0059,
"step": 4830
},
{
"epoch": 28.982035928143713,
"grad_norm": 0.23563134670257568,
"learning_rate": 5.677485731255545e-05,
"loss": 0.0061,
"step": 4840
},
{
"epoch": 29.04191616766467,
"grad_norm": 0.22537508606910706,
"learning_rate": 5.6610998458416296e-05,
"loss": 0.0082,
"step": 4850
},
{
"epoch": 29.101796407185628,
"grad_norm": 0.23789671063423157,
"learning_rate": 5.644706730745716e-05,
"loss": 0.0077,
"step": 4860
},
{
"epoch": 29.161676646706585,
"grad_norm": 0.21229776740074158,
"learning_rate": 5.628306565240287e-05,
"loss": 0.0065,
"step": 4870
},
{
"epoch": 29.221556886227546,
"grad_norm": 0.19470864534378052,
"learning_rate": 5.611899528674923e-05,
"loss": 0.0071,
"step": 4880
},
{
"epoch": 29.281437125748504,
"grad_norm": 0.1888543665409088,
"learning_rate": 5.595485800474349e-05,
"loss": 0.0066,
"step": 4890
},
{
"epoch": 29.34131736526946,
"grad_norm": 0.22629915177822113,
"learning_rate": 5.579065560136467e-05,
"loss": 0.0074,
"step": 4900
},
{
"epoch": 29.40119760479042,
"grad_norm": 0.16065622866153717,
"learning_rate": 5.562638987230392e-05,
"loss": 0.0058,
"step": 4910
},
{
"epoch": 29.461077844311376,
"grad_norm": 0.13282473385334015,
"learning_rate": 5.546206261394498e-05,
"loss": 0.006,
"step": 4920
},
{
"epoch": 29.520958083832337,
"grad_norm": 0.18144407868385315,
"learning_rate": 5.529767562334437e-05,
"loss": 0.0058,
"step": 4930
},
{
"epoch": 29.580838323353294,
"grad_norm": 0.1158028319478035,
"learning_rate": 5.5133230698211926e-05,
"loss": 0.0062,
"step": 4940
},
{
"epoch": 29.64071856287425,
"grad_norm": 0.22190764546394348,
"learning_rate": 5.496872963689096e-05,
"loss": 0.0071,
"step": 4950
},
{
"epoch": 29.70059880239521,
"grad_norm": 0.0977877601981163,
"learning_rate": 5.4804174238338756e-05,
"loss": 0.0065,
"step": 4960
},
{
"epoch": 29.760479041916167,
"grad_norm": 0.13936443626880646,
"learning_rate": 5.463956630210678e-05,
"loss": 0.0053,
"step": 4970
},
{
"epoch": 29.820359281437124,
"grad_norm": 0.1428326964378357,
"learning_rate": 5.4474907628321046e-05,
"loss": 0.006,
"step": 4980
},
{
"epoch": 29.880239520958085,
"grad_norm": 0.12146999686956406,
"learning_rate": 5.431020001766244e-05,
"loss": 0.0059,
"step": 4990
},
{
"epoch": 29.940119760479043,
"grad_norm": 0.12373974919319153,
"learning_rate": 5.4145445271346986e-05,
"loss": 0.0056,
"step": 5000
},
{
"epoch": 30.0,
"grad_norm": 0.10870573669672012,
"learning_rate": 5.398064519110622e-05,
"loss": 0.0095,
"step": 5010
},
{
"epoch": 30.059880239520957,
"grad_norm": 0.21035031974315643,
"learning_rate": 5.3815801579167394e-05,
"loss": 0.006,
"step": 5020
},
{
"epoch": 30.119760479041915,
"grad_norm": 0.16052226722240448,
"learning_rate": 5.365091623823382e-05,
"loss": 0.0048,
"step": 5030
},
{
"epoch": 30.179640718562876,
"grad_norm": 0.1395321637392044,
"learning_rate": 5.348599097146521e-05,
"loss": 0.0058,
"step": 5040
},
{
"epoch": 30.239520958083833,
"grad_norm": 0.14141885936260223,
"learning_rate": 5.3321027582457836e-05,
"loss": 0.0054,
"step": 5050
},
{
"epoch": 30.29940119760479,
"grad_norm": 0.2570410370826721,
"learning_rate": 5.315602787522491e-05,
"loss": 0.0064,
"step": 5060
},
{
"epoch": 30.35928143712575,
"grad_norm": 0.15067161619663239,
"learning_rate": 5.299099365417678e-05,
"loss": 0.0063,
"step": 5070
},
{
"epoch": 30.419161676646706,
"grad_norm": 0.25824785232543945,
"learning_rate": 5.2825926724101236e-05,
"loss": 0.0072,
"step": 5080
},
{
"epoch": 30.479041916167663,
"grad_norm": 0.22695602476596832,
"learning_rate": 5.26608288901438e-05,
"loss": 0.0105,
"step": 5090
},
{
"epoch": 30.538922155688624,
"grad_norm": 0.22027285397052765,
"learning_rate": 5.24957019577879e-05,
"loss": 0.0077,
"step": 5100
},
{
"epoch": 30.59880239520958,
"grad_norm": 0.20709674060344696,
"learning_rate": 5.2330547732835266e-05,
"loss": 0.01,
"step": 5110
},
{
"epoch": 30.65868263473054,
"grad_norm": 0.1574249565601349,
"learning_rate": 5.2165368021385996e-05,
"loss": 0.0074,
"step": 5120
},
{
"epoch": 30.718562874251496,
"grad_norm": 0.16116395592689514,
"learning_rate": 5.200016462981897e-05,
"loss": 0.0076,
"step": 5130
},
{
"epoch": 30.778443113772454,
"grad_norm": 0.21255122125148773,
"learning_rate": 5.1834939364772015e-05,
"loss": 0.0064,
"step": 5140
},
{
"epoch": 30.83832335329341,
"grad_norm": 0.10041582584381104,
"learning_rate": 5.166969403312214e-05,
"loss": 0.0048,
"step": 5150
},
{
"epoch": 30.898203592814372,
"grad_norm": 0.22462719678878784,
"learning_rate": 5.1504430441965844e-05,
"loss": 0.0065,
"step": 5160
},
{
"epoch": 30.95808383233533,
"grad_norm": 0.2707398533821106,
"learning_rate": 5.133915039859923e-05,
"loss": 0.0058,
"step": 5170
},
{
"epoch": 31.017964071856287,
"grad_norm": 0.15304575860500336,
"learning_rate": 5.1173855710498444e-05,
"loss": 0.0051,
"step": 5180
},
{
"epoch": 31.077844311377245,
"grad_norm": 0.12331144511699677,
"learning_rate": 5.100854818529967e-05,
"loss": 0.0055,
"step": 5190
},
{
"epoch": 31.137724550898202,
"grad_norm": 0.1182287186384201,
"learning_rate": 5.084322963077951e-05,
"loss": 0.006,
"step": 5200
},
{
"epoch": 31.197604790419163,
"grad_norm": 0.18593420088291168,
"learning_rate": 5.067790185483522e-05,
"loss": 0.0063,
"step": 5210
},
{
"epoch": 31.25748502994012,
"grad_norm": 0.28433340787887573,
"learning_rate": 5.0512566665464844e-05,
"loss": 0.0063,
"step": 5220
},
{
"epoch": 31.317365269461078,
"grad_norm": 0.20429112017154694,
"learning_rate": 5.034722587074755e-05,
"loss": 0.0055,
"step": 5230
},
{
"epoch": 31.377245508982035,
"grad_norm": 0.14675700664520264,
"learning_rate": 5.018188127882375e-05,
"loss": 0.0061,
"step": 5240
},
{
"epoch": 31.437125748502993,
"grad_norm": 0.1851230412721634,
"learning_rate": 5.0016534697875417e-05,
"loss": 0.0055,
"step": 5250
},
{
"epoch": 31.49700598802395,
"grad_norm": 0.17463643848896027,
"learning_rate": 4.9851187936106294e-05,
"loss": 0.0072,
"step": 5260
},
{
"epoch": 31.55688622754491,
"grad_norm": 0.2184799462556839,
"learning_rate": 4.968584280172206e-05,
"loss": 0.0062,
"step": 5270
},
{
"epoch": 31.61676646706587,
"grad_norm": 0.3347960114479065,
"learning_rate": 4.95205011029106e-05,
"loss": 0.0056,
"step": 5280
},
{
"epoch": 31.676646706586826,
"grad_norm": 0.22836114466190338,
"learning_rate": 4.935516464782227e-05,
"loss": 0.0076,
"step": 5290
},
{
"epoch": 31.736526946107784,
"grad_norm": 0.1752859652042389,
"learning_rate": 4.918983524455003e-05,
"loss": 0.0079,
"step": 5300
},
{
"epoch": 31.79640718562874,
"grad_norm": 0.22407758235931396,
"learning_rate": 4.9024514701109766e-05,
"loss": 0.0059,
"step": 5310
},
{
"epoch": 31.856287425149702,
"grad_norm": 0.1734432727098465,
"learning_rate": 4.885920482542043e-05,
"loss": 0.0079,
"step": 5320
},
{
"epoch": 31.91616766467066,
"grad_norm": 0.1189454197883606,
"learning_rate": 4.869390742528438e-05,
"loss": 0.0063,
"step": 5330
},
{
"epoch": 31.976047904191617,
"grad_norm": 0.11871115118265152,
"learning_rate": 4.852862430836744e-05,
"loss": 0.0065,
"step": 5340
},
{
"epoch": 32.035928143712574,
"grad_norm": 0.15404345095157623,
"learning_rate": 4.836335728217933e-05,
"loss": 0.0055,
"step": 5350
},
{
"epoch": 32.09580838323353,
"grad_norm": 0.1752903312444687,
"learning_rate": 4.819810815405379e-05,
"loss": 0.0051,
"step": 5360
},
{
"epoch": 32.15568862275449,
"grad_norm": 0.1483910083770752,
"learning_rate": 4.803287873112877e-05,
"loss": 0.0063,
"step": 5370
},
{
"epoch": 32.21556886227545,
"grad_norm": 0.19743403792381287,
"learning_rate": 4.786767082032681e-05,
"loss": 0.0052,
"step": 5380
},
{
"epoch": 32.275449101796404,
"grad_norm": 0.12940751016139984,
"learning_rate": 4.77024862283351e-05,
"loss": 0.0056,
"step": 5390
},
{
"epoch": 32.33532934131736,
"grad_norm": 0.1173861026763916,
"learning_rate": 4.753732676158593e-05,
"loss": 0.0046,
"step": 5400
},
{
"epoch": 32.395209580838326,
"grad_norm": 0.06855036318302155,
"learning_rate": 4.737219422623672e-05,
"loss": 0.0052,
"step": 5410
},
{
"epoch": 32.455089820359284,
"grad_norm": 0.1282402127981186,
"learning_rate": 4.720709042815044e-05,
"loss": 0.0044,
"step": 5420
},
{
"epoch": 32.51497005988024,
"grad_norm": 0.1340729296207428,
"learning_rate": 4.704201717287578e-05,
"loss": 0.0074,
"step": 5430
},
{
"epoch": 32.5748502994012,
"grad_norm": 0.14398351311683655,
"learning_rate": 4.6876976265627404e-05,
"loss": 0.0065,
"step": 5440
},
{
"epoch": 32.634730538922156,
"grad_norm": 0.1844923049211502,
"learning_rate": 4.671196951126626e-05,
"loss": 0.0056,
"step": 5450
},
{
"epoch": 32.69461077844311,
"grad_norm": 0.1355842649936676,
"learning_rate": 4.654699871427971e-05,
"loss": 0.0047,
"step": 5460
},
{
"epoch": 32.75449101796407,
"grad_norm": 0.15436507761478424,
"learning_rate": 4.6382065678762034e-05,
"loss": 0.0052,
"step": 5470
},
{
"epoch": 32.81437125748503,
"grad_norm": 0.1131393164396286,
"learning_rate": 4.6217172208394424e-05,
"loss": 0.005,
"step": 5480
},
{
"epoch": 32.874251497005986,
"grad_norm": 0.15003302693367004,
"learning_rate": 4.605232010642549e-05,
"loss": 0.0063,
"step": 5490
},
{
"epoch": 32.93413173652694,
"grad_norm": 0.134606271982193,
"learning_rate": 4.588751117565142e-05,
"loss": 0.0054,
"step": 5500
},
{
"epoch": 32.9940119760479,
"grad_norm": 0.1801108568906784,
"learning_rate": 4.5722747218396214e-05,
"loss": 0.0067,
"step": 5510
},
{
"epoch": 33.053892215568865,
"grad_norm": 0.188354030251503,
"learning_rate": 4.5558030036492194e-05,
"loss": 0.0065,
"step": 5520
},
{
"epoch": 33.11377245508982,
"grad_norm": 0.11942395567893982,
"learning_rate": 4.539336143125999e-05,
"loss": 0.0053,
"step": 5530
},
{
"epoch": 33.17365269461078,
"grad_norm": 0.12230503559112549,
"learning_rate": 4.522874320348916e-05,
"loss": 0.0061,
"step": 5540
},
{
"epoch": 33.23353293413174,
"grad_norm": 0.17051735520362854,
"learning_rate": 4.506417715341821e-05,
"loss": 0.0057,
"step": 5550
},
{
"epoch": 33.293413173652695,
"grad_norm": 0.13177450001239777,
"learning_rate": 4.489966508071511e-05,
"loss": 0.0066,
"step": 5560
},
{
"epoch": 33.35329341317365,
"grad_norm": 0.17847491800785065,
"learning_rate": 4.4735208784457575e-05,
"loss": 0.0059,
"step": 5570
},
{
"epoch": 33.41317365269461,
"grad_norm": 0.18818309903144836,
"learning_rate": 4.457081006311325e-05,
"loss": 0.0051,
"step": 5580
},
{
"epoch": 33.47305389221557,
"grad_norm": 0.17753636837005615,
"learning_rate": 4.440647071452027e-05,
"loss": 0.0056,
"step": 5590
},
{
"epoch": 33.532934131736525,
"grad_norm": 0.20678026974201202,
"learning_rate": 4.424219253586737e-05,
"loss": 0.0062,
"step": 5600
},
{
"epoch": 33.59281437125748,
"grad_norm": 0.13742487132549286,
"learning_rate": 4.407797732367443e-05,
"loss": 0.0056,
"step": 5610
},
{
"epoch": 33.65269461077844,
"grad_norm": 0.13119739294052124,
"learning_rate": 4.391382687377268e-05,
"loss": 0.0061,
"step": 5620
},
{
"epoch": 33.712574850299404,
"grad_norm": 0.20738616585731506,
"learning_rate": 4.374974298128512e-05,
"loss": 0.0062,
"step": 5630
},
{
"epoch": 33.77245508982036,
"grad_norm": 0.17375703155994415,
"learning_rate": 4.358572744060699e-05,
"loss": 0.0072,
"step": 5640
},
{
"epoch": 33.83233532934132,
"grad_norm": 0.310068815946579,
"learning_rate": 4.342178204538588e-05,
"loss": 0.0065,
"step": 5650
},
{
"epoch": 33.89221556886228,
"grad_norm": 0.19243353605270386,
"learning_rate": 4.325790858850241e-05,
"loss": 0.0052,
"step": 5660
},
{
"epoch": 33.952095808383234,
"grad_norm": 0.2731423079967499,
"learning_rate": 4.309410886205043e-05,
"loss": 0.0071,
"step": 5670
},
{
"epoch": 34.01197604790419,
"grad_norm": 0.21550576388835907,
"learning_rate": 4.293038465731752e-05,
"loss": 0.0054,
"step": 5680
},
{
"epoch": 34.07185628742515,
"grad_norm": 0.17855776846408844,
"learning_rate": 4.276673776476533e-05,
"loss": 0.0061,
"step": 5690
},
{
"epoch": 34.131736526946106,
"grad_norm": 0.31720563769340515,
"learning_rate": 4.260316997401007e-05,
"loss": 0.0064,
"step": 5700
},
{
"epoch": 34.191616766467064,
"grad_norm": 0.22748303413391113,
"learning_rate": 4.243968307380293e-05,
"loss": 0.0053,
"step": 5710
},
{
"epoch": 34.25149700598802,
"grad_norm": 0.09045732766389847,
"learning_rate": 4.22762788520104e-05,
"loss": 0.0051,
"step": 5720
},
{
"epoch": 34.31137724550898,
"grad_norm": 0.10661271959543228,
"learning_rate": 4.211295909559491e-05,
"loss": 0.0066,
"step": 5730
},
{
"epoch": 34.37125748502994,
"grad_norm": 0.16480374336242676,
"learning_rate": 4.194972559059511e-05,
"loss": 0.0049,
"step": 5740
},
{
"epoch": 34.4311377245509,
"grad_norm": 0.1724926382303238,
"learning_rate": 4.178658012210651e-05,
"loss": 0.0067,
"step": 5750
},
{
"epoch": 34.49101796407186,
"grad_norm": 0.10128136724233627,
"learning_rate": 4.162352447426177e-05,
"loss": 0.0066,
"step": 5760
},
{
"epoch": 34.550898203592816,
"grad_norm": 0.09300406277179718,
"learning_rate": 4.146056043021135e-05,
"loss": 0.0043,
"step": 5770
},
{
"epoch": 34.61077844311377,
"grad_norm": 0.13379859924316406,
"learning_rate": 4.1297689772103944e-05,
"loss": 0.0047,
"step": 5780
},
{
"epoch": 34.67065868263473,
"grad_norm": 0.2153003066778183,
"learning_rate": 4.113491428106694e-05,
"loss": 0.0051,
"step": 5790
},
{
"epoch": 34.73053892215569,
"grad_norm": 0.22840136289596558,
"learning_rate": 4.0972235737187055e-05,
"loss": 0.0055,
"step": 5800
},
{
"epoch": 34.790419161676645,
"grad_norm": 0.15960967540740967,
"learning_rate": 4.080965591949076e-05,
"loss": 0.006,
"step": 5810
},
{
"epoch": 34.8502994011976,
"grad_norm": 0.1118282824754715,
"learning_rate": 4.0647176605924924e-05,
"loss": 0.005,
"step": 5820
},
{
"epoch": 34.91017964071856,
"grad_norm": 0.22873201966285706,
"learning_rate": 4.0484799573337255e-05,
"loss": 0.0064,
"step": 5830
},
{
"epoch": 34.97005988023952,
"grad_norm": 0.08231984078884125,
"learning_rate": 4.032252659745699e-05,
"loss": 0.0047,
"step": 5840
},
{
"epoch": 35.02994011976048,
"grad_norm": 0.16873399913311005,
"learning_rate": 4.016035945287539e-05,
"loss": 0.0048,
"step": 5850
},
{
"epoch": 35.08982035928144,
"grad_norm": 0.23176051676273346,
"learning_rate": 3.999829991302635e-05,
"loss": 0.0067,
"step": 5860
},
{
"epoch": 35.1497005988024,
"grad_norm": 0.30874305963516235,
"learning_rate": 3.983634975016707e-05,
"loss": 0.0071,
"step": 5870
},
{
"epoch": 35.209580838323355,
"grad_norm": 0.23135434091091156,
"learning_rate": 3.967451073535854e-05,
"loss": 0.0054,
"step": 5880
},
{
"epoch": 35.26946107784431,
"grad_norm": 0.20160160958766937,
"learning_rate": 3.951278463844633e-05,
"loss": 0.0061,
"step": 5890
},
{
"epoch": 35.32934131736527,
"grad_norm": 0.10719034075737,
"learning_rate": 3.935117322804111e-05,
"loss": 0.0062,
"step": 5900
},
{
"epoch": 35.38922155688623,
"grad_norm": 0.14799779653549194,
"learning_rate": 3.918967827149938e-05,
"loss": 0.0068,
"step": 5910
},
{
"epoch": 35.449101796407184,
"grad_norm": 0.10325650125741959,
"learning_rate": 3.9028301534904094e-05,
"loss": 0.0059,
"step": 5920
},
{
"epoch": 35.50898203592814,
"grad_norm": 0.18317624926567078,
"learning_rate": 3.88670447830454e-05,
"loss": 0.0056,
"step": 5930
},
{
"epoch": 35.5688622754491,
"grad_norm": 0.16398105025291443,
"learning_rate": 3.870590977940132e-05,
"loss": 0.0046,
"step": 5940
},
{
"epoch": 35.62874251497006,
"grad_norm": 0.17062370479106903,
"learning_rate": 3.8544898286118404e-05,
"loss": 0.0065,
"step": 5950
},
{
"epoch": 35.68862275449102,
"grad_norm": 0.11496248841285706,
"learning_rate": 3.838401206399257e-05,
"loss": 0.0048,
"step": 5960
},
{
"epoch": 35.74850299401198,
"grad_norm": 0.2242138683795929,
"learning_rate": 3.822325287244975e-05,
"loss": 0.0057,
"step": 5970
},
{
"epoch": 35.808383233532936,
"grad_norm": 0.13879981637001038,
"learning_rate": 3.8062622469526725e-05,
"loss": 0.006,
"step": 5980
},
{
"epoch": 35.868263473053894,
"grad_norm": 0.218858003616333,
"learning_rate": 3.790212261185183e-05,
"loss": 0.0057,
"step": 5990
},
{
"epoch": 35.92814371257485,
"grad_norm": 0.17514842748641968,
"learning_rate": 3.7741755054625794e-05,
"loss": 0.0068,
"step": 6000
},
{
"epoch": 35.98802395209581,
"grad_norm": 0.13060270249843597,
"learning_rate": 3.758152155160255e-05,
"loss": 0.0069,
"step": 6010
},
{
"epoch": 36.047904191616766,
"grad_norm": 0.1274816393852234,
"learning_rate": 3.742142385506999e-05,
"loss": 0.0043,
"step": 6020
},
{
"epoch": 36.10778443113772,
"grad_norm": 0.21617907285690308,
"learning_rate": 3.72614637158309e-05,
"loss": 0.0056,
"step": 6030
},
{
"epoch": 36.16766467065868,
"grad_norm": 0.2352694272994995,
"learning_rate": 3.710164288318371e-05,
"loss": 0.0055,
"step": 6040
},
{
"epoch": 36.22754491017964,
"grad_norm": 0.19244319200515747,
"learning_rate": 3.694196310490345e-05,
"loss": 0.0069,
"step": 6050
},
{
"epoch": 36.287425149700596,
"grad_norm": 0.1213480681180954,
"learning_rate": 3.678242612722259e-05,
"loss": 0.0058,
"step": 6060
},
{
"epoch": 36.34730538922156,
"grad_norm": 0.14737939834594727,
"learning_rate": 3.6623033694811953e-05,
"loss": 0.0049,
"step": 6070
},
{
"epoch": 36.40718562874252,
"grad_norm": 0.26000404357910156,
"learning_rate": 3.6463787550761665e-05,
"loss": 0.0072,
"step": 6080
},
{
"epoch": 36.467065868263475,
"grad_norm": 0.13417655229568481,
"learning_rate": 3.630468943656202e-05,
"loss": 0.0065,
"step": 6090
},
{
"epoch": 36.52694610778443,
"grad_norm": 0.18231870234012604,
"learning_rate": 3.6145741092084523e-05,
"loss": 0.005,
"step": 6100
},
{
"epoch": 36.58682634730539,
"grad_norm": 0.20071330666542053,
"learning_rate": 3.598694425556278e-05,
"loss": 0.0078,
"step": 6110
},
{
"epoch": 36.64670658682635,
"grad_norm": 0.16619427502155304,
"learning_rate": 3.58283006635736e-05,
"loss": 0.0058,
"step": 6120
},
{
"epoch": 36.706586826347305,
"grad_norm": 0.19290019571781158,
"learning_rate": 3.566981205101781e-05,
"loss": 0.0062,
"step": 6130
},
{
"epoch": 36.76646706586826,
"grad_norm": 0.1627812683582306,
"learning_rate": 3.5511480151101556e-05,
"loss": 0.0052,
"step": 6140
},
{
"epoch": 36.82634730538922,
"grad_norm": 0.13674411177635193,
"learning_rate": 3.5353306695317104e-05,
"loss": 0.0058,
"step": 6150
},
{
"epoch": 36.88622754491018,
"grad_norm": 0.27093377709388733,
"learning_rate": 3.519529341342402e-05,
"loss": 0.0076,
"step": 6160
},
{
"epoch": 36.946107784431135,
"grad_norm": 0.1490563303232193,
"learning_rate": 3.503744203343026e-05,
"loss": 0.0051,
"step": 6170
},
{
"epoch": 37.0059880239521,
"grad_norm": 0.15254539251327515,
"learning_rate": 3.487975428157318e-05,
"loss": 0.0057,
"step": 6180
},
{
"epoch": 37.06586826347306,
"grad_norm": 0.2596532702445984,
"learning_rate": 3.472223188230083e-05,
"loss": 0.0055,
"step": 6190
},
{
"epoch": 37.125748502994014,
"grad_norm": 0.1940075159072876,
"learning_rate": 3.4564876558252866e-05,
"loss": 0.0051,
"step": 6200
},
{
"epoch": 37.18562874251497,
"grad_norm": 0.20725709199905396,
"learning_rate": 3.440769003024195e-05,
"loss": 0.0053,
"step": 6210
},
{
"epoch": 37.24550898203593,
"grad_norm": 0.24404729902744293,
"learning_rate": 3.425067401723477e-05,
"loss": 0.0059,
"step": 6220
},
{
"epoch": 37.30538922155689,
"grad_norm": 0.26057863235473633,
"learning_rate": 3.409383023633325e-05,
"loss": 0.006,
"step": 6230
},
{
"epoch": 37.365269461077844,
"grad_norm": 0.13587641716003418,
"learning_rate": 3.3937160402755894e-05,
"loss": 0.0049,
"step": 6240
},
{
"epoch": 37.4251497005988,
"grad_norm": 0.09946554899215698,
"learning_rate": 3.378066622981885e-05,
"loss": 0.005,
"step": 6250
},
{
"epoch": 37.48502994011976,
"grad_norm": 0.17789135873317719,
"learning_rate": 3.362434942891738e-05,
"loss": 0.0046,
"step": 6260
},
{
"epoch": 37.544910179640716,
"grad_norm": 0.15340478718280792,
"learning_rate": 3.346821170950693e-05,
"loss": 0.0054,
"step": 6270
},
{
"epoch": 37.604790419161674,
"grad_norm": 0.07373283058404922,
"learning_rate": 3.3312254779084585e-05,
"loss": 0.0049,
"step": 6280
},
{
"epoch": 37.66467065868264,
"grad_norm": 0.13324517011642456,
"learning_rate": 3.315648034317039e-05,
"loss": 0.0058,
"step": 6290
},
{
"epoch": 37.724550898203596,
"grad_norm": 0.11624062806367874,
"learning_rate": 3.3000890105288564e-05,
"loss": 0.0057,
"step": 6300
},
{
"epoch": 37.78443113772455,
"grad_norm": 0.11727824807167053,
"learning_rate": 3.284548576694908e-05,
"loss": 0.0074,
"step": 6310
},
{
"epoch": 37.84431137724551,
"grad_norm": 0.14204180240631104,
"learning_rate": 3.2690269027628815e-05,
"loss": 0.0054,
"step": 6320
},
{
"epoch": 37.90419161676647,
"grad_norm": 0.11533647775650024,
"learning_rate": 3.253524158475324e-05,
"loss": 0.0057,
"step": 6330
},
{
"epoch": 37.964071856287426,
"grad_norm": 0.1671261340379715,
"learning_rate": 3.238040513367757e-05,
"loss": 0.0058,
"step": 6340
},
{
"epoch": 38.02395209580838,
"grad_norm": 0.17862123250961304,
"learning_rate": 3.222576136766843e-05,
"loss": 0.0066,
"step": 6350
},
{
"epoch": 38.08383233532934,
"grad_norm": 0.18684110045433044,
"learning_rate": 3.2071311977885324e-05,
"loss": 0.005,
"step": 6360
},
{
"epoch": 38.1437125748503,
"grad_norm": 0.18058808147907257,
"learning_rate": 3.191705865336197e-05,
"loss": 0.006,
"step": 6370
},
{
"epoch": 38.203592814371255,
"grad_norm": 0.1999422162771225,
"learning_rate": 3.1763003080988075e-05,
"loss": 0.0056,
"step": 6380
},
{
"epoch": 38.26347305389221,
"grad_norm": 0.21451738476753235,
"learning_rate": 3.160914694549063e-05,
"loss": 0.0061,
"step": 6390
},
{
"epoch": 38.32335329341317,
"grad_norm": 0.2804773449897766,
"learning_rate": 3.145549192941573e-05,
"loss": 0.0058,
"step": 6400
},
{
"epoch": 38.383233532934135,
"grad_norm": 0.18604964017868042,
"learning_rate": 3.130203971310999e-05,
"loss": 0.0052,
"step": 6410
},
{
"epoch": 38.44311377245509,
"grad_norm": 0.19763803482055664,
"learning_rate": 3.114879197470225e-05,
"loss": 0.0052,
"step": 6420
},
{
"epoch": 38.50299401197605,
"grad_norm": 0.259390264749527,
"learning_rate": 3.0995750390085285e-05,
"loss": 0.0068,
"step": 6430
},
{
"epoch": 38.56287425149701,
"grad_norm": 0.10210886597633362,
"learning_rate": 3.084291663289728e-05,
"loss": 0.0051,
"step": 6440
},
{
"epoch": 38.622754491017965,
"grad_norm": 0.2174130380153656,
"learning_rate": 3.069029237450375e-05,
"loss": 0.0061,
"step": 6450
},
{
"epoch": 38.68263473053892,
"grad_norm": 0.12058410048484802,
"learning_rate": 3.053787928397911e-05,
"loss": 0.0043,
"step": 6460
},
{
"epoch": 38.74251497005988,
"grad_norm": 0.07876723259687424,
"learning_rate": 3.0385679028088526e-05,
"loss": 0.0049,
"step": 6470
},
{
"epoch": 38.80239520958084,
"grad_norm": 0.11566948890686035,
"learning_rate": 3.023369327126959e-05,
"loss": 0.0044,
"step": 6480
},
{
"epoch": 38.862275449101794,
"grad_norm": 0.09109222888946533,
"learning_rate": 3.0081923675614198e-05,
"loss": 0.0044,
"step": 6490
},
{
"epoch": 38.92215568862275,
"grad_norm": 0.11482509225606918,
"learning_rate": 2.993037190085034e-05,
"loss": 0.0044,
"step": 6500
},
{
"epoch": 38.98203592814371,
"grad_norm": 0.15522947907447815,
"learning_rate": 2.977903960432392e-05,
"loss": 0.0051,
"step": 6510
},
{
"epoch": 39.041916167664674,
"grad_norm": 0.19587208330631256,
"learning_rate": 2.9627928440980722e-05,
"loss": 0.0043,
"step": 6520
},
{
"epoch": 39.10179640718563,
"grad_norm": 0.15015973150730133,
"learning_rate": 2.9477040063348183e-05,
"loss": 0.0055,
"step": 6530
},
{
"epoch": 39.16167664670659,
"grad_norm": 0.13606302440166473,
"learning_rate": 2.9326376121517456e-05,
"loss": 0.0046,
"step": 6540
},
{
"epoch": 39.221556886227546,
"grad_norm": 0.12762711942195892,
"learning_rate": 2.9175938263125236e-05,
"loss": 0.0054,
"step": 6550
},
{
"epoch": 39.2814371257485,
"grad_norm": 0.2542532980442047,
"learning_rate": 2.9025728133335873e-05,
"loss": 0.0049,
"step": 6560
},
{
"epoch": 39.34131736526946,
"grad_norm": 0.125878244638443,
"learning_rate": 2.8875747374823288e-05,
"loss": 0.0053,
"step": 6570
},
{
"epoch": 39.40119760479042,
"grad_norm": 0.1674126535654068,
"learning_rate": 2.872599762775298e-05,
"loss": 0.006,
"step": 6580
},
{
"epoch": 39.461077844311376,
"grad_norm": 0.1522694081068039,
"learning_rate": 2.857648052976425e-05,
"loss": 0.0054,
"step": 6590
},
{
"epoch": 39.52095808383233,
"grad_norm": 0.2470812201499939,
"learning_rate": 2.8427197715952047e-05,
"loss": 0.0047,
"step": 6600
},
{
"epoch": 39.58083832335329,
"grad_norm": 0.20480410754680634,
"learning_rate": 2.8278150818849393e-05,
"loss": 0.006,
"step": 6610
},
{
"epoch": 39.64071856287425,
"grad_norm": 0.24696604907512665,
"learning_rate": 2.812934146840922e-05,
"loss": 0.0047,
"step": 6620
},
{
"epoch": 39.70059880239521,
"grad_norm": 0.11105551570653915,
"learning_rate": 2.7980771291986764e-05,
"loss": 0.0035,
"step": 6630
},
{
"epoch": 39.76047904191617,
"grad_norm": 0.12085378915071487,
"learning_rate": 2.783244191432167e-05,
"loss": 0.0049,
"step": 6640
},
{
"epoch": 39.82035928143713,
"grad_norm": 0.11907858401536942,
"learning_rate": 2.768435495752022e-05,
"loss": 0.0041,
"step": 6650
},
{
"epoch": 39.880239520958085,
"grad_norm": 0.23365391790866852,
"learning_rate": 2.753651204103771e-05,
"loss": 0.0064,
"step": 6660
},
{
"epoch": 39.94011976047904,
"grad_norm": 0.12890322506427765,
"learning_rate": 2.7388914781660523e-05,
"loss": 0.0045,
"step": 6670
},
{
"epoch": 40.0,
"grad_norm": 0.1367163509130478,
"learning_rate": 2.7241564793488693e-05,
"loss": 0.005,
"step": 6680
},
{
"epoch": 40.05988023952096,
"grad_norm": 0.12733227014541626,
"learning_rate": 2.7094463687918037e-05,
"loss": 0.0043,
"step": 6690
},
{
"epoch": 40.119760479041915,
"grad_norm": 0.10784728825092316,
"learning_rate": 2.694761307362268e-05,
"loss": 0.0054,
"step": 6700
},
{
"epoch": 40.17964071856287,
"grad_norm": 0.0882386788725853,
"learning_rate": 2.6801014556537467e-05,
"loss": 0.0043,
"step": 6710
},
{
"epoch": 40.23952095808383,
"grad_norm": 0.08688399940729141,
"learning_rate": 2.6654669739840243e-05,
"loss": 0.0045,
"step": 6720
},
{
"epoch": 40.29940119760479,
"grad_norm": 0.12476253509521484,
"learning_rate": 2.650858022393451e-05,
"loss": 0.0067,
"step": 6730
},
{
"epoch": 40.35928143712575,
"grad_norm": 0.10005974769592285,
"learning_rate": 2.6362747606431747e-05,
"loss": 0.0052,
"step": 6740
},
{
"epoch": 40.41916167664671,
"grad_norm": 0.15608790516853333,
"learning_rate": 2.6217173482134172e-05,
"loss": 0.0041,
"step": 6750
},
{
"epoch": 40.47904191616767,
"grad_norm": 0.1754852831363678,
"learning_rate": 2.6071859443017044e-05,
"loss": 0.0041,
"step": 6760
},
{
"epoch": 40.538922155688624,
"grad_norm": 0.08172114193439484,
"learning_rate": 2.5926807078211414e-05,
"loss": 0.0042,
"step": 6770
},
{
"epoch": 40.59880239520958,
"grad_norm": 0.09136908501386642,
"learning_rate": 2.5782017973986728e-05,
"loss": 0.0043,
"step": 6780
},
{
"epoch": 40.65868263473054,
"grad_norm": 0.18534158170223236,
"learning_rate": 2.5637493713733374e-05,
"loss": 0.0042,
"step": 6790
},
{
"epoch": 40.7185628742515,
"grad_norm": 0.15171270072460175,
"learning_rate": 2.549323587794559e-05,
"loss": 0.0042,
"step": 6800
},
{
"epoch": 40.778443113772454,
"grad_norm": 0.11207517981529236,
"learning_rate": 2.5349246044203895e-05,
"loss": 0.0039,
"step": 6810
},
{
"epoch": 40.83832335329341,
"grad_norm": 0.12690980732440948,
"learning_rate": 2.520552578715808e-05,
"loss": 0.0045,
"step": 6820
},
{
"epoch": 40.89820359281437,
"grad_norm": 0.09671933948993683,
"learning_rate": 2.506207667850981e-05,
"loss": 0.0049,
"step": 6830
},
{
"epoch": 40.958083832335326,
"grad_norm": 0.08247523754835129,
"learning_rate": 2.4918900286995555e-05,
"loss": 0.0051,
"step": 6840
},
{
"epoch": 41.01796407185629,
"grad_norm": 0.127054363489151,
"learning_rate": 2.4775998178369458e-05,
"loss": 0.0038,
"step": 6850
},
{
"epoch": 41.07784431137725,
"grad_norm": 0.12966464459896088,
"learning_rate": 2.4633371915386017e-05,
"loss": 0.0052,
"step": 6860
},
{
"epoch": 41.137724550898206,
"grad_norm": 0.13621652126312256,
"learning_rate": 2.4491023057783235e-05,
"loss": 0.0046,
"step": 6870
},
{
"epoch": 41.19760479041916,
"grad_norm": 0.13433954119682312,
"learning_rate": 2.4348953162265375e-05,
"loss": 0.0041,
"step": 6880
},
{
"epoch": 41.25748502994012,
"grad_norm": 0.14023637771606445,
"learning_rate": 2.420716378248607e-05,
"loss": 0.0045,
"step": 6890
},
{
"epoch": 41.31736526946108,
"grad_norm": 0.10107818245887756,
"learning_rate": 2.4065656469031266e-05,
"loss": 0.0042,
"step": 6900
},
{
"epoch": 41.377245508982035,
"grad_norm": 0.07907533645629883,
"learning_rate": 2.3924432769402268e-05,
"loss": 0.0056,
"step": 6910
},
{
"epoch": 41.43712574850299,
"grad_norm": 0.18893224000930786,
"learning_rate": 2.3783494227998844e-05,
"loss": 0.0043,
"step": 6920
},
{
"epoch": 41.49700598802395,
"grad_norm": 0.17834605276584625,
"learning_rate": 2.3642842386102264e-05,
"loss": 0.0051,
"step": 6930
},
{
"epoch": 41.55688622754491,
"grad_norm": 0.2308255434036255,
"learning_rate": 2.3502478781858567e-05,
"loss": 0.0051,
"step": 6940
},
{
"epoch": 41.616766467065865,
"grad_norm": 0.12349040061235428,
"learning_rate": 2.3362404950261628e-05,
"loss": 0.0042,
"step": 6950
},
{
"epoch": 41.67664670658683,
"grad_norm": 0.10960213094949722,
"learning_rate": 2.3222622423136458e-05,
"loss": 0.0039,
"step": 6960
},
{
"epoch": 41.73652694610779,
"grad_norm": 0.1554371863603592,
"learning_rate": 2.3083132729122332e-05,
"loss": 0.0043,
"step": 6970
},
{
"epoch": 41.796407185628745,
"grad_norm": 0.11700791865587234,
"learning_rate": 2.294393739365621e-05,
"loss": 0.0041,
"step": 6980
},
{
"epoch": 41.8562874251497,
"grad_norm": 0.1540096402168274,
"learning_rate": 2.2805037938956e-05,
"loss": 0.0037,
"step": 6990
},
{
"epoch": 41.91616766467066,
"grad_norm": 0.1582740843296051,
"learning_rate": 2.266643588400386e-05,
"loss": 0.0041,
"step": 7000
},
{
"epoch": 41.97604790419162,
"grad_norm": 0.10413503646850586,
"learning_rate": 2.252813274452969e-05,
"loss": 0.004,
"step": 7010
},
{
"epoch": 42.035928143712574,
"grad_norm": 0.07931295037269592,
"learning_rate": 2.2390130032994427e-05,
"loss": 0.004,
"step": 7020
},
{
"epoch": 42.09580838323353,
"grad_norm": 0.12863239645957947,
"learning_rate": 2.2252429258573633e-05,
"loss": 0.0042,
"step": 7030
},
{
"epoch": 42.15568862275449,
"grad_norm": 0.19994114339351654,
"learning_rate": 2.2115031927140904e-05,
"loss": 0.006,
"step": 7040
},
{
"epoch": 42.21556886227545,
"grad_norm": 0.14814303815364838,
"learning_rate": 2.1977939541251463e-05,
"loss": 0.0035,
"step": 7050
},
{
"epoch": 42.275449101796404,
"grad_norm": 0.12121354043483734,
"learning_rate": 2.1841153600125684e-05,
"loss": 0.0045,
"step": 7060
},
{
"epoch": 42.33532934131736,
"grad_norm": 0.17686079442501068,
"learning_rate": 2.170467559963267e-05,
"loss": 0.0037,
"step": 7070
},
{
"epoch": 42.395209580838326,
"grad_norm": 0.1135316789150238,
"learning_rate": 2.1568507032273982e-05,
"loss": 0.0035,
"step": 7080
},
{
"epoch": 42.455089820359284,
"grad_norm": 0.15939050912857056,
"learning_rate": 2.1432649387167264e-05,
"loss": 0.0047,
"step": 7090
},
{
"epoch": 42.51497005988024,
"grad_norm": 0.18091681599617004,
"learning_rate": 2.1297104150029973e-05,
"loss": 0.0041,
"step": 7100
},
{
"epoch": 42.5748502994012,
"grad_norm": 0.1854577362537384,
"learning_rate": 2.116187280316307e-05,
"loss": 0.0056,
"step": 7110
},
{
"epoch": 42.634730538922156,
"grad_norm": 0.2728532552719116,
"learning_rate": 2.1026956825434908e-05,
"loss": 0.0042,
"step": 7120
},
{
"epoch": 42.69461077844311,
"grad_norm": 0.18949194252490997,
"learning_rate": 2.0892357692265017e-05,
"loss": 0.0057,
"step": 7130
},
{
"epoch": 42.75449101796407,
"grad_norm": 0.2114059180021286,
"learning_rate": 2.0758076875607947e-05,
"loss": 0.0071,
"step": 7140
},
{
"epoch": 42.81437125748503,
"grad_norm": 0.1484140008687973,
"learning_rate": 2.0624115843937207e-05,
"loss": 0.0056,
"step": 7150
},
{
"epoch": 42.874251497005986,
"grad_norm": 0.2194991409778595,
"learning_rate": 2.0490476062229157e-05,
"loss": 0.0056,
"step": 7160
},
{
"epoch": 42.93413173652694,
"grad_norm": 0.24683687090873718,
"learning_rate": 2.035715899194704e-05,
"loss": 0.006,
"step": 7170
},
{
"epoch": 42.9940119760479,
"grad_norm": 0.1990325152873993,
"learning_rate": 2.022416609102499e-05,
"loss": 0.0053,
"step": 7180
},
{
"epoch": 43.053892215568865,
"grad_norm": 0.16312725841999054,
"learning_rate": 2.009149881385205e-05,
"loss": 0.0049,
"step": 7190
},
{
"epoch": 43.11377245508982,
"grad_norm": 0.28681010007858276,
"learning_rate": 1.995915861125634e-05,
"loss": 0.0054,
"step": 7200
},
{
"epoch": 43.17365269461078,
"grad_norm": 0.14880476891994476,
"learning_rate": 1.9827146930489065e-05,
"loss": 0.0038,
"step": 7210
},
{
"epoch": 43.23353293413174,
"grad_norm": 0.10214495658874512,
"learning_rate": 1.9695465215208848e-05,
"loss": 0.0039,
"step": 7220
},
{
"epoch": 43.293413173652695,
"grad_norm": 0.10757572948932648,
"learning_rate": 1.9564114905465813e-05,
"loss": 0.0052,
"step": 7230
},
{
"epoch": 43.35329341317365,
"grad_norm": 0.11554937809705734,
"learning_rate": 1.9433097437685936e-05,
"loss": 0.0045,
"step": 7240
},
{
"epoch": 43.41317365269461,
"grad_norm": 0.123574398458004,
"learning_rate": 1.930241424465521e-05,
"loss": 0.0048,
"step": 7250
},
{
"epoch": 43.47305389221557,
"grad_norm": 0.22432149946689606,
"learning_rate": 1.9172066755504115e-05,
"loss": 0.0048,
"step": 7260
},
{
"epoch": 43.532934131736525,
"grad_norm": 0.12835153937339783,
"learning_rate": 1.9042056395691914e-05,
"loss": 0.0043,
"step": 7270
},
{
"epoch": 43.59281437125748,
"grad_norm": 0.3164388835430145,
"learning_rate": 1.8912384586991066e-05,
"loss": 0.0054,
"step": 7280
},
{
"epoch": 43.65269461077844,
"grad_norm": 0.18250100314617157,
"learning_rate": 1.8783052747471717e-05,
"loss": 0.0039,
"step": 7290
},
{
"epoch": 43.712574850299404,
"grad_norm": 0.12120740115642548,
"learning_rate": 1.865406229148611e-05,
"loss": 0.004,
"step": 7300
},
{
"epoch": 43.77245508982036,
"grad_norm": 0.07131797075271606,
"learning_rate": 1.8525414629653233e-05,
"loss": 0.0042,
"step": 7310
},
{
"epoch": 43.83233532934132,
"grad_norm": 0.13069207966327667,
"learning_rate": 1.8397111168843255e-05,
"loss": 0.004,
"step": 7320
},
{
"epoch": 43.89221556886228,
"grad_norm": 0.111685611307621,
"learning_rate": 1.8269153312162323e-05,
"loss": 0.0045,
"step": 7330
},
{
"epoch": 43.952095808383234,
"grad_norm": 0.19609613716602325,
"learning_rate": 1.8141542458937054e-05,
"loss": 0.006,
"step": 7340
},
{
"epoch": 44.01197604790419,
"grad_norm": 0.060919344425201416,
"learning_rate": 1.8014280004699268e-05,
"loss": 0.0036,
"step": 7350
},
{
"epoch": 44.07185628742515,
"grad_norm": 0.1315314769744873,
"learning_rate": 1.788736734117078e-05,
"loss": 0.0042,
"step": 7360
},
{
"epoch": 44.131736526946106,
"grad_norm": 0.10033899545669556,
"learning_rate": 1.7760805856248152e-05,
"loss": 0.0037,
"step": 7370
},
{
"epoch": 44.191616766467064,
"grad_norm": 0.11199448257684708,
"learning_rate": 1.7634596933987518e-05,
"loss": 0.0053,
"step": 7380
},
{
"epoch": 44.25149700598802,
"grad_norm": 0.10838598757982254,
"learning_rate": 1.7508741954589404e-05,
"loss": 0.0037,
"step": 7390
},
{
"epoch": 44.31137724550898,
"grad_norm": 0.11036816984415054,
"learning_rate": 1.7383242294383717e-05,
"loss": 0.0046,
"step": 7400
},
{
"epoch": 44.37125748502994,
"grad_norm": 0.11777987331151962,
"learning_rate": 1.7258099325814632e-05,
"loss": 0.0035,
"step": 7410
},
{
"epoch": 44.4311377245509,
"grad_norm": 0.12325455248355865,
"learning_rate": 1.7133314417425594e-05,
"loss": 0.0044,
"step": 7420
},
{
"epoch": 44.49101796407186,
"grad_norm": 0.07163011282682419,
"learning_rate": 1.7008888933844408e-05,
"loss": 0.004,
"step": 7430
},
{
"epoch": 44.550898203592816,
"grad_norm": 0.08228765428066254,
"learning_rate": 1.6884824235768172e-05,
"loss": 0.0038,
"step": 7440
},
{
"epoch": 44.61077844311377,
"grad_norm": 0.08889100700616837,
"learning_rate": 1.6761121679948592e-05,
"loss": 0.0048,
"step": 7450
},
{
"epoch": 44.67065868263473,
"grad_norm": 0.09399433434009552,
"learning_rate": 1.663778261917695e-05,
"loss": 0.0038,
"step": 7460
},
{
"epoch": 44.73053892215569,
"grad_norm": 0.12770161032676697,
"learning_rate": 1.651480840226952e-05,
"loss": 0.004,
"step": 7470
},
{
"epoch": 44.790419161676645,
"grad_norm": 0.06503359228372574,
"learning_rate": 1.639220037405258e-05,
"loss": 0.0035,
"step": 7480
},
{
"epoch": 44.8502994011976,
"grad_norm": 0.1480531096458435,
"learning_rate": 1.6269959875347906e-05,
"loss": 0.0059,
"step": 7490
},
{
"epoch": 44.91017964071856,
"grad_norm": 0.09000187367200851,
"learning_rate": 1.614808824295802e-05,
"loss": 0.0032,
"step": 7500
},
{
"epoch": 44.97005988023952,
"grad_norm": 0.11256859451532364,
"learning_rate": 1.602658680965152e-05,
"loss": 0.0037,
"step": 7510
},
{
"epoch": 45.02994011976048,
"grad_norm": 0.11892662197351456,
"learning_rate": 1.5905456904148686e-05,
"loss": 0.0033,
"step": 7520
},
{
"epoch": 45.08982035928144,
"grad_norm": 0.0951654240489006,
"learning_rate": 1.57846998511067e-05,
"loss": 0.0038,
"step": 7530
},
{
"epoch": 45.1497005988024,
"grad_norm": 0.10214338451623917,
"learning_rate": 1.566431697110538e-05,
"loss": 0.003,
"step": 7540
},
{
"epoch": 45.209580838323355,
"grad_norm": 0.08763334155082703,
"learning_rate": 1.554430958063259e-05,
"loss": 0.0043,
"step": 7550
},
{
"epoch": 45.26946107784431,
"grad_norm": 0.17269887030124664,
"learning_rate": 1.5424678992069912e-05,
"loss": 0.0037,
"step": 7560
},
{
"epoch": 45.32934131736527,
"grad_norm": 0.07810787856578827,
"learning_rate": 1.5305426513678362e-05,
"loss": 0.0047,
"step": 7570
},
{
"epoch": 45.38922155688623,
"grad_norm": 0.10794202238321304,
"learning_rate": 1.518655344958388e-05,
"loss": 0.0045,
"step": 7580
},
{
"epoch": 45.449101796407184,
"grad_norm": 0.15913841128349304,
"learning_rate": 1.5068061099763275e-05,
"loss": 0.005,
"step": 7590
},
{
"epoch": 45.50898203592814,
"grad_norm": 0.14160914719104767,
"learning_rate": 1.494995076002988e-05,
"loss": 0.0049,
"step": 7600
},
{
"epoch": 45.5688622754491,
"grad_norm": 0.21142461895942688,
"learning_rate": 1.4832223722019456e-05,
"loss": 0.0045,
"step": 7610
},
{
"epoch": 45.62874251497006,
"grad_norm": 0.20032016932964325,
"learning_rate": 1.4714881273176035e-05,
"loss": 0.0064,
"step": 7620
},
{
"epoch": 45.68862275449102,
"grad_norm": 0.13826598227024078,
"learning_rate": 1.4597924696737835e-05,
"loss": 0.0033,
"step": 7630
},
{
"epoch": 45.74850299401198,
"grad_norm": 0.11065464466810226,
"learning_rate": 1.4481355271723252e-05,
"loss": 0.0037,
"step": 7640
},
{
"epoch": 45.808383233532936,
"grad_norm": 0.08762434870004654,
"learning_rate": 1.4365174272916809e-05,
"loss": 0.0035,
"step": 7650
},
{
"epoch": 45.868263473053894,
"grad_norm": 0.09540656208992004,
"learning_rate": 1.4249382970855319e-05,
"loss": 0.0054,
"step": 7660
},
{
"epoch": 45.92814371257485,
"grad_norm": 0.08577928692102432,
"learning_rate": 1.4133982631813903e-05,
"loss": 0.0034,
"step": 7670
},
{
"epoch": 45.98802395209581,
"grad_norm": 0.07898104190826416,
"learning_rate": 1.4018974517792194e-05,
"loss": 0.0036,
"step": 7680
},
{
"epoch": 46.047904191616766,
"grad_norm": 0.11655700206756592,
"learning_rate": 1.390435988650048e-05,
"loss": 0.0044,
"step": 7690
},
{
"epoch": 46.10778443113772,
"grad_norm": 0.108432337641716,
"learning_rate": 1.3790139991346006e-05,
"loss": 0.0045,
"step": 7700
},
{
"epoch": 46.16766467065868,
"grad_norm": 0.07035807520151138,
"learning_rate": 1.367631608141926e-05,
"loss": 0.0036,
"step": 7710
},
{
"epoch": 46.22754491017964,
"grad_norm": 0.15821093320846558,
"learning_rate": 1.3562889401480278e-05,
"loss": 0.005,
"step": 7720
},
{
"epoch": 46.287425149700596,
"grad_norm": 0.13470861315727234,
"learning_rate": 1.3449861191945074e-05,
"loss": 0.0031,
"step": 7730
},
{
"epoch": 46.34730538922156,
"grad_norm": 0.08224532008171082,
"learning_rate": 1.3337232688872009e-05,
"loss": 0.0036,
"step": 7740
},
{
"epoch": 46.40718562874252,
"grad_norm": 0.1143488958477974,
"learning_rate": 1.3225005123948364e-05,
"loss": 0.0041,
"step": 7750
},
{
"epoch": 46.467065868263475,
"grad_norm": 0.0909508690237999,
"learning_rate": 1.311317972447681e-05,
"loss": 0.0038,
"step": 7760
},
{
"epoch": 46.52694610778443,
"grad_norm": 0.08813850581645966,
"learning_rate": 1.3001757713361996e-05,
"loss": 0.0035,
"step": 7770
},
{
"epoch": 46.58682634730539,
"grad_norm": 0.07225538790225983,
"learning_rate": 1.2890740309097204e-05,
"loss": 0.0071,
"step": 7780
},
{
"epoch": 46.64670658682635,
"grad_norm": 0.16005778312683105,
"learning_rate": 1.2780128725750944e-05,
"loss": 0.0038,
"step": 7790
},
{
"epoch": 46.706586826347305,
"grad_norm": 0.08824863284826279,
"learning_rate": 1.266992417295379e-05,
"loss": 0.0043,
"step": 7800
},
{
"epoch": 46.76646706586826,
"grad_norm": 0.12088621407747269,
"learning_rate": 1.2560127855885073e-05,
"loss": 0.0036,
"step": 7810
},
{
"epoch": 46.82634730538922,
"grad_norm": 0.13209691643714905,
"learning_rate": 1.2450740975259745e-05,
"loss": 0.0044,
"step": 7820
},
{
"epoch": 46.88622754491018,
"grad_norm": 0.0907764658331871,
"learning_rate": 1.234176472731517e-05,
"loss": 0.0044,
"step": 7830
},
{
"epoch": 46.946107784431135,
"grad_norm": 0.10527919232845306,
"learning_rate": 1.2233200303798158e-05,
"loss": 0.0036,
"step": 7840
},
{
"epoch": 47.0059880239521,
"grad_norm": 0.10500690340995789,
"learning_rate": 1.2125048891951846e-05,
"loss": 0.004,
"step": 7850
},
{
"epoch": 47.06586826347306,
"grad_norm": 0.1854642629623413,
"learning_rate": 1.2017311674502745e-05,
"loss": 0.0046,
"step": 7860
},
{
"epoch": 47.125748502994014,
"grad_norm": 0.13455477356910706,
"learning_rate": 1.1909989829647822e-05,
"loss": 0.0031,
"step": 7870
},
{
"epoch": 47.18562874251497,
"grad_norm": 0.08324426412582397,
"learning_rate": 1.1803084531041553e-05,
"loss": 0.0032,
"step": 7880
},
{
"epoch": 47.24550898203593,
"grad_norm": 0.09490043669939041,
"learning_rate": 1.1696596947783162e-05,
"loss": 0.0027,
"step": 7890
},
{
"epoch": 47.30538922155689,
"grad_norm": 0.13496926426887512,
"learning_rate": 1.1590528244403803e-05,
"loss": 0.0031,
"step": 7900
},
{
"epoch": 47.365269461077844,
"grad_norm": 0.09702477604150772,
"learning_rate": 1.148487958085382e-05,
"loss": 0.0031,
"step": 7910
},
{
"epoch": 47.4251497005988,
"grad_norm": 0.20016546547412872,
"learning_rate": 1.1379652112490086e-05,
"loss": 0.0037,
"step": 7920
},
{
"epoch": 47.48502994011976,
"grad_norm": 0.07785689830780029,
"learning_rate": 1.1274846990063315e-05,
"loss": 0.0039,
"step": 7930
},
{
"epoch": 47.544910179640716,
"grad_norm": 0.09707417339086533,
"learning_rate": 1.117046535970554e-05,
"loss": 0.0043,
"step": 7940
},
{
"epoch": 47.604790419161674,
"grad_norm": 0.07220587879419327,
"learning_rate": 1.106650836291755e-05,
"loss": 0.0038,
"step": 7950
},
{
"epoch": 47.66467065868264,
"grad_norm": 0.1931801736354828,
"learning_rate": 1.0962977136556418e-05,
"loss": 0.0039,
"step": 7960
},
{
"epoch": 47.724550898203596,
"grad_norm": 0.08576804399490356,
"learning_rate": 1.0859872812823024e-05,
"loss": 0.0029,
"step": 7970
},
{
"epoch": 47.78443113772455,
"grad_norm": 0.14273039996623993,
"learning_rate": 1.0757196519249747e-05,
"loss": 0.0038,
"step": 7980
},
{
"epoch": 47.84431137724551,
"grad_norm": 0.12143658846616745,
"learning_rate": 1.0654949378688077e-05,
"loss": 0.0036,
"step": 7990
},
{
"epoch": 47.90419161676647,
"grad_norm": 0.16703766584396362,
"learning_rate": 1.0553132509296376e-05,
"loss": 0.0039,
"step": 8000
},
{
"epoch": 47.964071856287426,
"grad_norm": 0.052701886743307114,
"learning_rate": 1.0451747024527613e-05,
"loss": 0.0029,
"step": 8010
},
{
"epoch": 48.02395209580838,
"grad_norm": 0.10135592520236969,
"learning_rate": 1.0350794033117189e-05,
"loss": 0.0033,
"step": 8020
},
{
"epoch": 48.08383233532934,
"grad_norm": 0.1006510853767395,
"learning_rate": 1.0250274639070856e-05,
"loss": 0.0026,
"step": 8030
},
{
"epoch": 48.1437125748503,
"grad_norm": 0.07589986175298691,
"learning_rate": 1.0150189941652599e-05,
"loss": 0.0035,
"step": 8040
},
{
"epoch": 48.203592814371255,
"grad_norm": 0.1506795436143875,
"learning_rate": 1.0050541035372635e-05,
"loss": 0.0039,
"step": 8050
},
{
"epoch": 48.26347305389221,
"grad_norm": 0.0997994989156723,
"learning_rate": 9.951329009975458e-06,
"loss": 0.0031,
"step": 8060
},
{
"epoch": 48.32335329341317,
"grad_norm": 0.08480530232191086,
"learning_rate": 9.852554950427845e-06,
"loss": 0.0036,
"step": 8070
},
{
"epoch": 48.383233532934135,
"grad_norm": 0.08194699138402939,
"learning_rate": 9.754219936907105e-06,
"loss": 0.0032,
"step": 8080
},
{
"epoch": 48.44311377245509,
"grad_norm": 0.05077745020389557,
"learning_rate": 9.656325044789194e-06,
"loss": 0.0033,
"step": 8090
},
{
"epoch": 48.50299401197605,
"grad_norm": 0.05620262771844864,
"learning_rate": 9.55887134463697e-06,
"loss": 0.003,
"step": 8100
},
{
"epoch": 48.56287425149701,
"grad_norm": 0.11098481714725494,
"learning_rate": 9.461859902188475e-06,
"loss": 0.0046,
"step": 8110
},
{
"epoch": 48.622754491017965,
"grad_norm": 0.1423126608133316,
"learning_rate": 9.365291778345303e-06,
"loss": 0.0037,
"step": 8120
},
{
"epoch": 48.68263473053892,
"grad_norm": 0.07402303069829941,
"learning_rate": 9.269168029160991e-06,
"loss": 0.0047,
"step": 8130
},
{
"epoch": 48.74251497005988,
"grad_norm": 0.17480172216892242,
"learning_rate": 9.173489705829447e-06,
"loss": 0.0032,
"step": 8140
},
{
"epoch": 48.80239520958084,
"grad_norm": 0.07544239610433578,
"learning_rate": 9.078257854673516e-06,
"loss": 0.0029,
"step": 8150
},
{
"epoch": 48.862275449101794,
"grad_norm": 0.14797523617744446,
"learning_rate": 8.983473517133429e-06,
"loss": 0.0037,
"step": 8160
},
{
"epoch": 48.92215568862275,
"grad_norm": 0.09308101236820221,
"learning_rate": 8.889137729755537e-06,
"loss": 0.0038,
"step": 8170
},
{
"epoch": 48.98203592814371,
"grad_norm": 0.07997914403676987,
"learning_rate": 8.79525152418087e-06,
"loss": 0.0046,
"step": 8180
},
{
"epoch": 49.041916167664674,
"grad_norm": 0.06806595623493195,
"learning_rate": 8.701815927133961e-06,
"loss": 0.0048,
"step": 8190
},
{
"epoch": 49.10179640718563,
"grad_norm": 0.14955078065395355,
"learning_rate": 8.608831960411534e-06,
"loss": 0.0043,
"step": 8200
},
{
"epoch": 49.16167664670659,
"grad_norm": 0.10370472818613052,
"learning_rate": 8.516300640871321e-06,
"loss": 0.0032,
"step": 8210
},
{
"epoch": 49.221556886227546,
"grad_norm": 0.07083608955144882,
"learning_rate": 8.424222980421038e-06,
"loss": 0.0037,
"step": 8220
},
{
"epoch": 49.2814371257485,
"grad_norm": 0.11663700640201569,
"learning_rate": 8.332599986007184e-06,
"loss": 0.0027,
"step": 8230
},
{
"epoch": 49.34131736526946,
"grad_norm": 0.06833653151988983,
"learning_rate": 8.241432659604203e-06,
"loss": 0.0033,
"step": 8240
},
{
"epoch": 49.40119760479042,
"grad_norm": 0.07409913092851639,
"learning_rate": 8.150721998203331e-06,
"loss": 0.0036,
"step": 8250
},
{
"epoch": 49.461077844311376,
"grad_norm": 0.05694573000073433,
"learning_rate": 8.06046899380184e-06,
"loss": 0.0028,
"step": 8260
},
{
"epoch": 49.52095808383233,
"grad_norm": 0.08030648529529572,
"learning_rate": 7.970674633392133e-06,
"loss": 0.0036,
"step": 8270
},
{
"epoch": 49.58083832335329,
"grad_norm": 0.06093934178352356,
"learning_rate": 7.881339898950924e-06,
"loss": 0.0025,
"step": 8280
},
{
"epoch": 49.64071856287425,
"grad_norm": 0.13092978298664093,
"learning_rate": 7.792465767428597e-06,
"loss": 0.0031,
"step": 8290
},
{
"epoch": 49.70059880239521,
"grad_norm": 0.14689597487449646,
"learning_rate": 7.704053210738376e-06,
"loss": 0.0039,
"step": 8300
},
{
"epoch": 49.76047904191617,
"grad_norm": 0.040117938071489334,
"learning_rate": 7.6161031957458494e-06,
"loss": 0.0038,
"step": 8310
},
{
"epoch": 49.82035928143713,
"grad_norm": 0.15532176196575165,
"learning_rate": 7.5286166842582605e-06,
"loss": 0.0038,
"step": 8320
},
{
"epoch": 49.880239520958085,
"grad_norm": 0.0856887623667717,
"learning_rate": 7.4415946330140814e-06,
"loss": 0.0042,
"step": 8330
},
{
"epoch": 49.94011976047904,
"grad_norm": 0.03667578473687172,
"learning_rate": 7.3550379936725644e-06,
"loss": 0.0033,
"step": 8340
},
{
"epoch": 50.0,
"grad_norm": 0.08501570671796799,
"learning_rate": 7.2689477128032035e-06,
"loss": 0.0034,
"step": 8350
},
{
"epoch": 50.05988023952096,
"grad_norm": 0.0592077262699604,
"learning_rate": 7.183324731875551e-06,
"loss": 0.0033,
"step": 8360
},
{
"epoch": 50.119760479041915,
"grad_norm": 0.09360985457897186,
"learning_rate": 7.098169987248782e-06,
"loss": 0.0023,
"step": 8370
},
{
"epoch": 50.17964071856287,
"grad_norm": 0.16940154135227203,
"learning_rate": 7.013484410161553e-06,
"loss": 0.0039,
"step": 8380
},
{
"epoch": 50.23952095808383,
"grad_norm": 0.12633410096168518,
"learning_rate": 6.92926892672176e-06,
"loss": 0.0032,
"step": 8390
},
{
"epoch": 50.29940119760479,
"grad_norm": 0.08256031572818756,
"learning_rate": 6.845524457896446e-06,
"loss": 0.0053,
"step": 8400
},
{
"epoch": 50.35928143712575,
"grad_norm": 0.0617823489010334,
"learning_rate": 6.7622519195017165e-06,
"loss": 0.0029,
"step": 8410
},
{
"epoch": 50.41916167664671,
"grad_norm": 0.05179158225655556,
"learning_rate": 6.679452222192684e-06,
"loss": 0.0025,
"step": 8420
},
{
"epoch": 50.47904191616767,
"grad_norm": 0.05615438148379326,
"learning_rate": 6.597126271453579e-06,
"loss": 0.0033,
"step": 8430
},
{
"epoch": 50.538922155688624,
"grad_norm": 0.09643064439296722,
"learning_rate": 6.51527496758782e-06,
"loss": 0.0027,
"step": 8440
},
{
"epoch": 50.59880239520958,
"grad_norm": 0.12618598341941833,
"learning_rate": 6.433899205708155e-06,
"loss": 0.0048,
"step": 8450
},
{
"epoch": 50.65868263473054,
"grad_norm": 0.13306772708892822,
"learning_rate": 6.352999875726856e-06,
"loss": 0.0033,
"step": 8460
},
{
"epoch": 50.7185628742515,
"grad_norm": 0.1367420256137848,
"learning_rate": 6.272577862346052e-06,
"loss": 0.004,
"step": 8470
},
{
"epoch": 50.778443113772454,
"grad_norm": 0.07138808816671371,
"learning_rate": 6.192634045047996e-06,
"loss": 0.0043,
"step": 8480
},
{
"epoch": 50.83832335329341,
"grad_norm": 0.07425539195537567,
"learning_rate": 6.113169298085458e-06,
"loss": 0.0044,
"step": 8490
},
{
"epoch": 50.89820359281437,
"grad_norm": 0.13559626042842865,
"learning_rate": 6.034184490472195e-06,
"loss": 0.0039,
"step": 8500
},
{
"epoch": 50.958083832335326,
"grad_norm": 0.11653909832239151,
"learning_rate": 5.955680485973386e-06,
"loss": 0.0049,
"step": 8510
},
{
"epoch": 51.01796407185629,
"grad_norm": 0.09567855298519135,
"learning_rate": 5.877658143096265e-06,
"loss": 0.0029,
"step": 8520
},
{
"epoch": 51.07784431137725,
"grad_norm": 0.2062114030122757,
"learning_rate": 5.800118315080661e-06,
"loss": 0.0039,
"step": 8530
},
{
"epoch": 51.137724550898206,
"grad_norm": 0.06853962689638138,
"learning_rate": 5.723061849889716e-06,
"loss": 0.0028,
"step": 8540
},
{
"epoch": 51.19760479041916,
"grad_norm": 0.08331948518753052,
"learning_rate": 5.646489590200604e-06,
"loss": 0.0033,
"step": 8550
},
{
"epoch": 51.25748502994012,
"grad_norm": 0.03685355931520462,
"learning_rate": 5.570402373395256e-06,
"loss": 0.0031,
"step": 8560
},
{
"epoch": 51.31736526946108,
"grad_norm": 0.05594068765640259,
"learning_rate": 5.494801031551305e-06,
"loss": 0.0039,
"step": 8570
},
{
"epoch": 51.377245508982035,
"grad_norm": 0.08362322300672531,
"learning_rate": 5.41968639143291e-06,
"loss": 0.0027,
"step": 8580
},
{
"epoch": 51.43712574850299,
"grad_norm": 0.14236518740653992,
"learning_rate": 5.345059274481751e-06,
"loss": 0.0033,
"step": 8590
},
{
"epoch": 51.49700598802395,
"grad_norm": 0.09911030530929565,
"learning_rate": 5.270920496808002e-06,
"loss": 0.0035,
"step": 8600
},
{
"epoch": 51.55688622754491,
"grad_norm": 0.08180464059114456,
"learning_rate": 5.1972708691814695e-06,
"loss": 0.004,
"step": 8610
},
{
"epoch": 51.616766467065865,
"grad_norm": 0.07401694357395172,
"learning_rate": 5.124111197022674e-06,
"loss": 0.0041,
"step": 8620
},
{
"epoch": 51.67664670658683,
"grad_norm": 0.23763364553451538,
"learning_rate": 5.051442280394081e-06,
"loss": 0.0038,
"step": 8630
},
{
"epoch": 51.73652694610779,
"grad_norm": 0.061368055641651154,
"learning_rate": 4.979264913991322e-06,
"loss": 0.0047,
"step": 8640
},
{
"epoch": 51.796407185628745,
"grad_norm": 0.10266540199518204,
"learning_rate": 4.907579887134489e-06,
"loss": 0.0038,
"step": 8650
},
{
"epoch": 51.8562874251497,
"grad_norm": 0.15884973108768463,
"learning_rate": 4.836387983759572e-06,
"loss": 0.003,
"step": 8660
},
{
"epoch": 51.91616766467066,
"grad_norm": 0.09529859572649002,
"learning_rate": 4.765689982409816e-06,
"loss": 0.0042,
"step": 8670
},
{
"epoch": 51.97604790419162,
"grad_norm": 0.06333234161138535,
"learning_rate": 4.695486656227233e-06,
"loss": 0.004,
"step": 8680
},
{
"epoch": 52.035928143712574,
"grad_norm": 0.06246256083250046,
"learning_rate": 4.625778772944156e-06,
"loss": 0.0042,
"step": 8690
},
{
"epoch": 52.09580838323353,
"grad_norm": 0.13045619428157806,
"learning_rate": 4.556567094874825e-06,
"loss": 0.0033,
"step": 8700
},
{
"epoch": 52.15568862275449,
"grad_norm": 0.08753932267427444,
"learning_rate": 4.487852378907059e-06,
"loss": 0.0034,
"step": 8710
},
{
"epoch": 52.21556886227545,
"grad_norm": 0.07123212516307831,
"learning_rate": 4.419635376493986e-06,
"loss": 0.0032,
"step": 8720
},
{
"epoch": 52.275449101796404,
"grad_norm": 0.07041085511445999,
"learning_rate": 4.351916833645825e-06,
"loss": 0.004,
"step": 8730
},
{
"epoch": 52.33532934131736,
"grad_norm": 0.0965387374162674,
"learning_rate": 4.284697490921691e-06,
"loss": 0.0034,
"step": 8740
},
{
"epoch": 52.395209580838326,
"grad_norm": 0.12906357645988464,
"learning_rate": 4.2179780834215585e-06,
"loss": 0.0028,
"step": 8750
},
{
"epoch": 52.455089820359284,
"grad_norm": 0.06360188126564026,
"learning_rate": 4.151759340778178e-06,
"loss": 0.0035,
"step": 8760
},
{
"epoch": 52.51497005988024,
"grad_norm": 0.05023306608200073,
"learning_rate": 4.086041987149109e-06,
"loss": 0.0027,
"step": 8770
},
{
"epoch": 52.5748502994012,
"grad_norm": 0.08533389866352081,
"learning_rate": 4.020826741208811e-06,
"loss": 0.003,
"step": 8780
},
{
"epoch": 52.634730538922156,
"grad_norm": 0.04934442415833473,
"learning_rate": 3.956114316140746e-06,
"loss": 0.003,
"step": 8790
},
{
"epoch": 52.69461077844311,
"grad_norm": 0.12529414892196655,
"learning_rate": 3.891905419629643e-06,
"loss": 0.0049,
"step": 8800
},
{
"epoch": 52.75449101796407,
"grad_norm": 0.1008111760020256,
"learning_rate": 3.8282007538536946e-06,
"loss": 0.0026,
"step": 8810
},
{
"epoch": 52.81437125748503,
"grad_norm": 0.1696193516254425,
"learning_rate": 3.7650010154769265e-06,
"loss": 0.0028,
"step": 8820
},
{
"epoch": 52.874251497005986,
"grad_norm": 0.07706660777330399,
"learning_rate": 3.7023068956415608e-06,
"loss": 0.0033,
"step": 8830
},
{
"epoch": 52.93413173652694,
"grad_norm": 0.09362232685089111,
"learning_rate": 3.6401190799604303e-06,
"loss": 0.0047,
"step": 8840
},
{
"epoch": 52.9940119760479,
"grad_norm": 0.09532835334539413,
"learning_rate": 3.578438248509536e-06,
"loss": 0.0036,
"step": 8850
},
{
"epoch": 53.053892215568865,
"grad_norm": 0.11764580011367798,
"learning_rate": 3.5172650758205583e-06,
"loss": 0.003,
"step": 8860
},
{
"epoch": 53.11377245508982,
"grad_norm": 0.15160970389842987,
"learning_rate": 3.45660023087353e-06,
"loss": 0.0041,
"step": 8870
},
{
"epoch": 53.17365269461078,
"grad_norm": 0.09766502678394318,
"learning_rate": 3.3964443770894528e-06,
"loss": 0.0028,
"step": 8880
},
{
"epoch": 53.23353293413174,
"grad_norm": 0.04131095111370087,
"learning_rate": 3.3367981723231245e-06,
"loss": 0.0026,
"step": 8890
},
{
"epoch": 53.293413173652695,
"grad_norm": 0.08767053484916687,
"learning_rate": 3.2776622688558746e-06,
"loss": 0.0035,
"step": 8900
},
{
"epoch": 53.35329341317365,
"grad_norm": 0.08918289840221405,
"learning_rate": 3.2190373133884677e-06,
"loss": 0.0032,
"step": 8910
},
{
"epoch": 53.41317365269461,
"grad_norm": 0.06048423424363136,
"learning_rate": 3.1609239470340446e-06,
"loss": 0.0032,
"step": 8920
},
{
"epoch": 53.47305389221557,
"grad_norm": 0.10308495163917542,
"learning_rate": 3.1033228053110373e-06,
"loss": 0.0032,
"step": 8930
},
{
"epoch": 53.532934131736525,
"grad_norm": 0.08775908499956131,
"learning_rate": 3.0462345181363314e-06,
"loss": 0.0027,
"step": 8940
},
{
"epoch": 53.59281437125748,
"grad_norm": 0.13678984344005585,
"learning_rate": 2.9896597098182654e-06,
"loss": 0.0028,
"step": 8950
},
{
"epoch": 53.65269461077844,
"grad_norm": 0.03553814813494682,
"learning_rate": 2.933598999049891e-06,
"loss": 0.0025,
"step": 8960
},
{
"epoch": 53.712574850299404,
"grad_norm": 0.05184748023748398,
"learning_rate": 2.8780529989021697e-06,
"loss": 0.003,
"step": 8970
},
{
"epoch": 53.77245508982036,
"grad_norm": 0.06308116763830185,
"learning_rate": 2.823022316817242e-06,
"loss": 0.0032,
"step": 8980
},
{
"epoch": 53.83233532934132,
"grad_norm": 0.05940672382712364,
"learning_rate": 2.7685075546018456e-06,
"loss": 0.0026,
"step": 8990
},
{
"epoch": 53.89221556886228,
"grad_norm": 0.07400498539209366,
"learning_rate": 2.7145093084206598e-06,
"loss": 0.0027,
"step": 9000
},
{
"epoch": 53.952095808383234,
"grad_norm": 0.06883690506219864,
"learning_rate": 2.661028168789892e-06,
"loss": 0.0029,
"step": 9010
},
{
"epoch": 54.01197604790419,
"grad_norm": 0.08570241928100586,
"learning_rate": 2.6080647205706855e-06,
"loss": 0.0025,
"step": 9020
},
{
"epoch": 54.07185628742515,
"grad_norm": 0.1345815509557724,
"learning_rate": 2.555619542962834e-06,
"loss": 0.0027,
"step": 9030
},
{
"epoch": 54.131736526946106,
"grad_norm": 0.08837680518627167,
"learning_rate": 2.503693209498409e-06,
"loss": 0.003,
"step": 9040
},
{
"epoch": 54.191616766467064,
"grad_norm": 0.061925455927848816,
"learning_rate": 2.452286288035449e-06,
"loss": 0.0038,
"step": 9050
},
{
"epoch": 54.25149700598802,
"grad_norm": 0.042738113552331924,
"learning_rate": 2.4013993407518363e-06,
"loss": 0.0031,
"step": 9060
},
{
"epoch": 54.31137724550898,
"grad_norm": 0.04965038225054741,
"learning_rate": 2.351032924139063e-06,
"loss": 0.0036,
"step": 9070
},
{
"epoch": 54.37125748502994,
"grad_norm": 0.11647455394268036,
"learning_rate": 2.30118758899619e-06,
"loss": 0.0031,
"step": 9080
},
{
"epoch": 54.4311377245509,
"grad_norm": 0.09770586341619492,
"learning_rate": 2.2518638804238157e-06,
"loss": 0.0027,
"step": 9090
},
{
"epoch": 54.49101796407186,
"grad_norm": 0.05404093861579895,
"learning_rate": 2.203062337818118e-06,
"loss": 0.0035,
"step": 9100
},
{
"epoch": 54.550898203592816,
"grad_norm": 0.05813749134540558,
"learning_rate": 2.1547834948649483e-06,
"loss": 0.0028,
"step": 9110
},
{
"epoch": 54.61077844311377,
"grad_norm": 0.04653799906373024,
"learning_rate": 2.1070278795340017e-06,
"loss": 0.0028,
"step": 9120
},
{
"epoch": 54.67065868263473,
"grad_norm": 0.06094174087047577,
"learning_rate": 2.059796014073029e-06,
"loss": 0.0038,
"step": 9130
},
{
"epoch": 54.73053892215569,
"grad_norm": 0.04083485156297684,
"learning_rate": 2.01308841500214e-06,
"loss": 0.0024,
"step": 9140
},
{
"epoch": 54.790419161676645,
"grad_norm": 0.07090050727128983,
"learning_rate": 1.9669055931081704e-06,
"loss": 0.0024,
"step": 9150
},
{
"epoch": 54.8502994011976,
"grad_norm": 0.052495718002319336,
"learning_rate": 1.9212480534390507e-06,
"loss": 0.0029,
"step": 9160
},
{
"epoch": 54.91017964071856,
"grad_norm": 0.08185823261737823,
"learning_rate": 1.8761162952983246e-06,
"loss": 0.0044,
"step": 9170
},
{
"epoch": 54.97005988023952,
"grad_norm": 0.037968676537275314,
"learning_rate": 1.8315108122396618e-06,
"loss": 0.0034,
"step": 9180
},
{
"epoch": 55.02994011976048,
"grad_norm": 0.07475395500659943,
"learning_rate": 1.787432092061475e-06,
"loss": 0.0038,
"step": 9190
},
{
"epoch": 55.08982035928144,
"grad_norm": 0.0627773180603981,
"learning_rate": 1.743880616801602e-06,
"loss": 0.0036,
"step": 9200
},
{
"epoch": 55.1497005988024,
"grad_norm": 0.05314618721604347,
"learning_rate": 1.7008568627319865e-06,
"loss": 0.0032,
"step": 9210
},
{
"epoch": 55.209580838323355,
"grad_norm": 0.12232566624879837,
"learning_rate": 1.6583613003535226e-06,
"loss": 0.0034,
"step": 9220
},
{
"epoch": 55.26946107784431,
"grad_norm": 0.09394638240337372,
"learning_rate": 1.6163943943908522e-06,
"loss": 0.0029,
"step": 9230
},
{
"epoch": 55.32934131736527,
"grad_norm": 0.07310517132282257,
"learning_rate": 1.5749566037873476e-06,
"loss": 0.0022,
"step": 9240
},
{
"epoch": 55.38922155688623,
"grad_norm": 0.10899892449378967,
"learning_rate": 1.5340483817000428e-06,
"loss": 0.0042,
"step": 9250
},
{
"epoch": 55.449101796407184,
"grad_norm": 0.06273479759693146,
"learning_rate": 1.4936701754947101e-06,
"loss": 0.0032,
"step": 9260
},
{
"epoch": 55.50898203592814,
"grad_norm": 0.05700475722551346,
"learning_rate": 1.4538224267409361e-06,
"loss": 0.005,
"step": 9270
},
{
"epoch": 55.5688622754491,
"grad_norm": 0.03717967867851257,
"learning_rate": 1.414505571207314e-06,
"loss": 0.0023,
"step": 9280
},
{
"epoch": 55.62874251497006,
"grad_norm": 0.048519451171159744,
"learning_rate": 1.3757200388566816e-06,
"loss": 0.0026,
"step": 9290
},
{
"epoch": 55.68862275449102,
"grad_norm": 0.15759176015853882,
"learning_rate": 1.3374662538414074e-06,
"loss": 0.0032,
"step": 9300
},
{
"epoch": 55.74850299401198,
"grad_norm": 0.06813961267471313,
"learning_rate": 1.2997446344987617e-06,
"loss": 0.0028,
"step": 9310
},
{
"epoch": 55.808383233532936,
"grad_norm": 0.050866179168224335,
"learning_rate": 1.262555593346315e-06,
"loss": 0.0036,
"step": 9320
},
{
"epoch": 55.868263473053894,
"grad_norm": 0.10787027329206467,
"learning_rate": 1.2258995370774685e-06,
"loss": 0.0035,
"step": 9330
},
{
"epoch": 55.92814371257485,
"grad_norm": 0.08167394995689392,
"learning_rate": 1.1897768665569798e-06,
"loss": 0.0026,
"step": 9340
},
{
"epoch": 55.98802395209581,
"grad_norm": 0.0572650209069252,
"learning_rate": 1.1541879768165954e-06,
"loss": 0.003,
"step": 9350
},
{
"epoch": 56.047904191616766,
"grad_norm": 0.1302272528409958,
"learning_rate": 1.1191332570507085e-06,
"loss": 0.0037,
"step": 9360
},
{
"epoch": 56.10778443113772,
"grad_norm": 0.06131012365221977,
"learning_rate": 1.0846130906121132e-06,
"loss": 0.0029,
"step": 9370
},
{
"epoch": 56.16766467065868,
"grad_norm": 0.07775352895259857,
"learning_rate": 1.0506278550078131e-06,
"loss": 0.0027,
"step": 9380
},
{
"epoch": 56.22754491017964,
"grad_norm": 0.05495336651802063,
"learning_rate": 1.0171779218949185e-06,
"loss": 0.0023,
"step": 9390
},
{
"epoch": 56.287425149700596,
"grad_norm": 0.10397523641586304,
"learning_rate": 9.842636570765174e-07,
"loss": 0.0027,
"step": 9400
},
{
"epoch": 56.34730538922156,
"grad_norm": 0.10846229642629623,
"learning_rate": 9.518854204977612e-07,
"loss": 0.0052,
"step": 9410
},
{
"epoch": 56.40718562874252,
"grad_norm": 0.081324003636837,
"learning_rate": 9.200435662418349e-07,
"loss": 0.0042,
"step": 9420
},
{
"epoch": 56.467065868263475,
"grad_norm": 0.1041228249669075,
"learning_rate": 8.887384425261658e-07,
"loss": 0.0033,
"step": 9430
},
{
"epoch": 56.52694610778443,
"grad_norm": 0.10407817363739014,
"learning_rate": 8.579703916985648e-07,
"loss": 0.0041,
"step": 9440
},
{
"epoch": 56.58682634730539,
"grad_norm": 0.10266060382127762,
"learning_rate": 8.277397502335194e-07,
"loss": 0.0039,
"step": 9450
},
{
"epoch": 56.64670658682635,
"grad_norm": 0.1146034225821495,
"learning_rate": 7.980468487284675e-07,
"loss": 0.0034,
"step": 9460
},
{
"epoch": 56.706586826347305,
"grad_norm": 0.05145533010363579,
"learning_rate": 7.688920119002297e-07,
"loss": 0.0027,
"step": 9470
},
{
"epoch": 56.76646706586826,
"grad_norm": 0.04698356240987778,
"learning_rate": 7.402755585814269e-07,
"loss": 0.0023,
"step": 9480
},
{
"epoch": 56.82634730538922,
"grad_norm": 0.11943868547677994,
"learning_rate": 7.121978017170073e-07,
"loss": 0.0031,
"step": 9490
},
{
"epoch": 56.88622754491018,
"grad_norm": 0.07906266301870346,
"learning_rate": 6.846590483608306e-07,
"loss": 0.0025,
"step": 9500
},
{
"epoch": 56.946107784431135,
"grad_norm": 0.05585436522960663,
"learning_rate": 6.576595996722834e-07,
"loss": 0.0026,
"step": 9510
},
{
"epoch": 57.0059880239521,
"grad_norm": 0.15105164051055908,
"learning_rate": 6.311997509130141e-07,
"loss": 0.003,
"step": 9520
},
{
"epoch": 57.06586826347306,
"grad_norm": 0.04525098204612732,
"learning_rate": 6.052797914436803e-07,
"loss": 0.0026,
"step": 9530
},
{
"epoch": 57.125748502994014,
"grad_norm": 0.031021978706121445,
"learning_rate": 5.799000047208181e-07,
"loss": 0.0024,
"step": 9540
},
{
"epoch": 57.18562874251497,
"grad_norm": 0.04009797051548958,
"learning_rate": 5.550606682937054e-07,
"loss": 0.0038,
"step": 9550
},
{
"epoch": 57.24550898203593,
"grad_norm": 0.05139565467834473,
"learning_rate": 5.307620538013481e-07,
"loss": 0.003,
"step": 9560
},
{
"epoch": 57.30538922155689,
"grad_norm": 0.05232622101902962,
"learning_rate": 5.070044269694874e-07,
"loss": 0.0029,
"step": 9570
},
{
"epoch": 57.365269461077844,
"grad_norm": 0.06974203884601593,
"learning_rate": 4.837880476077417e-07,
"loss": 0.0022,
"step": 9580
},
{
"epoch": 57.4251497005988,
"grad_norm": 0.04489697888493538,
"learning_rate": 4.6111316960670835e-07,
"loss": 0.0027,
"step": 9590
},
{
"epoch": 57.48502994011976,
"grad_norm": 0.07406875491142273,
"learning_rate": 4.389800409352218e-07,
"loss": 0.0028,
"step": 9600
},
{
"epoch": 57.544910179640716,
"grad_norm": 0.09496568888425827,
"learning_rate": 4.173889036376277e-07,
"loss": 0.0025,
"step": 9610
},
{
"epoch": 57.604790419161674,
"grad_norm": 0.03899621590971947,
"learning_rate": 3.963399938311463e-07,
"loss": 0.0031,
"step": 9620
},
{
"epoch": 57.66467065868264,
"grad_norm": 0.07910379767417908,
"learning_rate": 3.7583354170328545e-07,
"loss": 0.0039,
"step": 9630
},
{
"epoch": 57.724550898203596,
"grad_norm": 0.05029458552598953,
"learning_rate": 3.558697715093207e-07,
"loss": 0.0028,
"step": 9640
},
{
"epoch": 57.78443113772455,
"grad_norm": 0.08895451575517654,
"learning_rate": 3.3644890156983576e-07,
"loss": 0.0024,
"step": 9650
},
{
"epoch": 57.84431137724551,
"grad_norm": 0.08851414918899536,
"learning_rate": 3.175711442683638e-07,
"loss": 0.0025,
"step": 9660
},
{
"epoch": 57.90419161676647,
"grad_norm": 0.09583312273025513,
"learning_rate": 2.9923670604902197e-07,
"loss": 0.0034,
"step": 9670
},
{
"epoch": 57.964071856287426,
"grad_norm": 0.050059035420417786,
"learning_rate": 2.814457874143028e-07,
"loss": 0.0039,
"step": 9680
},
{
"epoch": 58.02395209580838,
"grad_norm": 0.02851296216249466,
"learning_rate": 2.641985829228366e-07,
"loss": 0.0023,
"step": 9690
},
{
"epoch": 58.08383233532934,
"grad_norm": 0.04403981566429138,
"learning_rate": 2.474952811872877e-07,
"loss": 0.0026,
"step": 9700
},
{
"epoch": 58.1437125748503,
"grad_norm": 0.04316142201423645,
"learning_rate": 2.3133606487228397e-07,
"loss": 0.0025,
"step": 9710
},
{
"epoch": 58.203592814371255,
"grad_norm": 0.07886752486228943,
"learning_rate": 2.157211106924295e-07,
"loss": 0.0026,
"step": 9720
},
{
"epoch": 58.26347305389221,
"grad_norm": 0.05740463361144066,
"learning_rate": 2.006505894103672e-07,
"loss": 0.0034,
"step": 9730
},
{
"epoch": 58.32335329341317,
"grad_norm": 0.059307705610990524,
"learning_rate": 1.8612466583489696e-07,
"loss": 0.0023,
"step": 9740
},
{
"epoch": 58.383233532934135,
"grad_norm": 0.06218918040394783,
"learning_rate": 1.7214349881918834e-07,
"loss": 0.0029,
"step": 9750
},
{
"epoch": 58.44311377245509,
"grad_norm": 0.12438742071390152,
"learning_rate": 1.5870724125904845e-07,
"loss": 0.0035,
"step": 9760
},
{
"epoch": 58.50299401197605,
"grad_norm": 0.16784633696079254,
"learning_rate": 1.4581604009124006e-07,
"loss": 0.0035,
"step": 9770
},
{
"epoch": 58.56287425149701,
"grad_norm": 0.12060746550559998,
"learning_rate": 1.334700362918717e-07,
"loss": 0.0024,
"step": 9780
},
{
"epoch": 58.622754491017965,
"grad_norm": 0.05394333228468895,
"learning_rate": 1.2166936487486015e-07,
"loss": 0.004,
"step": 9790
},
{
"epoch": 58.68263473053892,
"grad_norm": 0.04444008693099022,
"learning_rate": 1.1041415489045914e-07,
"loss": 0.0034,
"step": 9800
},
{
"epoch": 58.74251497005988,
"grad_norm": 0.0819375291466713,
"learning_rate": 9.970452942384412e-08,
"loss": 0.0043,
"step": 9810
},
{
"epoch": 58.80239520958084,
"grad_norm": 0.10424412786960602,
"learning_rate": 8.954060559375754e-08,
"loss": 0.0028,
"step": 9820
},
{
"epoch": 58.862275449101794,
"grad_norm": 0.16958363354206085,
"learning_rate": 7.99224945512489e-08,
"loss": 0.0028,
"step": 9830
},
{
"epoch": 58.92215568862275,
"grad_norm": 0.08972541242837906,
"learning_rate": 7.085030147843675e-08,
"loss": 0.0023,
"step": 9840
},
{
"epoch": 58.98203592814371,
"grad_norm": 0.10610095411539078,
"learning_rate": 6.232412558736523e-08,
"loss": 0.0043,
"step": 9850
},
{
"epoch": 59.041916167664674,
"grad_norm": 0.03575646132230759,
"learning_rate": 5.434406011893822e-08,
"loss": 0.0034,
"step": 9860
},
{
"epoch": 59.10179640718563,
"grad_norm": 0.07221001386642456,
"learning_rate": 4.6910192341864664e-08,
"loss": 0.0038,
"step": 9870
},
{
"epoch": 59.16167664670659,
"grad_norm": 0.09040188789367676,
"learning_rate": 4.0022603551737035e-08,
"loss": 0.0024,
"step": 9880
},
{
"epoch": 59.221556886227546,
"grad_norm": 0.16115228831768036,
"learning_rate": 3.3681369070120985e-08,
"loss": 0.0035,
"step": 9890
},
{
"epoch": 59.2814371257485,
"grad_norm": 0.039209216833114624,
"learning_rate": 2.7886558243744866e-08,
"loss": 0.0024,
"step": 9900
},
{
"epoch": 59.34131736526946,
"grad_norm": 0.07305026054382324,
"learning_rate": 2.2638234443722596e-08,
"loss": 0.0024,
"step": 9910
},
{
"epoch": 59.40119760479042,
"grad_norm": 0.07122573256492615,
"learning_rate": 1.7936455064887504e-08,
"loss": 0.0035,
"step": 9920
},
{
"epoch": 59.461077844311376,
"grad_norm": 0.03151412680745125,
"learning_rate": 1.378127152514841e-08,
"loss": 0.0025,
"step": 9930
},
{
"epoch": 59.52095808383233,
"grad_norm": 0.1414758712053299,
"learning_rate": 1.0172729264917857e-08,
"loss": 0.0056,
"step": 9940
},
{
"epoch": 59.58083832335329,
"grad_norm": 0.037605319172143936,
"learning_rate": 7.1108677466458215e-09,
"loss": 0.0037,
"step": 9950
},
{
"epoch": 59.64071856287425,
"grad_norm": 0.047965340316295624,
"learning_rate": 4.595720454353414e-09,
"loss": 0.0029,
"step": 9960
},
{
"epoch": 59.70059880239521,
"grad_norm": 0.03662804514169693,
"learning_rate": 2.627314893294264e-09,
"loss": 0.0024,
"step": 9970
},
{
"epoch": 59.76047904191617,
"grad_norm": 0.04171596094965935,
"learning_rate": 1.2056725896270048e-09,
"loss": 0.0029,
"step": 9980
},
{
"epoch": 59.82035928143713,
"grad_norm": 0.09089730679988861,
"learning_rate": 3.308090902098826e-10,
"loss": 0.0029,
"step": 9990
},
{
"epoch": 59.880239520958085,
"grad_norm": 0.029615165665745735,
"learning_rate": 2.7339624120159555e-12,
"loss": 0.0023,
"step": 10000
},
{
"epoch": 59.880239520958085,
"step": 10000,
"total_flos": 0.0,
"train_loss": 0.011509704336524009,
"train_runtime": 18787.7837,
"train_samples_per_second": 212.904,
"train_steps_per_second": 0.532
}
],
"logging_steps": 10,
"max_steps": 10000,
"num_input_tokens_seen": 0,
"num_train_epochs": 60,
"save_steps": 800,
"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": 50,
"trial_name": null,
"trial_params": null
}