svjack's picture
Upload folder using huggingface_hub
eca60d4 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.352,
"eval_steps": 500,
"global_step": 4400,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0004,
"grad_norm": 4.773154258728027,
"learning_rate": 4.9999997806754614e-05,
"loss": 2.4535,
"step": 5
},
{
"epoch": 0.0008,
"grad_norm": 5.585668563842773,
"learning_rate": 4.999999122701883e-05,
"loss": 1.9042,
"step": 10
},
{
"epoch": 0.0012,
"grad_norm": 2.7855770587921143,
"learning_rate": 4.99999802607938e-05,
"loss": 1.6398,
"step": 15
},
{
"epoch": 0.0016,
"grad_norm": 2.4066975116729736,
"learning_rate": 4.9999964908081455e-05,
"loss": 1.3978,
"step": 20
},
{
"epoch": 0.002,
"grad_norm": 1.6574153900146484,
"learning_rate": 4.999994516888449e-05,
"loss": 1.3416,
"step": 25
},
{
"epoch": 0.0024,
"grad_norm": 1.886969804763794,
"learning_rate": 4.999992104320636e-05,
"loss": 1.4384,
"step": 30
},
{
"epoch": 0.0028,
"grad_norm": 1.656772494316101,
"learning_rate": 4.99998925310513e-05,
"loss": 1.3529,
"step": 35
},
{
"epoch": 0.0032,
"grad_norm": 1.6747465133666992,
"learning_rate": 4.999985963242432e-05,
"loss": 1.3142,
"step": 40
},
{
"epoch": 0.0036,
"grad_norm": 1.633796215057373,
"learning_rate": 4.999982234733118e-05,
"loss": 1.3162,
"step": 45
},
{
"epoch": 0.004,
"grad_norm": 1.5636812448501587,
"learning_rate": 4.999978067577844e-05,
"loss": 1.3443,
"step": 50
},
{
"epoch": 0.0044,
"grad_norm": 1.6511592864990234,
"learning_rate": 4.9999734617773405e-05,
"loss": 1.2582,
"step": 55
},
{
"epoch": 0.0048,
"grad_norm": 1.6226695775985718,
"learning_rate": 4.999968417332415e-05,
"loss": 1.3449,
"step": 60
},
{
"epoch": 0.0052,
"grad_norm": 1.502573847770691,
"learning_rate": 4.9999629342439524e-05,
"loss": 1.2836,
"step": 65
},
{
"epoch": 0.0056,
"grad_norm": 1.5057326555252075,
"learning_rate": 4.999957012512916e-05,
"loss": 1.2437,
"step": 70
},
{
"epoch": 0.006,
"grad_norm": 1.627768874168396,
"learning_rate": 4.999950652140343e-05,
"loss": 1.2318,
"step": 75
},
{
"epoch": 0.0064,
"grad_norm": 1.7765494585037231,
"learning_rate": 4.999943853127351e-05,
"loss": 1.3253,
"step": 80
},
{
"epoch": 0.0068,
"grad_norm": 1.7718794345855713,
"learning_rate": 4.999936615475133e-05,
"loss": 1.3195,
"step": 85
},
{
"epoch": 0.0072,
"grad_norm": 1.2835720777511597,
"learning_rate": 4.999928939184958e-05,
"loss": 1.2973,
"step": 90
},
{
"epoch": 0.0076,
"grad_norm": 1.77462637424469,
"learning_rate": 4.999920824258173e-05,
"loss": 1.2883,
"step": 95
},
{
"epoch": 0.008,
"grad_norm": 1.3049944639205933,
"learning_rate": 4.999912270696202e-05,
"loss": 1.346,
"step": 100
},
{
"epoch": 0.0084,
"grad_norm": 1.7110176086425781,
"learning_rate": 4.9999032785005464e-05,
"loss": 1.3119,
"step": 105
},
{
"epoch": 0.0088,
"grad_norm": 1.6607234477996826,
"learning_rate": 4.9998938476727826e-05,
"loss": 1.2937,
"step": 110
},
{
"epoch": 0.0092,
"grad_norm": 1.6394551992416382,
"learning_rate": 4.999883978214567e-05,
"loss": 1.2376,
"step": 115
},
{
"epoch": 0.0096,
"grad_norm": 1.8243571519851685,
"learning_rate": 4.9998736701276295e-05,
"loss": 1.3009,
"step": 120
},
{
"epoch": 0.01,
"grad_norm": 1.9557856321334839,
"learning_rate": 4.999862923413781e-05,
"loss": 1.2827,
"step": 125
},
{
"epoch": 0.0104,
"grad_norm": 1.4470913410186768,
"learning_rate": 4.999851738074904e-05,
"loss": 1.3158,
"step": 130
},
{
"epoch": 0.0108,
"grad_norm": 1.4409664869308472,
"learning_rate": 4.999840114112965e-05,
"loss": 1.2554,
"step": 135
},
{
"epoch": 0.0112,
"grad_norm": 1.5476871728897095,
"learning_rate": 4.99982805153e-05,
"loss": 1.2788,
"step": 140
},
{
"epoch": 0.0116,
"grad_norm": 1.5238358974456787,
"learning_rate": 4.999815550328128e-05,
"loss": 1.22,
"step": 145
},
{
"epoch": 0.012,
"grad_norm": 1.388649344444275,
"learning_rate": 4.9998026105095405e-05,
"loss": 1.2836,
"step": 150
},
{
"epoch": 0.0124,
"grad_norm": 1.6821857690811157,
"learning_rate": 4.999789232076509e-05,
"loss": 1.2498,
"step": 155
},
{
"epoch": 0.0128,
"grad_norm": 1.7721805572509766,
"learning_rate": 4.9997754150313815e-05,
"loss": 1.2757,
"step": 160
},
{
"epoch": 0.0132,
"grad_norm": 1.5760489702224731,
"learning_rate": 4.99976115937658e-05,
"loss": 1.2378,
"step": 165
},
{
"epoch": 0.0136,
"grad_norm": 1.7729449272155762,
"learning_rate": 4.999746465114609e-05,
"loss": 1.2126,
"step": 170
},
{
"epoch": 0.014,
"grad_norm": 1.813071370124817,
"learning_rate": 4.999731332248044e-05,
"loss": 1.2379,
"step": 175
},
{
"epoch": 0.0144,
"grad_norm": 1.7245049476623535,
"learning_rate": 4.999715760779541e-05,
"loss": 1.2449,
"step": 180
},
{
"epoch": 0.0148,
"grad_norm": 1.8477286100387573,
"learning_rate": 4.999699750711833e-05,
"loss": 1.2934,
"step": 185
},
{
"epoch": 0.0152,
"grad_norm": 1.4089431762695312,
"learning_rate": 4.9996833020477285e-05,
"loss": 1.2313,
"step": 190
},
{
"epoch": 0.0156,
"grad_norm": 2.9127607345581055,
"learning_rate": 4.999666414790113e-05,
"loss": 1.2328,
"step": 195
},
{
"epoch": 0.016,
"grad_norm": 1.9049687385559082,
"learning_rate": 4.9996490889419514e-05,
"loss": 1.2497,
"step": 200
},
{
"epoch": 0.0164,
"grad_norm": 2.1356122493743896,
"learning_rate": 4.9996313245062823e-05,
"loss": 1.2877,
"step": 205
},
{
"epoch": 0.0168,
"grad_norm": 1.6691316366195679,
"learning_rate": 4.999613121486222e-05,
"loss": 1.2884,
"step": 210
},
{
"epoch": 0.0172,
"grad_norm": 1.8735146522521973,
"learning_rate": 4.999594479884965e-05,
"loss": 1.2742,
"step": 215
},
{
"epoch": 0.0176,
"grad_norm": 1.6481894254684448,
"learning_rate": 4.999575399705783e-05,
"loss": 1.2408,
"step": 220
},
{
"epoch": 0.018,
"grad_norm": 1.7049623727798462,
"learning_rate": 4.999555880952023e-05,
"loss": 1.1997,
"step": 225
},
{
"epoch": 0.0184,
"grad_norm": 1.9054642915725708,
"learning_rate": 4.999535923627109e-05,
"loss": 1.2414,
"step": 230
},
{
"epoch": 0.0188,
"grad_norm": 1.798621416091919,
"learning_rate": 4.999515527734545e-05,
"loss": 1.2848,
"step": 235
},
{
"epoch": 0.0192,
"grad_norm": 1.767776608467102,
"learning_rate": 4.999494693277907e-05,
"loss": 1.2626,
"step": 240
},
{
"epoch": 0.0196,
"grad_norm": 2.2002179622650146,
"learning_rate": 4.999473420260853e-05,
"loss": 1.2582,
"step": 245
},
{
"epoch": 0.02,
"grad_norm": 1.9857462644577026,
"learning_rate": 4.999451708687114e-05,
"loss": 1.2702,
"step": 250
},
{
"epoch": 0.0204,
"grad_norm": 1.5707662105560303,
"learning_rate": 4.9994295585605e-05,
"loss": 1.1929,
"step": 255
},
{
"epoch": 0.0208,
"grad_norm": 1.5665488243103027,
"learning_rate": 4.999406969884897e-05,
"loss": 1.2231,
"step": 260
},
{
"epoch": 0.0212,
"grad_norm": 1.508448600769043,
"learning_rate": 4.9993839426642685e-05,
"loss": 1.2621,
"step": 265
},
{
"epoch": 0.0216,
"grad_norm": 1.6147902011871338,
"learning_rate": 4.999360476902656e-05,
"loss": 1.2124,
"step": 270
},
{
"epoch": 0.022,
"grad_norm": 1.9145556688308716,
"learning_rate": 4.999336572604175e-05,
"loss": 1.278,
"step": 275
},
{
"epoch": 0.0224,
"grad_norm": 1.7562631368637085,
"learning_rate": 4.999312229773022e-05,
"loss": 1.2608,
"step": 280
},
{
"epoch": 0.0228,
"grad_norm": 1.7679455280303955,
"learning_rate": 4.9992874484134653e-05,
"loss": 1.2849,
"step": 285
},
{
"epoch": 0.0232,
"grad_norm": 1.5522981882095337,
"learning_rate": 4.999262228529855e-05,
"loss": 1.2251,
"step": 290
},
{
"epoch": 0.0236,
"grad_norm": 1.8789103031158447,
"learning_rate": 4.999236570126616e-05,
"loss": 1.2721,
"step": 295
},
{
"epoch": 0.024,
"grad_norm": 1.6382054090499878,
"learning_rate": 4.99921047320825e-05,
"loss": 1.2494,
"step": 300
},
{
"epoch": 0.0244,
"grad_norm": 1.7209609746932983,
"learning_rate": 4.999183937779336e-05,
"loss": 1.214,
"step": 305
},
{
"epoch": 0.0248,
"grad_norm": 2.084233283996582,
"learning_rate": 4.99915696384453e-05,
"loss": 1.2649,
"step": 310
},
{
"epoch": 0.0252,
"grad_norm": 1.9279961585998535,
"learning_rate": 4.9991295514085644e-05,
"loss": 1.1784,
"step": 315
},
{
"epoch": 0.0256,
"grad_norm": 2.1031174659729004,
"learning_rate": 4.9991017004762496e-05,
"loss": 1.2838,
"step": 320
},
{
"epoch": 0.026,
"grad_norm": 1.5954639911651611,
"learning_rate": 4.999073411052472e-05,
"loss": 1.3073,
"step": 325
},
{
"epoch": 0.0264,
"grad_norm": 1.612317681312561,
"learning_rate": 4.9990446831421955e-05,
"loss": 1.2914,
"step": 330
},
{
"epoch": 0.0268,
"grad_norm": 1.5913759469985962,
"learning_rate": 4.99901551675046e-05,
"loss": 1.3132,
"step": 335
},
{
"epoch": 0.0272,
"grad_norm": 1.8079237937927246,
"learning_rate": 4.998985911882384e-05,
"loss": 1.1934,
"step": 340
},
{
"epoch": 0.0276,
"grad_norm": 1.9866459369659424,
"learning_rate": 4.998955868543161e-05,
"loss": 1.2486,
"step": 345
},
{
"epoch": 0.028,
"grad_norm": 1.791840672492981,
"learning_rate": 4.998925386738063e-05,
"loss": 1.2049,
"step": 350
},
{
"epoch": 0.0284,
"grad_norm": 1.7033882141113281,
"learning_rate": 4.998894466472438e-05,
"loss": 1.2323,
"step": 355
},
{
"epoch": 0.0288,
"grad_norm": 1.525989294052124,
"learning_rate": 4.998863107751711e-05,
"loss": 1.2439,
"step": 360
},
{
"epoch": 0.0292,
"grad_norm": 1.5799592733383179,
"learning_rate": 4.9988313105813856e-05,
"loss": 1.2807,
"step": 365
},
{
"epoch": 0.0296,
"grad_norm": 1.7438101768493652,
"learning_rate": 4.9987990749670395e-05,
"loss": 1.2605,
"step": 370
},
{
"epoch": 0.03,
"grad_norm": 1.538445234298706,
"learning_rate": 4.998766400914329e-05,
"loss": 1.2185,
"step": 375
},
{
"epoch": 0.0304,
"grad_norm": 1.7448121309280396,
"learning_rate": 4.998733288428987e-05,
"loss": 1.2511,
"step": 380
},
{
"epoch": 0.0308,
"grad_norm": 1.7485535144805908,
"learning_rate": 4.9986997375168246e-05,
"loss": 1.2266,
"step": 385
},
{
"epoch": 0.0312,
"grad_norm": 1.970339059829712,
"learning_rate": 4.9986657481837277e-05,
"loss": 1.2363,
"step": 390
},
{
"epoch": 0.0316,
"grad_norm": 1.883518934249878,
"learning_rate": 4.9986313204356594e-05,
"loss": 1.2034,
"step": 395
},
{
"epoch": 0.032,
"grad_norm": 1.6562483310699463,
"learning_rate": 4.9985964542786614e-05,
"loss": 1.2578,
"step": 400
},
{
"epoch": 0.0324,
"grad_norm": 2.5307445526123047,
"learning_rate": 4.99856114971885e-05,
"loss": 1.2506,
"step": 405
},
{
"epoch": 0.0328,
"grad_norm": 1.5744280815124512,
"learning_rate": 4.9985254067624215e-05,
"loss": 1.3237,
"step": 410
},
{
"epoch": 0.0332,
"grad_norm": 1.7964937686920166,
"learning_rate": 4.9984892254156465e-05,
"loss": 1.2538,
"step": 415
},
{
"epoch": 0.0336,
"grad_norm": 1.790570855140686,
"learning_rate": 4.998452605684874e-05,
"loss": 1.3154,
"step": 420
},
{
"epoch": 0.034,
"grad_norm": 1.8018208742141724,
"learning_rate": 4.998415547576527e-05,
"loss": 1.2145,
"step": 425
},
{
"epoch": 0.0344,
"grad_norm": 1.8359650373458862,
"learning_rate": 4.998378051097111e-05,
"loss": 1.2926,
"step": 430
},
{
"epoch": 0.0348,
"grad_norm": 1.9998548030853271,
"learning_rate": 4.9983401162532025e-05,
"loss": 1.2439,
"step": 435
},
{
"epoch": 0.0352,
"grad_norm": 1.9759939908981323,
"learning_rate": 4.998301743051459e-05,
"loss": 1.2516,
"step": 440
},
{
"epoch": 0.0356,
"grad_norm": 1.7635678052902222,
"learning_rate": 4.9982629314986126e-05,
"loss": 1.2549,
"step": 445
},
{
"epoch": 0.036,
"grad_norm": 1.8558741807937622,
"learning_rate": 4.998223681601473e-05,
"loss": 1.278,
"step": 450
},
{
"epoch": 0.0364,
"grad_norm": 1.6500812768936157,
"learning_rate": 4.998183993366928e-05,
"loss": 1.2406,
"step": 455
},
{
"epoch": 0.0368,
"grad_norm": 2.186829090118408,
"learning_rate": 4.998143866801942e-05,
"loss": 1.2514,
"step": 460
},
{
"epoch": 0.0372,
"grad_norm": 1.806112289428711,
"learning_rate": 4.998103301913552e-05,
"loss": 1.2641,
"step": 465
},
{
"epoch": 0.0376,
"grad_norm": 1.7315857410430908,
"learning_rate": 4.9980622987088795e-05,
"loss": 1.2743,
"step": 470
},
{
"epoch": 0.038,
"grad_norm": 1.6048853397369385,
"learning_rate": 4.998020857195117e-05,
"loss": 1.2765,
"step": 475
},
{
"epoch": 0.0384,
"grad_norm": 1.7586766481399536,
"learning_rate": 4.997978977379536e-05,
"loss": 1.2237,
"step": 480
},
{
"epoch": 0.0388,
"grad_norm": 1.8284968137741089,
"learning_rate": 4.997936659269486e-05,
"loss": 1.2165,
"step": 485
},
{
"epoch": 0.0392,
"grad_norm": 1.761220097541809,
"learning_rate": 4.9978939028723894e-05,
"loss": 1.2629,
"step": 490
},
{
"epoch": 0.0396,
"grad_norm": 1.9978684186935425,
"learning_rate": 4.9978507081957494e-05,
"loss": 1.2974,
"step": 495
},
{
"epoch": 0.04,
"grad_norm": 1.830068588256836,
"learning_rate": 4.997807075247146e-05,
"loss": 1.2658,
"step": 500
},
{
"epoch": 0.0404,
"grad_norm": 1.7393221855163574,
"learning_rate": 4.9977630040342346e-05,
"loss": 1.23,
"step": 505
},
{
"epoch": 0.0408,
"grad_norm": 1.7979861497879028,
"learning_rate": 4.9977184945647473e-05,
"loss": 1.266,
"step": 510
},
{
"epoch": 0.0412,
"grad_norm": 1.6707987785339355,
"learning_rate": 4.9976735468464935e-05,
"loss": 1.2253,
"step": 515
},
{
"epoch": 0.0416,
"grad_norm": 1.7217047214508057,
"learning_rate": 4.997628160887361e-05,
"loss": 1.2137,
"step": 520
},
{
"epoch": 0.042,
"grad_norm": 1.889464259147644,
"learning_rate": 4.9975823366953124e-05,
"loss": 1.2829,
"step": 525
},
{
"epoch": 0.0424,
"grad_norm": 1.7830469608306885,
"learning_rate": 4.997536074278387e-05,
"loss": 1.2531,
"step": 530
},
{
"epoch": 0.0428,
"grad_norm": 1.875157356262207,
"learning_rate": 4.9974893736447045e-05,
"loss": 1.2991,
"step": 535
},
{
"epoch": 0.0432,
"grad_norm": 1.7998117208480835,
"learning_rate": 4.997442234802456e-05,
"loss": 1.3188,
"step": 540
},
{
"epoch": 0.0436,
"grad_norm": 1.8334897756576538,
"learning_rate": 4.997394657759915e-05,
"loss": 1.2815,
"step": 545
},
{
"epoch": 0.044,
"grad_norm": 1.8215042352676392,
"learning_rate": 4.9973466425254286e-05,
"loss": 1.2606,
"step": 550
},
{
"epoch": 0.0444,
"grad_norm": 1.7977797985076904,
"learning_rate": 4.997298189107421e-05,
"loss": 1.2369,
"step": 555
},
{
"epoch": 0.0448,
"grad_norm": 1.560219407081604,
"learning_rate": 4.997249297514394e-05,
"loss": 1.2351,
"step": 560
},
{
"epoch": 0.0452,
"grad_norm": 1.5830539464950562,
"learning_rate": 4.997199967754925e-05,
"loss": 1.221,
"step": 565
},
{
"epoch": 0.0456,
"grad_norm": 1.8532063961029053,
"learning_rate": 4.997150199837671e-05,
"loss": 1.1822,
"step": 570
},
{
"epoch": 0.046,
"grad_norm": 2.081393003463745,
"learning_rate": 4.997099993771365e-05,
"loss": 1.2456,
"step": 575
},
{
"epoch": 0.0464,
"grad_norm": 1.8640109300613403,
"learning_rate": 4.997049349564814e-05,
"loss": 1.2186,
"step": 580
},
{
"epoch": 0.0468,
"grad_norm": 1.730169415473938,
"learning_rate": 4.996998267226905e-05,
"loss": 1.2369,
"step": 585
},
{
"epoch": 0.0472,
"grad_norm": 1.847095251083374,
"learning_rate": 4.996946746766601e-05,
"loss": 1.2597,
"step": 590
},
{
"epoch": 0.0476,
"grad_norm": 1.756137490272522,
"learning_rate": 4.9968947881929414e-05,
"loss": 1.1965,
"step": 595
},
{
"epoch": 0.048,
"grad_norm": 1.890035629272461,
"learning_rate": 4.996842391515044e-05,
"loss": 1.2345,
"step": 600
},
{
"epoch": 0.0484,
"grad_norm": 1.7195771932601929,
"learning_rate": 4.996789556742101e-05,
"loss": 1.248,
"step": 605
},
{
"epoch": 0.0488,
"grad_norm": 1.7649211883544922,
"learning_rate": 4.996736283883382e-05,
"loss": 1.2504,
"step": 610
},
{
"epoch": 0.0492,
"grad_norm": 1.7328550815582275,
"learning_rate": 4.9966825729482364e-05,
"loss": 1.3122,
"step": 615
},
{
"epoch": 0.0496,
"grad_norm": 1.6901662349700928,
"learning_rate": 4.996628423946087e-05,
"loss": 1.2508,
"step": 620
},
{
"epoch": 0.05,
"grad_norm": 1.7033767700195312,
"learning_rate": 4.996573836886435e-05,
"loss": 1.2145,
"step": 625
},
{
"epoch": 0.0504,
"grad_norm": 1.4727296829223633,
"learning_rate": 4.996518811778858e-05,
"loss": 1.2779,
"step": 630
},
{
"epoch": 0.0508,
"grad_norm": 1.6590018272399902,
"learning_rate": 4.9964633486330116e-05,
"loss": 1.2722,
"step": 635
},
{
"epoch": 0.0512,
"grad_norm": 2.6924021244049072,
"learning_rate": 4.996407447458626e-05,
"loss": 1.2182,
"step": 640
},
{
"epoch": 0.0516,
"grad_norm": 1.8809802532196045,
"learning_rate": 4.99635110826551e-05,
"loss": 1.2368,
"step": 645
},
{
"epoch": 0.052,
"grad_norm": 1.8109803199768066,
"learning_rate": 4.99629433106355e-05,
"loss": 1.2447,
"step": 650
},
{
"epoch": 0.0524,
"grad_norm": 1.6945043802261353,
"learning_rate": 4.996237115862706e-05,
"loss": 1.1895,
"step": 655
},
{
"epoch": 0.0528,
"grad_norm": 1.5174620151519775,
"learning_rate": 4.99617946267302e-05,
"loss": 1.2276,
"step": 660
},
{
"epoch": 0.0532,
"grad_norm": 1.7584028244018555,
"learning_rate": 4.9961213715046045e-05,
"loss": 1.2308,
"step": 665
},
{
"epoch": 0.0536,
"grad_norm": 2.0159788131713867,
"learning_rate": 4.996062842367654e-05,
"loss": 1.2044,
"step": 670
},
{
"epoch": 0.054,
"grad_norm": 1.7064976692199707,
"learning_rate": 4.996003875272438e-05,
"loss": 1.2071,
"step": 675
},
{
"epoch": 0.0544,
"grad_norm": 2.149348258972168,
"learning_rate": 4.995944470229302e-05,
"loss": 1.2218,
"step": 680
},
{
"epoch": 0.0548,
"grad_norm": 1.5453240871429443,
"learning_rate": 4.9958846272486704e-05,
"loss": 1.2287,
"step": 685
},
{
"epoch": 0.0552,
"grad_norm": 1.5038634538650513,
"learning_rate": 4.9958243463410414e-05,
"loss": 1.2153,
"step": 690
},
{
"epoch": 0.0556,
"grad_norm": 1.830925464630127,
"learning_rate": 4.995763627516994e-05,
"loss": 1.2849,
"step": 695
},
{
"epoch": 0.056,
"grad_norm": 1.650992512702942,
"learning_rate": 4.9957024707871806e-05,
"loss": 1.1965,
"step": 700
},
{
"epoch": 0.0564,
"grad_norm": 1.9642475843429565,
"learning_rate": 4.995640876162332e-05,
"loss": 1.2489,
"step": 705
},
{
"epoch": 0.0568,
"grad_norm": 1.6701452732086182,
"learning_rate": 4.9955788436532545e-05,
"loss": 1.1833,
"step": 710
},
{
"epoch": 0.0572,
"grad_norm": 1.8591731786727905,
"learning_rate": 4.9955163732708346e-05,
"loss": 1.2699,
"step": 715
},
{
"epoch": 0.0576,
"grad_norm": 1.772956371307373,
"learning_rate": 4.995453465026032e-05,
"loss": 1.2582,
"step": 720
},
{
"epoch": 0.058,
"grad_norm": 1.8155946731567383,
"learning_rate": 4.9953901189298845e-05,
"loss": 1.2256,
"step": 725
},
{
"epoch": 0.0584,
"grad_norm": 1.7217077016830444,
"learning_rate": 4.9953263349935074e-05,
"loss": 1.1909,
"step": 730
},
{
"epoch": 0.0588,
"grad_norm": 1.8732049465179443,
"learning_rate": 4.995262113228091e-05,
"loss": 1.2023,
"step": 735
},
{
"epoch": 0.0592,
"grad_norm": 1.96453857421875,
"learning_rate": 4.995197453644905e-05,
"loss": 1.2066,
"step": 740
},
{
"epoch": 0.0596,
"grad_norm": 1.7250394821166992,
"learning_rate": 4.9951323562552934e-05,
"loss": 1.2038,
"step": 745
},
{
"epoch": 0.06,
"grad_norm": 1.5440521240234375,
"learning_rate": 4.995066821070679e-05,
"loss": 1.2509,
"step": 750
},
{
"epoch": 0.0604,
"grad_norm": 1.8293278217315674,
"learning_rate": 4.99500084810256e-05,
"loss": 1.2952,
"step": 755
},
{
"epoch": 0.0608,
"grad_norm": 1.5876418352127075,
"learning_rate": 4.994934437362513e-05,
"loss": 1.1556,
"step": 760
},
{
"epoch": 0.0612,
"grad_norm": 1.4117156267166138,
"learning_rate": 4.994867588862189e-05,
"loss": 1.189,
"step": 765
},
{
"epoch": 0.0616,
"grad_norm": 1.6715811491012573,
"learning_rate": 4.994800302613318e-05,
"loss": 1.1876,
"step": 770
},
{
"epoch": 0.062,
"grad_norm": 1.551499366760254,
"learning_rate": 4.994732578627706e-05,
"loss": 1.2369,
"step": 775
},
{
"epoch": 0.0624,
"grad_norm": 1.9798293113708496,
"learning_rate": 4.9946644169172355e-05,
"loss": 1.214,
"step": 780
},
{
"epoch": 0.0628,
"grad_norm": 1.97709321975708,
"learning_rate": 4.994595817493867e-05,
"loss": 1.2118,
"step": 785
},
{
"epoch": 0.0632,
"grad_norm": 1.8808115720748901,
"learning_rate": 4.9945267803696364e-05,
"loss": 1.223,
"step": 790
},
{
"epoch": 0.0636,
"grad_norm": 2.002140998840332,
"learning_rate": 4.9944573055566556e-05,
"loss": 1.1949,
"step": 795
},
{
"epoch": 0.064,
"grad_norm": 1.5803587436676025,
"learning_rate": 4.994387393067117e-05,
"loss": 1.2275,
"step": 800
},
{
"epoch": 0.0644,
"grad_norm": 1.8434706926345825,
"learning_rate": 4.9943170429132855e-05,
"loss": 1.1766,
"step": 805
},
{
"epoch": 0.0648,
"grad_norm": 1.7833911180496216,
"learning_rate": 4.9942462551075056e-05,
"loss": 1.2238,
"step": 810
},
{
"epoch": 0.0652,
"grad_norm": 1.8478978872299194,
"learning_rate": 4.994175029662198e-05,
"loss": 1.2024,
"step": 815
},
{
"epoch": 0.0656,
"grad_norm": 1.597944974899292,
"learning_rate": 4.994103366589859e-05,
"loss": 1.261,
"step": 820
},
{
"epoch": 0.066,
"grad_norm": 1.515336275100708,
"learning_rate": 4.994031265903063e-05,
"loss": 1.2248,
"step": 825
},
{
"epoch": 0.0664,
"grad_norm": 1.7543236017227173,
"learning_rate": 4.9939587276144616e-05,
"loss": 1.216,
"step": 830
},
{
"epoch": 0.0668,
"grad_norm": 1.6050115823745728,
"learning_rate": 4.993885751736781e-05,
"loss": 1.1822,
"step": 835
},
{
"epoch": 0.0672,
"grad_norm": 1.7370154857635498,
"learning_rate": 4.993812338282826e-05,
"loss": 1.278,
"step": 840
},
{
"epoch": 0.0676,
"grad_norm": 1.7703272104263306,
"learning_rate": 4.993738487265478e-05,
"loss": 1.2005,
"step": 845
},
{
"epoch": 0.068,
"grad_norm": 1.6980682611465454,
"learning_rate": 4.993664198697694e-05,
"loss": 1.1899,
"step": 850
},
{
"epoch": 0.0684,
"grad_norm": 1.7821124792099,
"learning_rate": 4.99358947259251e-05,
"loss": 1.219,
"step": 855
},
{
"epoch": 0.0688,
"grad_norm": 1.5321515798568726,
"learning_rate": 4.993514308963036e-05,
"loss": 1.2443,
"step": 860
},
{
"epoch": 0.0692,
"grad_norm": 1.5153166055679321,
"learning_rate": 4.993438707822462e-05,
"loss": 1.2262,
"step": 865
},
{
"epoch": 0.0696,
"grad_norm": 1.7875345945358276,
"learning_rate": 4.993362669184051e-05,
"loss": 1.2437,
"step": 870
},
{
"epoch": 0.07,
"grad_norm": 1.6037136316299438,
"learning_rate": 4.9932861930611454e-05,
"loss": 1.2757,
"step": 875
},
{
"epoch": 0.0704,
"grad_norm": 1.4552197456359863,
"learning_rate": 4.993209279467164e-05,
"loss": 1.1969,
"step": 880
},
{
"epoch": 0.0708,
"grad_norm": 1.4519599676132202,
"learning_rate": 4.993131928415602e-05,
"loss": 1.2608,
"step": 885
},
{
"epoch": 0.0712,
"grad_norm": 1.5513790845870972,
"learning_rate": 4.993054139920032e-05,
"loss": 1.2101,
"step": 890
},
{
"epoch": 0.0716,
"grad_norm": 1.6675373315811157,
"learning_rate": 4.9929759139941e-05,
"loss": 1.2782,
"step": 895
},
{
"epoch": 0.072,
"grad_norm": 1.501469373703003,
"learning_rate": 4.992897250651535e-05,
"loss": 1.1617,
"step": 900
},
{
"epoch": 0.0724,
"grad_norm": 1.6184465885162354,
"learning_rate": 4.992818149906138e-05,
"loss": 1.2491,
"step": 905
},
{
"epoch": 0.0728,
"grad_norm": 1.8921860456466675,
"learning_rate": 4.992738611771787e-05,
"loss": 1.2323,
"step": 910
},
{
"epoch": 0.0732,
"grad_norm": 1.698572039604187,
"learning_rate": 4.992658636262438e-05,
"loss": 1.2315,
"step": 915
},
{
"epoch": 0.0736,
"grad_norm": 1.8895962238311768,
"learning_rate": 4.992578223392124e-05,
"loss": 1.2285,
"step": 920
},
{
"epoch": 0.074,
"grad_norm": 1.4930404424667358,
"learning_rate": 4.992497373174955e-05,
"loss": 1.2123,
"step": 925
},
{
"epoch": 0.0744,
"grad_norm": 1.791544795036316,
"learning_rate": 4.992416085625115e-05,
"loss": 1.2318,
"step": 930
},
{
"epoch": 0.0748,
"grad_norm": 1.7196767330169678,
"learning_rate": 4.9923343607568684e-05,
"loss": 1.1746,
"step": 935
},
{
"epoch": 0.0752,
"grad_norm": 1.732494592666626,
"learning_rate": 4.992252198584554e-05,
"loss": 1.2146,
"step": 940
},
{
"epoch": 0.0756,
"grad_norm": 1.6859017610549927,
"learning_rate": 4.992169599122587e-05,
"loss": 1.2061,
"step": 945
},
{
"epoch": 0.076,
"grad_norm": 1.8825091123580933,
"learning_rate": 4.9920865623854615e-05,
"loss": 1.1909,
"step": 950
},
{
"epoch": 0.0764,
"grad_norm": 1.9849854707717896,
"learning_rate": 4.9920030883877476e-05,
"loss": 1.2249,
"step": 955
},
{
"epoch": 0.0768,
"grad_norm": 2.0635294914245605,
"learning_rate": 4.9919191771440905e-05,
"loss": 1.2154,
"step": 960
},
{
"epoch": 0.0772,
"grad_norm": 1.7828255891799927,
"learning_rate": 4.991834828669213e-05,
"loss": 1.2488,
"step": 965
},
{
"epoch": 0.0776,
"grad_norm": 1.5887553691864014,
"learning_rate": 4.991750042977916e-05,
"loss": 1.1906,
"step": 970
},
{
"epoch": 0.078,
"grad_norm": 1.6549077033996582,
"learning_rate": 4.991664820085074e-05,
"loss": 1.2324,
"step": 975
},
{
"epoch": 0.0784,
"grad_norm": 1.843769907951355,
"learning_rate": 4.991579160005644e-05,
"loss": 1.2576,
"step": 980
},
{
"epoch": 0.0788,
"grad_norm": 1.4578025341033936,
"learning_rate": 4.991493062754651e-05,
"loss": 1.2846,
"step": 985
},
{
"epoch": 0.0792,
"grad_norm": 1.8506391048431396,
"learning_rate": 4.991406528347206e-05,
"loss": 1.2192,
"step": 990
},
{
"epoch": 0.0796,
"grad_norm": 1.5859005451202393,
"learning_rate": 4.991319556798488e-05,
"loss": 1.2622,
"step": 995
},
{
"epoch": 0.08,
"grad_norm": 1.7537639141082764,
"learning_rate": 4.991232148123761e-05,
"loss": 1.2302,
"step": 1000
},
{
"epoch": 0.0804,
"grad_norm": 1.6842198371887207,
"learning_rate": 4.99114430233836e-05,
"loss": 1.2561,
"step": 1005
},
{
"epoch": 0.0808,
"grad_norm": 1.5184071063995361,
"learning_rate": 4.991056019457697e-05,
"loss": 1.223,
"step": 1010
},
{
"epoch": 0.0812,
"grad_norm": 1.5156010389328003,
"learning_rate": 4.990967299497264e-05,
"loss": 1.2002,
"step": 1015
},
{
"epoch": 0.0816,
"grad_norm": 1.7296911478042603,
"learning_rate": 4.990878142472628e-05,
"loss": 1.2095,
"step": 1020
},
{
"epoch": 0.082,
"grad_norm": 1.5455466508865356,
"learning_rate": 4.990788548399432e-05,
"loss": 1.2368,
"step": 1025
},
{
"epoch": 0.0824,
"grad_norm": 1.8350162506103516,
"learning_rate": 4.990698517293395e-05,
"loss": 1.2379,
"step": 1030
},
{
"epoch": 0.0828,
"grad_norm": 1.8894885778427124,
"learning_rate": 4.9906080491703146e-05,
"loss": 1.1947,
"step": 1035
},
{
"epoch": 0.0832,
"grad_norm": 1.892716646194458,
"learning_rate": 4.990517144046064e-05,
"loss": 1.2111,
"step": 1040
},
{
"epoch": 0.0836,
"grad_norm": 1.657734751701355,
"learning_rate": 4.990425801936594e-05,
"loss": 1.2507,
"step": 1045
},
{
"epoch": 0.084,
"grad_norm": 1.800045132637024,
"learning_rate": 4.990334022857932e-05,
"loss": 1.2333,
"step": 1050
},
{
"epoch": 0.0844,
"grad_norm": 1.6811712980270386,
"learning_rate": 4.990241806826179e-05,
"loss": 1.2072,
"step": 1055
},
{
"epoch": 0.0848,
"grad_norm": 1.667485237121582,
"learning_rate": 4.9901491538575185e-05,
"loss": 1.1987,
"step": 1060
},
{
"epoch": 0.0852,
"grad_norm": 1.4977103471755981,
"learning_rate": 4.9900560639682045e-05,
"loss": 1.2305,
"step": 1065
},
{
"epoch": 0.0856,
"grad_norm": 1.898008942604065,
"learning_rate": 4.9899625371745726e-05,
"loss": 1.1646,
"step": 1070
},
{
"epoch": 0.086,
"grad_norm": 1.67731511592865,
"learning_rate": 4.989868573493032e-05,
"loss": 1.1999,
"step": 1075
},
{
"epoch": 0.0864,
"grad_norm": 1.7164236307144165,
"learning_rate": 4.9897741729400705e-05,
"loss": 1.2169,
"step": 1080
},
{
"epoch": 0.0868,
"grad_norm": 1.7378337383270264,
"learning_rate": 4.9896793355322495e-05,
"loss": 1.2711,
"step": 1085
},
{
"epoch": 0.0872,
"grad_norm": 1.4327822923660278,
"learning_rate": 4.989584061286211e-05,
"loss": 1.1496,
"step": 1090
},
{
"epoch": 0.0876,
"grad_norm": 1.9170866012573242,
"learning_rate": 4.989488350218671e-05,
"loss": 1.2445,
"step": 1095
},
{
"epoch": 0.088,
"grad_norm": 1.6763935089111328,
"learning_rate": 4.9893922023464236e-05,
"loss": 1.1969,
"step": 1100
},
{
"epoch": 0.0884,
"grad_norm": 1.7465934753417969,
"learning_rate": 4.989295617686337e-05,
"loss": 1.1862,
"step": 1105
},
{
"epoch": 0.0888,
"grad_norm": 1.5775673389434814,
"learning_rate": 4.9891985962553606e-05,
"loss": 1.2265,
"step": 1110
},
{
"epoch": 0.0892,
"grad_norm": 2.123936176300049,
"learning_rate": 4.989101138070516e-05,
"loss": 1.2077,
"step": 1115
},
{
"epoch": 0.0896,
"grad_norm": 1.6794697046279907,
"learning_rate": 4.989003243148904e-05,
"loss": 1.1972,
"step": 1120
},
{
"epoch": 0.09,
"grad_norm": 1.656134843826294,
"learning_rate": 4.9889049115077005e-05,
"loss": 1.2433,
"step": 1125
},
{
"epoch": 0.0904,
"grad_norm": 1.6735204458236694,
"learning_rate": 4.988806143164159e-05,
"loss": 1.2143,
"step": 1130
},
{
"epoch": 0.0908,
"grad_norm": 1.779127597808838,
"learning_rate": 4.9887069381356094e-05,
"loss": 1.2113,
"step": 1135
},
{
"epoch": 0.0912,
"grad_norm": 1.6463018655776978,
"learning_rate": 4.988607296439458e-05,
"loss": 1.1745,
"step": 1140
},
{
"epoch": 0.0916,
"grad_norm": 1.8336529731750488,
"learning_rate": 4.988507218093189e-05,
"loss": 1.2708,
"step": 1145
},
{
"epoch": 0.092,
"grad_norm": 1.718851089477539,
"learning_rate": 4.98840670311436e-05,
"loss": 1.2398,
"step": 1150
},
{
"epoch": 0.0924,
"grad_norm": 1.5948344469070435,
"learning_rate": 4.988305751520609e-05,
"loss": 1.2493,
"step": 1155
},
{
"epoch": 0.0928,
"grad_norm": 1.7609024047851562,
"learning_rate": 4.988204363329648e-05,
"loss": 1.2151,
"step": 1160
},
{
"epoch": 0.0932,
"grad_norm": 1.860521674156189,
"learning_rate": 4.988102538559268e-05,
"loss": 1.2635,
"step": 1165
},
{
"epoch": 0.0936,
"grad_norm": 1.8233569860458374,
"learning_rate": 4.988000277227334e-05,
"loss": 1.213,
"step": 1170
},
{
"epoch": 0.094,
"grad_norm": 1.875961184501648,
"learning_rate": 4.987897579351788e-05,
"loss": 1.249,
"step": 1175
},
{
"epoch": 0.0944,
"grad_norm": 1.5737124681472778,
"learning_rate": 4.987794444950651e-05,
"loss": 1.245,
"step": 1180
},
{
"epoch": 0.0948,
"grad_norm": 1.6306177377700806,
"learning_rate": 4.9876908740420175e-05,
"loss": 1.1596,
"step": 1185
},
{
"epoch": 0.0952,
"grad_norm": 1.912415862083435,
"learning_rate": 4.9875868666440604e-05,
"loss": 1.1932,
"step": 1190
},
{
"epoch": 0.0956,
"grad_norm": 1.7863068580627441,
"learning_rate": 4.9874824227750305e-05,
"loss": 1.166,
"step": 1195
},
{
"epoch": 0.096,
"grad_norm": 1.4861044883728027,
"learning_rate": 4.987377542453251e-05,
"loss": 1.2109,
"step": 1200
},
{
"epoch": 0.0964,
"grad_norm": 1.8580693006515503,
"learning_rate": 4.987272225697125e-05,
"loss": 1.2005,
"step": 1205
},
{
"epoch": 0.0968,
"grad_norm": 1.5345180034637451,
"learning_rate": 4.9871664725251314e-05,
"loss": 1.2178,
"step": 1210
},
{
"epoch": 0.0972,
"grad_norm": 1.6024930477142334,
"learning_rate": 4.987060282955826e-05,
"loss": 1.1837,
"step": 1215
},
{
"epoch": 0.0976,
"grad_norm": 1.8284636735916138,
"learning_rate": 4.986953657007841e-05,
"loss": 1.2051,
"step": 1220
},
{
"epoch": 0.098,
"grad_norm": 1.7044328451156616,
"learning_rate": 4.986846594699883e-05,
"loss": 1.1946,
"step": 1225
},
{
"epoch": 0.0984,
"grad_norm": 1.8590692281723022,
"learning_rate": 4.98673909605074e-05,
"loss": 1.215,
"step": 1230
},
{
"epoch": 0.0988,
"grad_norm": 1.660416603088379,
"learning_rate": 4.986631161079272e-05,
"loss": 1.2124,
"step": 1235
},
{
"epoch": 0.0992,
"grad_norm": 1.5783928632736206,
"learning_rate": 4.986522789804417e-05,
"loss": 1.2048,
"step": 1240
},
{
"epoch": 0.0996,
"grad_norm": 1.6372880935668945,
"learning_rate": 4.9864139822451905e-05,
"loss": 1.1847,
"step": 1245
},
{
"epoch": 0.1,
"grad_norm": 1.6335861682891846,
"learning_rate": 4.9863047384206835e-05,
"loss": 1.201,
"step": 1250
},
{
"epoch": 0.1004,
"grad_norm": 1.7448561191558838,
"learning_rate": 4.9861950583500636e-05,
"loss": 1.2208,
"step": 1255
},
{
"epoch": 0.1008,
"grad_norm": 1.6538008451461792,
"learning_rate": 4.9860849420525766e-05,
"loss": 1.3027,
"step": 1260
},
{
"epoch": 0.1012,
"grad_norm": 1.5223474502563477,
"learning_rate": 4.9859743895475416e-05,
"loss": 1.21,
"step": 1265
},
{
"epoch": 0.1016,
"grad_norm": 2.5573835372924805,
"learning_rate": 4.985863400854358e-05,
"loss": 1.2278,
"step": 1270
},
{
"epoch": 0.102,
"grad_norm": 1.7527281045913696,
"learning_rate": 4.9857519759924974e-05,
"loss": 1.2594,
"step": 1275
},
{
"epoch": 0.1024,
"grad_norm": 1.5259441137313843,
"learning_rate": 4.9856401149815126e-05,
"loss": 1.1852,
"step": 1280
},
{
"epoch": 0.1028,
"grad_norm": 1.6664983034133911,
"learning_rate": 4.985527817841029e-05,
"loss": 1.2257,
"step": 1285
},
{
"epoch": 0.1032,
"grad_norm": 1.7659004926681519,
"learning_rate": 4.985415084590752e-05,
"loss": 1.1816,
"step": 1290
},
{
"epoch": 0.1036,
"grad_norm": 1.7176613807678223,
"learning_rate": 4.9853019152504607e-05,
"loss": 1.2328,
"step": 1295
},
{
"epoch": 0.104,
"grad_norm": 1.7339608669281006,
"learning_rate": 4.985188309840012e-05,
"loss": 1.3235,
"step": 1300
},
{
"epoch": 0.1044,
"grad_norm": 1.5208759307861328,
"learning_rate": 4.985074268379338e-05,
"loss": 1.2024,
"step": 1305
},
{
"epoch": 0.1048,
"grad_norm": 1.7345401048660278,
"learning_rate": 4.98495979088845e-05,
"loss": 1.2577,
"step": 1310
},
{
"epoch": 0.1052,
"grad_norm": 1.6194533109664917,
"learning_rate": 4.984844877387433e-05,
"loss": 1.1717,
"step": 1315
},
{
"epoch": 0.1056,
"grad_norm": 1.9546769857406616,
"learning_rate": 4.9847295278964514e-05,
"loss": 1.2002,
"step": 1320
},
{
"epoch": 0.106,
"grad_norm": 1.7305984497070312,
"learning_rate": 4.984613742435742e-05,
"loss": 1.2273,
"step": 1325
},
{
"epoch": 0.1064,
"grad_norm": 1.9428414106369019,
"learning_rate": 4.9844975210256217e-05,
"loss": 1.1932,
"step": 1330
},
{
"epoch": 0.1068,
"grad_norm": 1.6974749565124512,
"learning_rate": 4.984380863686482e-05,
"loss": 1.207,
"step": 1335
},
{
"epoch": 0.1072,
"grad_norm": 1.6390514373779297,
"learning_rate": 4.984263770438793e-05,
"loss": 1.1626,
"step": 1340
},
{
"epoch": 0.1076,
"grad_norm": 1.3905149698257446,
"learning_rate": 4.9841462413030995e-05,
"loss": 1.2294,
"step": 1345
},
{
"epoch": 0.108,
"grad_norm": 1.6545757055282593,
"learning_rate": 4.984028276300021e-05,
"loss": 1.2109,
"step": 1350
},
{
"epoch": 0.1084,
"grad_norm": 1.6141990423202515,
"learning_rate": 4.983909875450258e-05,
"loss": 1.2104,
"step": 1355
},
{
"epoch": 0.1088,
"grad_norm": 1.4683756828308105,
"learning_rate": 4.9837910387745845e-05,
"loss": 1.1762,
"step": 1360
},
{
"epoch": 0.1092,
"grad_norm": 1.566309928894043,
"learning_rate": 4.983671766293851e-05,
"loss": 1.2136,
"step": 1365
},
{
"epoch": 0.1096,
"grad_norm": 1.6710453033447266,
"learning_rate": 4.9835520580289854e-05,
"loss": 1.2212,
"step": 1370
},
{
"epoch": 0.11,
"grad_norm": 1.7142356634140015,
"learning_rate": 4.983431914000991e-05,
"loss": 1.1812,
"step": 1375
},
{
"epoch": 0.1104,
"grad_norm": 1.8193789720535278,
"learning_rate": 4.98331133423095e-05,
"loss": 1.2175,
"step": 1380
},
{
"epoch": 0.1108,
"grad_norm": 1.675352931022644,
"learning_rate": 4.9831903187400166e-05,
"loss": 1.2099,
"step": 1385
},
{
"epoch": 0.1112,
"grad_norm": 1.8119478225708008,
"learning_rate": 4.9830688675494265e-05,
"loss": 1.2313,
"step": 1390
},
{
"epoch": 0.1116,
"grad_norm": 1.5995213985443115,
"learning_rate": 4.982946980680488e-05,
"loss": 1.2507,
"step": 1395
},
{
"epoch": 0.112,
"grad_norm": 1.719618797302246,
"learning_rate": 4.982824658154589e-05,
"loss": 1.1767,
"step": 1400
},
{
"epoch": 0.1124,
"grad_norm": 1.6213531494140625,
"learning_rate": 4.982701899993189e-05,
"loss": 1.2055,
"step": 1405
},
{
"epoch": 0.1128,
"grad_norm": 1.788510799407959,
"learning_rate": 4.9825787062178315e-05,
"loss": 1.1826,
"step": 1410
},
{
"epoch": 0.1132,
"grad_norm": 1.8999608755111694,
"learning_rate": 4.982455076850129e-05,
"loss": 1.2034,
"step": 1415
},
{
"epoch": 0.1136,
"grad_norm": 1.538346767425537,
"learning_rate": 4.982331011911774e-05,
"loss": 1.1955,
"step": 1420
},
{
"epoch": 0.114,
"grad_norm": 1.760783314704895,
"learning_rate": 4.982206511424534e-05,
"loss": 1.2228,
"step": 1425
},
{
"epoch": 0.1144,
"grad_norm": 1.7535641193389893,
"learning_rate": 4.982081575410256e-05,
"loss": 1.1827,
"step": 1430
},
{
"epoch": 0.1148,
"grad_norm": 1.7111599445343018,
"learning_rate": 4.98195620389086e-05,
"loss": 1.2226,
"step": 1435
},
{
"epoch": 0.1152,
"grad_norm": 1.7336255311965942,
"learning_rate": 4.981830396888344e-05,
"loss": 1.2594,
"step": 1440
},
{
"epoch": 0.1156,
"grad_norm": 1.9409478902816772,
"learning_rate": 4.981704154424781e-05,
"loss": 1.2773,
"step": 1445
},
{
"epoch": 0.116,
"grad_norm": 1.6650772094726562,
"learning_rate": 4.9815774765223226e-05,
"loss": 1.1894,
"step": 1450
},
{
"epoch": 0.1164,
"grad_norm": 1.776209831237793,
"learning_rate": 4.9814503632031954e-05,
"loss": 1.2409,
"step": 1455
},
{
"epoch": 0.1168,
"grad_norm": 1.806114912033081,
"learning_rate": 4.981322814489703e-05,
"loss": 1.2251,
"step": 1460
},
{
"epoch": 0.1172,
"grad_norm": 1.8370269536972046,
"learning_rate": 4.9811948304042234e-05,
"loss": 1.1519,
"step": 1465
},
{
"epoch": 0.1176,
"grad_norm": 1.9353857040405273,
"learning_rate": 4.981066410969215e-05,
"loss": 1.2432,
"step": 1470
},
{
"epoch": 0.118,
"grad_norm": 1.5059784650802612,
"learning_rate": 4.980937556207208e-05,
"loss": 1.1641,
"step": 1475
},
{
"epoch": 0.1184,
"grad_norm": 1.5918395519256592,
"learning_rate": 4.980808266140813e-05,
"loss": 1.2464,
"step": 1480
},
{
"epoch": 0.1188,
"grad_norm": 1.6992462873458862,
"learning_rate": 4.980678540792715e-05,
"loss": 1.2007,
"step": 1485
},
{
"epoch": 0.1192,
"grad_norm": 1.7134628295898438,
"learning_rate": 4.980548380185674e-05,
"loss": 1.1925,
"step": 1490
},
{
"epoch": 0.1196,
"grad_norm": 1.8130481243133545,
"learning_rate": 4.9804177843425295e-05,
"loss": 1.1775,
"step": 1495
},
{
"epoch": 0.12,
"grad_norm": 1.811142086982727,
"learning_rate": 4.980286753286195e-05,
"loss": 1.2813,
"step": 1500
},
{
"epoch": 0.1204,
"grad_norm": 1.7274268865585327,
"learning_rate": 4.980155287039662e-05,
"loss": 1.2213,
"step": 1505
},
{
"epoch": 0.1208,
"grad_norm": 1.8031953573226929,
"learning_rate": 4.980023385625996e-05,
"loss": 1.1459,
"step": 1510
},
{
"epoch": 0.1212,
"grad_norm": 1.6211178302764893,
"learning_rate": 4.979891049068342e-05,
"loss": 1.1575,
"step": 1515
},
{
"epoch": 0.1216,
"grad_norm": 1.4958020448684692,
"learning_rate": 4.979758277389919e-05,
"loss": 1.1516,
"step": 1520
},
{
"epoch": 0.122,
"grad_norm": 1.8061717748641968,
"learning_rate": 4.9796250706140224e-05,
"loss": 1.2135,
"step": 1525
},
{
"epoch": 0.1224,
"grad_norm": 1.4984862804412842,
"learning_rate": 4.979491428764026e-05,
"loss": 1.2702,
"step": 1530
},
{
"epoch": 0.1228,
"grad_norm": 1.9237560033798218,
"learning_rate": 4.979357351863377e-05,
"loss": 1.2653,
"step": 1535
},
{
"epoch": 0.1232,
"grad_norm": 1.7026283740997314,
"learning_rate": 4.979222839935602e-05,
"loss": 1.2467,
"step": 1540
},
{
"epoch": 0.1236,
"grad_norm": 1.523640751838684,
"learning_rate": 4.979087893004302e-05,
"loss": 1.2177,
"step": 1545
},
{
"epoch": 0.124,
"grad_norm": 1.5386415719985962,
"learning_rate": 4.9789525110931545e-05,
"loss": 1.1528,
"step": 1550
},
{
"epoch": 0.1244,
"grad_norm": 1.6132177114486694,
"learning_rate": 4.9788166942259135e-05,
"loss": 1.1925,
"step": 1555
},
{
"epoch": 0.1248,
"grad_norm": 1.8370720148086548,
"learning_rate": 4.9786804424264085e-05,
"loss": 1.1702,
"step": 1560
},
{
"epoch": 0.1252,
"grad_norm": 1.78260338306427,
"learning_rate": 4.978543755718549e-05,
"loss": 1.2394,
"step": 1565
},
{
"epoch": 0.1256,
"grad_norm": 1.3837835788726807,
"learning_rate": 4.978406634126315e-05,
"loss": 1.2256,
"step": 1570
},
{
"epoch": 0.126,
"grad_norm": 1.776090383529663,
"learning_rate": 4.978269077673767e-05,
"loss": 1.2859,
"step": 1575
},
{
"epoch": 0.1264,
"grad_norm": 1.423762321472168,
"learning_rate": 4.9781310863850405e-05,
"loss": 1.2486,
"step": 1580
},
{
"epoch": 0.1268,
"grad_norm": 1.5128095149993896,
"learning_rate": 4.977992660284347e-05,
"loss": 1.188,
"step": 1585
},
{
"epoch": 0.1272,
"grad_norm": 1.7844202518463135,
"learning_rate": 4.977853799395976e-05,
"loss": 1.1579,
"step": 1590
},
{
"epoch": 0.1276,
"grad_norm": 1.7243590354919434,
"learning_rate": 4.9777145037442906e-05,
"loss": 1.2631,
"step": 1595
},
{
"epoch": 0.128,
"grad_norm": 1.7235385179519653,
"learning_rate": 4.977574773353732e-05,
"loss": 1.144,
"step": 1600
},
{
"epoch": 0.1284,
"grad_norm": 1.522979974746704,
"learning_rate": 4.9774346082488176e-05,
"loss": 1.2471,
"step": 1605
},
{
"epoch": 0.1288,
"grad_norm": 1.3776745796203613,
"learning_rate": 4.9772940084541405e-05,
"loss": 1.222,
"step": 1610
},
{
"epoch": 0.1292,
"grad_norm": 1.7246540784835815,
"learning_rate": 4.97715297399437e-05,
"loss": 1.1857,
"step": 1615
},
{
"epoch": 0.1296,
"grad_norm": 1.747706413269043,
"learning_rate": 4.977011504894252e-05,
"loss": 1.1973,
"step": 1620
},
{
"epoch": 0.13,
"grad_norm": 1.6644729375839233,
"learning_rate": 4.976869601178609e-05,
"loss": 1.1559,
"step": 1625
},
{
"epoch": 0.1304,
"grad_norm": 1.5264381170272827,
"learning_rate": 4.9767272628723396e-05,
"loss": 1.256,
"step": 1630
},
{
"epoch": 0.1308,
"grad_norm": 1.5906404256820679,
"learning_rate": 4.9765844900004176e-05,
"loss": 1.1486,
"step": 1635
},
{
"epoch": 0.1312,
"grad_norm": 1.4463717937469482,
"learning_rate": 4.9764412825878943e-05,
"loss": 1.2871,
"step": 1640
},
{
"epoch": 0.1316,
"grad_norm": 1.727552056312561,
"learning_rate": 4.976297640659897e-05,
"loss": 1.178,
"step": 1645
},
{
"epoch": 0.132,
"grad_norm": 1.5339679718017578,
"learning_rate": 4.976153564241628e-05,
"loss": 1.2556,
"step": 1650
},
{
"epoch": 0.1324,
"grad_norm": 1.5562660694122314,
"learning_rate": 4.9760090533583686e-05,
"loss": 1.1758,
"step": 1655
},
{
"epoch": 0.1328,
"grad_norm": 1.6301156282424927,
"learning_rate": 4.975864108035474e-05,
"loss": 1.2246,
"step": 1660
},
{
"epoch": 0.1332,
"grad_norm": 1.9611403942108154,
"learning_rate": 4.975718728298375e-05,
"loss": 1.2752,
"step": 1665
},
{
"epoch": 0.1336,
"grad_norm": 1.4898903369903564,
"learning_rate": 4.975572914172582e-05,
"loss": 1.1814,
"step": 1670
},
{
"epoch": 0.134,
"grad_norm": 1.7393544912338257,
"learning_rate": 4.975426665683678e-05,
"loss": 1.207,
"step": 1675
},
{
"epoch": 0.1344,
"grad_norm": 1.5046724081039429,
"learning_rate": 4.975279982857324e-05,
"loss": 1.2219,
"step": 1680
},
{
"epoch": 0.1348,
"grad_norm": 1.8701568841934204,
"learning_rate": 4.9751328657192565e-05,
"loss": 1.2013,
"step": 1685
},
{
"epoch": 0.1352,
"grad_norm": 1.7540148496627808,
"learning_rate": 4.97498531429529e-05,
"loss": 1.1784,
"step": 1690
},
{
"epoch": 0.1356,
"grad_norm": 2.0447638034820557,
"learning_rate": 4.974837328611312e-05,
"loss": 1.2023,
"step": 1695
},
{
"epoch": 0.136,
"grad_norm": 1.6013314723968506,
"learning_rate": 4.9746889086932895e-05,
"loss": 1.2658,
"step": 1700
},
{
"epoch": 0.1364,
"grad_norm": 1.83162522315979,
"learning_rate": 4.974540054567264e-05,
"loss": 1.2172,
"step": 1705
},
{
"epoch": 0.1368,
"grad_norm": 1.6083216667175293,
"learning_rate": 4.9743907662593524e-05,
"loss": 1.2236,
"step": 1710
},
{
"epoch": 0.1372,
"grad_norm": 1.5732282400131226,
"learning_rate": 4.97424104379575e-05,
"loss": 1.1822,
"step": 1715
},
{
"epoch": 0.1376,
"grad_norm": 1.6815297603607178,
"learning_rate": 4.974090887202726e-05,
"loss": 1.2386,
"step": 1720
},
{
"epoch": 0.138,
"grad_norm": 1.5495154857635498,
"learning_rate": 4.9739402965066276e-05,
"loss": 1.1664,
"step": 1725
},
{
"epoch": 0.1384,
"grad_norm": 1.4523470401763916,
"learning_rate": 4.9737892717338774e-05,
"loss": 1.1526,
"step": 1730
},
{
"epoch": 0.1388,
"grad_norm": 1.6146246194839478,
"learning_rate": 4.973637812910973e-05,
"loss": 1.1802,
"step": 1735
},
{
"epoch": 0.1392,
"grad_norm": 1.5565937757492065,
"learning_rate": 4.9734859200644905e-05,
"loss": 1.2272,
"step": 1740
},
{
"epoch": 0.1396,
"grad_norm": 1.7188202142715454,
"learning_rate": 4.9733335932210814e-05,
"loss": 1.3008,
"step": 1745
},
{
"epoch": 0.14,
"grad_norm": 1.7676734924316406,
"learning_rate": 4.9731808324074717e-05,
"loss": 1.2122,
"step": 1750
},
{
"epoch": 0.1404,
"grad_norm": 1.3866500854492188,
"learning_rate": 4.973027637650464e-05,
"loss": 1.2475,
"step": 1755
},
{
"epoch": 0.1408,
"grad_norm": 1.736222267150879,
"learning_rate": 4.97287400897694e-05,
"loss": 1.1491,
"step": 1760
},
{
"epoch": 0.1412,
"grad_norm": 1.7289378643035889,
"learning_rate": 4.972719946413854e-05,
"loss": 1.2638,
"step": 1765
},
{
"epoch": 0.1416,
"grad_norm": 1.5107754468917847,
"learning_rate": 4.972565449988239e-05,
"loss": 1.2414,
"step": 1770
},
{
"epoch": 0.142,
"grad_norm": 1.6020653247833252,
"learning_rate": 4.972410519727201e-05,
"loss": 1.154,
"step": 1775
},
{
"epoch": 0.1424,
"grad_norm": 1.57801353931427,
"learning_rate": 4.972255155657925e-05,
"loss": 1.2212,
"step": 1780
},
{
"epoch": 0.1428,
"grad_norm": 1.7181590795516968,
"learning_rate": 4.972099357807671e-05,
"loss": 1.2248,
"step": 1785
},
{
"epoch": 0.1432,
"grad_norm": 1.356872320175171,
"learning_rate": 4.9719431262037755e-05,
"loss": 1.0836,
"step": 1790
},
{
"epoch": 0.1436,
"grad_norm": 1.4767268896102905,
"learning_rate": 4.9717864608736506e-05,
"loss": 1.2183,
"step": 1795
},
{
"epoch": 0.144,
"grad_norm": 1.496610164642334,
"learning_rate": 4.971629361844785e-05,
"loss": 1.2099,
"step": 1800
},
{
"epoch": 0.1444,
"grad_norm": 1.6755746603012085,
"learning_rate": 4.971471829144743e-05,
"loss": 1.1769,
"step": 1805
},
{
"epoch": 0.1448,
"grad_norm": 1.4553226232528687,
"learning_rate": 4.9713138628011654e-05,
"loss": 1.2173,
"step": 1810
},
{
"epoch": 0.1452,
"grad_norm": 1.7685695886611938,
"learning_rate": 4.971155462841769e-05,
"loss": 1.1878,
"step": 1815
},
{
"epoch": 0.1456,
"grad_norm": 1.6360673904418945,
"learning_rate": 4.9709966292943455e-05,
"loss": 1.1759,
"step": 1820
},
{
"epoch": 0.146,
"grad_norm": 1.7544503211975098,
"learning_rate": 4.9708373621867656e-05,
"loss": 1.1912,
"step": 1825
},
{
"epoch": 0.1464,
"grad_norm": 1.6585785150527954,
"learning_rate": 4.9706776615469716e-05,
"loss": 1.2538,
"step": 1830
},
{
"epoch": 0.1468,
"grad_norm": 1.6937992572784424,
"learning_rate": 4.970517527402988e-05,
"loss": 1.1707,
"step": 1835
},
{
"epoch": 0.1472,
"grad_norm": 1.5035377740859985,
"learning_rate": 4.970356959782909e-05,
"loss": 1.0796,
"step": 1840
},
{
"epoch": 0.1476,
"grad_norm": 1.5920889377593994,
"learning_rate": 4.970195958714909e-05,
"loss": 1.2042,
"step": 1845
},
{
"epoch": 0.148,
"grad_norm": 1.7095478773117065,
"learning_rate": 4.970034524227238e-05,
"loss": 1.2424,
"step": 1850
},
{
"epoch": 0.1484,
"grad_norm": 1.565622091293335,
"learning_rate": 4.96987265634822e-05,
"loss": 1.2442,
"step": 1855
},
{
"epoch": 0.1488,
"grad_norm": 1.5227761268615723,
"learning_rate": 4.9697103551062556e-05,
"loss": 1.2286,
"step": 1860
},
{
"epoch": 0.1492,
"grad_norm": 1.8364589214324951,
"learning_rate": 4.9695476205298235e-05,
"loss": 1.2409,
"step": 1865
},
{
"epoch": 0.1496,
"grad_norm": 1.476838231086731,
"learning_rate": 4.969384452647477e-05,
"loss": 1.2491,
"step": 1870
},
{
"epoch": 0.15,
"grad_norm": 1.617569088935852,
"learning_rate": 4.9692208514878444e-05,
"loss": 1.1515,
"step": 1875
},
{
"epoch": 0.1504,
"grad_norm": 1.623694658279419,
"learning_rate": 4.969056817079633e-05,
"loss": 1.2829,
"step": 1880
},
{
"epoch": 0.1508,
"grad_norm": 1.8498085737228394,
"learning_rate": 4.968892349451621e-05,
"loss": 1.2083,
"step": 1885
},
{
"epoch": 0.1512,
"grad_norm": 1.544599175453186,
"learning_rate": 4.968727448632669e-05,
"loss": 1.1898,
"step": 1890
},
{
"epoch": 0.1516,
"grad_norm": 1.7508872747421265,
"learning_rate": 4.968562114651709e-05,
"loss": 1.2359,
"step": 1895
},
{
"epoch": 0.152,
"grad_norm": 1.6149662733078003,
"learning_rate": 4.968396347537751e-05,
"loss": 1.2036,
"step": 1900
},
{
"epoch": 0.1524,
"grad_norm": 1.5705629587173462,
"learning_rate": 4.96823014731988e-05,
"loss": 1.2044,
"step": 1905
},
{
"epoch": 0.1528,
"grad_norm": 1.7372088432312012,
"learning_rate": 4.9680635140272575e-05,
"loss": 1.3014,
"step": 1910
},
{
"epoch": 0.1532,
"grad_norm": 1.4629701375961304,
"learning_rate": 4.967896447689121e-05,
"loss": 1.2249,
"step": 1915
},
{
"epoch": 0.1536,
"grad_norm": 1.6758002042770386,
"learning_rate": 4.967728948334784e-05,
"loss": 1.2519,
"step": 1920
},
{
"epoch": 0.154,
"grad_norm": 1.6669517755508423,
"learning_rate": 4.967561015993635e-05,
"loss": 1.2515,
"step": 1925
},
{
"epoch": 0.1544,
"grad_norm": 1.556141972541809,
"learning_rate": 4.9673926506951404e-05,
"loss": 1.1648,
"step": 1930
},
{
"epoch": 0.1548,
"grad_norm": 1.5720252990722656,
"learning_rate": 4.967223852468842e-05,
"loss": 1.1668,
"step": 1935
},
{
"epoch": 0.1552,
"grad_norm": 1.5914875268936157,
"learning_rate": 4.967054621344356e-05,
"loss": 1.2718,
"step": 1940
},
{
"epoch": 0.1556,
"grad_norm": 1.8460205793380737,
"learning_rate": 4.966884957351375e-05,
"loss": 1.215,
"step": 1945
},
{
"epoch": 0.156,
"grad_norm": 1.4069819450378418,
"learning_rate": 4.96671486051967e-05,
"loss": 1.2576,
"step": 1950
},
{
"epoch": 0.1564,
"grad_norm": 1.8129643201828003,
"learning_rate": 4.966544330879085e-05,
"loss": 1.1778,
"step": 1955
},
{
"epoch": 0.1568,
"grad_norm": 1.6528633832931519,
"learning_rate": 4.966373368459541e-05,
"loss": 1.1827,
"step": 1960
},
{
"epoch": 0.1572,
"grad_norm": 1.5496547222137451,
"learning_rate": 4.966201973291036e-05,
"loss": 1.16,
"step": 1965
},
{
"epoch": 0.1576,
"grad_norm": 1.7578569650650024,
"learning_rate": 4.966030145403642e-05,
"loss": 1.2164,
"step": 1970
},
{
"epoch": 0.158,
"grad_norm": 1.6833785772323608,
"learning_rate": 4.9658578848275076e-05,
"loss": 1.2729,
"step": 1975
},
{
"epoch": 0.1584,
"grad_norm": 1.4320454597473145,
"learning_rate": 4.965685191592859e-05,
"loss": 1.2141,
"step": 1980
},
{
"epoch": 0.1588,
"grad_norm": 1.5332552194595337,
"learning_rate": 4.9655120657299945e-05,
"loss": 1.2568,
"step": 1985
},
{
"epoch": 0.1592,
"grad_norm": 1.4154534339904785,
"learning_rate": 4.965338507269294e-05,
"loss": 1.2167,
"step": 1990
},
{
"epoch": 0.1596,
"grad_norm": 1.6980712413787842,
"learning_rate": 4.965164516241206e-05,
"loss": 1.178,
"step": 1995
},
{
"epoch": 0.16,
"grad_norm": 1.452364444732666,
"learning_rate": 4.964990092676263e-05,
"loss": 1.2442,
"step": 2000
},
{
"epoch": 0.1604,
"grad_norm": 1.5581384897232056,
"learning_rate": 4.964815236605066e-05,
"loss": 1.1763,
"step": 2005
},
{
"epoch": 0.1608,
"grad_norm": 1.5522022247314453,
"learning_rate": 4.964639948058297e-05,
"loss": 1.2057,
"step": 2010
},
{
"epoch": 0.1612,
"grad_norm": 1.8759955167770386,
"learning_rate": 4.964464227066712e-05,
"loss": 1.2072,
"step": 2015
},
{
"epoch": 0.1616,
"grad_norm": 1.6909701824188232,
"learning_rate": 4.964288073661142e-05,
"loss": 1.1904,
"step": 2020
},
{
"epoch": 0.162,
"grad_norm": 1.7072064876556396,
"learning_rate": 4.9641114878724956e-05,
"loss": 1.255,
"step": 2025
},
{
"epoch": 0.1624,
"grad_norm": 1.663236379623413,
"learning_rate": 4.963934469731756e-05,
"loss": 1.2245,
"step": 2030
},
{
"epoch": 0.1628,
"grad_norm": 1.6193203926086426,
"learning_rate": 4.963757019269983e-05,
"loss": 1.2019,
"step": 2035
},
{
"epoch": 0.1632,
"grad_norm": 3.217263698577881,
"learning_rate": 4.963579136518312e-05,
"loss": 1.1895,
"step": 2040
},
{
"epoch": 0.1636,
"grad_norm": 1.704827904701233,
"learning_rate": 4.963400821507954e-05,
"loss": 1.1823,
"step": 2045
},
{
"epoch": 0.164,
"grad_norm": 1.611045002937317,
"learning_rate": 4.9632220742701965e-05,
"loss": 1.2057,
"step": 2050
},
{
"epoch": 0.1644,
"grad_norm": 1.6475915908813477,
"learning_rate": 4.963042894836403e-05,
"loss": 1.2055,
"step": 2055
},
{
"epoch": 0.1648,
"grad_norm": 1.3033784627914429,
"learning_rate": 4.96286328323801e-05,
"loss": 1.3009,
"step": 2060
},
{
"epoch": 0.1652,
"grad_norm": 1.928715705871582,
"learning_rate": 4.962683239506534e-05,
"loss": 1.1779,
"step": 2065
},
{
"epoch": 0.1656,
"grad_norm": 1.7251336574554443,
"learning_rate": 4.962502763673565e-05,
"loss": 1.2536,
"step": 2070
},
{
"epoch": 0.166,
"grad_norm": 1.4346777200698853,
"learning_rate": 4.9623218557707694e-05,
"loss": 1.2126,
"step": 2075
},
{
"epoch": 0.1664,
"grad_norm": 1.4237911701202393,
"learning_rate": 4.96214051582989e-05,
"loss": 1.1766,
"step": 2080
},
{
"epoch": 0.1668,
"grad_norm": 2.051907539367676,
"learning_rate": 4.961958743882742e-05,
"loss": 1.2302,
"step": 2085
},
{
"epoch": 0.1672,
"grad_norm": 1.6428817510604858,
"learning_rate": 4.961776539961222e-05,
"loss": 1.164,
"step": 2090
},
{
"epoch": 0.1676,
"grad_norm": 1.492691993713379,
"learning_rate": 4.961593904097297e-05,
"loss": 1.2141,
"step": 2095
},
{
"epoch": 0.168,
"grad_norm": 1.70375394821167,
"learning_rate": 4.9614108363230135e-05,
"loss": 1.2091,
"step": 2100
},
{
"epoch": 0.1684,
"grad_norm": 1.7503658533096313,
"learning_rate": 4.961227336670493e-05,
"loss": 1.1867,
"step": 2105
},
{
"epoch": 0.1688,
"grad_norm": 1.8217905759811401,
"learning_rate": 4.961043405171931e-05,
"loss": 1.1937,
"step": 2110
},
{
"epoch": 0.1692,
"grad_norm": 1.8402591943740845,
"learning_rate": 4.9608590418596016e-05,
"loss": 1.1849,
"step": 2115
},
{
"epoch": 0.1696,
"grad_norm": 1.6328750848770142,
"learning_rate": 4.960674246765851e-05,
"loss": 1.2358,
"step": 2120
},
{
"epoch": 0.17,
"grad_norm": 1.8972320556640625,
"learning_rate": 4.960489019923105e-05,
"loss": 1.1965,
"step": 2125
},
{
"epoch": 0.1704,
"grad_norm": 1.7403242588043213,
"learning_rate": 4.9603033613638626e-05,
"loss": 1.2007,
"step": 2130
},
{
"epoch": 0.1708,
"grad_norm": 1.811218023300171,
"learning_rate": 4.9601172711207005e-05,
"loss": 1.1738,
"step": 2135
},
{
"epoch": 0.1712,
"grad_norm": 1.8417551517486572,
"learning_rate": 4.959930749226269e-05,
"loss": 1.2474,
"step": 2140
},
{
"epoch": 0.1716,
"grad_norm": 1.7173854112625122,
"learning_rate": 4.9597437957132955e-05,
"loss": 1.1534,
"step": 2145
},
{
"epoch": 0.172,
"grad_norm": 1.5625674724578857,
"learning_rate": 4.959556410614582e-05,
"loss": 1.1943,
"step": 2150
},
{
"epoch": 0.1724,
"grad_norm": 1.8985899686813354,
"learning_rate": 4.959368593963007e-05,
"loss": 1.1962,
"step": 2155
},
{
"epoch": 0.1728,
"grad_norm": 1.7198944091796875,
"learning_rate": 4.959180345791528e-05,
"loss": 1.2023,
"step": 2160
},
{
"epoch": 0.1732,
"grad_norm": 1.4986478090286255,
"learning_rate": 4.95899166613317e-05,
"loss": 1.2129,
"step": 2165
},
{
"epoch": 0.1736,
"grad_norm": 1.4849358797073364,
"learning_rate": 4.958802555021042e-05,
"loss": 1.1964,
"step": 2170
},
{
"epoch": 0.174,
"grad_norm": 1.5313656330108643,
"learning_rate": 4.958613012488324e-05,
"loss": 1.227,
"step": 2175
},
{
"epoch": 0.1744,
"grad_norm": 1.6085180044174194,
"learning_rate": 4.958423038568274e-05,
"loss": 1.1662,
"step": 2180
},
{
"epoch": 0.1748,
"grad_norm": 1.8227916955947876,
"learning_rate": 4.9582326332942244e-05,
"loss": 1.2102,
"step": 2185
},
{
"epoch": 0.1752,
"grad_norm": 1.7549911737442017,
"learning_rate": 4.958041796699583e-05,
"loss": 1.1915,
"step": 2190
},
{
"epoch": 0.1756,
"grad_norm": 2.0257728099823,
"learning_rate": 4.957850528817834e-05,
"loss": 1.1737,
"step": 2195
},
{
"epoch": 0.176,
"grad_norm": 1.7481350898742676,
"learning_rate": 4.9576588296825386e-05,
"loss": 1.2165,
"step": 2200
},
{
"epoch": 0.1764,
"grad_norm": 1.3570259809494019,
"learning_rate": 4.957466699327331e-05,
"loss": 1.2375,
"step": 2205
},
{
"epoch": 0.1768,
"grad_norm": 1.782902717590332,
"learning_rate": 4.957274137785922e-05,
"loss": 1.2078,
"step": 2210
},
{
"epoch": 0.1772,
"grad_norm": 1.4159862995147705,
"learning_rate": 4.9570811450921e-05,
"loss": 1.1476,
"step": 2215
},
{
"epoch": 0.1776,
"grad_norm": 1.5632498264312744,
"learning_rate": 4.956887721279726e-05,
"loss": 1.2646,
"step": 2220
},
{
"epoch": 0.178,
"grad_norm": 1.5948270559310913,
"learning_rate": 4.9566938663827377e-05,
"loss": 1.1947,
"step": 2225
},
{
"epoch": 0.1784,
"grad_norm": 1.6205294132232666,
"learning_rate": 4.95649958043515e-05,
"loss": 1.1819,
"step": 2230
},
{
"epoch": 0.1788,
"grad_norm": 1.8377573490142822,
"learning_rate": 4.9563048634710516e-05,
"loss": 1.202,
"step": 2235
},
{
"epoch": 0.1792,
"grad_norm": 1.8803945779800415,
"learning_rate": 4.956109715524608e-05,
"loss": 1.2488,
"step": 2240
},
{
"epoch": 0.1796,
"grad_norm": 1.6330218315124512,
"learning_rate": 4.9559141366300594e-05,
"loss": 1.168,
"step": 2245
},
{
"epoch": 0.18,
"grad_norm": 1.9217214584350586,
"learning_rate": 4.9557181268217227e-05,
"loss": 1.2207,
"step": 2250
},
{
"epoch": 0.1804,
"grad_norm": 1.7866103649139404,
"learning_rate": 4.9555216861339876e-05,
"loss": 1.2106,
"step": 2255
},
{
"epoch": 0.1808,
"grad_norm": 1.7389551401138306,
"learning_rate": 4.955324814601324e-05,
"loss": 1.176,
"step": 2260
},
{
"epoch": 0.1812,
"grad_norm": 1.8251265287399292,
"learning_rate": 4.955127512258273e-05,
"loss": 1.1974,
"step": 2265
},
{
"epoch": 0.1816,
"grad_norm": 1.402008056640625,
"learning_rate": 4.954929779139455e-05,
"loss": 1.2318,
"step": 2270
},
{
"epoch": 0.182,
"grad_norm": 1.4329712390899658,
"learning_rate": 4.954731615279563e-05,
"loss": 1.1501,
"step": 2275
},
{
"epoch": 0.1824,
"grad_norm": 1.6272238492965698,
"learning_rate": 4.9545330207133664e-05,
"loss": 1.1612,
"step": 2280
},
{
"epoch": 0.1828,
"grad_norm": 1.5852447748184204,
"learning_rate": 4.954333995475712e-05,
"loss": 1.2153,
"step": 2285
},
{
"epoch": 0.1832,
"grad_norm": 1.7610340118408203,
"learning_rate": 4.9541345396015193e-05,
"loss": 1.2303,
"step": 2290
},
{
"epoch": 0.1836,
"grad_norm": 1.5253504514694214,
"learning_rate": 4.953934653125786e-05,
"loss": 1.1946,
"step": 2295
},
{
"epoch": 0.184,
"grad_norm": 1.3086066246032715,
"learning_rate": 4.953734336083583e-05,
"loss": 1.1768,
"step": 2300
},
{
"epoch": 0.1844,
"grad_norm": 1.6699544191360474,
"learning_rate": 4.9535335885100575e-05,
"loss": 1.1911,
"step": 2305
},
{
"epoch": 0.1848,
"grad_norm": 1.6090809106826782,
"learning_rate": 4.953332410440435e-05,
"loss": 1.1878,
"step": 2310
},
{
"epoch": 0.1852,
"grad_norm": 1.6638312339782715,
"learning_rate": 4.953130801910011e-05,
"loss": 1.2607,
"step": 2315
},
{
"epoch": 0.1856,
"grad_norm": 1.8563107252120972,
"learning_rate": 4.952928762954161e-05,
"loss": 1.1796,
"step": 2320
},
{
"epoch": 0.186,
"grad_norm": 1.6669505834579468,
"learning_rate": 4.952726293608335e-05,
"loss": 1.2501,
"step": 2325
},
{
"epoch": 0.1864,
"grad_norm": 1.6013710498809814,
"learning_rate": 4.952523393908059e-05,
"loss": 1.2109,
"step": 2330
},
{
"epoch": 0.1868,
"grad_norm": 1.4567033052444458,
"learning_rate": 4.952320063888932e-05,
"loss": 1.182,
"step": 2335
},
{
"epoch": 0.1872,
"grad_norm": 1.8492445945739746,
"learning_rate": 4.952116303586631e-05,
"loss": 1.144,
"step": 2340
},
{
"epoch": 0.1876,
"grad_norm": 1.6545616388320923,
"learning_rate": 4.951912113036908e-05,
"loss": 1.2085,
"step": 2345
},
{
"epoch": 0.188,
"grad_norm": 2.060640573501587,
"learning_rate": 4.951707492275589e-05,
"loss": 1.2282,
"step": 2350
},
{
"epoch": 0.1884,
"grad_norm": 1.481137752532959,
"learning_rate": 4.951502441338578e-05,
"loss": 1.2097,
"step": 2355
},
{
"epoch": 0.1888,
"grad_norm": 1.6217643022537231,
"learning_rate": 4.951296960261853e-05,
"loss": 1.1946,
"step": 2360
},
{
"epoch": 0.1892,
"grad_norm": 1.6603466272354126,
"learning_rate": 4.9510910490814666e-05,
"loss": 1.1917,
"step": 2365
},
{
"epoch": 0.1896,
"grad_norm": 1.6394786834716797,
"learning_rate": 4.9508847078335495e-05,
"loss": 1.2794,
"step": 2370
},
{
"epoch": 0.19,
"grad_norm": 1.5222853422164917,
"learning_rate": 4.9506779365543046e-05,
"loss": 1.2223,
"step": 2375
},
{
"epoch": 0.1904,
"grad_norm": 1.431310772895813,
"learning_rate": 4.9504707352800125e-05,
"loss": 1.2508,
"step": 2380
},
{
"epoch": 0.1908,
"grad_norm": 1.5400367975234985,
"learning_rate": 4.950263104047031e-05,
"loss": 1.2497,
"step": 2385
},
{
"epoch": 0.1912,
"grad_norm": 1.5690429210662842,
"learning_rate": 4.950055042891786e-05,
"loss": 1.2056,
"step": 2390
},
{
"epoch": 0.1916,
"grad_norm": 1.645679235458374,
"learning_rate": 4.949846551850788e-05,
"loss": 1.1822,
"step": 2395
},
{
"epoch": 0.192,
"grad_norm": 1.8032221794128418,
"learning_rate": 4.949637630960617e-05,
"loss": 1.2102,
"step": 2400
},
{
"epoch": 0.1924,
"grad_norm": 1.5847824811935425,
"learning_rate": 4.949428280257932e-05,
"loss": 1.2044,
"step": 2405
},
{
"epoch": 0.1928,
"grad_norm": 1.5460349321365356,
"learning_rate": 4.9492184997794624e-05,
"loss": 1.1794,
"step": 2410
},
{
"epoch": 0.1932,
"grad_norm": 1.4651161432266235,
"learning_rate": 4.949008289562019e-05,
"loss": 1.1904,
"step": 2415
},
{
"epoch": 0.1936,
"grad_norm": 1.5283023118972778,
"learning_rate": 4.948797649642484e-05,
"loss": 1.2205,
"step": 2420
},
{
"epoch": 0.194,
"grad_norm": 1.733211874961853,
"learning_rate": 4.948586580057816e-05,
"loss": 1.2324,
"step": 2425
},
{
"epoch": 0.1944,
"grad_norm": 1.4706699848175049,
"learning_rate": 4.94837508084505e-05,
"loss": 1.17,
"step": 2430
},
{
"epoch": 0.1948,
"grad_norm": 1.7306286096572876,
"learning_rate": 4.948163152041295e-05,
"loss": 1.2305,
"step": 2435
},
{
"epoch": 0.1952,
"grad_norm": 1.6754021644592285,
"learning_rate": 4.9479507936837364e-05,
"loss": 1.2448,
"step": 2440
},
{
"epoch": 0.1956,
"grad_norm": 1.6080971956253052,
"learning_rate": 4.9477380058096343e-05,
"loss": 1.2214,
"step": 2445
},
{
"epoch": 0.196,
"grad_norm": 1.6312017440795898,
"learning_rate": 4.947524788456325e-05,
"loss": 1.2026,
"step": 2450
},
{
"epoch": 0.1964,
"grad_norm": 1.5370644330978394,
"learning_rate": 4.947311141661218e-05,
"loss": 1.2314,
"step": 2455
},
{
"epoch": 0.1968,
"grad_norm": 1.5101724863052368,
"learning_rate": 4.947097065461801e-05,
"loss": 1.1738,
"step": 2460
},
{
"epoch": 0.1972,
"grad_norm": 1.5563039779663086,
"learning_rate": 4.946882559895635e-05,
"loss": 1.196,
"step": 2465
},
{
"epoch": 0.1976,
"grad_norm": 1.8207404613494873,
"learning_rate": 4.9466676250003576e-05,
"loss": 1.2136,
"step": 2470
},
{
"epoch": 0.198,
"grad_norm": 1.5161926746368408,
"learning_rate": 4.9464522608136805e-05,
"loss": 1.2426,
"step": 2475
},
{
"epoch": 0.1984,
"grad_norm": 1.722732663154602,
"learning_rate": 4.946236467373392e-05,
"loss": 1.1381,
"step": 2480
},
{
"epoch": 0.1988,
"grad_norm": 1.4065109491348267,
"learning_rate": 4.946020244717355e-05,
"loss": 1.2316,
"step": 2485
},
{
"epoch": 0.1992,
"grad_norm": 1.5711966753005981,
"learning_rate": 4.945803592883509e-05,
"loss": 1.1682,
"step": 2490
},
{
"epoch": 0.1996,
"grad_norm": 1.6523547172546387,
"learning_rate": 4.945586511909865e-05,
"loss": 1.1923,
"step": 2495
},
{
"epoch": 0.2,
"grad_norm": 1.4545037746429443,
"learning_rate": 4.9453690018345144e-05,
"loss": 1.2123,
"step": 2500
},
{
"epoch": 0.2004,
"grad_norm": 1.5507062673568726,
"learning_rate": 4.9451510626956196e-05,
"loss": 1.1574,
"step": 2505
},
{
"epoch": 0.2008,
"grad_norm": 1.5979762077331543,
"learning_rate": 4.944932694531422e-05,
"loss": 1.2123,
"step": 2510
},
{
"epoch": 0.2012,
"grad_norm": 1.7365368604660034,
"learning_rate": 4.944713897380235e-05,
"loss": 1.1493,
"step": 2515
},
{
"epoch": 0.2016,
"grad_norm": 1.5356484651565552,
"learning_rate": 4.9444946712804494e-05,
"loss": 1.1729,
"step": 2520
},
{
"epoch": 0.202,
"grad_norm": 1.5858573913574219,
"learning_rate": 4.9442750162705295e-05,
"loss": 1.2178,
"step": 2525
},
{
"epoch": 0.2024,
"grad_norm": 1.4931632280349731,
"learning_rate": 4.9440549323890176e-05,
"loss": 1.2411,
"step": 2530
},
{
"epoch": 0.2028,
"grad_norm": 1.933543086051941,
"learning_rate": 4.943834419674529e-05,
"loss": 1.2217,
"step": 2535
},
{
"epoch": 0.2032,
"grad_norm": 2.18145489692688,
"learning_rate": 4.943613478165753e-05,
"loss": 1.1514,
"step": 2540
},
{
"epoch": 0.2036,
"grad_norm": 1.430503249168396,
"learning_rate": 4.943392107901458e-05,
"loss": 1.2466,
"step": 2545
},
{
"epoch": 0.204,
"grad_norm": 1.3851200342178345,
"learning_rate": 4.943170308920484e-05,
"loss": 1.2234,
"step": 2550
},
{
"epoch": 0.2044,
"grad_norm": 1.4635603427886963,
"learning_rate": 4.942948081261749e-05,
"loss": 1.1526,
"step": 2555
},
{
"epoch": 0.2048,
"grad_norm": 1.6983256340026855,
"learning_rate": 4.9427254249642444e-05,
"loss": 1.2161,
"step": 2560
},
{
"epoch": 0.2052,
"grad_norm": 1.7087597846984863,
"learning_rate": 4.942502340067038e-05,
"loss": 1.2492,
"step": 2565
},
{
"epoch": 0.2056,
"grad_norm": 1.786859393119812,
"learning_rate": 4.9422788266092715e-05,
"loss": 1.1799,
"step": 2570
},
{
"epoch": 0.206,
"grad_norm": 1.4722228050231934,
"learning_rate": 4.942054884630162e-05,
"loss": 1.2395,
"step": 2575
},
{
"epoch": 0.2064,
"grad_norm": 1.6726198196411133,
"learning_rate": 4.941830514169004e-05,
"loss": 1.148,
"step": 2580
},
{
"epoch": 0.2068,
"grad_norm": 1.4107918739318848,
"learning_rate": 4.941605715265164e-05,
"loss": 1.1613,
"step": 2585
},
{
"epoch": 0.2072,
"grad_norm": 1.531030297279358,
"learning_rate": 4.941380487958086e-05,
"loss": 1.3184,
"step": 2590
},
{
"epoch": 0.2076,
"grad_norm": 1.9271130561828613,
"learning_rate": 4.941154832287288e-05,
"loss": 1.2318,
"step": 2595
},
{
"epoch": 0.208,
"grad_norm": 1.493416428565979,
"learning_rate": 4.940928748292363e-05,
"loss": 1.1343,
"step": 2600
},
{
"epoch": 0.2084,
"grad_norm": 1.7135025262832642,
"learning_rate": 4.9407022360129796e-05,
"loss": 1.1987,
"step": 2605
},
{
"epoch": 0.2088,
"grad_norm": 1.576484203338623,
"learning_rate": 4.9404752954888824e-05,
"loss": 1.2837,
"step": 2610
},
{
"epoch": 0.2092,
"grad_norm": 1.5237318277359009,
"learning_rate": 4.9402479267598887e-05,
"loss": 1.1382,
"step": 2615
},
{
"epoch": 0.2096,
"grad_norm": 1.70023512840271,
"learning_rate": 4.940020129865895e-05,
"loss": 1.2325,
"step": 2620
},
{
"epoch": 0.21,
"grad_norm": 1.4940378665924072,
"learning_rate": 4.939791904846869e-05,
"loss": 1.1978,
"step": 2625
},
{
"epoch": 0.2104,
"grad_norm": 1.4972904920578003,
"learning_rate": 4.939563251742855e-05,
"loss": 1.1918,
"step": 2630
},
{
"epoch": 0.2108,
"grad_norm": 1.4454622268676758,
"learning_rate": 4.939334170593972e-05,
"loss": 1.1975,
"step": 2635
},
{
"epoch": 0.2112,
"grad_norm": 1.844054102897644,
"learning_rate": 4.939104661440415e-05,
"loss": 1.1543,
"step": 2640
},
{
"epoch": 0.2116,
"grad_norm": 1.446661114692688,
"learning_rate": 4.938874724322454e-05,
"loss": 1.2034,
"step": 2645
},
{
"epoch": 0.212,
"grad_norm": 1.7901402711868286,
"learning_rate": 4.938644359280433e-05,
"loss": 1.1945,
"step": 2650
},
{
"epoch": 0.2124,
"grad_norm": 1.6717280149459839,
"learning_rate": 4.938413566354772e-05,
"loss": 1.1632,
"step": 2655
},
{
"epoch": 0.2128,
"grad_norm": 1.4415397644042969,
"learning_rate": 4.938182345585966e-05,
"loss": 1.1729,
"step": 2660
},
{
"epoch": 0.2132,
"grad_norm": 1.5009517669677734,
"learning_rate": 4.937950697014585e-05,
"loss": 1.298,
"step": 2665
},
{
"epoch": 0.2136,
"grad_norm": 1.6571409702301025,
"learning_rate": 4.937718620681273e-05,
"loss": 1.2179,
"step": 2670
},
{
"epoch": 0.214,
"grad_norm": 1.4815192222595215,
"learning_rate": 4.937486116626752e-05,
"loss": 1.175,
"step": 2675
},
{
"epoch": 0.2144,
"grad_norm": 1.3653024435043335,
"learning_rate": 4.9372531848918145e-05,
"loss": 1.123,
"step": 2680
},
{
"epoch": 0.2148,
"grad_norm": 1.743571162223816,
"learning_rate": 4.937019825517333e-05,
"loss": 1.224,
"step": 2685
},
{
"epoch": 0.2152,
"grad_norm": 1.633367657661438,
"learning_rate": 4.936786038544251e-05,
"loss": 1.1604,
"step": 2690
},
{
"epoch": 0.2156,
"grad_norm": 1.5236365795135498,
"learning_rate": 4.936551824013589e-05,
"loss": 1.255,
"step": 2695
},
{
"epoch": 0.216,
"grad_norm": 1.6469926834106445,
"learning_rate": 4.9363171819664434e-05,
"loss": 1.2126,
"step": 2700
},
{
"epoch": 0.2164,
"grad_norm": 1.4393495321273804,
"learning_rate": 4.936082112443983e-05,
"loss": 1.2172,
"step": 2705
},
{
"epoch": 0.2168,
"grad_norm": 1.4337791204452515,
"learning_rate": 4.935846615487453e-05,
"loss": 1.2352,
"step": 2710
},
{
"epoch": 0.2172,
"grad_norm": 1.488175630569458,
"learning_rate": 4.935610691138175e-05,
"loss": 1.1805,
"step": 2715
},
{
"epoch": 0.2176,
"grad_norm": 1.7446844577789307,
"learning_rate": 4.935374339437543e-05,
"loss": 1.1901,
"step": 2720
},
{
"epoch": 0.218,
"grad_norm": 1.6607255935668945,
"learning_rate": 4.935137560427027e-05,
"loss": 1.2369,
"step": 2725
},
{
"epoch": 0.2184,
"grad_norm": 1.4664397239685059,
"learning_rate": 4.934900354148173e-05,
"loss": 1.1617,
"step": 2730
},
{
"epoch": 0.2188,
"grad_norm": 1.5030437707901,
"learning_rate": 4.934662720642601e-05,
"loss": 1.2486,
"step": 2735
},
{
"epoch": 0.2192,
"grad_norm": 1.3806397914886475,
"learning_rate": 4.934424659952006e-05,
"loss": 1.1869,
"step": 2740
},
{
"epoch": 0.2196,
"grad_norm": 1.8397371768951416,
"learning_rate": 4.934186172118157e-05,
"loss": 1.2266,
"step": 2745
},
{
"epoch": 0.22,
"grad_norm": 1.6181316375732422,
"learning_rate": 4.933947257182901e-05,
"loss": 1.223,
"step": 2750
},
{
"epoch": 0.2204,
"grad_norm": 1.459485650062561,
"learning_rate": 4.933707915188156e-05,
"loss": 1.2411,
"step": 2755
},
{
"epoch": 0.2208,
"grad_norm": 1.567919373512268,
"learning_rate": 4.933468146175918e-05,
"loss": 1.2363,
"step": 2760
},
{
"epoch": 0.2212,
"grad_norm": 1.3925971984863281,
"learning_rate": 4.9332279501882564e-05,
"loss": 1.2627,
"step": 2765
},
{
"epoch": 0.2216,
"grad_norm": 1.4231473207473755,
"learning_rate": 4.932987327267316e-05,
"loss": 1.1886,
"step": 2770
},
{
"epoch": 0.222,
"grad_norm": 1.6479936838150024,
"learning_rate": 4.9327462774553166e-05,
"loss": 1.1865,
"step": 2775
},
{
"epoch": 0.2224,
"grad_norm": 1.5866742134094238,
"learning_rate": 4.9325048007945526e-05,
"loss": 1.1956,
"step": 2780
},
{
"epoch": 0.2228,
"grad_norm": 1.6373844146728516,
"learning_rate": 4.932262897327393e-05,
"loss": 1.1883,
"step": 2785
},
{
"epoch": 0.2232,
"grad_norm": 1.3438984155654907,
"learning_rate": 4.9320205670962814e-05,
"loss": 1.2066,
"step": 2790
},
{
"epoch": 0.2236,
"grad_norm": 1.5680036544799805,
"learning_rate": 4.93177781014374e-05,
"loss": 1.2162,
"step": 2795
},
{
"epoch": 0.224,
"grad_norm": 1.7673883438110352,
"learning_rate": 4.9315346265123594e-05,
"loss": 1.1993,
"step": 2800
},
{
"epoch": 0.2244,
"grad_norm": 1.8897387981414795,
"learning_rate": 4.93129101624481e-05,
"loss": 1.2517,
"step": 2805
},
{
"epoch": 0.2248,
"grad_norm": 1.4127836227416992,
"learning_rate": 4.931046979383835e-05,
"loss": 1.2182,
"step": 2810
},
{
"epoch": 0.2252,
"grad_norm": 1.5710681676864624,
"learning_rate": 4.930802515972255e-05,
"loss": 1.1595,
"step": 2815
},
{
"epoch": 0.2256,
"grad_norm": 1.7648258209228516,
"learning_rate": 4.9305576260529607e-05,
"loss": 1.2714,
"step": 2820
},
{
"epoch": 0.226,
"grad_norm": 1.628089427947998,
"learning_rate": 4.930312309668922e-05,
"loss": 1.1958,
"step": 2825
},
{
"epoch": 0.2264,
"grad_norm": 1.6491187810897827,
"learning_rate": 4.930066566863182e-05,
"loss": 1.2086,
"step": 2830
},
{
"epoch": 0.2268,
"grad_norm": 1.5382400751113892,
"learning_rate": 4.929820397678858e-05,
"loss": 1.2069,
"step": 2835
},
{
"epoch": 0.2272,
"grad_norm": NaN,
"learning_rate": 4.929623155367841e-05,
"loss": 1.1623,
"step": 2840
},
{
"epoch": 0.2276,
"grad_norm": 1.6124461889266968,
"learning_rate": 4.929376218810963e-05,
"loss": 1.1087,
"step": 2845
},
{
"epoch": 0.228,
"grad_norm": 1.5880391597747803,
"learning_rate": 4.9291288559966286e-05,
"loss": 1.1652,
"step": 2850
},
{
"epoch": 0.2284,
"grad_norm": 1.4969943761825562,
"learning_rate": 4.928881066968241e-05,
"loss": 1.1668,
"step": 2855
},
{
"epoch": 0.2288,
"grad_norm": 1.5591754913330078,
"learning_rate": 4.928632851769277e-05,
"loss": 1.2072,
"step": 2860
},
{
"epoch": 0.2292,
"grad_norm": 1.5494965314865112,
"learning_rate": 4.9283842104432886e-05,
"loss": 1.225,
"step": 2865
},
{
"epoch": 0.2296,
"grad_norm": 1.4970649480819702,
"learning_rate": 4.928135143033903e-05,
"loss": 1.2435,
"step": 2870
},
{
"epoch": 0.23,
"grad_norm": 1.5952671766281128,
"learning_rate": 4.9278856495848195e-05,
"loss": 1.1857,
"step": 2875
},
{
"epoch": 0.2304,
"grad_norm": 1.7178370952606201,
"learning_rate": 4.927635730139816e-05,
"loss": 1.2055,
"step": 2880
},
{
"epoch": 0.2308,
"grad_norm": 1.5079060792922974,
"learning_rate": 4.927385384742742e-05,
"loss": 1.2028,
"step": 2885
},
{
"epoch": 0.2312,
"grad_norm": 1.504225730895996,
"learning_rate": 4.927134613437524e-05,
"loss": 1.1848,
"step": 2890
},
{
"epoch": 0.2316,
"grad_norm": 1.7504538297653198,
"learning_rate": 4.9268834162681607e-05,
"loss": 1.1649,
"step": 2895
},
{
"epoch": 0.232,
"grad_norm": 1.7637404203414917,
"learning_rate": 4.9266317932787294e-05,
"loss": 1.1629,
"step": 2900
},
{
"epoch": 0.2324,
"grad_norm": 1.6753672361373901,
"learning_rate": 4.926379744513378e-05,
"loss": 1.252,
"step": 2905
},
{
"epoch": 0.2328,
"grad_norm": 1.7061619758605957,
"learning_rate": 4.9261272700163315e-05,
"loss": 1.2134,
"step": 2910
},
{
"epoch": 0.2332,
"grad_norm": 1.7861872911453247,
"learning_rate": 4.9258743698318886e-05,
"loss": 1.2394,
"step": 2915
},
{
"epoch": 0.2336,
"grad_norm": 1.9671781063079834,
"learning_rate": 4.925621044004424e-05,
"loss": 1.2327,
"step": 2920
},
{
"epoch": 0.234,
"grad_norm": 1.5310274362564087,
"learning_rate": 4.9253672925783846e-05,
"loss": 1.2464,
"step": 2925
},
{
"epoch": 0.2344,
"grad_norm": 1.5653115510940552,
"learning_rate": 4.925113115598295e-05,
"loss": 1.252,
"step": 2930
},
{
"epoch": 0.2348,
"grad_norm": 1.6528226137161255,
"learning_rate": 4.9248585131087524e-05,
"loss": 1.1158,
"step": 2935
},
{
"epoch": 0.2352,
"grad_norm": 1.6093913316726685,
"learning_rate": 4.9246034851544294e-05,
"loss": 1.1969,
"step": 2940
},
{
"epoch": 0.2356,
"grad_norm": 1.875658631324768,
"learning_rate": 4.9243480317800727e-05,
"loss": 1.167,
"step": 2945
},
{
"epoch": 0.236,
"grad_norm": 1.6207784414291382,
"learning_rate": 4.9240921530305043e-05,
"loss": 1.1764,
"step": 2950
},
{
"epoch": 0.2364,
"grad_norm": 1.3430836200714111,
"learning_rate": 4.9238358489506206e-05,
"loss": 1.1961,
"step": 2955
},
{
"epoch": 0.2368,
"grad_norm": 1.6542601585388184,
"learning_rate": 4.9235791195853935e-05,
"loss": 1.178,
"step": 2960
},
{
"epoch": 0.2372,
"grad_norm": 1.4790018796920776,
"learning_rate": 4.9233219649798676e-05,
"loss": 1.1947,
"step": 2965
},
{
"epoch": 0.2376,
"grad_norm": 1.6997478008270264,
"learning_rate": 4.923064385179163e-05,
"loss": 1.1895,
"step": 2970
},
{
"epoch": 0.238,
"grad_norm": 1.570786476135254,
"learning_rate": 4.9228063802284755e-05,
"loss": 1.1722,
"step": 2975
},
{
"epoch": 0.2384,
"grad_norm": 1.4543014764785767,
"learning_rate": 4.922547950173073e-05,
"loss": 1.153,
"step": 2980
},
{
"epoch": 0.2388,
"grad_norm": 1.5781867504119873,
"learning_rate": 4.922289095058301e-05,
"loss": 1.2221,
"step": 2985
},
{
"epoch": 0.2392,
"grad_norm": 1.5818521976470947,
"learning_rate": 4.922029814929577e-05,
"loss": 1.1816,
"step": 2990
},
{
"epoch": 0.2396,
"grad_norm": 1.3536361455917358,
"learning_rate": 4.921770109832396e-05,
"loss": 1.2118,
"step": 2995
},
{
"epoch": 0.24,
"grad_norm": 1.9095213413238525,
"learning_rate": 4.9215099798123245e-05,
"loss": 1.1636,
"step": 3000
},
{
"epoch": 0.2404,
"grad_norm": 1.721266746520996,
"learning_rate": 4.9212494249150046e-05,
"loss": 1.1963,
"step": 3005
},
{
"epoch": 0.2408,
"grad_norm": 1.4719969034194946,
"learning_rate": 4.920988445186153e-05,
"loss": 1.1835,
"step": 3010
},
{
"epoch": 0.2412,
"grad_norm": 1.5968965291976929,
"learning_rate": 4.920727040671562e-05,
"loss": 1.2169,
"step": 3015
},
{
"epoch": 0.2416,
"grad_norm": 1.3878967761993408,
"learning_rate": 4.920465211417097e-05,
"loss": 1.17,
"step": 3020
},
{
"epoch": 0.242,
"grad_norm": 1.4749703407287598,
"learning_rate": 4.9202029574686986e-05,
"loss": 1.2224,
"step": 3025
},
{
"epoch": 0.2424,
"grad_norm": 1.751265525817871,
"learning_rate": 4.9199402788723817e-05,
"loss": 1.2115,
"step": 3030
},
{
"epoch": 0.2428,
"grad_norm": 1.902597427368164,
"learning_rate": 4.919677175674236e-05,
"loss": 1.2174,
"step": 3035
},
{
"epoch": 0.2432,
"grad_norm": 1.5894381999969482,
"learning_rate": 4.919413647920426e-05,
"loss": 1.1873,
"step": 3040
},
{
"epoch": 0.2436,
"grad_norm": 1.6837565898895264,
"learning_rate": 4.9191496956571886e-05,
"loss": 1.2807,
"step": 3045
},
{
"epoch": 0.244,
"grad_norm": 1.464011788368225,
"learning_rate": 4.9188853189308385e-05,
"loss": 1.1782,
"step": 3050
},
{
"epoch": 0.2444,
"grad_norm": 1.8131499290466309,
"learning_rate": 4.918620517787762e-05,
"loss": 1.1536,
"step": 3055
},
{
"epoch": 0.2448,
"grad_norm": 1.5556432008743286,
"learning_rate": 4.9183552922744204e-05,
"loss": 1.2281,
"step": 3060
},
{
"epoch": 0.2452,
"grad_norm": 1.51846444606781,
"learning_rate": 4.9180896424373526e-05,
"loss": 1.2087,
"step": 3065
},
{
"epoch": 0.2456,
"grad_norm": 1.334592580795288,
"learning_rate": 4.917823568323167e-05,
"loss": 1.1934,
"step": 3070
},
{
"epoch": 0.246,
"grad_norm": 1.6459736824035645,
"learning_rate": 4.9175570699785494e-05,
"loss": 1.2278,
"step": 3075
},
{
"epoch": 0.2464,
"grad_norm": 1.4319044351577759,
"learning_rate": 4.9172901474502606e-05,
"loss": 1.1466,
"step": 3080
},
{
"epoch": 0.2468,
"grad_norm": 1.677406907081604,
"learning_rate": 4.917022800785134e-05,
"loss": 1.1745,
"step": 3085
},
{
"epoch": 0.2472,
"grad_norm": 1.3851629495620728,
"learning_rate": 4.916755030030077e-05,
"loss": 1.1146,
"step": 3090
},
{
"epoch": 0.2476,
"grad_norm": 1.5272161960601807,
"learning_rate": 4.9164868352320746e-05,
"loss": 1.2117,
"step": 3095
},
{
"epoch": 0.248,
"grad_norm": 1.5623695850372314,
"learning_rate": 4.916218216438183e-05,
"loss": 1.1319,
"step": 3100
},
{
"epoch": 0.2484,
"grad_norm": 1.4115298986434937,
"learning_rate": 4.915949173695534e-05,
"loss": 1.1782,
"step": 3105
},
{
"epoch": 0.2488,
"grad_norm": 1.6355260610580444,
"learning_rate": 4.915679707051334e-05,
"loss": 1.2008,
"step": 3110
},
{
"epoch": 0.2492,
"grad_norm": 1.600929856300354,
"learning_rate": 4.915409816552864e-05,
"loss": 1.2382,
"step": 3115
},
{
"epoch": 0.2496,
"grad_norm": 1.574831247329712,
"learning_rate": 4.915139502247478e-05,
"loss": 1.188,
"step": 3120
},
{
"epoch": 0.25,
"grad_norm": 1.426468014717102,
"learning_rate": 4.914868764182605e-05,
"loss": 1.0933,
"step": 3125
},
{
"epoch": 0.2504,
"grad_norm": 1.7503633499145508,
"learning_rate": 4.914597602405749e-05,
"loss": 1.1729,
"step": 3130
},
{
"epoch": 0.2508,
"grad_norm": 1.7526756525039673,
"learning_rate": 4.914326016964489e-05,
"loss": 1.1639,
"step": 3135
},
{
"epoch": 0.2512,
"grad_norm": 1.611013412475586,
"learning_rate": 4.9140540079064757e-05,
"loss": 1.193,
"step": 3140
},
{
"epoch": 0.2516,
"grad_norm": 1.6564090251922607,
"learning_rate": 4.913781575279437e-05,
"loss": 1.2075,
"step": 3145
},
{
"epoch": 0.252,
"grad_norm": 1.5011568069458008,
"learning_rate": 4.913508719131173e-05,
"loss": 1.2225,
"step": 3150
},
{
"epoch": 0.2524,
"grad_norm": 1.7857328653335571,
"learning_rate": 4.913235439509559e-05,
"loss": 1.1788,
"step": 3155
},
{
"epoch": 0.2528,
"grad_norm": 1.6372023820877075,
"learning_rate": 4.912961736462545e-05,
"loss": 1.2139,
"step": 3160
},
{
"epoch": 0.2532,
"grad_norm": 1.658705234527588,
"learning_rate": 4.912687610038155e-05,
"loss": 1.1624,
"step": 3165
},
{
"epoch": 0.2536,
"grad_norm": 1.5441612005233765,
"learning_rate": 4.9124130602844856e-05,
"loss": 1.2376,
"step": 3170
},
{
"epoch": 0.254,
"grad_norm": 1.6848093271255493,
"learning_rate": 4.91213808724971e-05,
"loss": 1.1688,
"step": 3175
},
{
"epoch": 0.2544,
"grad_norm": 1.510003924369812,
"learning_rate": 4.911862690982076e-05,
"loss": 1.162,
"step": 3180
},
{
"epoch": 0.2548,
"grad_norm": 1.604901909828186,
"learning_rate": 4.9115868715299044e-05,
"loss": 1.1569,
"step": 3185
},
{
"epoch": 0.2552,
"grad_norm": 1.606873869895935,
"learning_rate": 4.911310628941588e-05,
"loss": 1.2192,
"step": 3190
},
{
"epoch": 0.2556,
"grad_norm": 1.5792936086654663,
"learning_rate": 4.9110339632655986e-05,
"loss": 1.2065,
"step": 3195
},
{
"epoch": 0.256,
"grad_norm": 1.4216783046722412,
"learning_rate": 4.9107568745504796e-05,
"loss": 1.1822,
"step": 3200
},
{
"epoch": 0.2564,
"grad_norm": 1.6753329038619995,
"learning_rate": 4.910479362844848e-05,
"loss": 1.2008,
"step": 3205
},
{
"epoch": 0.2568,
"grad_norm": 1.7560926675796509,
"learning_rate": 4.9102014281973965e-05,
"loss": 1.2367,
"step": 3210
},
{
"epoch": 0.2572,
"grad_norm": 1.5583642721176147,
"learning_rate": 4.90992307065689e-05,
"loss": 1.0914,
"step": 3215
},
{
"epoch": 0.2576,
"grad_norm": 1.4227246046066284,
"learning_rate": 4.909644290272172e-05,
"loss": 1.1597,
"step": 3220
},
{
"epoch": 0.258,
"grad_norm": 1.5329896211624146,
"learning_rate": 4.909365087092154e-05,
"loss": 1.1979,
"step": 3225
},
{
"epoch": 0.2584,
"grad_norm": 1.5972647666931152,
"learning_rate": 4.909085461165827e-05,
"loss": 1.1652,
"step": 3230
},
{
"epoch": 0.2588,
"grad_norm": 1.2990952730178833,
"learning_rate": 4.9088054125422536e-05,
"loss": 1.1712,
"step": 3235
},
{
"epoch": 0.2592,
"grad_norm": 1.9855566024780273,
"learning_rate": 4.90852494127057e-05,
"loss": 1.1496,
"step": 3240
},
{
"epoch": 0.2596,
"grad_norm": 1.7270143032073975,
"learning_rate": 4.9082440473999894e-05,
"loss": 1.1812,
"step": 3245
},
{
"epoch": 0.26,
"grad_norm": 1.8106093406677246,
"learning_rate": 4.9079627309797954e-05,
"loss": 1.1809,
"step": 3250
},
{
"epoch": 0.2604,
"grad_norm": 1.5340815782546997,
"learning_rate": 4.907680992059349e-05,
"loss": 1.1683,
"step": 3255
},
{
"epoch": 0.2608,
"grad_norm": 2.0770516395568848,
"learning_rate": 4.9073988306880835e-05,
"loss": 1.2156,
"step": 3260
},
{
"epoch": 0.2612,
"grad_norm": 1.7493189573287964,
"learning_rate": 4.907116246915507e-05,
"loss": 1.1812,
"step": 3265
},
{
"epoch": 0.2616,
"grad_norm": 1.8361352682113647,
"learning_rate": 4.906833240791202e-05,
"loss": 1.1727,
"step": 3270
},
{
"epoch": 0.262,
"grad_norm": 1.3989719152450562,
"learning_rate": 4.906549812364824e-05,
"loss": 1.2014,
"step": 3275
},
{
"epoch": 0.2624,
"grad_norm": 1.319180965423584,
"learning_rate": 4.906265961686103e-05,
"loss": 1.1225,
"step": 3280
},
{
"epoch": 0.2628,
"grad_norm": 1.345543622970581,
"learning_rate": 4.9059816888048435e-05,
"loss": 1.1486,
"step": 3285
},
{
"epoch": 0.2632,
"grad_norm": 1.6714309453964233,
"learning_rate": 4.9056969937709244e-05,
"loss": 1.2246,
"step": 3290
},
{
"epoch": 0.2636,
"grad_norm": 1.8523310422897339,
"learning_rate": 4.905411876634298e-05,
"loss": 1.2018,
"step": 3295
},
{
"epoch": 0.264,
"grad_norm": 2.405902147293091,
"learning_rate": 4.905126337444991e-05,
"loss": 1.1864,
"step": 3300
},
{
"epoch": 0.2644,
"grad_norm": 1.8036085367202759,
"learning_rate": 4.9048403762531034e-05,
"loss": 1.1622,
"step": 3305
},
{
"epoch": 0.2648,
"grad_norm": 1.7046128511428833,
"learning_rate": 4.90455399310881e-05,
"loss": 1.236,
"step": 3310
},
{
"epoch": 0.2652,
"grad_norm": 1.393291711807251,
"learning_rate": 4.90426718806236e-05,
"loss": 1.1616,
"step": 3315
},
{
"epoch": 0.2656,
"grad_norm": 1.621147632598877,
"learning_rate": 4.903979961164076e-05,
"loss": 1.1271,
"step": 3320
},
{
"epoch": 0.266,
"grad_norm": 1.672818899154663,
"learning_rate": 4.903692312464354e-05,
"loss": 1.2407,
"step": 3325
},
{
"epoch": 0.2664,
"grad_norm": 1.5829147100448608,
"learning_rate": 4.9034042420136656e-05,
"loss": 1.2192,
"step": 3330
},
{
"epoch": 0.2668,
"grad_norm": 1.6565492153167725,
"learning_rate": 4.903115749862555e-05,
"loss": 1.1928,
"step": 3335
},
{
"epoch": 0.2672,
"grad_norm": 1.6932697296142578,
"learning_rate": 4.9028268360616415e-05,
"loss": 1.155,
"step": 3340
},
{
"epoch": 0.2676,
"grad_norm": 1.8302545547485352,
"learning_rate": 4.902537500661617e-05,
"loss": 1.2264,
"step": 3345
},
{
"epoch": 0.268,
"grad_norm": 1.6364758014678955,
"learning_rate": 4.9022477437132485e-05,
"loss": 1.2153,
"step": 3350
},
{
"epoch": 0.2684,
"grad_norm": 1.8215218782424927,
"learning_rate": 4.9019575652673776e-05,
"loss": 1.1692,
"step": 3355
},
{
"epoch": 0.2688,
"grad_norm": 1.5692654848098755,
"learning_rate": 4.901666965374917e-05,
"loss": 1.2065,
"step": 3360
},
{
"epoch": 0.2692,
"grad_norm": 1.7172425985336304,
"learning_rate": 4.901375944086857e-05,
"loss": 1.1476,
"step": 3365
},
{
"epoch": 0.2696,
"grad_norm": 1.4633954763412476,
"learning_rate": 4.901084501454259e-05,
"loss": 1.1985,
"step": 3370
},
{
"epoch": 0.27,
"grad_norm": 1.598120927810669,
"learning_rate": 4.900792637528261e-05,
"loss": 1.1644,
"step": 3375
},
{
"epoch": 0.2704,
"grad_norm": 1.7177152633666992,
"learning_rate": 4.900500352360071e-05,
"loss": 1.1758,
"step": 3380
},
{
"epoch": 0.2708,
"grad_norm": 1.6786895990371704,
"learning_rate": 4.900207646000975e-05,
"loss": 1.2129,
"step": 3385
},
{
"epoch": 0.2712,
"grad_norm": 1.7196831703186035,
"learning_rate": 4.89991451850233e-05,
"loss": 1.2125,
"step": 3390
},
{
"epoch": 0.2716,
"grad_norm": 2.0396034717559814,
"learning_rate": 4.899620969915568e-05,
"loss": 1.2289,
"step": 3395
},
{
"epoch": 0.272,
"grad_norm": 1.5949326753616333,
"learning_rate": 4.899327000292197e-05,
"loss": 1.1816,
"step": 3400
},
{
"epoch": 0.2724,
"grad_norm": 1.5548696517944336,
"learning_rate": 4.8990326096837945e-05,
"loss": 1.1429,
"step": 3405
},
{
"epoch": 0.2728,
"grad_norm": 1.6374534368515015,
"learning_rate": 4.8987377981420154e-05,
"loss": 1.1942,
"step": 3410
},
{
"epoch": 0.2732,
"grad_norm": 1.7425651550292969,
"learning_rate": 4.898442565718586e-05,
"loss": 1.1463,
"step": 3415
},
{
"epoch": 0.2736,
"grad_norm": 1.5861971378326416,
"learning_rate": 4.898146912465309e-05,
"loss": 1.1314,
"step": 3420
},
{
"epoch": 0.274,
"grad_norm": 1.7639867067337036,
"learning_rate": 4.897850838434059e-05,
"loss": 1.1952,
"step": 3425
},
{
"epoch": 0.2744,
"grad_norm": 1.6477775573730469,
"learning_rate": 4.897554343676785e-05,
"loss": 1.1891,
"step": 3430
},
{
"epoch": 0.2748,
"grad_norm": 1.676500678062439,
"learning_rate": 4.897257428245511e-05,
"loss": 1.1931,
"step": 3435
},
{
"epoch": 0.2752,
"grad_norm": 1.694056749343872,
"learning_rate": 4.896960092192332e-05,
"loss": 1.1777,
"step": 3440
},
{
"epoch": 0.2756,
"grad_norm": 1.711790680885315,
"learning_rate": 4.896662335569419e-05,
"loss": 1.1927,
"step": 3445
},
{
"epoch": 0.276,
"grad_norm": 1.548268437385559,
"learning_rate": 4.896364158429016e-05,
"loss": 1.1411,
"step": 3450
},
{
"epoch": 0.2764,
"grad_norm": 1.5789943933486938,
"learning_rate": 4.896065560823442e-05,
"loss": 1.2414,
"step": 3455
},
{
"epoch": 0.2768,
"grad_norm": 1.7220913171768188,
"learning_rate": 4.895766542805088e-05,
"loss": 1.1892,
"step": 3460
},
{
"epoch": 0.2772,
"grad_norm": 1.6338775157928467,
"learning_rate": 4.895467104426421e-05,
"loss": 1.1721,
"step": 3465
},
{
"epoch": 0.2776,
"grad_norm": 1.5051183700561523,
"learning_rate": 4.895167245739978e-05,
"loss": 1.1903,
"step": 3470
},
{
"epoch": 0.278,
"grad_norm": 1.6972531080245972,
"learning_rate": 4.894866966798374e-05,
"loss": 1.2405,
"step": 3475
},
{
"epoch": 0.2784,
"grad_norm": 1.515919804573059,
"learning_rate": 4.8945662676542946e-05,
"loss": 1.1592,
"step": 3480
},
{
"epoch": 0.2788,
"grad_norm": 1.4641064405441284,
"learning_rate": 4.894265148360502e-05,
"loss": 1.1888,
"step": 3485
},
{
"epoch": 0.2792,
"grad_norm": 2.053290843963623,
"learning_rate": 4.893963608969828e-05,
"loss": 1.1829,
"step": 3490
},
{
"epoch": 0.2796,
"grad_norm": 1.4748057126998901,
"learning_rate": 4.8936616495351826e-05,
"loss": 1.1498,
"step": 3495
},
{
"epoch": 0.28,
"grad_norm": 1.6499887704849243,
"learning_rate": 4.8933592701095466e-05,
"loss": 1.2141,
"step": 3500
},
{
"epoch": 0.2804,
"grad_norm": 1.6166945695877075,
"learning_rate": 4.8930564707459756e-05,
"loss": 1.1658,
"step": 3505
},
{
"epoch": 0.2808,
"grad_norm": 1.8569326400756836,
"learning_rate": 4.892753251497599e-05,
"loss": 1.2115,
"step": 3510
},
{
"epoch": 0.2812,
"grad_norm": 1.6417903900146484,
"learning_rate": 4.89244961241762e-05,
"loss": 1.1523,
"step": 3515
},
{
"epoch": 0.2816,
"grad_norm": 1.4864392280578613,
"learning_rate": 4.892145553559313e-05,
"loss": 1.1845,
"step": 3520
},
{
"epoch": 0.282,
"grad_norm": 1.427742600440979,
"learning_rate": 4.89184107497603e-05,
"loss": 1.2047,
"step": 3525
},
{
"epoch": 0.2824,
"grad_norm": 1.2292075157165527,
"learning_rate": 4.891536176721194e-05,
"loss": 1.1957,
"step": 3530
},
{
"epoch": 0.2828,
"grad_norm": 1.3831216096878052,
"learning_rate": 4.891230858848302e-05,
"loss": 1.1987,
"step": 3535
},
{
"epoch": 0.2832,
"grad_norm": 1.2832379341125488,
"learning_rate": 4.8909251214109264e-05,
"loss": 1.209,
"step": 3540
},
{
"epoch": 0.2836,
"grad_norm": 1.6963448524475098,
"learning_rate": 4.89061896446271e-05,
"loss": 1.1535,
"step": 3545
},
{
"epoch": 0.284,
"grad_norm": 1.6399435997009277,
"learning_rate": 4.890312388057371e-05,
"loss": 1.1641,
"step": 3550
},
{
"epoch": 0.2844,
"grad_norm": 1.6266658306121826,
"learning_rate": 4.890005392248702e-05,
"loss": 1.1824,
"step": 3555
},
{
"epoch": 0.2848,
"grad_norm": 1.6541725397109985,
"learning_rate": 4.889697977090569e-05,
"loss": 1.1722,
"step": 3560
},
{
"epoch": 0.2852,
"grad_norm": 1.7126740217208862,
"learning_rate": 4.8893901426369097e-05,
"loss": 1.1279,
"step": 3565
},
{
"epoch": 0.2856,
"grad_norm": 1.5441510677337646,
"learning_rate": 4.889081888941738e-05,
"loss": 1.2244,
"step": 3570
},
{
"epoch": 0.286,
"grad_norm": 1.7929022312164307,
"learning_rate": 4.888773216059138e-05,
"loss": 1.1526,
"step": 3575
},
{
"epoch": 0.2864,
"grad_norm": 1.5796688795089722,
"learning_rate": 4.88846412404327e-05,
"loss": 1.1694,
"step": 3580
},
{
"epoch": 0.2868,
"grad_norm": 1.8234366178512573,
"learning_rate": 4.888154612948369e-05,
"loss": 1.2442,
"step": 3585
},
{
"epoch": 0.2872,
"grad_norm": 1.587524652481079,
"learning_rate": 4.887844682828739e-05,
"loss": 1.1758,
"step": 3590
},
{
"epoch": 0.2876,
"grad_norm": 1.4562510251998901,
"learning_rate": 4.887534333738762e-05,
"loss": 1.1729,
"step": 3595
},
{
"epoch": 0.288,
"grad_norm": 1.8747830390930176,
"learning_rate": 4.887223565732892e-05,
"loss": 1.1426,
"step": 3600
},
{
"epoch": 0.2884,
"grad_norm": 1.7252371311187744,
"learning_rate": 4.886912378865655e-05,
"loss": 1.2046,
"step": 3605
},
{
"epoch": 0.2888,
"grad_norm": 1.5491045713424683,
"learning_rate": 4.8866007731916516e-05,
"loss": 1.2523,
"step": 3610
},
{
"epoch": 0.2892,
"grad_norm": 1.66487455368042,
"learning_rate": 4.886288748765557e-05,
"loss": 1.1938,
"step": 3615
},
{
"epoch": 0.2896,
"grad_norm": 1.7166435718536377,
"learning_rate": 4.885976305642119e-05,
"loss": 1.1753,
"step": 3620
},
{
"epoch": 0.29,
"grad_norm": 1.592100977897644,
"learning_rate": 4.885663443876158e-05,
"loss": 1.148,
"step": 3625
},
{
"epoch": 0.2904,
"grad_norm": 1.4008827209472656,
"learning_rate": 4.885350163522568e-05,
"loss": 1.2096,
"step": 3630
},
{
"epoch": 0.2908,
"grad_norm": 1.8386391401290894,
"learning_rate": 4.8850364646363196e-05,
"loss": 1.1847,
"step": 3635
},
{
"epoch": 0.2912,
"grad_norm": 1.4619133472442627,
"learning_rate": 4.884722347272452e-05,
"loss": 1.1385,
"step": 3640
},
{
"epoch": 0.2916,
"grad_norm": 1.4952281713485718,
"learning_rate": 4.88440781148608e-05,
"loss": 1.1568,
"step": 3645
},
{
"epoch": 0.292,
"grad_norm": 1.7998713254928589,
"learning_rate": 4.884092857332393e-05,
"loss": 1.2424,
"step": 3650
},
{
"epoch": 0.2924,
"grad_norm": 1.4490965604782104,
"learning_rate": 4.883777484866653e-05,
"loss": 1.2045,
"step": 3655
},
{
"epoch": 0.2928,
"grad_norm": 1.3473284244537354,
"learning_rate": 4.8834616941441944e-05,
"loss": 1.1838,
"step": 3660
},
{
"epoch": 0.2932,
"grad_norm": 1.6016281843185425,
"learning_rate": 4.883145485220426e-05,
"loss": 1.1829,
"step": 3665
},
{
"epoch": 0.2936,
"grad_norm": 1.5061883926391602,
"learning_rate": 4.88282885815083e-05,
"loss": 1.1409,
"step": 3670
},
{
"epoch": 0.294,
"grad_norm": 1.6041043996810913,
"learning_rate": 4.8825118129909596e-05,
"loss": 1.171,
"step": 3675
},
{
"epoch": 0.2944,
"grad_norm": 1.563822865486145,
"learning_rate": 4.882194349796446e-05,
"loss": 1.229,
"step": 3680
},
{
"epoch": 0.2948,
"grad_norm": 1.5734202861785889,
"learning_rate": 4.8818764686229904e-05,
"loss": 1.2496,
"step": 3685
},
{
"epoch": 0.2952,
"grad_norm": 1.5884230136871338,
"learning_rate": 4.881558169526368e-05,
"loss": 1.2271,
"step": 3690
},
{
"epoch": 0.2956,
"grad_norm": 1.6446903944015503,
"learning_rate": 4.881239452562427e-05,
"loss": 1.196,
"step": 3695
},
{
"epoch": 0.296,
"grad_norm": 1.6241570711135864,
"learning_rate": 4.880920317787091e-05,
"loss": 1.2396,
"step": 3700
},
{
"epoch": 0.2964,
"grad_norm": 1.3656667470932007,
"learning_rate": 4.8806007652563526e-05,
"loss": 1.2361,
"step": 3705
},
{
"epoch": 0.2968,
"grad_norm": 1.456030011177063,
"learning_rate": 4.880280795026283e-05,
"loss": 1.1698,
"step": 3710
},
{
"epoch": 0.2972,
"grad_norm": 1.6222285032272339,
"learning_rate": 4.879960407153022e-05,
"loss": 1.2176,
"step": 3715
},
{
"epoch": 0.2976,
"grad_norm": 1.7688349485397339,
"learning_rate": 4.879639601692786e-05,
"loss": 1.1966,
"step": 3720
},
{
"epoch": 0.298,
"grad_norm": 1.7796573638916016,
"learning_rate": 4.879318378701863e-05,
"loss": 1.2249,
"step": 3725
},
{
"epoch": 0.2984,
"grad_norm": 1.565400242805481,
"learning_rate": 4.8789967382366144e-05,
"loss": 1.1471,
"step": 3730
},
{
"epoch": 0.2988,
"grad_norm": 1.5751515626907349,
"learning_rate": 4.8786746803534754e-05,
"loss": 1.1918,
"step": 3735
},
{
"epoch": 0.2992,
"grad_norm": 1.7343270778656006,
"learning_rate": 4.878352205108955e-05,
"loss": 1.209,
"step": 3740
},
{
"epoch": 0.2996,
"grad_norm": 1.8670405149459839,
"learning_rate": 4.878029312559633e-05,
"loss": 1.2088,
"step": 3745
},
{
"epoch": 0.3,
"grad_norm": 1.6701018810272217,
"learning_rate": 4.8777060027621644e-05,
"loss": 1.1447,
"step": 3750
},
{
"epoch": 0.3004,
"grad_norm": 1.8615483045578003,
"learning_rate": 4.877382275773278e-05,
"loss": 1.2075,
"step": 3755
},
{
"epoch": 0.3008,
"grad_norm": 1.3462893962860107,
"learning_rate": 4.877058131649774e-05,
"loss": 1.13,
"step": 3760
},
{
"epoch": 0.3012,
"grad_norm": 1.4814949035644531,
"learning_rate": 4.876733570448527e-05,
"loss": 1.2166,
"step": 3765
},
{
"epoch": 0.3016,
"grad_norm": 1.3817886114120483,
"learning_rate": 4.8764085922264843e-05,
"loss": 1.1571,
"step": 3770
},
{
"epoch": 0.302,
"grad_norm": 1.8147492408752441,
"learning_rate": 4.8760831970406664e-05,
"loss": 1.1623,
"step": 3775
},
{
"epoch": 0.3024,
"grad_norm": 1.6408344507217407,
"learning_rate": 4.875757384948167e-05,
"loss": 1.2238,
"step": 3780
},
{
"epoch": 0.3028,
"grad_norm": 1.627832293510437,
"learning_rate": 4.875431156006152e-05,
"loss": 1.1451,
"step": 3785
},
{
"epoch": 0.3032,
"grad_norm": 1.5649932622909546,
"learning_rate": 4.8751045102718626e-05,
"loss": 1.1593,
"step": 3790
},
{
"epoch": 0.3036,
"grad_norm": 1.6648448705673218,
"learning_rate": 4.8747774478026124e-05,
"loss": 1.1864,
"step": 3795
},
{
"epoch": 0.304,
"grad_norm": 1.5145492553710938,
"learning_rate": 4.8744499686557864e-05,
"loss": 1.1968,
"step": 3800
},
{
"epoch": 0.3044,
"grad_norm": 1.4245922565460205,
"learning_rate": 4.874122072888845e-05,
"loss": 1.2392,
"step": 3805
},
{
"epoch": 0.3048,
"grad_norm": 1.7010691165924072,
"learning_rate": 4.87379376055932e-05,
"loss": 1.2011,
"step": 3810
},
{
"epoch": 0.3052,
"grad_norm": 1.9582581520080566,
"learning_rate": 4.873465031724816e-05,
"loss": 1.2047,
"step": 3815
},
{
"epoch": 0.3056,
"grad_norm": 1.9121880531311035,
"learning_rate": 4.873135886443014e-05,
"loss": 1.1488,
"step": 3820
},
{
"epoch": 0.306,
"grad_norm": 1.608534812927246,
"learning_rate": 4.8728063247716646e-05,
"loss": 1.1653,
"step": 3825
},
{
"epoch": 0.3064,
"grad_norm": 1.6228535175323486,
"learning_rate": 4.872476346768593e-05,
"loss": 1.1882,
"step": 3830
},
{
"epoch": 0.3068,
"grad_norm": 1.4595372676849365,
"learning_rate": 4.8721459524916955e-05,
"loss": 1.1683,
"step": 3835
},
{
"epoch": 0.3072,
"grad_norm": 1.4573352336883545,
"learning_rate": 4.8718151419989445e-05,
"loss": 1.1298,
"step": 3840
},
{
"epoch": 0.3076,
"grad_norm": 1.5250157117843628,
"learning_rate": 4.8714839153483823e-05,
"loss": 1.2049,
"step": 3845
},
{
"epoch": 0.308,
"grad_norm": 1.4276963472366333,
"learning_rate": 4.8711522725981276e-05,
"loss": 1.1412,
"step": 3850
},
{
"epoch": 0.3084,
"grad_norm": 1.514461636543274,
"learning_rate": 4.8708202138063695e-05,
"loss": 1.1392,
"step": 3855
},
{
"epoch": 0.3088,
"grad_norm": 1.6108927726745605,
"learning_rate": 4.8704877390313716e-05,
"loss": 1.201,
"step": 3860
},
{
"epoch": 0.3092,
"grad_norm": 1.9335185289382935,
"learning_rate": 4.870154848331469e-05,
"loss": 1.159,
"step": 3865
},
{
"epoch": 0.3096,
"grad_norm": 1.754297137260437,
"learning_rate": 4.86982154176507e-05,
"loss": 1.1432,
"step": 3870
},
{
"epoch": 0.31,
"grad_norm": 1.6593637466430664,
"learning_rate": 4.869487819390657e-05,
"loss": 1.2357,
"step": 3875
},
{
"epoch": 0.3104,
"grad_norm": 1.3409785032272339,
"learning_rate": 4.8691536812667856e-05,
"loss": 1.1105,
"step": 3880
},
{
"epoch": 0.3108,
"grad_norm": 1.4268968105316162,
"learning_rate": 4.8688191274520825e-05,
"loss": 1.1857,
"step": 3885
},
{
"epoch": 0.3112,
"grad_norm": 1.5844037532806396,
"learning_rate": 4.8684841580052496e-05,
"loss": 1.1087,
"step": 3890
},
{
"epoch": 0.3116,
"grad_norm": 1.4313236474990845,
"learning_rate": 4.86814877298506e-05,
"loss": 1.2033,
"step": 3895
},
{
"epoch": 0.312,
"grad_norm": 1.4828399419784546,
"learning_rate": 4.8678129724503584e-05,
"loss": 1.1876,
"step": 3900
},
{
"epoch": 0.3124,
"grad_norm": 1.517876148223877,
"learning_rate": 4.867476756460067e-05,
"loss": 1.2316,
"step": 3905
},
{
"epoch": 0.3128,
"grad_norm": 1.5256012678146362,
"learning_rate": 4.867140125073177e-05,
"loss": 1.2264,
"step": 3910
},
{
"epoch": 0.3132,
"grad_norm": 1.4679909944534302,
"learning_rate": 4.866803078348753e-05,
"loss": 1.1672,
"step": 3915
},
{
"epoch": 0.3136,
"grad_norm": 1.3713257312774658,
"learning_rate": 4.866465616345934e-05,
"loss": 1.2037,
"step": 3920
},
{
"epoch": 0.314,
"grad_norm": 1.5898433923721313,
"learning_rate": 4.8661277391239306e-05,
"loss": 1.122,
"step": 3925
},
{
"epoch": 0.3144,
"grad_norm": 1.6622995138168335,
"learning_rate": 4.8657894467420276e-05,
"loss": 1.1442,
"step": 3930
},
{
"epoch": 0.3148,
"grad_norm": 1.6419644355773926,
"learning_rate": 4.865450739259579e-05,
"loss": 1.1647,
"step": 3935
},
{
"epoch": 0.3152,
"grad_norm": 1.4513815641403198,
"learning_rate": 4.865111616736017e-05,
"loss": 1.1515,
"step": 3940
},
{
"epoch": 0.3156,
"grad_norm": 1.7421190738677979,
"learning_rate": 4.8647720792308425e-05,
"loss": 1.2137,
"step": 3945
},
{
"epoch": 0.316,
"grad_norm": 1.4491236209869385,
"learning_rate": 4.8644321268036314e-05,
"loss": 1.1752,
"step": 3950
},
{
"epoch": 0.3164,
"grad_norm": 1.8071634769439697,
"learning_rate": 4.8640917595140314e-05,
"loss": 1.1821,
"step": 3955
},
{
"epoch": 0.3168,
"grad_norm": 1.7549405097961426,
"learning_rate": 4.863750977421762e-05,
"loss": 1.2574,
"step": 3960
},
{
"epoch": 0.3172,
"grad_norm": 1.4997444152832031,
"learning_rate": 4.8634097805866185e-05,
"loss": 1.1735,
"step": 3965
},
{
"epoch": 0.3176,
"grad_norm": 1.421034336090088,
"learning_rate": 4.863068169068467e-05,
"loss": 1.1464,
"step": 3970
},
{
"epoch": 0.318,
"grad_norm": 1.5631002187728882,
"learning_rate": 4.862726142927245e-05,
"loss": 1.2239,
"step": 3975
},
{
"epoch": 0.3184,
"grad_norm": 1.3816300630569458,
"learning_rate": 4.8623837022229656e-05,
"loss": 1.2619,
"step": 3980
},
{
"epoch": 0.3188,
"grad_norm": 1.437118649482727,
"learning_rate": 4.862040847015713e-05,
"loss": 1.1265,
"step": 3985
},
{
"epoch": 0.3192,
"grad_norm": 1.6327508687973022,
"learning_rate": 4.8616975773656445e-05,
"loss": 1.1522,
"step": 3990
},
{
"epoch": 0.3196,
"grad_norm": 1.5986108779907227,
"learning_rate": 4.8613538933329906e-05,
"loss": 1.1543,
"step": 3995
},
{
"epoch": 0.32,
"grad_norm": 1.4712469577789307,
"learning_rate": 4.861009794978053e-05,
"loss": 1.1363,
"step": 4000
},
{
"epoch": 0.3204,
"grad_norm": 1.5345813035964966,
"learning_rate": 4.860665282361207e-05,
"loss": 1.2319,
"step": 4005
},
{
"epoch": 0.3208,
"grad_norm": 1.425037145614624,
"learning_rate": 4.860320355542901e-05,
"loss": 1.2544,
"step": 4010
},
{
"epoch": 0.3212,
"grad_norm": 1.7960505485534668,
"learning_rate": 4.859975014583657e-05,
"loss": 1.1738,
"step": 4015
},
{
"epoch": 0.3216,
"grad_norm": 1.6393851041793823,
"learning_rate": 4.8596292595440654e-05,
"loss": 1.154,
"step": 4020
},
{
"epoch": 0.322,
"grad_norm": 1.3774423599243164,
"learning_rate": 4.8592830904847956e-05,
"loss": 1.1987,
"step": 4025
},
{
"epoch": 0.3224,
"grad_norm": 1.455966591835022,
"learning_rate": 4.8589365074665836e-05,
"loss": 1.1879,
"step": 4030
},
{
"epoch": 0.3228,
"grad_norm": 1.541073203086853,
"learning_rate": 4.858589510550243e-05,
"loss": 1.174,
"step": 4035
},
{
"epoch": 0.3232,
"grad_norm": 1.5856127738952637,
"learning_rate": 4.858242099796656e-05,
"loss": 1.2052,
"step": 4040
},
{
"epoch": 0.3236,
"grad_norm": 1.6346274614334106,
"learning_rate": 4.85789427526678e-05,
"loss": 1.1631,
"step": 4045
},
{
"epoch": 0.324,
"grad_norm": 1.6323809623718262,
"learning_rate": 4.8575460370216444e-05,
"loss": 1.1798,
"step": 4050
},
{
"epoch": 0.3244,
"grad_norm": 1.6759973764419556,
"learning_rate": 4.85719738512235e-05,
"loss": 1.2269,
"step": 4055
},
{
"epoch": 0.3248,
"grad_norm": 1.4718966484069824,
"learning_rate": 4.856848319630072e-05,
"loss": 1.1528,
"step": 4060
},
{
"epoch": 0.3252,
"grad_norm": 1.486632227897644,
"learning_rate": 4.856498840606056e-05,
"loss": 1.1459,
"step": 4065
},
{
"epoch": 0.3256,
"grad_norm": 1.5227134227752686,
"learning_rate": 4.8561489481116234e-05,
"loss": 1.2153,
"step": 4070
},
{
"epoch": 0.326,
"grad_norm": 1.5857396125793457,
"learning_rate": 4.8557986422081654e-05,
"loss": 1.1898,
"step": 4075
},
{
"epoch": 0.3264,
"grad_norm": 1.5316146612167358,
"learning_rate": 4.855447922957146e-05,
"loss": 1.1493,
"step": 4080
},
{
"epoch": 0.3268,
"grad_norm": 1.4671624898910522,
"learning_rate": 4.855096790420103e-05,
"loss": 1.1801,
"step": 4085
},
{
"epoch": 0.3272,
"grad_norm": 1.4884071350097656,
"learning_rate": 4.854745244658645e-05,
"loss": 1.1678,
"step": 4090
},
{
"epoch": 0.3276,
"grad_norm": 1.5596617460250854,
"learning_rate": 4.8543932857344545e-05,
"loss": 1.1672,
"step": 4095
},
{
"epoch": 0.328,
"grad_norm": 1.524793267250061,
"learning_rate": 4.854040913709287e-05,
"loss": 1.199,
"step": 4100
},
{
"epoch": 0.3284,
"grad_norm": 1.5925273895263672,
"learning_rate": 4.8536881286449684e-05,
"loss": 1.148,
"step": 4105
},
{
"epoch": 0.3288,
"grad_norm": 1.3768569231033325,
"learning_rate": 4.853334930603399e-05,
"loss": 1.1311,
"step": 4110
},
{
"epoch": 0.3292,
"grad_norm": 1.8312355279922485,
"learning_rate": 4.852981319646551e-05,
"loss": 1.1435,
"step": 4115
},
{
"epoch": 0.3296,
"grad_norm": 1.4960983991622925,
"learning_rate": 4.8526272958364674e-05,
"loss": 1.2346,
"step": 4120
},
{
"epoch": 0.33,
"grad_norm": 1.4610507488250732,
"learning_rate": 4.852272859235267e-05,
"loss": 1.145,
"step": 4125
},
{
"epoch": 0.3304,
"grad_norm": 1.437524437904358,
"learning_rate": 4.8519180099051375e-05,
"loss": 1.1523,
"step": 4130
},
{
"epoch": 0.3308,
"grad_norm": 1.7281017303466797,
"learning_rate": 4.8515627479083415e-05,
"loss": 1.1713,
"step": 4135
},
{
"epoch": 0.3312,
"grad_norm": 1.7618118524551392,
"learning_rate": 4.851207073307213e-05,
"loss": 1.2128,
"step": 4140
},
{
"epoch": 0.3316,
"grad_norm": 1.6273146867752075,
"learning_rate": 4.8508509861641585e-05,
"loss": 1.1961,
"step": 4145
},
{
"epoch": 0.332,
"grad_norm": 1.548926830291748,
"learning_rate": 4.850494486541657e-05,
"loss": 1.2366,
"step": 4150
},
{
"epoch": 0.3324,
"grad_norm": 1.465455412864685,
"learning_rate": 4.8501375745022595e-05,
"loss": 1.1614,
"step": 4155
},
{
"epoch": 0.3328,
"grad_norm": 1.7306607961654663,
"learning_rate": 4.8497802501085895e-05,
"loss": 1.2314,
"step": 4160
},
{
"epoch": 0.3332,
"grad_norm": 1.5650298595428467,
"learning_rate": 4.8494225134233434e-05,
"loss": 1.1667,
"step": 4165
},
{
"epoch": 0.3336,
"grad_norm": 1.6374757289886475,
"learning_rate": 4.84906436450929e-05,
"loss": 1.158,
"step": 4170
},
{
"epoch": 0.334,
"grad_norm": 1.8572663068771362,
"learning_rate": 4.848705803429269e-05,
"loss": 1.2126,
"step": 4175
},
{
"epoch": 0.3344,
"grad_norm": 2.0607059001922607,
"learning_rate": 4.8483468302461943e-05,
"loss": 1.256,
"step": 4180
},
{
"epoch": 0.3348,
"grad_norm": 1.6313749551773071,
"learning_rate": 4.84798744502305e-05,
"loss": 1.2197,
"step": 4185
},
{
"epoch": 0.3352,
"grad_norm": 1.4325757026672363,
"learning_rate": 4.847627647822896e-05,
"loss": 1.1577,
"step": 4190
},
{
"epoch": 0.3356,
"grad_norm": 1.4574649333953857,
"learning_rate": 4.847267438708859e-05,
"loss": 1.1881,
"step": 4195
},
{
"epoch": 0.336,
"grad_norm": 1.6589447259902954,
"learning_rate": 4.8469068177441435e-05,
"loss": 1.1362,
"step": 4200
},
{
"epoch": 0.3364,
"grad_norm": 1.4486385583877563,
"learning_rate": 4.8465457849920226e-05,
"loss": 1.1839,
"step": 4205
},
{
"epoch": 0.3368,
"grad_norm": 1.6947062015533447,
"learning_rate": 4.8461843405158446e-05,
"loss": 1.1522,
"step": 4210
},
{
"epoch": 0.3372,
"grad_norm": 2.0725297927856445,
"learning_rate": 4.845822484379027e-05,
"loss": 1.1357,
"step": 4215
},
{
"epoch": 0.3376,
"grad_norm": 1.6981180906295776,
"learning_rate": 4.8454602166450615e-05,
"loss": 1.2319,
"step": 4220
},
{
"epoch": 0.338,
"grad_norm": 1.5875322818756104,
"learning_rate": 4.8450975373775107e-05,
"loss": 1.167,
"step": 4225
},
{
"epoch": 0.3384,
"grad_norm": 1.6823283433914185,
"learning_rate": 4.844734446640011e-05,
"loss": 1.1926,
"step": 4230
},
{
"epoch": 0.3388,
"grad_norm": 1.5389896631240845,
"learning_rate": 4.844370944496271e-05,
"loss": 1.1144,
"step": 4235
},
{
"epoch": 0.3392,
"grad_norm": 1.5896904468536377,
"learning_rate": 4.844007031010068e-05,
"loss": 1.1776,
"step": 4240
},
{
"epoch": 0.3396,
"grad_norm": 1.4404852390289307,
"learning_rate": 4.843642706245257e-05,
"loss": 1.2146,
"step": 4245
},
{
"epoch": 0.34,
"grad_norm": 1.3411272764205933,
"learning_rate": 4.84327797026576e-05,
"loss": 1.1603,
"step": 4250
},
{
"epoch": 0.3404,
"grad_norm": 1.722702980041504,
"learning_rate": 4.842912823135575e-05,
"loss": 1.2679,
"step": 4255
},
{
"epoch": 0.3408,
"grad_norm": 1.692683219909668,
"learning_rate": 4.84254726491877e-05,
"loss": 1.1717,
"step": 4260
},
{
"epoch": 0.3412,
"grad_norm": 1.814393162727356,
"learning_rate": 4.842181295679486e-05,
"loss": 1.1264,
"step": 4265
},
{
"epoch": 0.3416,
"grad_norm": 1.421974539756775,
"learning_rate": 4.841814915481935e-05,
"loss": 1.1899,
"step": 4270
},
{
"epoch": 0.342,
"grad_norm": 1.519344449043274,
"learning_rate": 4.841448124390403e-05,
"loss": 1.2029,
"step": 4275
},
{
"epoch": 0.3424,
"grad_norm": 1.3807154893875122,
"learning_rate": 4.841080922469247e-05,
"loss": 1.1751,
"step": 4280
},
{
"epoch": 0.3428,
"grad_norm": 1.5611778497695923,
"learning_rate": 4.840713309782895e-05,
"loss": 1.1801,
"step": 4285
},
{
"epoch": 0.3432,
"grad_norm": 1.7587612867355347,
"learning_rate": 4.840345286395849e-05,
"loss": 1.1458,
"step": 4290
},
{
"epoch": 0.3436,
"grad_norm": 2.1552114486694336,
"learning_rate": 4.839976852372682e-05,
"loss": 1.1537,
"step": 4295
},
{
"epoch": 0.344,
"grad_norm": 1.483056902885437,
"learning_rate": 4.8396080077780394e-05,
"loss": 1.1974,
"step": 4300
},
{
"epoch": 0.3444,
"grad_norm": 1.3012534379959106,
"learning_rate": 4.839238752676638e-05,
"loss": 1.2229,
"step": 4305
},
{
"epoch": 0.3448,
"grad_norm": 1.6024060249328613,
"learning_rate": 4.838869087133269e-05,
"loss": 1.1466,
"step": 4310
},
{
"epoch": 0.3452,
"grad_norm": 2.0319061279296875,
"learning_rate": 4.8384990112127915e-05,
"loss": 1.1838,
"step": 4315
},
{
"epoch": 0.3456,
"grad_norm": 1.35572350025177,
"learning_rate": 4.83812852498014e-05,
"loss": 1.1846,
"step": 4320
},
{
"epoch": 0.346,
"grad_norm": 1.7080681324005127,
"learning_rate": 4.83775762850032e-05,
"loss": 1.151,
"step": 4325
},
{
"epoch": 0.3464,
"grad_norm": 1.3965811729431152,
"learning_rate": 4.837386321838408e-05,
"loss": 1.3165,
"step": 4330
},
{
"epoch": 0.3468,
"grad_norm": 1.673601746559143,
"learning_rate": 4.837014605059555e-05,
"loss": 1.2241,
"step": 4335
},
{
"epoch": 0.3472,
"grad_norm": 1.2754968404769897,
"learning_rate": 4.83664247822898e-05,
"loss": 1.1173,
"step": 4340
},
{
"epoch": 0.3476,
"grad_norm": 1.8693870306015015,
"learning_rate": 4.836269941411978e-05,
"loss": 1.1775,
"step": 4345
},
{
"epoch": 0.348,
"grad_norm": 1.5695033073425293,
"learning_rate": 4.835896994673914e-05,
"loss": 1.1773,
"step": 4350
},
{
"epoch": 0.3484,
"grad_norm": 1.314810037612915,
"learning_rate": 4.835523638080224e-05,
"loss": 1.2331,
"step": 4355
},
{
"epoch": 0.3488,
"grad_norm": 1.2816500663757324,
"learning_rate": 4.835149871696418e-05,
"loss": 1.1959,
"step": 4360
},
{
"epoch": 0.3492,
"grad_norm": 1.5477250814437866,
"learning_rate": 4.8347756955880765e-05,
"loss": 1.2451,
"step": 4365
},
{
"epoch": 0.3496,
"grad_norm": 1.6063194274902344,
"learning_rate": 4.834401109820853e-05,
"loss": 1.1905,
"step": 4370
},
{
"epoch": 0.35,
"grad_norm": 1.7513134479522705,
"learning_rate": 4.834026114460471e-05,
"loss": 1.2092,
"step": 4375
},
{
"epoch": 0.3504,
"grad_norm": 1.578026533126831,
"learning_rate": 4.833650709572728e-05,
"loss": 1.1645,
"step": 4380
},
{
"epoch": 0.3508,
"grad_norm": 1.597808599472046,
"learning_rate": 4.833274895223492e-05,
"loss": 1.2233,
"step": 4385
},
{
"epoch": 0.3512,
"grad_norm": 1.4159451723098755,
"learning_rate": 4.8328986714787034e-05,
"loss": 1.1685,
"step": 4390
},
{
"epoch": 0.3516,
"grad_norm": 1.6277220249176025,
"learning_rate": 4.832522038404374e-05,
"loss": 1.2059,
"step": 4395
},
{
"epoch": 0.352,
"grad_norm": 1.6037828922271729,
"learning_rate": 4.832144996066588e-05,
"loss": 1.1512,
"step": 4400
}
],
"logging_steps": 5,
"max_steps": 37500,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 100,
"total_flos": 6.908065048162468e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}