ChatLaw2E_plain_7B / trainer_state.json
chestnutlzj's picture
Upload folder using huggingface_hub
17758cf verified
Raw
History Blame Contribute Delete
343 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.9999067555596999,
"eval_steps": 500,
"global_step": 10724,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 5.524007747080166,
"learning_rate": 9.999994636280297e-06,
"loss": 1.823,
"step": 5
},
{
"epoch": 0.0,
"grad_norm": 3.751582456194241,
"learning_rate": 9.999978545132696e-06,
"loss": 1.3441,
"step": 10
},
{
"epoch": 0.0,
"grad_norm": 3.3490766596115806,
"learning_rate": 9.999951726591716e-06,
"loss": 1.3161,
"step": 15
},
{
"epoch": 0.0,
"grad_norm": 3.2220183943018883,
"learning_rate": 9.999914180714902e-06,
"loss": 1.3822,
"step": 20
},
{
"epoch": 0.0,
"grad_norm": 3.7449433236910914,
"learning_rate": 9.999865907582806e-06,
"loss": 1.4161,
"step": 25
},
{
"epoch": 0.01,
"grad_norm": 4.13884314404206,
"learning_rate": 9.999806907298994e-06,
"loss": 1.325,
"step": 30
},
{
"epoch": 0.01,
"grad_norm": 3.1735066858647407,
"learning_rate": 9.999737179990057e-06,
"loss": 1.3145,
"step": 35
},
{
"epoch": 0.01,
"grad_norm": 4.044320591378858,
"learning_rate": 9.999656725805588e-06,
"loss": 1.3288,
"step": 40
},
{
"epoch": 0.01,
"grad_norm": 2.945673848593178,
"learning_rate": 9.999565544918204e-06,
"loss": 1.3304,
"step": 45
},
{
"epoch": 0.01,
"grad_norm": 3.0178952115872986,
"learning_rate": 9.99946363752353e-06,
"loss": 1.302,
"step": 50
},
{
"epoch": 0.01,
"grad_norm": 2.908785287683549,
"learning_rate": 9.999351003840209e-06,
"loss": 1.3062,
"step": 55
},
{
"epoch": 0.01,
"grad_norm": 3.1033835338024884,
"learning_rate": 9.999227644109894e-06,
"loss": 1.3064,
"step": 60
},
{
"epoch": 0.01,
"grad_norm": 3.061377074716208,
"learning_rate": 9.999093558597254e-06,
"loss": 1.325,
"step": 65
},
{
"epoch": 0.01,
"grad_norm": 2.336953651662858,
"learning_rate": 9.998948747589966e-06,
"loss": 1.2808,
"step": 70
},
{
"epoch": 0.01,
"grad_norm": 2.9142448487551142,
"learning_rate": 9.998793211398721e-06,
"loss": 1.3047,
"step": 75
},
{
"epoch": 0.01,
"grad_norm": 2.6943468676524334,
"learning_rate": 9.998626950357218e-06,
"loss": 1.3182,
"step": 80
},
{
"epoch": 0.02,
"grad_norm": 2.979676737020957,
"learning_rate": 9.99844996482217e-06,
"loss": 1.312,
"step": 85
},
{
"epoch": 0.02,
"grad_norm": 3.4798768099858544,
"learning_rate": 9.998262255173298e-06,
"loss": 1.279,
"step": 90
},
{
"epoch": 0.02,
"grad_norm": 3.034106416493002,
"learning_rate": 9.998063821813328e-06,
"loss": 1.3215,
"step": 95
},
{
"epoch": 0.02,
"grad_norm": 3.1680260266966953,
"learning_rate": 9.997854665168001e-06,
"loss": 1.3095,
"step": 100
},
{
"epoch": 0.02,
"grad_norm": 2.9582365796044368,
"learning_rate": 9.997634785686054e-06,
"loss": 1.3602,
"step": 105
},
{
"epoch": 0.02,
"grad_norm": 3.4201546672414294,
"learning_rate": 9.997404183839241e-06,
"loss": 1.3675,
"step": 110
},
{
"epoch": 0.02,
"grad_norm": 3.2960605448255067,
"learning_rate": 9.997162860122313e-06,
"loss": 1.3517,
"step": 115
},
{
"epoch": 0.02,
"grad_norm": 2.774691284711409,
"learning_rate": 9.996910815053027e-06,
"loss": 1.3123,
"step": 120
},
{
"epoch": 0.02,
"grad_norm": 2.7634540383113446,
"learning_rate": 9.996648049172143e-06,
"loss": 1.2638,
"step": 125
},
{
"epoch": 0.02,
"grad_norm": 2.6841122836813094,
"learning_rate": 9.996374563043422e-06,
"loss": 1.3476,
"step": 130
},
{
"epoch": 0.03,
"grad_norm": 3.2063540307974288,
"learning_rate": 9.996090357253625e-06,
"loss": 1.3489,
"step": 135
},
{
"epoch": 0.03,
"grad_norm": 2.5801773637103023,
"learning_rate": 9.995795432412513e-06,
"loss": 1.3069,
"step": 140
},
{
"epoch": 0.03,
"grad_norm": 2.860331572440809,
"learning_rate": 9.995489789152844e-06,
"loss": 1.3342,
"step": 145
},
{
"epoch": 0.03,
"grad_norm": 2.9984719758607397,
"learning_rate": 9.99517342813037e-06,
"loss": 1.3705,
"step": 150
},
{
"epoch": 0.03,
"grad_norm": 3.230814648227745,
"learning_rate": 9.994846350023842e-06,
"loss": 1.3816,
"step": 155
},
{
"epoch": 0.03,
"grad_norm": 2.6882982691959696,
"learning_rate": 9.994508555534999e-06,
"loss": 1.2859,
"step": 160
},
{
"epoch": 0.03,
"grad_norm": 2.7701863434403355,
"learning_rate": 9.994160045388576e-06,
"loss": 1.264,
"step": 165
},
{
"epoch": 0.03,
"grad_norm": 2.837077069983557,
"learning_rate": 9.9938008203323e-06,
"loss": 1.2557,
"step": 170
},
{
"epoch": 0.03,
"grad_norm": 2.7592433235464915,
"learning_rate": 9.993430881136883e-06,
"loss": 1.3556,
"step": 175
},
{
"epoch": 0.03,
"grad_norm": 2.618947907449918,
"learning_rate": 9.993050228596021e-06,
"loss": 1.3118,
"step": 180
},
{
"epoch": 0.03,
"grad_norm": 3.2327219520967563,
"learning_rate": 9.992658863526405e-06,
"loss": 1.2953,
"step": 185
},
{
"epoch": 0.04,
"grad_norm": 2.917865999401656,
"learning_rate": 9.992256786767702e-06,
"loss": 1.3146,
"step": 190
},
{
"epoch": 0.04,
"grad_norm": 2.611540955710515,
"learning_rate": 9.99184399918256e-06,
"loss": 1.3002,
"step": 195
},
{
"epoch": 0.04,
"grad_norm": 2.975342658891056,
"learning_rate": 9.991420501656615e-06,
"loss": 1.2903,
"step": 200
},
{
"epoch": 0.04,
"grad_norm": 3.1866127183127277,
"learning_rate": 9.990986295098472e-06,
"loss": 1.3178,
"step": 205
},
{
"epoch": 0.04,
"grad_norm": 3.1225984655510652,
"learning_rate": 9.990541380439716e-06,
"loss": 1.3263,
"step": 210
},
{
"epoch": 0.04,
"grad_norm": 2.4693895609084078,
"learning_rate": 9.990085758634907e-06,
"loss": 1.2732,
"step": 215
},
{
"epoch": 0.04,
"grad_norm": 3.529179960181463,
"learning_rate": 9.989619430661576e-06,
"loss": 1.2542,
"step": 220
},
{
"epoch": 0.04,
"grad_norm": 2.769607377162276,
"learning_rate": 9.989142397520225e-06,
"loss": 1.2751,
"step": 225
},
{
"epoch": 0.04,
"grad_norm": 2.6943100281562233,
"learning_rate": 9.98865466023432e-06,
"loss": 1.3003,
"step": 230
},
{
"epoch": 0.04,
"grad_norm": 3.418579661198453,
"learning_rate": 9.9881562198503e-06,
"loss": 1.2373,
"step": 235
},
{
"epoch": 0.04,
"grad_norm": 5.452061410443576,
"learning_rate": 9.987647077437559e-06,
"loss": 1.2977,
"step": 240
},
{
"epoch": 0.05,
"grad_norm": 2.718344890262994,
"learning_rate": 9.987127234088456e-06,
"loss": 1.2952,
"step": 245
},
{
"epoch": 0.05,
"grad_norm": 2.872673940218012,
"learning_rate": 9.986596690918308e-06,
"loss": 1.297,
"step": 250
},
{
"epoch": 0.05,
"grad_norm": 4.033704916837394,
"learning_rate": 9.986055449065394e-06,
"loss": 1.2778,
"step": 255
},
{
"epoch": 0.05,
"grad_norm": 2.6759149455624804,
"learning_rate": 9.985503509690937e-06,
"loss": 1.2974,
"step": 260
},
{
"epoch": 0.05,
"grad_norm": 2.9791514906697585,
"learning_rate": 9.984940873979116e-06,
"loss": 1.3069,
"step": 265
},
{
"epoch": 0.05,
"grad_norm": 2.7904291459421575,
"learning_rate": 9.984367543137062e-06,
"loss": 1.2749,
"step": 270
},
{
"epoch": 0.05,
"grad_norm": 2.8139332500165652,
"learning_rate": 9.98378351839485e-06,
"loss": 1.2622,
"step": 275
},
{
"epoch": 0.05,
"grad_norm": 3.057619432752788,
"learning_rate": 9.983188801005492e-06,
"loss": 1.2835,
"step": 280
},
{
"epoch": 0.05,
"grad_norm": 2.794308837656347,
"learning_rate": 9.982583392244954e-06,
"loss": 1.278,
"step": 285
},
{
"epoch": 0.05,
"grad_norm": 3.0284278220138034,
"learning_rate": 9.981967293412128e-06,
"loss": 1.2832,
"step": 290
},
{
"epoch": 0.06,
"grad_norm": 2.5950675242644285,
"learning_rate": 9.98134050582885e-06,
"loss": 1.2421,
"step": 295
},
{
"epoch": 0.06,
"grad_norm": 3.0925652328309123,
"learning_rate": 9.980703030839884e-06,
"loss": 1.2869,
"step": 300
},
{
"epoch": 0.06,
"grad_norm": 2.6030703692470243,
"learning_rate": 9.980054869812923e-06,
"loss": 1.2463,
"step": 305
},
{
"epoch": 0.06,
"grad_norm": 2.7199426489230674,
"learning_rate": 9.979396024138593e-06,
"loss": 1.2865,
"step": 310
},
{
"epoch": 0.06,
"grad_norm": 2.8797052662634712,
"learning_rate": 9.978726495230434e-06,
"loss": 1.3188,
"step": 315
},
{
"epoch": 0.06,
"grad_norm": 2.8935014338459344,
"learning_rate": 9.978046284524917e-06,
"loss": 1.3377,
"step": 320
},
{
"epoch": 0.06,
"grad_norm": 2.779176704862691,
"learning_rate": 9.977355393481423e-06,
"loss": 1.3213,
"step": 325
},
{
"epoch": 0.06,
"grad_norm": 2.6820244249317375,
"learning_rate": 9.976653823582253e-06,
"loss": 1.259,
"step": 330
},
{
"epoch": 0.06,
"grad_norm": 3.0357143479356665,
"learning_rate": 9.975941576332611e-06,
"loss": 1.2904,
"step": 335
},
{
"epoch": 0.06,
"grad_norm": 2.8019467918942915,
"learning_rate": 9.975218653260621e-06,
"loss": 1.3113,
"step": 340
},
{
"epoch": 0.06,
"grad_norm": 2.3048841868601495,
"learning_rate": 9.974485055917303e-06,
"loss": 1.3132,
"step": 345
},
{
"epoch": 0.07,
"grad_norm": 2.5081602045955065,
"learning_rate": 9.973740785876582e-06,
"loss": 1.3331,
"step": 350
},
{
"epoch": 0.07,
"grad_norm": 2.679225954751873,
"learning_rate": 9.97298584473528e-06,
"loss": 1.3206,
"step": 355
},
{
"epoch": 0.07,
"grad_norm": 2.7120364585174075,
"learning_rate": 9.972220234113114e-06,
"loss": 1.2441,
"step": 360
},
{
"epoch": 0.07,
"grad_norm": 3.0678883501219705,
"learning_rate": 9.971443955652692e-06,
"loss": 1.2323,
"step": 365
},
{
"epoch": 0.07,
"grad_norm": 2.5544818482029843,
"learning_rate": 9.97065701101951e-06,
"loss": 1.2432,
"step": 370
},
{
"epoch": 0.07,
"grad_norm": 2.6731233037881257,
"learning_rate": 9.969859401901948e-06,
"loss": 1.2668,
"step": 375
},
{
"epoch": 0.07,
"grad_norm": 2.804128919360195,
"learning_rate": 9.969051130011268e-06,
"loss": 1.3494,
"step": 380
},
{
"epoch": 0.07,
"grad_norm": 3.079258704000447,
"learning_rate": 9.968232197081609e-06,
"loss": 1.2553,
"step": 385
},
{
"epoch": 0.07,
"grad_norm": 2.7829735470494263,
"learning_rate": 9.967402604869976e-06,
"loss": 1.3145,
"step": 390
},
{
"epoch": 0.07,
"grad_norm": 2.8363957159184623,
"learning_rate": 9.966562355156254e-06,
"loss": 1.2802,
"step": 395
},
{
"epoch": 0.07,
"grad_norm": 3.090152084812277,
"learning_rate": 9.965711449743187e-06,
"loss": 1.3289,
"step": 400
},
{
"epoch": 0.08,
"grad_norm": 2.8212887732560223,
"learning_rate": 9.964849890456382e-06,
"loss": 1.2879,
"step": 405
},
{
"epoch": 0.08,
"grad_norm": 3.2716434155310394,
"learning_rate": 9.963977679144304e-06,
"loss": 1.311,
"step": 410
},
{
"epoch": 0.08,
"grad_norm": 2.8169159980839464,
"learning_rate": 9.963094817678273e-06,
"loss": 1.2731,
"step": 415
},
{
"epoch": 0.08,
"grad_norm": 3.1670684683630705,
"learning_rate": 9.962201307952455e-06,
"loss": 1.2683,
"step": 420
},
{
"epoch": 0.08,
"grad_norm": 2.62866536431656,
"learning_rate": 9.961297151883869e-06,
"loss": 1.2799,
"step": 425
},
{
"epoch": 0.08,
"grad_norm": 3.362938035312287,
"learning_rate": 9.960382351412365e-06,
"loss": 1.3027,
"step": 430
},
{
"epoch": 0.08,
"grad_norm": 2.6267772082519825,
"learning_rate": 9.95945690850064e-06,
"loss": 1.2544,
"step": 435
},
{
"epoch": 0.08,
"grad_norm": 2.773463036278164,
"learning_rate": 9.958520825134221e-06,
"loss": 1.2887,
"step": 440
},
{
"epoch": 0.08,
"grad_norm": 2.911616224637602,
"learning_rate": 9.957574103321462e-06,
"loss": 1.2402,
"step": 445
},
{
"epoch": 0.08,
"grad_norm": 2.395016720915871,
"learning_rate": 9.956616745093541e-06,
"loss": 1.2492,
"step": 450
},
{
"epoch": 0.08,
"grad_norm": 2.4830984531007423,
"learning_rate": 9.955648752504463e-06,
"loss": 1.243,
"step": 455
},
{
"epoch": 0.09,
"grad_norm": 2.8437350034534052,
"learning_rate": 9.954670127631042e-06,
"loss": 1.3052,
"step": 460
},
{
"epoch": 0.09,
"grad_norm": 2.794782444316208,
"learning_rate": 9.953680872572906e-06,
"loss": 1.2655,
"step": 465
},
{
"epoch": 0.09,
"grad_norm": 2.622763840389913,
"learning_rate": 9.95268098945249e-06,
"loss": 1.3403,
"step": 470
},
{
"epoch": 0.09,
"grad_norm": 2.963439619215708,
"learning_rate": 9.95167048041503e-06,
"loss": 1.3105,
"step": 475
},
{
"epoch": 0.09,
"grad_norm": 2.6699317361231447,
"learning_rate": 9.950649347628564e-06,
"loss": 1.255,
"step": 480
},
{
"epoch": 0.09,
"grad_norm": 2.909028049137236,
"learning_rate": 9.949617593283916e-06,
"loss": 1.3104,
"step": 485
},
{
"epoch": 0.09,
"grad_norm": 2.5792044118119613,
"learning_rate": 9.948575219594704e-06,
"loss": 1.2582,
"step": 490
},
{
"epoch": 0.09,
"grad_norm": 2.981241252777049,
"learning_rate": 9.94752222879733e-06,
"loss": 1.3307,
"step": 495
},
{
"epoch": 0.09,
"grad_norm": 2.8980802115787023,
"learning_rate": 9.94645862315097e-06,
"loss": 1.2995,
"step": 500
},
{
"epoch": 0.09,
"grad_norm": 2.7552380730072485,
"learning_rate": 9.945384404937581e-06,
"loss": 1.2996,
"step": 505
},
{
"epoch": 0.1,
"grad_norm": 3.182612212241565,
"learning_rate": 9.94429957646188e-06,
"loss": 1.3282,
"step": 510
},
{
"epoch": 0.1,
"grad_norm": 2.893316970193462,
"learning_rate": 9.943204140051357e-06,
"loss": 1.289,
"step": 515
},
{
"epoch": 0.1,
"grad_norm": 3.0853061089675307,
"learning_rate": 9.942098098056255e-06,
"loss": 1.3116,
"step": 520
},
{
"epoch": 0.1,
"grad_norm": 3.2597008125913147,
"learning_rate": 9.940981452849577e-06,
"loss": 1.2695,
"step": 525
},
{
"epoch": 0.1,
"grad_norm": 2.803486275623243,
"learning_rate": 9.939854206827069e-06,
"loss": 1.2988,
"step": 530
},
{
"epoch": 0.1,
"grad_norm": 2.7856894140636626,
"learning_rate": 9.938716362407226e-06,
"loss": 1.2213,
"step": 535
},
{
"epoch": 0.1,
"grad_norm": 3.1781516944496344,
"learning_rate": 9.937567922031277e-06,
"loss": 1.2921,
"step": 540
},
{
"epoch": 0.1,
"grad_norm": 2.70081195002388,
"learning_rate": 9.936408888163187e-06,
"loss": 1.2939,
"step": 545
},
{
"epoch": 0.1,
"grad_norm": 2.571513269019906,
"learning_rate": 9.93523926328965e-06,
"loss": 1.2787,
"step": 550
},
{
"epoch": 0.1,
"grad_norm": 2.5347818693158395,
"learning_rate": 9.934059049920085e-06,
"loss": 1.2396,
"step": 555
},
{
"epoch": 0.1,
"grad_norm": 2.927026090981205,
"learning_rate": 9.932868250586619e-06,
"loss": 1.3322,
"step": 560
},
{
"epoch": 0.11,
"grad_norm": 3.0072023467515714,
"learning_rate": 9.931666867844103e-06,
"loss": 1.2505,
"step": 565
},
{
"epoch": 0.11,
"grad_norm": 2.5798486811625816,
"learning_rate": 9.930454904270087e-06,
"loss": 1.3168,
"step": 570
},
{
"epoch": 0.11,
"grad_norm": 2.606273714286121,
"learning_rate": 9.929232362464825e-06,
"loss": 1.2929,
"step": 575
},
{
"epoch": 0.11,
"grad_norm": 2.862057851885314,
"learning_rate": 9.927999245051263e-06,
"loss": 1.2221,
"step": 580
},
{
"epoch": 0.11,
"grad_norm": 3.0533100345061666,
"learning_rate": 9.926755554675043e-06,
"loss": 1.2824,
"step": 585
},
{
"epoch": 0.11,
"grad_norm": 2.8268549184677196,
"learning_rate": 9.925501294004485e-06,
"loss": 1.2689,
"step": 590
},
{
"epoch": 0.11,
"grad_norm": 2.9474244570654515,
"learning_rate": 9.924236465730592e-06,
"loss": 1.266,
"step": 595
},
{
"epoch": 0.11,
"grad_norm": 2.5171703874147315,
"learning_rate": 9.922961072567037e-06,
"loss": 1.2945,
"step": 600
},
{
"epoch": 0.11,
"grad_norm": 2.8323155471229073,
"learning_rate": 9.92167511725016e-06,
"loss": 1.2497,
"step": 605
},
{
"epoch": 0.11,
"grad_norm": 2.6443126596211632,
"learning_rate": 9.920378602538963e-06,
"loss": 1.2388,
"step": 610
},
{
"epoch": 0.11,
"grad_norm": 2.77248510868743,
"learning_rate": 9.919071531215104e-06,
"loss": 1.2655,
"step": 615
},
{
"epoch": 0.12,
"grad_norm": 3.526397598316819,
"learning_rate": 9.917753906082885e-06,
"loss": 1.302,
"step": 620
},
{
"epoch": 0.12,
"grad_norm": 3.015236540785339,
"learning_rate": 9.916425729969259e-06,
"loss": 1.2072,
"step": 625
},
{
"epoch": 0.12,
"grad_norm": 2.8038695983344937,
"learning_rate": 9.915087005723809e-06,
"loss": 1.2924,
"step": 630
},
{
"epoch": 0.12,
"grad_norm": 2.5898365904275034,
"learning_rate": 9.913737736218751e-06,
"loss": 1.2357,
"step": 635
},
{
"epoch": 0.12,
"grad_norm": 2.4969045444120566,
"learning_rate": 9.912377924348931e-06,
"loss": 1.2601,
"step": 640
},
{
"epoch": 0.12,
"grad_norm": 2.8599851914324357,
"learning_rate": 9.911007573031803e-06,
"loss": 1.2779,
"step": 645
},
{
"epoch": 0.12,
"grad_norm": 2.53590231137642,
"learning_rate": 9.909626685207444e-06,
"loss": 1.2425,
"step": 650
},
{
"epoch": 0.12,
"grad_norm": 2.7844674380752457,
"learning_rate": 9.908235263838529e-06,
"loss": 1.2628,
"step": 655
},
{
"epoch": 0.12,
"grad_norm": 2.921537190563611,
"learning_rate": 9.906833311910335e-06,
"loss": 1.2918,
"step": 660
},
{
"epoch": 0.12,
"grad_norm": 2.674269569684917,
"learning_rate": 9.905420832430736e-06,
"loss": 1.2317,
"step": 665
},
{
"epoch": 0.12,
"grad_norm": 2.7972975778914493,
"learning_rate": 9.903997828430188e-06,
"loss": 1.2706,
"step": 670
},
{
"epoch": 0.13,
"grad_norm": 3.0857826068674234,
"learning_rate": 9.902564302961727e-06,
"loss": 1.2953,
"step": 675
},
{
"epoch": 0.13,
"grad_norm": 2.353182975912364,
"learning_rate": 9.901120259100969e-06,
"loss": 1.3039,
"step": 680
},
{
"epoch": 0.13,
"grad_norm": 2.5050793033453953,
"learning_rate": 9.89966569994609e-06,
"loss": 1.2548,
"step": 685
},
{
"epoch": 0.13,
"grad_norm": 2.8182846692065966,
"learning_rate": 9.898200628617829e-06,
"loss": 1.3022,
"step": 690
},
{
"epoch": 0.13,
"grad_norm": 3.6265366655340845,
"learning_rate": 9.896725048259478e-06,
"loss": 1.241,
"step": 695
},
{
"epoch": 0.13,
"grad_norm": 2.6434221137993625,
"learning_rate": 9.895238962036878e-06,
"loss": 1.275,
"step": 700
},
{
"epoch": 0.13,
"grad_norm": 2.5173535448609283,
"learning_rate": 9.893742373138408e-06,
"loss": 1.2568,
"step": 705
},
{
"epoch": 0.13,
"grad_norm": 2.7542295966472348,
"learning_rate": 9.892235284774985e-06,
"loss": 1.2868,
"step": 710
},
{
"epoch": 0.13,
"grad_norm": 2.368577938425068,
"learning_rate": 9.890717700180043e-06,
"loss": 1.2245,
"step": 715
},
{
"epoch": 0.13,
"grad_norm": 2.5278687487866884,
"learning_rate": 9.889189622609545e-06,
"loss": 1.2583,
"step": 720
},
{
"epoch": 0.14,
"grad_norm": 2.702147632129278,
"learning_rate": 9.887651055341961e-06,
"loss": 1.2428,
"step": 725
},
{
"epoch": 0.14,
"grad_norm": 2.35390476449974,
"learning_rate": 9.886102001678271e-06,
"loss": 1.282,
"step": 730
},
{
"epoch": 0.14,
"grad_norm": 3.0658442781629756,
"learning_rate": 9.884542464941948e-06,
"loss": 1.29,
"step": 735
},
{
"epoch": 0.14,
"grad_norm": 3.5482489583081733,
"learning_rate": 9.882972448478962e-06,
"loss": 1.2043,
"step": 740
},
{
"epoch": 0.14,
"grad_norm": 2.8661887660145897,
"learning_rate": 9.881391955657762e-06,
"loss": 1.2325,
"step": 745
},
{
"epoch": 0.14,
"grad_norm": 2.7066006373702063,
"learning_rate": 9.879800989869277e-06,
"loss": 1.2436,
"step": 750
},
{
"epoch": 0.14,
"grad_norm": 2.7536179864413666,
"learning_rate": 9.878199554526904e-06,
"loss": 1.2886,
"step": 755
},
{
"epoch": 0.14,
"grad_norm": 2.4756211385632683,
"learning_rate": 9.876587653066504e-06,
"loss": 1.3146,
"step": 760
},
{
"epoch": 0.14,
"grad_norm": 2.678588724807592,
"learning_rate": 9.874965288946395e-06,
"loss": 1.2565,
"step": 765
},
{
"epoch": 0.14,
"grad_norm": 2.6248055714027645,
"learning_rate": 9.873332465647333e-06,
"loss": 1.2838,
"step": 770
},
{
"epoch": 0.14,
"grad_norm": 2.933089391512721,
"learning_rate": 9.871689186672527e-06,
"loss": 1.3046,
"step": 775
},
{
"epoch": 0.15,
"grad_norm": 2.8054595245249194,
"learning_rate": 9.870035455547607e-06,
"loss": 1.3532,
"step": 780
},
{
"epoch": 0.15,
"grad_norm": 3.029754362133909,
"learning_rate": 9.868371275820636e-06,
"loss": 1.2358,
"step": 785
},
{
"epoch": 0.15,
"grad_norm": 2.381829122673505,
"learning_rate": 9.866696651062092e-06,
"loss": 1.2673,
"step": 790
},
{
"epoch": 0.15,
"grad_norm": 2.9578049809020284,
"learning_rate": 9.86501158486486e-06,
"loss": 1.3057,
"step": 795
},
{
"epoch": 0.15,
"grad_norm": 3.2079875550888652,
"learning_rate": 9.86331608084423e-06,
"loss": 1.2836,
"step": 800
},
{
"epoch": 0.15,
"grad_norm": 2.7394380027483316,
"learning_rate": 9.861610142637885e-06,
"loss": 1.2967,
"step": 805
},
{
"epoch": 0.15,
"grad_norm": 2.757020347162176,
"learning_rate": 9.859893773905897e-06,
"loss": 1.327,
"step": 810
},
{
"epoch": 0.15,
"grad_norm": 2.7386946320924723,
"learning_rate": 9.858166978330711e-06,
"loss": 1.3324,
"step": 815
},
{
"epoch": 0.15,
"grad_norm": 2.487600070733838,
"learning_rate": 9.856429759617148e-06,
"loss": 1.243,
"step": 820
},
{
"epoch": 0.15,
"grad_norm": 2.4859094980688643,
"learning_rate": 9.854682121492388e-06,
"loss": 1.2884,
"step": 825
},
{
"epoch": 0.15,
"grad_norm": 2.671185046699724,
"learning_rate": 9.852924067705968e-06,
"loss": 1.2814,
"step": 830
},
{
"epoch": 0.16,
"grad_norm": 2.842082690479725,
"learning_rate": 9.851155602029774e-06,
"loss": 1.1763,
"step": 835
},
{
"epoch": 0.16,
"grad_norm": 2.590003326493348,
"learning_rate": 9.849376728258024e-06,
"loss": 1.2732,
"step": 840
},
{
"epoch": 0.16,
"grad_norm": 3.755687197612357,
"learning_rate": 9.847587450207273e-06,
"loss": 1.3001,
"step": 845
},
{
"epoch": 0.16,
"grad_norm": 3.078220599557541,
"learning_rate": 9.845787771716391e-06,
"loss": 1.297,
"step": 850
},
{
"epoch": 0.16,
"grad_norm": 2.8978172028530995,
"learning_rate": 9.843977696646572e-06,
"loss": 1.2412,
"step": 855
},
{
"epoch": 0.16,
"grad_norm": 2.5403657098108647,
"learning_rate": 9.842157228881305e-06,
"loss": 1.2634,
"step": 860
},
{
"epoch": 0.16,
"grad_norm": 2.95924694930617,
"learning_rate": 9.840326372326386e-06,
"loss": 1.2423,
"step": 865
},
{
"epoch": 0.16,
"grad_norm": 2.709630812317427,
"learning_rate": 9.838485130909892e-06,
"loss": 1.3462,
"step": 870
},
{
"epoch": 0.16,
"grad_norm": 2.8155472202298384,
"learning_rate": 9.836633508582185e-06,
"loss": 1.2516,
"step": 875
},
{
"epoch": 0.16,
"grad_norm": 2.500070860342406,
"learning_rate": 9.834771509315899e-06,
"loss": 1.2849,
"step": 880
},
{
"epoch": 0.17,
"grad_norm": 2.3553653961885845,
"learning_rate": 9.832899137105931e-06,
"loss": 1.2972,
"step": 885
},
{
"epoch": 0.17,
"grad_norm": 2.739911552383276,
"learning_rate": 9.831016395969432e-06,
"loss": 1.2898,
"step": 890
},
{
"epoch": 0.17,
"grad_norm": 2.7979814034906996,
"learning_rate": 9.829123289945803e-06,
"loss": 1.2352,
"step": 895
},
{
"epoch": 0.17,
"grad_norm": 2.7756637703151266,
"learning_rate": 9.827219823096677e-06,
"loss": 1.2838,
"step": 900
},
{
"epoch": 0.17,
"grad_norm": 3.147380497985643,
"learning_rate": 9.82530599950592e-06,
"loss": 1.3009,
"step": 905
},
{
"epoch": 0.17,
"grad_norm": 2.6195588988426817,
"learning_rate": 9.823381823279617e-06,
"loss": 1.2542,
"step": 910
},
{
"epoch": 0.17,
"grad_norm": 3.0294691695407687,
"learning_rate": 9.821447298546063e-06,
"loss": 1.2529,
"step": 915
},
{
"epoch": 0.17,
"grad_norm": 3.0406219437265136,
"learning_rate": 9.819502429455762e-06,
"loss": 1.2788,
"step": 920
},
{
"epoch": 0.17,
"grad_norm": 2.751447227490008,
"learning_rate": 9.817547220181404e-06,
"loss": 1.2474,
"step": 925
},
{
"epoch": 0.17,
"grad_norm": 2.7920501109572946,
"learning_rate": 9.815581674917866e-06,
"loss": 1.2426,
"step": 930
},
{
"epoch": 0.17,
"grad_norm": 2.702602399095983,
"learning_rate": 9.813605797882204e-06,
"loss": 1.2792,
"step": 935
},
{
"epoch": 0.18,
"grad_norm": 2.5213037151582913,
"learning_rate": 9.811619593313636e-06,
"loss": 1.2396,
"step": 940
},
{
"epoch": 0.18,
"grad_norm": 2.5410701879815263,
"learning_rate": 9.809623065473544e-06,
"loss": 1.332,
"step": 945
},
{
"epoch": 0.18,
"grad_norm": 2.754615306176169,
"learning_rate": 9.807616218645448e-06,
"loss": 1.3013,
"step": 950
},
{
"epoch": 0.18,
"grad_norm": 3.0527201906235515,
"learning_rate": 9.805599057135017e-06,
"loss": 1.2555,
"step": 955
},
{
"epoch": 0.18,
"grad_norm": 2.7211770353555464,
"learning_rate": 9.803571585270047e-06,
"loss": 1.2898,
"step": 960
},
{
"epoch": 0.18,
"grad_norm": 2.8451338452669073,
"learning_rate": 9.801533807400455e-06,
"loss": 1.3269,
"step": 965
},
{
"epoch": 0.18,
"grad_norm": 2.375893987239719,
"learning_rate": 9.799485727898265e-06,
"loss": 1.2695,
"step": 970
},
{
"epoch": 0.18,
"grad_norm": 2.749006570276283,
"learning_rate": 9.79742735115761e-06,
"loss": 1.2496,
"step": 975
},
{
"epoch": 0.18,
"grad_norm": 2.505279680209459,
"learning_rate": 9.795358681594712e-06,
"loss": 1.2445,
"step": 980
},
{
"epoch": 0.18,
"grad_norm": 2.582391358707717,
"learning_rate": 9.793279723647874e-06,
"loss": 1.2596,
"step": 985
},
{
"epoch": 0.18,
"grad_norm": 2.5671931470502467,
"learning_rate": 9.79119048177748e-06,
"loss": 1.2719,
"step": 990
},
{
"epoch": 0.19,
"grad_norm": 2.8817921609658854,
"learning_rate": 9.789090960465968e-06,
"loss": 1.2621,
"step": 995
},
{
"epoch": 0.19,
"grad_norm": 3.0026356453735605,
"learning_rate": 9.786981164217838e-06,
"loss": 1.3099,
"step": 1000
},
{
"epoch": 0.19,
"grad_norm": 2.575472157187206,
"learning_rate": 9.784861097559632e-06,
"loss": 1.255,
"step": 1005
},
{
"epoch": 0.19,
"grad_norm": 2.8771804234712754,
"learning_rate": 9.782730765039927e-06,
"loss": 1.2492,
"step": 1010
},
{
"epoch": 0.19,
"grad_norm": 2.7679377180963827,
"learning_rate": 9.780590171229327e-06,
"loss": 1.2866,
"step": 1015
},
{
"epoch": 0.19,
"grad_norm": 2.616337211838346,
"learning_rate": 9.778439320720448e-06,
"loss": 1.2376,
"step": 1020
},
{
"epoch": 0.19,
"grad_norm": 2.482034813414732,
"learning_rate": 9.776278218127914e-06,
"loss": 1.2445,
"step": 1025
},
{
"epoch": 0.19,
"grad_norm": 2.7329915919806758,
"learning_rate": 9.774106868088346e-06,
"loss": 1.2521,
"step": 1030
},
{
"epoch": 0.19,
"grad_norm": 2.680070713070951,
"learning_rate": 9.771925275260348e-06,
"loss": 1.3296,
"step": 1035
},
{
"epoch": 0.19,
"grad_norm": 2.8977960820861597,
"learning_rate": 9.769733444324501e-06,
"loss": 1.2053,
"step": 1040
},
{
"epoch": 0.19,
"grad_norm": 2.977796667138015,
"learning_rate": 9.767531379983353e-06,
"loss": 1.293,
"step": 1045
},
{
"epoch": 0.2,
"grad_norm": 3.4348023333586823,
"learning_rate": 9.765319086961405e-06,
"loss": 1.2839,
"step": 1050
},
{
"epoch": 0.2,
"grad_norm": 2.408646301347671,
"learning_rate": 9.763096570005103e-06,
"loss": 1.2475,
"step": 1055
},
{
"epoch": 0.2,
"grad_norm": 2.4851841489693043,
"learning_rate": 9.760863833882834e-06,
"loss": 1.2663,
"step": 1060
},
{
"epoch": 0.2,
"grad_norm": 2.862755821551652,
"learning_rate": 9.758620883384906e-06,
"loss": 1.3317,
"step": 1065
},
{
"epoch": 0.2,
"grad_norm": 2.4831453533426915,
"learning_rate": 9.756367723323538e-06,
"loss": 1.2525,
"step": 1070
},
{
"epoch": 0.2,
"grad_norm": 2.570370036777222,
"learning_rate": 9.754104358532863e-06,
"loss": 1.2796,
"step": 1075
},
{
"epoch": 0.2,
"grad_norm": 2.583982409024191,
"learning_rate": 9.7518307938689e-06,
"loss": 1.3011,
"step": 1080
},
{
"epoch": 0.2,
"grad_norm": 2.6843240127817904,
"learning_rate": 9.749547034209552e-06,
"loss": 1.2811,
"step": 1085
},
{
"epoch": 0.2,
"grad_norm": 2.3089147503032557,
"learning_rate": 9.747253084454601e-06,
"loss": 1.2071,
"step": 1090
},
{
"epoch": 0.2,
"grad_norm": 2.3584969392089796,
"learning_rate": 9.74494894952569e-06,
"loss": 1.2372,
"step": 1095
},
{
"epoch": 0.21,
"grad_norm": 2.393274524242671,
"learning_rate": 9.742634634366308e-06,
"loss": 1.245,
"step": 1100
},
{
"epoch": 0.21,
"grad_norm": 2.7920124624113787,
"learning_rate": 9.740310143941792e-06,
"loss": 1.3044,
"step": 1105
},
{
"epoch": 0.21,
"grad_norm": 2.8358988005907597,
"learning_rate": 9.73797548323931e-06,
"loss": 1.2091,
"step": 1110
},
{
"epoch": 0.21,
"grad_norm": 2.672057798702065,
"learning_rate": 9.735630657267846e-06,
"loss": 1.2464,
"step": 1115
},
{
"epoch": 0.21,
"grad_norm": 2.9456854326651194,
"learning_rate": 9.733275671058195e-06,
"loss": 1.2838,
"step": 1120
},
{
"epoch": 0.21,
"grad_norm": 2.8896784051857574,
"learning_rate": 9.730910529662954e-06,
"loss": 1.2381,
"step": 1125
},
{
"epoch": 0.21,
"grad_norm": 2.769582071620601,
"learning_rate": 9.728535238156504e-06,
"loss": 1.3079,
"step": 1130
},
{
"epoch": 0.21,
"grad_norm": 2.4653295633356302,
"learning_rate": 9.726149801635002e-06,
"loss": 1.278,
"step": 1135
},
{
"epoch": 0.21,
"grad_norm": 2.941156988269561,
"learning_rate": 9.723754225216378e-06,
"loss": 1.2622,
"step": 1140
},
{
"epoch": 0.21,
"grad_norm": 2.637713701891722,
"learning_rate": 9.721348514040307e-06,
"loss": 1.2717,
"step": 1145
},
{
"epoch": 0.21,
"grad_norm": 2.7517262009540175,
"learning_rate": 9.718932673268218e-06,
"loss": 1.2582,
"step": 1150
},
{
"epoch": 0.22,
"grad_norm": 2.9577500419309435,
"learning_rate": 9.716506708083264e-06,
"loss": 1.2748,
"step": 1155
},
{
"epoch": 0.22,
"grad_norm": 3.090524775310035,
"learning_rate": 9.714070623690326e-06,
"loss": 1.2764,
"step": 1160
},
{
"epoch": 0.22,
"grad_norm": 2.3932563623472327,
"learning_rate": 9.711624425315993e-06,
"loss": 1.2321,
"step": 1165
},
{
"epoch": 0.22,
"grad_norm": 2.8559504713397175,
"learning_rate": 9.709168118208553e-06,
"loss": 1.2493,
"step": 1170
},
{
"epoch": 0.22,
"grad_norm": 2.7551996217047545,
"learning_rate": 9.706701707637984e-06,
"loss": 1.2613,
"step": 1175
},
{
"epoch": 0.22,
"grad_norm": 2.966463340002555,
"learning_rate": 9.704225198895943e-06,
"loss": 1.2348,
"step": 1180
},
{
"epoch": 0.22,
"grad_norm": 2.5657038224244744,
"learning_rate": 9.701738597295743e-06,
"loss": 1.2136,
"step": 1185
},
{
"epoch": 0.22,
"grad_norm": 2.5586347071410445,
"learning_rate": 9.699241908172364e-06,
"loss": 1.2914,
"step": 1190
},
{
"epoch": 0.22,
"grad_norm": 2.5434709087968197,
"learning_rate": 9.696735136882419e-06,
"loss": 1.2659,
"step": 1195
},
{
"epoch": 0.22,
"grad_norm": 2.7061026705057185,
"learning_rate": 9.694218288804154e-06,
"loss": 1.2609,
"step": 1200
},
{
"epoch": 0.22,
"grad_norm": 2.396398638179384,
"learning_rate": 9.691691369337439e-06,
"loss": 1.2483,
"step": 1205
},
{
"epoch": 0.23,
"grad_norm": 2.438971895626567,
"learning_rate": 9.689154383903749e-06,
"loss": 1.2568,
"step": 1210
},
{
"epoch": 0.23,
"grad_norm": 2.801969219455209,
"learning_rate": 9.686607337946155e-06,
"loss": 1.189,
"step": 1215
},
{
"epoch": 0.23,
"grad_norm": 2.715494965332097,
"learning_rate": 9.684050236929309e-06,
"loss": 1.2282,
"step": 1220
},
{
"epoch": 0.23,
"grad_norm": 2.43406850408372,
"learning_rate": 9.681483086339448e-06,
"loss": 1.2654,
"step": 1225
},
{
"epoch": 0.23,
"grad_norm": 2.637848687572683,
"learning_rate": 9.678905891684359e-06,
"loss": 1.2365,
"step": 1230
},
{
"epoch": 0.23,
"grad_norm": 2.816634445745627,
"learning_rate": 9.676318658493378e-06,
"loss": 1.2892,
"step": 1235
},
{
"epoch": 0.23,
"grad_norm": 2.59188291753139,
"learning_rate": 9.673721392317388e-06,
"loss": 1.2818,
"step": 1240
},
{
"epoch": 0.23,
"grad_norm": 2.6589472617669214,
"learning_rate": 9.67111409872879e-06,
"loss": 1.2525,
"step": 1245
},
{
"epoch": 0.23,
"grad_norm": 2.418049084852354,
"learning_rate": 9.668496783321499e-06,
"loss": 1.2528,
"step": 1250
},
{
"epoch": 0.23,
"grad_norm": 2.6134422197628124,
"learning_rate": 9.665869451710938e-06,
"loss": 1.2489,
"step": 1255
},
{
"epoch": 0.23,
"grad_norm": 2.8277369949630207,
"learning_rate": 9.663232109534011e-06,
"loss": 1.2425,
"step": 1260
},
{
"epoch": 0.24,
"grad_norm": 2.6677083847071454,
"learning_rate": 9.660584762449106e-06,
"loss": 1.2528,
"step": 1265
},
{
"epoch": 0.24,
"grad_norm": 2.901015204481418,
"learning_rate": 9.657927416136072e-06,
"loss": 1.2856,
"step": 1270
},
{
"epoch": 0.24,
"grad_norm": 2.783653207223188,
"learning_rate": 9.655260076296213e-06,
"loss": 1.2194,
"step": 1275
},
{
"epoch": 0.24,
"grad_norm": 2.5728825133639455,
"learning_rate": 9.652582748652278e-06,
"loss": 1.2669,
"step": 1280
},
{
"epoch": 0.24,
"grad_norm": 2.730634926238819,
"learning_rate": 9.649895438948438e-06,
"loss": 1.2591,
"step": 1285
},
{
"epoch": 0.24,
"grad_norm": 2.7910418990912684,
"learning_rate": 9.647198152950285e-06,
"loss": 1.2911,
"step": 1290
},
{
"epoch": 0.24,
"grad_norm": 2.6279219550939485,
"learning_rate": 9.644490896444812e-06,
"loss": 1.2364,
"step": 1295
},
{
"epoch": 0.24,
"grad_norm": 2.6929143816353007,
"learning_rate": 9.641773675240408e-06,
"loss": 1.2472,
"step": 1300
},
{
"epoch": 0.24,
"grad_norm": 2.7126488640148536,
"learning_rate": 9.639046495166833e-06,
"loss": 1.2981,
"step": 1305
},
{
"epoch": 0.24,
"grad_norm": 2.275713238425623,
"learning_rate": 9.636309362075222e-06,
"loss": 1.3011,
"step": 1310
},
{
"epoch": 0.25,
"grad_norm": 2.813094238557557,
"learning_rate": 9.63356228183806e-06,
"loss": 1.2446,
"step": 1315
},
{
"epoch": 0.25,
"grad_norm": 3.0356903685468164,
"learning_rate": 9.630805260349176e-06,
"loss": 1.3118,
"step": 1320
},
{
"epoch": 0.25,
"grad_norm": 2.6472645381899635,
"learning_rate": 9.628038303523725e-06,
"loss": 1.2682,
"step": 1325
},
{
"epoch": 0.25,
"grad_norm": 2.512006211451414,
"learning_rate": 9.625261417298179e-06,
"loss": 1.2472,
"step": 1330
},
{
"epoch": 0.25,
"grad_norm": 2.488773622482499,
"learning_rate": 9.622474607630315e-06,
"loss": 1.2376,
"step": 1335
},
{
"epoch": 0.25,
"grad_norm": 2.685018435556931,
"learning_rate": 9.619677880499196e-06,
"loss": 1.3117,
"step": 1340
},
{
"epoch": 0.25,
"grad_norm": 2.5059232088783756,
"learning_rate": 9.61687124190517e-06,
"loss": 1.2424,
"step": 1345
},
{
"epoch": 0.25,
"grad_norm": 3.0030330704239327,
"learning_rate": 9.614054697869846e-06,
"loss": 1.2325,
"step": 1350
},
{
"epoch": 0.25,
"grad_norm": 2.7185700535574324,
"learning_rate": 9.611228254436082e-06,
"loss": 1.2709,
"step": 1355
},
{
"epoch": 0.25,
"grad_norm": 2.6550063124078167,
"learning_rate": 9.60839191766798e-06,
"loss": 1.2554,
"step": 1360
},
{
"epoch": 0.25,
"grad_norm": 2.3924990329834506,
"learning_rate": 9.605545693650869e-06,
"loss": 1.2411,
"step": 1365
},
{
"epoch": 0.26,
"grad_norm": 2.9253114263002726,
"learning_rate": 9.602689588491284e-06,
"loss": 1.2349,
"step": 1370
},
{
"epoch": 0.26,
"grad_norm": 3.1305134098499097,
"learning_rate": 9.599823608316962e-06,
"loss": 1.1982,
"step": 1375
},
{
"epoch": 0.26,
"grad_norm": 2.494931249541086,
"learning_rate": 9.596947759276835e-06,
"loss": 1.2799,
"step": 1380
},
{
"epoch": 0.26,
"grad_norm": 2.3982225168241644,
"learning_rate": 9.594062047541e-06,
"loss": 1.2257,
"step": 1385
},
{
"epoch": 0.26,
"grad_norm": 2.586775286217663,
"learning_rate": 9.591166479300714e-06,
"loss": 1.2032,
"step": 1390
},
{
"epoch": 0.26,
"grad_norm": 2.797391466557641,
"learning_rate": 9.588261060768388e-06,
"loss": 1.1992,
"step": 1395
},
{
"epoch": 0.26,
"grad_norm": 3.2692807339475976,
"learning_rate": 9.585345798177557e-06,
"loss": 1.2955,
"step": 1400
},
{
"epoch": 0.26,
"grad_norm": 2.4762035990234392,
"learning_rate": 9.582420697782884e-06,
"loss": 1.328,
"step": 1405
},
{
"epoch": 0.26,
"grad_norm": 2.88773258550614,
"learning_rate": 9.579485765860138e-06,
"loss": 1.2713,
"step": 1410
},
{
"epoch": 0.26,
"grad_norm": 2.537654166999561,
"learning_rate": 9.576541008706177e-06,
"loss": 1.2273,
"step": 1415
},
{
"epoch": 0.26,
"grad_norm": 2.5099612429483527,
"learning_rate": 9.573586432638947e-06,
"loss": 1.251,
"step": 1420
},
{
"epoch": 0.27,
"grad_norm": 2.4565468231653558,
"learning_rate": 9.570622043997447e-06,
"loss": 1.2887,
"step": 1425
},
{
"epoch": 0.27,
"grad_norm": 2.7085428069726394,
"learning_rate": 9.567647849141744e-06,
"loss": 1.2744,
"step": 1430
},
{
"epoch": 0.27,
"grad_norm": 2.649022261193647,
"learning_rate": 9.564663854452933e-06,
"loss": 1.3101,
"step": 1435
},
{
"epoch": 0.27,
"grad_norm": 2.584187921682549,
"learning_rate": 9.561670066333142e-06,
"loss": 1.2482,
"step": 1440
},
{
"epoch": 0.27,
"grad_norm": 2.6769977815059693,
"learning_rate": 9.558666491205503e-06,
"loss": 1.2361,
"step": 1445
},
{
"epoch": 0.27,
"grad_norm": 2.463200002777827,
"learning_rate": 9.555653135514152e-06,
"loss": 1.2859,
"step": 1450
},
{
"epoch": 0.27,
"grad_norm": 2.8051780394700536,
"learning_rate": 9.552630005724209e-06,
"loss": 1.2653,
"step": 1455
},
{
"epoch": 0.27,
"grad_norm": 2.7025858761551316,
"learning_rate": 9.549597108321759e-06,
"loss": 1.1986,
"step": 1460
},
{
"epoch": 0.27,
"grad_norm": 2.448947238174015,
"learning_rate": 9.546554449813847e-06,
"loss": 1.281,
"step": 1465
},
{
"epoch": 0.27,
"grad_norm": 2.583763069734913,
"learning_rate": 9.543502036728463e-06,
"loss": 1.2564,
"step": 1470
},
{
"epoch": 0.28,
"grad_norm": 2.8131660018770575,
"learning_rate": 9.54043987561452e-06,
"loss": 1.2581,
"step": 1475
},
{
"epoch": 0.28,
"grad_norm": 2.611621933823635,
"learning_rate": 9.537367973041847e-06,
"loss": 1.244,
"step": 1480
},
{
"epoch": 0.28,
"grad_norm": 2.725299231142621,
"learning_rate": 9.534286335601175e-06,
"loss": 1.2628,
"step": 1485
},
{
"epoch": 0.28,
"grad_norm": 2.5387805053467734,
"learning_rate": 9.53119496990412e-06,
"loss": 1.2373,
"step": 1490
},
{
"epoch": 0.28,
"grad_norm": 2.5030631132275074,
"learning_rate": 9.52809388258317e-06,
"loss": 1.307,
"step": 1495
},
{
"epoch": 0.28,
"grad_norm": 2.867164940207714,
"learning_rate": 9.524983080291667e-06,
"loss": 1.247,
"step": 1500
},
{
"epoch": 0.28,
"grad_norm": 2.7835274007016695,
"learning_rate": 9.521862569703802e-06,
"loss": 1.2873,
"step": 1505
},
{
"epoch": 0.28,
"grad_norm": 2.4127815440210476,
"learning_rate": 9.518732357514592e-06,
"loss": 1.2288,
"step": 1510
},
{
"epoch": 0.28,
"grad_norm": 2.490311452425188,
"learning_rate": 9.515592450439872e-06,
"loss": 1.2596,
"step": 1515
},
{
"epoch": 0.28,
"grad_norm": 2.5853819055443448,
"learning_rate": 9.512442855216274e-06,
"loss": 1.2479,
"step": 1520
},
{
"epoch": 0.28,
"grad_norm": 2.387948927852224,
"learning_rate": 9.509283578601212e-06,
"loss": 1.2669,
"step": 1525
},
{
"epoch": 0.29,
"grad_norm": 3.0927210433687904,
"learning_rate": 9.50611462737288e-06,
"loss": 1.2585,
"step": 1530
},
{
"epoch": 0.29,
"grad_norm": 2.467133574190019,
"learning_rate": 9.502936008330222e-06,
"loss": 1.2901,
"step": 1535
},
{
"epoch": 0.29,
"grad_norm": 2.6739108212719427,
"learning_rate": 9.499747728292928e-06,
"loss": 1.2252,
"step": 1540
},
{
"epoch": 0.29,
"grad_norm": 2.5205586130534834,
"learning_rate": 9.496549794101413e-06,
"loss": 1.2726,
"step": 1545
},
{
"epoch": 0.29,
"grad_norm": 2.604524536073613,
"learning_rate": 9.493342212616807e-06,
"loss": 1.2466,
"step": 1550
},
{
"epoch": 0.29,
"grad_norm": 2.4525722518831237,
"learning_rate": 9.490124990720936e-06,
"loss": 1.286,
"step": 1555
},
{
"epoch": 0.29,
"grad_norm": 2.56958620162423,
"learning_rate": 9.486898135316315e-06,
"loss": 1.2298,
"step": 1560
},
{
"epoch": 0.29,
"grad_norm": 2.9199217646471243,
"learning_rate": 9.483661653326118e-06,
"loss": 1.2662,
"step": 1565
},
{
"epoch": 0.29,
"grad_norm": 2.764973994546948,
"learning_rate": 9.48041555169418e-06,
"loss": 1.2296,
"step": 1570
},
{
"epoch": 0.29,
"grad_norm": 2.758405862424743,
"learning_rate": 9.477159837384973e-06,
"loss": 1.2885,
"step": 1575
},
{
"epoch": 0.29,
"grad_norm": 2.8682635603750635,
"learning_rate": 9.47389451738359e-06,
"loss": 1.2252,
"step": 1580
},
{
"epoch": 0.3,
"grad_norm": 2.796661944415381,
"learning_rate": 9.470619598695739e-06,
"loss": 1.3119,
"step": 1585
},
{
"epoch": 0.3,
"grad_norm": 2.188147528697614,
"learning_rate": 9.467335088347717e-06,
"loss": 1.2391,
"step": 1590
},
{
"epoch": 0.3,
"grad_norm": 2.779350124162447,
"learning_rate": 9.4640409933864e-06,
"loss": 1.2321,
"step": 1595
},
{
"epoch": 0.3,
"grad_norm": 2.530631252303528,
"learning_rate": 9.460737320879232e-06,
"loss": 1.2691,
"step": 1600
},
{
"epoch": 0.3,
"grad_norm": 2.418655276721412,
"learning_rate": 9.4574240779142e-06,
"loss": 1.2405,
"step": 1605
},
{
"epoch": 0.3,
"grad_norm": 2.5254333263722573,
"learning_rate": 9.454101271599823e-06,
"loss": 1.2432,
"step": 1610
},
{
"epoch": 0.3,
"grad_norm": 2.617128013213717,
"learning_rate": 9.45076890906515e-06,
"loss": 1.2873,
"step": 1615
},
{
"epoch": 0.3,
"grad_norm": 2.4365283243913654,
"learning_rate": 9.447426997459717e-06,
"loss": 1.2496,
"step": 1620
},
{
"epoch": 0.3,
"grad_norm": 2.659554267254664,
"learning_rate": 9.444075543953559e-06,
"loss": 1.2188,
"step": 1625
},
{
"epoch": 0.3,
"grad_norm": 2.962413555940077,
"learning_rate": 9.440714555737177e-06,
"loss": 1.3516,
"step": 1630
},
{
"epoch": 0.3,
"grad_norm": 2.5097728308245077,
"learning_rate": 9.437344040021532e-06,
"loss": 1.2091,
"step": 1635
},
{
"epoch": 0.31,
"grad_norm": 2.659238025294085,
"learning_rate": 9.433964004038023e-06,
"loss": 1.2308,
"step": 1640
},
{
"epoch": 0.31,
"grad_norm": 2.4819889409692504,
"learning_rate": 9.430574455038477e-06,
"loss": 1.1989,
"step": 1645
},
{
"epoch": 0.31,
"grad_norm": 2.6088722399364057,
"learning_rate": 9.427175400295132e-06,
"loss": 1.2804,
"step": 1650
},
{
"epoch": 0.31,
"grad_norm": 2.7713314708309613,
"learning_rate": 9.423766847100615e-06,
"loss": 1.269,
"step": 1655
},
{
"epoch": 0.31,
"grad_norm": 2.4942039927782065,
"learning_rate": 9.420348802767939e-06,
"loss": 1.2649,
"step": 1660
},
{
"epoch": 0.31,
"grad_norm": 2.5025712121307104,
"learning_rate": 9.416921274630474e-06,
"loss": 1.3082,
"step": 1665
},
{
"epoch": 0.31,
"grad_norm": 2.5408229636779085,
"learning_rate": 9.413484270041942e-06,
"loss": 1.2438,
"step": 1670
},
{
"epoch": 0.31,
"grad_norm": 2.945770428536271,
"learning_rate": 9.410037796376395e-06,
"loss": 1.2459,
"step": 1675
},
{
"epoch": 0.31,
"grad_norm": 2.6674498043140993,
"learning_rate": 9.406581861028199e-06,
"loss": 1.2352,
"step": 1680
},
{
"epoch": 0.31,
"grad_norm": 2.8933373067619743,
"learning_rate": 9.40311647141202e-06,
"loss": 1.2846,
"step": 1685
},
{
"epoch": 0.32,
"grad_norm": 2.5378972208134,
"learning_rate": 9.399641634962814e-06,
"loss": 1.2392,
"step": 1690
},
{
"epoch": 0.32,
"grad_norm": 2.905367330524848,
"learning_rate": 9.396157359135797e-06,
"loss": 1.2788,
"step": 1695
},
{
"epoch": 0.32,
"grad_norm": 2.5530777021198086,
"learning_rate": 9.392663651406442e-06,
"loss": 1.2617,
"step": 1700
},
{
"epoch": 0.32,
"grad_norm": 2.7538523238416275,
"learning_rate": 9.389160519270455e-06,
"loss": 1.2558,
"step": 1705
},
{
"epoch": 0.32,
"grad_norm": 2.5680458173433918,
"learning_rate": 9.385647970243765e-06,
"loss": 1.27,
"step": 1710
},
{
"epoch": 0.32,
"grad_norm": 2.980880170661701,
"learning_rate": 9.382126011862504e-06,
"loss": 1.2391,
"step": 1715
},
{
"epoch": 0.32,
"grad_norm": 2.7881725751657913,
"learning_rate": 9.378594651682988e-06,
"loss": 1.2394,
"step": 1720
},
{
"epoch": 0.32,
"grad_norm": 2.514336389736585,
"learning_rate": 9.375053897281712e-06,
"loss": 1.2701,
"step": 1725
},
{
"epoch": 0.32,
"grad_norm": 2.5630163523861142,
"learning_rate": 9.371503756255318e-06,
"loss": 1.2299,
"step": 1730
},
{
"epoch": 0.32,
"grad_norm": 2.8156056466588,
"learning_rate": 9.367944236220592e-06,
"loss": 1.2195,
"step": 1735
},
{
"epoch": 0.32,
"grad_norm": 2.7751652875235866,
"learning_rate": 9.36437534481444e-06,
"loss": 1.2178,
"step": 1740
},
{
"epoch": 0.33,
"grad_norm": 2.56505063504321,
"learning_rate": 9.360797089693876e-06,
"loss": 1.1842,
"step": 1745
},
{
"epoch": 0.33,
"grad_norm": 2.528699332800653,
"learning_rate": 9.357209478536004e-06,
"loss": 1.2836,
"step": 1750
},
{
"epoch": 0.33,
"grad_norm": 2.381664933642811,
"learning_rate": 9.353612519037998e-06,
"loss": 1.2564,
"step": 1755
},
{
"epoch": 0.33,
"grad_norm": 2.2616227243734466,
"learning_rate": 9.350006218917093e-06,
"loss": 1.206,
"step": 1760
},
{
"epoch": 0.33,
"grad_norm": 3.215182355715651,
"learning_rate": 9.346390585910564e-06,
"loss": 1.2162,
"step": 1765
},
{
"epoch": 0.33,
"grad_norm": 2.4068984775708535,
"learning_rate": 9.342765627775701e-06,
"loss": 1.256,
"step": 1770
},
{
"epoch": 0.33,
"grad_norm": 3.111790616048246,
"learning_rate": 9.339131352289813e-06,
"loss": 1.244,
"step": 1775
},
{
"epoch": 0.33,
"grad_norm": 2.497956759599184,
"learning_rate": 9.335487767250196e-06,
"loss": 1.2527,
"step": 1780
},
{
"epoch": 0.33,
"grad_norm": 2.8488643685890254,
"learning_rate": 9.331834880474112e-06,
"loss": 1.2307,
"step": 1785
},
{
"epoch": 0.33,
"grad_norm": 2.5441328642850043,
"learning_rate": 9.328172699798787e-06,
"loss": 1.2508,
"step": 1790
},
{
"epoch": 0.33,
"grad_norm": 2.905091377061353,
"learning_rate": 9.32450123308139e-06,
"loss": 1.2705,
"step": 1795
},
{
"epoch": 0.34,
"grad_norm": 2.441237879674571,
"learning_rate": 9.320820488199001e-06,
"loss": 1.2472,
"step": 1800
},
{
"epoch": 0.34,
"grad_norm": 2.478636338912326,
"learning_rate": 9.317130473048619e-06,
"loss": 1.2011,
"step": 1805
},
{
"epoch": 0.34,
"grad_norm": 2.9292520527141965,
"learning_rate": 9.313431195547125e-06,
"loss": 1.2806,
"step": 1810
},
{
"epoch": 0.34,
"grad_norm": 2.707691316832115,
"learning_rate": 9.309722663631273e-06,
"loss": 1.2255,
"step": 1815
},
{
"epoch": 0.34,
"grad_norm": 2.6886924956046268,
"learning_rate": 9.306004885257675e-06,
"loss": 1.1939,
"step": 1820
},
{
"epoch": 0.34,
"grad_norm": 2.537382921658179,
"learning_rate": 9.302277868402778e-06,
"loss": 1.2764,
"step": 1825
},
{
"epoch": 0.34,
"grad_norm": 2.6757765536638063,
"learning_rate": 9.298541621062852e-06,
"loss": 1.2118,
"step": 1830
},
{
"epoch": 0.34,
"grad_norm": 3.1550215206551084,
"learning_rate": 9.29479615125397e-06,
"loss": 1.2447,
"step": 1835
},
{
"epoch": 0.34,
"grad_norm": 2.438740255473035,
"learning_rate": 9.291041467011993e-06,
"loss": 1.1826,
"step": 1840
},
{
"epoch": 0.34,
"grad_norm": 2.351000695867108,
"learning_rate": 9.287277576392553e-06,
"loss": 1.2527,
"step": 1845
},
{
"epoch": 0.35,
"grad_norm": 2.4930032778748723,
"learning_rate": 9.283504487471025e-06,
"loss": 1.2268,
"step": 1850
},
{
"epoch": 0.35,
"grad_norm": 2.6221734670109296,
"learning_rate": 9.279722208342531e-06,
"loss": 1.2498,
"step": 1855
},
{
"epoch": 0.35,
"grad_norm": 2.498523147654453,
"learning_rate": 9.275930747121904e-06,
"loss": 1.2466,
"step": 1860
},
{
"epoch": 0.35,
"grad_norm": 3.063307098020864,
"learning_rate": 9.272130111943677e-06,
"loss": 1.2658,
"step": 1865
},
{
"epoch": 0.35,
"grad_norm": 2.3256287817017895,
"learning_rate": 9.268320310962067e-06,
"loss": 1.2377,
"step": 1870
},
{
"epoch": 0.35,
"grad_norm": 2.5407897240944775,
"learning_rate": 9.264501352350956e-06,
"loss": 1.2475,
"step": 1875
},
{
"epoch": 0.35,
"grad_norm": 2.5140612955200616,
"learning_rate": 9.260673244303876e-06,
"loss": 1.2515,
"step": 1880
},
{
"epoch": 0.35,
"grad_norm": 2.5361195689296188,
"learning_rate": 9.25683599503398e-06,
"loss": 1.2869,
"step": 1885
},
{
"epoch": 0.35,
"grad_norm": 2.43917003129886,
"learning_rate": 9.252989612774045e-06,
"loss": 1.2741,
"step": 1890
},
{
"epoch": 0.35,
"grad_norm": 2.4002807651389135,
"learning_rate": 9.249134105776438e-06,
"loss": 1.2317,
"step": 1895
},
{
"epoch": 0.35,
"grad_norm": 2.8840395380192585,
"learning_rate": 9.2452694823131e-06,
"loss": 1.2351,
"step": 1900
},
{
"epoch": 0.36,
"grad_norm": 2.4386304087371546,
"learning_rate": 9.241395750675535e-06,
"loss": 1.2785,
"step": 1905
},
{
"epoch": 0.36,
"grad_norm": 2.6751487281863238,
"learning_rate": 9.237512919174784e-06,
"loss": 1.2338,
"step": 1910
},
{
"epoch": 0.36,
"grad_norm": 2.607104816785263,
"learning_rate": 9.233620996141421e-06,
"loss": 1.2751,
"step": 1915
},
{
"epoch": 0.36,
"grad_norm": 2.4848314069202497,
"learning_rate": 9.229719989925516e-06,
"loss": 1.2258,
"step": 1920
},
{
"epoch": 0.36,
"grad_norm": 2.496832970791497,
"learning_rate": 9.22580990889663e-06,
"loss": 1.1904,
"step": 1925
},
{
"epoch": 0.36,
"grad_norm": 2.751264104084154,
"learning_rate": 9.221890761443797e-06,
"loss": 1.2238,
"step": 1930
},
{
"epoch": 0.36,
"grad_norm": 2.361121060540872,
"learning_rate": 9.2179625559755e-06,
"loss": 1.2844,
"step": 1935
},
{
"epoch": 0.36,
"grad_norm": 2.336073008697271,
"learning_rate": 9.214025300919654e-06,
"loss": 1.2414,
"step": 1940
},
{
"epoch": 0.36,
"grad_norm": 2.568163211952336,
"learning_rate": 9.21007900472359e-06,
"loss": 1.2547,
"step": 1945
},
{
"epoch": 0.36,
"grad_norm": 2.065600385985879,
"learning_rate": 9.206123675854046e-06,
"loss": 1.2857,
"step": 1950
},
{
"epoch": 0.36,
"grad_norm": 2.611039306343727,
"learning_rate": 9.202159322797127e-06,
"loss": 1.2271,
"step": 1955
},
{
"epoch": 0.37,
"grad_norm": 2.5230585835313497,
"learning_rate": 9.198185954058305e-06,
"loss": 1.2082,
"step": 1960
},
{
"epoch": 0.37,
"grad_norm": 2.8770021273735913,
"learning_rate": 9.194203578162395e-06,
"loss": 1.2284,
"step": 1965
},
{
"epoch": 0.37,
"grad_norm": 2.981346910969795,
"learning_rate": 9.190212203653536e-06,
"loss": 1.2703,
"step": 1970
},
{
"epoch": 0.37,
"grad_norm": 2.476339506574488,
"learning_rate": 9.186211839095173e-06,
"loss": 1.2289,
"step": 1975
},
{
"epoch": 0.37,
"grad_norm": 2.716079606788581,
"learning_rate": 9.182202493070042e-06,
"loss": 1.2389,
"step": 1980
},
{
"epoch": 0.37,
"grad_norm": 2.665652883562263,
"learning_rate": 9.178184174180144e-06,
"loss": 1.2321,
"step": 1985
},
{
"epoch": 0.37,
"grad_norm": 2.523860194811177,
"learning_rate": 9.174156891046735e-06,
"loss": 1.2759,
"step": 1990
},
{
"epoch": 0.37,
"grad_norm": 2.8324143392540377,
"learning_rate": 9.170120652310302e-06,
"loss": 1.2357,
"step": 1995
},
{
"epoch": 0.37,
"grad_norm": 3.480803939898195,
"learning_rate": 9.166075466630544e-06,
"loss": 1.2853,
"step": 2000
},
{
"epoch": 0.37,
"grad_norm": 2.534748993502781,
"learning_rate": 9.16202134268636e-06,
"loss": 1.2719,
"step": 2005
},
{
"epoch": 0.37,
"grad_norm": 2.270094668164831,
"learning_rate": 9.157958289175823e-06,
"loss": 1.2653,
"step": 2010
},
{
"epoch": 0.38,
"grad_norm": 2.7071336774041996,
"learning_rate": 9.153886314816167e-06,
"loss": 1.2794,
"step": 2015
},
{
"epoch": 0.38,
"grad_norm": 2.5214070416956895,
"learning_rate": 9.14980542834376e-06,
"loss": 1.2486,
"step": 2020
},
{
"epoch": 0.38,
"grad_norm": 2.1348372380758938,
"learning_rate": 9.145715638514097e-06,
"loss": 1.2588,
"step": 2025
},
{
"epoch": 0.38,
"grad_norm": 2.3960872799493917,
"learning_rate": 9.141616954101774e-06,
"loss": 1.2476,
"step": 2030
},
{
"epoch": 0.38,
"grad_norm": 2.3568100815228568,
"learning_rate": 9.137509383900465e-06,
"loss": 1.2592,
"step": 2035
},
{
"epoch": 0.38,
"grad_norm": 2.7326127821248325,
"learning_rate": 9.133392936722914e-06,
"loss": 1.2395,
"step": 2040
},
{
"epoch": 0.38,
"grad_norm": 2.4530898848045104,
"learning_rate": 9.129267621400908e-06,
"loss": 1.2348,
"step": 2045
},
{
"epoch": 0.38,
"grad_norm": 2.5224023797948827,
"learning_rate": 9.125133446785262e-06,
"loss": 1.2263,
"step": 2050
},
{
"epoch": 0.38,
"grad_norm": 2.582762768877811,
"learning_rate": 9.120990421745798e-06,
"loss": 1.2129,
"step": 2055
},
{
"epoch": 0.38,
"grad_norm": 2.687243860565364,
"learning_rate": 9.116838555171324e-06,
"loss": 1.1914,
"step": 2060
},
{
"epoch": 0.39,
"grad_norm": 2.541625235096613,
"learning_rate": 9.112677855969621e-06,
"loss": 1.2614,
"step": 2065
},
{
"epoch": 0.39,
"grad_norm": 2.553790871372534,
"learning_rate": 9.108508333067417e-06,
"loss": 1.2676,
"step": 2070
},
{
"epoch": 0.39,
"grad_norm": 2.5187101896462605,
"learning_rate": 9.104329995410373e-06,
"loss": 1.2615,
"step": 2075
},
{
"epoch": 0.39,
"grad_norm": 2.472597277322467,
"learning_rate": 9.100142851963065e-06,
"loss": 1.2858,
"step": 2080
},
{
"epoch": 0.39,
"grad_norm": 2.6967302147065566,
"learning_rate": 9.095946911708956e-06,
"loss": 1.251,
"step": 2085
},
{
"epoch": 0.39,
"grad_norm": 2.6563237373548394,
"learning_rate": 9.091742183650385e-06,
"loss": 1.2245,
"step": 2090
},
{
"epoch": 0.39,
"grad_norm": 2.8826970023720215,
"learning_rate": 9.087528676808546e-06,
"loss": 1.2095,
"step": 2095
},
{
"epoch": 0.39,
"grad_norm": 2.758147091936185,
"learning_rate": 9.083306400223465e-06,
"loss": 1.2625,
"step": 2100
},
{
"epoch": 0.39,
"grad_norm": 2.708953811891164,
"learning_rate": 9.079075362953986e-06,
"loss": 1.2543,
"step": 2105
},
{
"epoch": 0.39,
"grad_norm": 2.7842323272866554,
"learning_rate": 9.074835574077751e-06,
"loss": 1.1901,
"step": 2110
},
{
"epoch": 0.39,
"grad_norm": 2.3340312098539564,
"learning_rate": 9.07058704269117e-06,
"loss": 1.237,
"step": 2115
},
{
"epoch": 0.4,
"grad_norm": 2.201022209456728,
"learning_rate": 9.066329777909424e-06,
"loss": 1.2393,
"step": 2120
},
{
"epoch": 0.4,
"grad_norm": 2.9721479460201854,
"learning_rate": 9.062063788866413e-06,
"loss": 1.2357,
"step": 2125
},
{
"epoch": 0.4,
"grad_norm": 2.477950514807996,
"learning_rate": 9.057789084714772e-06,
"loss": 1.2783,
"step": 2130
},
{
"epoch": 0.4,
"grad_norm": 2.6801686742333914,
"learning_rate": 9.053505674625824e-06,
"loss": 1.2518,
"step": 2135
},
{
"epoch": 0.4,
"grad_norm": 2.284459164256047,
"learning_rate": 9.049213567789575e-06,
"loss": 1.2178,
"step": 2140
},
{
"epoch": 0.4,
"grad_norm": 2.5271344856024984,
"learning_rate": 9.044912773414686e-06,
"loss": 1.1993,
"step": 2145
},
{
"epoch": 0.4,
"grad_norm": 2.4050161954609566,
"learning_rate": 9.040603300728463e-06,
"loss": 1.2511,
"step": 2150
},
{
"epoch": 0.4,
"grad_norm": 2.7704594988994393,
"learning_rate": 9.036285158976823e-06,
"loss": 1.1996,
"step": 2155
},
{
"epoch": 0.4,
"grad_norm": 2.468996954771303,
"learning_rate": 9.031958357424288e-06,
"loss": 1.2108,
"step": 2160
},
{
"epoch": 0.4,
"grad_norm": 2.68457819828506,
"learning_rate": 9.02762290535396e-06,
"loss": 1.2252,
"step": 2165
},
{
"epoch": 0.4,
"grad_norm": 2.3714659097894177,
"learning_rate": 9.023278812067499e-06,
"loss": 1.2419,
"step": 2170
},
{
"epoch": 0.41,
"grad_norm": 2.8938246248834525,
"learning_rate": 9.018926086885103e-06,
"loss": 1.2567,
"step": 2175
},
{
"epoch": 0.41,
"grad_norm": 2.5030355446737245,
"learning_rate": 9.01456473914549e-06,
"loss": 1.1975,
"step": 2180
},
{
"epoch": 0.41,
"grad_norm": 2.5309960309803676,
"learning_rate": 9.010194778205882e-06,
"loss": 1.3214,
"step": 2185
},
{
"epoch": 0.41,
"grad_norm": 2.392058513782499,
"learning_rate": 9.005816213441974e-06,
"loss": 1.2852,
"step": 2190
},
{
"epoch": 0.41,
"grad_norm": 2.4868847080941237,
"learning_rate": 9.001429054247926e-06,
"loss": 1.2317,
"step": 2195
},
{
"epoch": 0.41,
"grad_norm": 3.0581127639032872,
"learning_rate": 8.997033310036335e-06,
"loss": 1.2517,
"step": 2200
},
{
"epoch": 0.41,
"grad_norm": 2.4248422954221525,
"learning_rate": 8.992628990238215e-06,
"loss": 1.2476,
"step": 2205
},
{
"epoch": 0.41,
"grad_norm": 2.294214303014934,
"learning_rate": 8.98821610430298e-06,
"loss": 1.2481,
"step": 2210
},
{
"epoch": 0.41,
"grad_norm": 2.826945710941023,
"learning_rate": 8.983794661698427e-06,
"loss": 1.262,
"step": 2215
},
{
"epoch": 0.41,
"grad_norm": 2.513263172269817,
"learning_rate": 8.979364671910705e-06,
"loss": 1.2564,
"step": 2220
},
{
"epoch": 0.41,
"grad_norm": 2.388911911230159,
"learning_rate": 8.974926144444305e-06,
"loss": 1.2869,
"step": 2225
},
{
"epoch": 0.42,
"grad_norm": 2.776503134174594,
"learning_rate": 8.970479088822032e-06,
"loss": 1.2186,
"step": 2230
},
{
"epoch": 0.42,
"grad_norm": 2.4937959196461836,
"learning_rate": 8.96602351458499e-06,
"loss": 1.2154,
"step": 2235
},
{
"epoch": 0.42,
"grad_norm": 2.6962622054584355,
"learning_rate": 8.961559431292562e-06,
"loss": 1.2443,
"step": 2240
},
{
"epoch": 0.42,
"grad_norm": 2.5956355309208545,
"learning_rate": 8.957086848522381e-06,
"loss": 1.2554,
"step": 2245
},
{
"epoch": 0.42,
"grad_norm": 2.7594593072064195,
"learning_rate": 8.952605775870323e-06,
"loss": 1.2676,
"step": 2250
},
{
"epoch": 0.42,
"grad_norm": 2.714032226160169,
"learning_rate": 8.948116222950471e-06,
"loss": 1.2117,
"step": 2255
},
{
"epoch": 0.42,
"grad_norm": 2.9432371600954874,
"learning_rate": 8.943618199395111e-06,
"loss": 1.3193,
"step": 2260
},
{
"epoch": 0.42,
"grad_norm": 2.5517853164601925,
"learning_rate": 8.939111714854693e-06,
"loss": 1.2122,
"step": 2265
},
{
"epoch": 0.42,
"grad_norm": 2.337294494695929,
"learning_rate": 8.934596778997831e-06,
"loss": 1.2581,
"step": 2270
},
{
"epoch": 0.42,
"grad_norm": 2.6000119365477756,
"learning_rate": 8.93007340151126e-06,
"loss": 1.2595,
"step": 2275
},
{
"epoch": 0.43,
"grad_norm": 2.491496550690855,
"learning_rate": 8.925541592099832e-06,
"loss": 1.2675,
"step": 2280
},
{
"epoch": 0.43,
"grad_norm": 2.9739305294116725,
"learning_rate": 8.92100136048649e-06,
"loss": 1.2279,
"step": 2285
},
{
"epoch": 0.43,
"grad_norm": 2.697112416494898,
"learning_rate": 8.916452716412248e-06,
"loss": 1.2281,
"step": 2290
},
{
"epoch": 0.43,
"grad_norm": 2.9089535182374626,
"learning_rate": 8.911895669636164e-06,
"loss": 1.234,
"step": 2295
},
{
"epoch": 0.43,
"grad_norm": 2.381662264130669,
"learning_rate": 8.907330229935327e-06,
"loss": 1.2221,
"step": 2300
},
{
"epoch": 0.43,
"grad_norm": 2.5219995591332633,
"learning_rate": 8.902756407104836e-06,
"loss": 1.2423,
"step": 2305
},
{
"epoch": 0.43,
"grad_norm": 2.354949055377429,
"learning_rate": 8.898174210957768e-06,
"loss": 1.2576,
"step": 2310
},
{
"epoch": 0.43,
"grad_norm": 2.7733201062503143,
"learning_rate": 8.89358365132517e-06,
"loss": 1.2588,
"step": 2315
},
{
"epoch": 0.43,
"grad_norm": 2.7596778146792897,
"learning_rate": 8.888984738056035e-06,
"loss": 1.249,
"step": 2320
},
{
"epoch": 0.43,
"grad_norm": 2.892780169742038,
"learning_rate": 8.884377481017273e-06,
"loss": 1.2028,
"step": 2325
},
{
"epoch": 0.43,
"grad_norm": 2.606206110122719,
"learning_rate": 8.8797618900937e-06,
"loss": 1.1971,
"step": 2330
},
{
"epoch": 0.44,
"grad_norm": 2.7230716452282238,
"learning_rate": 8.875137975188008e-06,
"loss": 1.2664,
"step": 2335
},
{
"epoch": 0.44,
"grad_norm": 2.774643601960243,
"learning_rate": 8.87050574622075e-06,
"loss": 1.1894,
"step": 2340
},
{
"epoch": 0.44,
"grad_norm": 2.5722147356763716,
"learning_rate": 8.865865213130322e-06,
"loss": 1.2562,
"step": 2345
},
{
"epoch": 0.44,
"grad_norm": 2.5301045858382305,
"learning_rate": 8.861216385872926e-06,
"loss": 1.2211,
"step": 2350
},
{
"epoch": 0.44,
"grad_norm": 2.5336960944371865,
"learning_rate": 8.85655927442257e-06,
"loss": 1.2428,
"step": 2355
},
{
"epoch": 0.44,
"grad_norm": 2.9864398074133374,
"learning_rate": 8.851893888771024e-06,
"loss": 1.2466,
"step": 2360
},
{
"epoch": 0.44,
"grad_norm": 2.4719774947680957,
"learning_rate": 8.84722023892782e-06,
"loss": 1.2146,
"step": 2365
},
{
"epoch": 0.44,
"grad_norm": 2.7847705036136907,
"learning_rate": 8.842538334920217e-06,
"loss": 1.2395,
"step": 2370
},
{
"epoch": 0.44,
"grad_norm": 2.7115968083330313,
"learning_rate": 8.837848186793182e-06,
"loss": 1.2483,
"step": 2375
},
{
"epoch": 0.44,
"grad_norm": 2.498538404354106,
"learning_rate": 8.833149804609372e-06,
"loss": 1.2014,
"step": 2380
},
{
"epoch": 0.44,
"grad_norm": 3.0546576296254027,
"learning_rate": 8.828443198449109e-06,
"loss": 1.2135,
"step": 2385
},
{
"epoch": 0.45,
"grad_norm": 2.8485505002879803,
"learning_rate": 8.82372837841036e-06,
"loss": 1.2318,
"step": 2390
},
{
"epoch": 0.45,
"grad_norm": 2.571389509311176,
"learning_rate": 8.819005354608712e-06,
"loss": 1.1785,
"step": 2395
},
{
"epoch": 0.45,
"grad_norm": 2.7341603256447065,
"learning_rate": 8.814274137177359e-06,
"loss": 1.2882,
"step": 2400
},
{
"epoch": 0.45,
"grad_norm": 2.38673842882881,
"learning_rate": 8.809534736267068e-06,
"loss": 1.2341,
"step": 2405
},
{
"epoch": 0.45,
"grad_norm": 2.5157200762234413,
"learning_rate": 8.804787162046165e-06,
"loss": 1.226,
"step": 2410
},
{
"epoch": 0.45,
"grad_norm": 2.4512225231854035,
"learning_rate": 8.800031424700518e-06,
"loss": 1.3022,
"step": 2415
},
{
"epoch": 0.45,
"grad_norm": 3.018409907978064,
"learning_rate": 8.795267534433497e-06,
"loss": 1.2353,
"step": 2420
},
{
"epoch": 0.45,
"grad_norm": 2.336307047503812,
"learning_rate": 8.790495501465976e-06,
"loss": 1.244,
"step": 2425
},
{
"epoch": 0.45,
"grad_norm": 2.6173817118831764,
"learning_rate": 8.78571533603629e-06,
"loss": 1.2408,
"step": 2430
},
{
"epoch": 0.45,
"grad_norm": 3.082589792469222,
"learning_rate": 8.780927048400231e-06,
"loss": 1.2414,
"step": 2435
},
{
"epoch": 0.46,
"grad_norm": 2.6145325587133597,
"learning_rate": 8.77613064883101e-06,
"loss": 1.1795,
"step": 2440
},
{
"epoch": 0.46,
"grad_norm": 2.3165502789649275,
"learning_rate": 8.771326147619243e-06,
"loss": 1.2152,
"step": 2445
},
{
"epoch": 0.46,
"grad_norm": 2.5107426100104853,
"learning_rate": 8.766513555072929e-06,
"loss": 1.2438,
"step": 2450
},
{
"epoch": 0.46,
"grad_norm": 2.3252364646501835,
"learning_rate": 8.761692881517428e-06,
"loss": 1.206,
"step": 2455
},
{
"epoch": 0.46,
"grad_norm": 2.369573589297014,
"learning_rate": 8.756864137295436e-06,
"loss": 1.2475,
"step": 2460
},
{
"epoch": 0.46,
"grad_norm": 2.4304676652220514,
"learning_rate": 8.752027332766965e-06,
"loss": 1.2198,
"step": 2465
},
{
"epoch": 0.46,
"grad_norm": 3.029950181599185,
"learning_rate": 8.747182478309321e-06,
"loss": 1.2473,
"step": 2470
},
{
"epoch": 0.46,
"grad_norm": 2.3276833420238536,
"learning_rate": 8.742329584317082e-06,
"loss": 1.2786,
"step": 2475
},
{
"epoch": 0.46,
"grad_norm": 2.5955319528618688,
"learning_rate": 8.73746866120207e-06,
"loss": 1.2491,
"step": 2480
},
{
"epoch": 0.46,
"grad_norm": 2.271857973340684,
"learning_rate": 8.732599719393341e-06,
"loss": 1.28,
"step": 2485
},
{
"epoch": 0.46,
"grad_norm": 2.4977245907120675,
"learning_rate": 8.727722769337147e-06,
"loss": 1.2397,
"step": 2490
},
{
"epoch": 0.47,
"grad_norm": 2.389057171322813,
"learning_rate": 8.722837821496925e-06,
"loss": 1.2456,
"step": 2495
},
{
"epoch": 0.47,
"grad_norm": 2.6678877706649873,
"learning_rate": 8.717944886353276e-06,
"loss": 1.2869,
"step": 2500
},
{
"epoch": 0.47,
"grad_norm": 2.4737604548224987,
"learning_rate": 8.713043974403928e-06,
"loss": 1.2634,
"step": 2505
},
{
"epoch": 0.47,
"grad_norm": 2.5974619042954354,
"learning_rate": 8.708135096163728e-06,
"loss": 1.2797,
"step": 2510
},
{
"epoch": 0.47,
"grad_norm": 2.4140156710174474,
"learning_rate": 8.70321826216462e-06,
"loss": 1.2333,
"step": 2515
},
{
"epoch": 0.47,
"grad_norm": 2.735433147498583,
"learning_rate": 8.698293482955605e-06,
"loss": 1.2615,
"step": 2520
},
{
"epoch": 0.47,
"grad_norm": 2.717654768808532,
"learning_rate": 8.693360769102743e-06,
"loss": 1.2237,
"step": 2525
},
{
"epoch": 0.47,
"grad_norm": 2.436937033475975,
"learning_rate": 8.688420131189111e-06,
"loss": 1.2109,
"step": 2530
},
{
"epoch": 0.47,
"grad_norm": 2.3731683394062744,
"learning_rate": 8.683471579814782e-06,
"loss": 1.2538,
"step": 2535
},
{
"epoch": 0.47,
"grad_norm": 2.423610268780273,
"learning_rate": 8.67851512559682e-06,
"loss": 1.2546,
"step": 2540
},
{
"epoch": 0.47,
"grad_norm": 2.484873214909234,
"learning_rate": 8.673550779169236e-06,
"loss": 1.1855,
"step": 2545
},
{
"epoch": 0.48,
"grad_norm": 2.4105445019685785,
"learning_rate": 8.668578551182972e-06,
"loss": 1.2437,
"step": 2550
},
{
"epoch": 0.48,
"grad_norm": 2.3073902118692793,
"learning_rate": 8.663598452305884e-06,
"loss": 1.1848,
"step": 2555
},
{
"epoch": 0.48,
"grad_norm": 2.485153537624552,
"learning_rate": 8.658610493222714e-06,
"loss": 1.203,
"step": 2560
},
{
"epoch": 0.48,
"grad_norm": 2.552191735957582,
"learning_rate": 8.65361468463507e-06,
"loss": 1.2728,
"step": 2565
},
{
"epoch": 0.48,
"grad_norm": 2.517770218713454,
"learning_rate": 8.648611037261397e-06,
"loss": 1.2595,
"step": 2570
},
{
"epoch": 0.48,
"grad_norm": 2.520499429177203,
"learning_rate": 8.643599561836958e-06,
"loss": 1.2534,
"step": 2575
},
{
"epoch": 0.48,
"grad_norm": 2.682134328033429,
"learning_rate": 8.638580269113817e-06,
"loss": 1.2171,
"step": 2580
},
{
"epoch": 0.48,
"grad_norm": 2.855280608865029,
"learning_rate": 8.6335531698608e-06,
"loss": 1.3234,
"step": 2585
},
{
"epoch": 0.48,
"grad_norm": 2.4942587653229733,
"learning_rate": 8.628518274863494e-06,
"loss": 1.1928,
"step": 2590
},
{
"epoch": 0.48,
"grad_norm": 2.4240939844890934,
"learning_rate": 8.623475594924202e-06,
"loss": 1.2482,
"step": 2595
},
{
"epoch": 0.48,
"grad_norm": 2.643886198160013,
"learning_rate": 8.61842514086193e-06,
"loss": 1.2556,
"step": 2600
},
{
"epoch": 0.49,
"grad_norm": 3.0430371576127198,
"learning_rate": 8.61336692351237e-06,
"loss": 1.2242,
"step": 2605
},
{
"epoch": 0.49,
"grad_norm": 2.914966940001587,
"learning_rate": 8.608300953727865e-06,
"loss": 1.2378,
"step": 2610
},
{
"epoch": 0.49,
"grad_norm": 2.0564710269311353,
"learning_rate": 8.603227242377392e-06,
"loss": 1.2633,
"step": 2615
},
{
"epoch": 0.49,
"grad_norm": 2.664747765412154,
"learning_rate": 8.598145800346537e-06,
"loss": 1.2316,
"step": 2620
},
{
"epoch": 0.49,
"grad_norm": 2.590698026636923,
"learning_rate": 8.593056638537472e-06,
"loss": 1.2533,
"step": 2625
},
{
"epoch": 0.49,
"grad_norm": 2.720013435674094,
"learning_rate": 8.58795976786893e-06,
"loss": 1.2326,
"step": 2630
},
{
"epoch": 0.49,
"grad_norm": 2.587306223223574,
"learning_rate": 8.582855199276189e-06,
"loss": 1.2626,
"step": 2635
},
{
"epoch": 0.49,
"grad_norm": 2.396094522080749,
"learning_rate": 8.577742943711037e-06,
"loss": 1.2044,
"step": 2640
},
{
"epoch": 0.49,
"grad_norm": 2.632459801832667,
"learning_rate": 8.572623012141756e-06,
"loss": 1.2194,
"step": 2645
},
{
"epoch": 0.49,
"grad_norm": 2.5802264347271486,
"learning_rate": 8.5674954155531e-06,
"loss": 1.2498,
"step": 2650
},
{
"epoch": 0.5,
"grad_norm": 2.326993564179618,
"learning_rate": 8.562360164946261e-06,
"loss": 1.2104,
"step": 2655
},
{
"epoch": 0.5,
"grad_norm": 2.4358311112489166,
"learning_rate": 8.55721727133886e-06,
"loss": 1.2509,
"step": 2660
},
{
"epoch": 0.5,
"grad_norm": 2.432075090103631,
"learning_rate": 8.552066745764911e-06,
"loss": 1.1837,
"step": 2665
},
{
"epoch": 0.5,
"grad_norm": 2.3984144854914065,
"learning_rate": 8.546908599274807e-06,
"loss": 1.2726,
"step": 2670
},
{
"epoch": 0.5,
"grad_norm": 2.5690105626973687,
"learning_rate": 8.541742842935286e-06,
"loss": 1.2625,
"step": 2675
},
{
"epoch": 0.5,
"grad_norm": 2.5583136858764464,
"learning_rate": 8.536569487829416e-06,
"loss": 1.263,
"step": 2680
},
{
"epoch": 0.5,
"grad_norm": 2.429437667412415,
"learning_rate": 8.531388545056569e-06,
"loss": 1.2279,
"step": 2685
},
{
"epoch": 0.5,
"grad_norm": 2.6678935500473546,
"learning_rate": 8.526200025732395e-06,
"loss": 1.2647,
"step": 2690
},
{
"epoch": 0.5,
"grad_norm": 2.6772674553916387,
"learning_rate": 8.521003940988796e-06,
"loss": 1.2381,
"step": 2695
},
{
"epoch": 0.5,
"grad_norm": 2.6232131162921357,
"learning_rate": 8.515800301973916e-06,
"loss": 1.2192,
"step": 2700
},
{
"epoch": 0.5,
"grad_norm": 2.508584977632273,
"learning_rate": 8.51058911985209e-06,
"loss": 1.2486,
"step": 2705
},
{
"epoch": 0.51,
"grad_norm": 2.4972313904885626,
"learning_rate": 8.505370405803854e-06,
"loss": 1.2121,
"step": 2710
},
{
"epoch": 0.51,
"grad_norm": 2.876194235114713,
"learning_rate": 8.50014417102589e-06,
"loss": 1.2434,
"step": 2715
},
{
"epoch": 0.51,
"grad_norm": 2.4126645230813786,
"learning_rate": 8.494910426731026e-06,
"loss": 1.1677,
"step": 2720
},
{
"epoch": 0.51,
"grad_norm": 2.842765954994177,
"learning_rate": 8.489669184148196e-06,
"loss": 1.3138,
"step": 2725
},
{
"epoch": 0.51,
"grad_norm": 2.7561841143747348,
"learning_rate": 8.48442045452242e-06,
"loss": 1.2539,
"step": 2730
},
{
"epoch": 0.51,
"grad_norm": 2.4533831653730243,
"learning_rate": 8.479164249114787e-06,
"loss": 1.21,
"step": 2735
},
{
"epoch": 0.51,
"grad_norm": 3.0366686445653324,
"learning_rate": 8.473900579202419e-06,
"loss": 1.262,
"step": 2740
},
{
"epoch": 0.51,
"grad_norm": 2.511447651995392,
"learning_rate": 8.468629456078456e-06,
"loss": 1.2612,
"step": 2745
},
{
"epoch": 0.51,
"grad_norm": 2.5858036767898906,
"learning_rate": 8.463350891052032e-06,
"loss": 1.2127,
"step": 2750
},
{
"epoch": 0.51,
"grad_norm": 2.856917866346965,
"learning_rate": 8.458064895448242e-06,
"loss": 1.2439,
"step": 2755
},
{
"epoch": 0.51,
"grad_norm": 2.8503466560810464,
"learning_rate": 8.452771480608126e-06,
"loss": 1.2404,
"step": 2760
},
{
"epoch": 0.52,
"grad_norm": 2.724946308949215,
"learning_rate": 8.447470657888642e-06,
"loss": 1.2191,
"step": 2765
},
{
"epoch": 0.52,
"grad_norm": 3.0293778238961124,
"learning_rate": 8.442162438662637e-06,
"loss": 1.2632,
"step": 2770
},
{
"epoch": 0.52,
"grad_norm": 2.3428716544118227,
"learning_rate": 8.436846834318836e-06,
"loss": 1.2473,
"step": 2775
},
{
"epoch": 0.52,
"grad_norm": 2.4162322848129127,
"learning_rate": 8.431523856261803e-06,
"loss": 1.2763,
"step": 2780
},
{
"epoch": 0.52,
"grad_norm": 2.3173281978902582,
"learning_rate": 8.426193515911922e-06,
"loss": 1.2339,
"step": 2785
},
{
"epoch": 0.52,
"grad_norm": 2.41881861854618,
"learning_rate": 8.420855824705372e-06,
"loss": 1.2428,
"step": 2790
},
{
"epoch": 0.52,
"grad_norm": 2.4540386228110567,
"learning_rate": 8.415510794094108e-06,
"loss": 1.251,
"step": 2795
},
{
"epoch": 0.52,
"grad_norm": 2.7791688006237814,
"learning_rate": 8.410158435545825e-06,
"loss": 1.2333,
"step": 2800
},
{
"epoch": 0.52,
"grad_norm": 2.2050497673913547,
"learning_rate": 8.404798760543947e-06,
"loss": 1.2943,
"step": 2805
},
{
"epoch": 0.52,
"grad_norm": 2.513060240203504,
"learning_rate": 8.399431780587588e-06,
"loss": 1.2542,
"step": 2810
},
{
"epoch": 0.52,
"grad_norm": 2.4705441230982337,
"learning_rate": 8.39405750719154e-06,
"loss": 1.2266,
"step": 2815
},
{
"epoch": 0.53,
"grad_norm": 2.3602531470830628,
"learning_rate": 8.388675951886242e-06,
"loss": 1.1894,
"step": 2820
},
{
"epoch": 0.53,
"grad_norm": 2.5625025420117247,
"learning_rate": 8.383287126217756e-06,
"loss": 1.2414,
"step": 2825
},
{
"epoch": 0.53,
"grad_norm": 2.270771349307043,
"learning_rate": 8.377891041747742e-06,
"loss": 1.2786,
"step": 2830
},
{
"epoch": 0.53,
"grad_norm": 2.457830364346707,
"learning_rate": 8.372487710053433e-06,
"loss": 1.275,
"step": 2835
},
{
"epoch": 0.53,
"grad_norm": 2.5075337353947913,
"learning_rate": 8.367077142727612e-06,
"loss": 1.2273,
"step": 2840
},
{
"epoch": 0.53,
"grad_norm": 2.864871208968174,
"learning_rate": 8.361659351378585e-06,
"loss": 1.1916,
"step": 2845
},
{
"epoch": 0.53,
"grad_norm": 2.478234689243336,
"learning_rate": 8.35623434763016e-06,
"loss": 1.2507,
"step": 2850
},
{
"epoch": 0.53,
"grad_norm": 2.6409866161664124,
"learning_rate": 8.350802143121614e-06,
"loss": 1.1816,
"step": 2855
},
{
"epoch": 0.53,
"grad_norm": 2.708641058912023,
"learning_rate": 8.345362749507677e-06,
"loss": 1.1989,
"step": 2860
},
{
"epoch": 0.53,
"grad_norm": 2.512060087338582,
"learning_rate": 8.339916178458502e-06,
"loss": 1.2063,
"step": 2865
},
{
"epoch": 0.54,
"grad_norm": 2.312677281400599,
"learning_rate": 8.334462441659644e-06,
"loss": 1.2793,
"step": 2870
},
{
"epoch": 0.54,
"grad_norm": 2.583586314697303,
"learning_rate": 8.329001550812024e-06,
"loss": 1.2189,
"step": 2875
},
{
"epoch": 0.54,
"grad_norm": 2.463821721761164,
"learning_rate": 8.32353351763192e-06,
"loss": 1.2308,
"step": 2880
},
{
"epoch": 0.54,
"grad_norm": 2.3329831134769887,
"learning_rate": 8.31805835385093e-06,
"loss": 1.3068,
"step": 2885
},
{
"epoch": 0.54,
"grad_norm": 2.3890079406752887,
"learning_rate": 8.312576071215954e-06,
"loss": 1.2802,
"step": 2890
},
{
"epoch": 0.54,
"grad_norm": 2.4337592685605864,
"learning_rate": 8.30708668148916e-06,
"loss": 1.2084,
"step": 2895
},
{
"epoch": 0.54,
"grad_norm": 2.979980545845011,
"learning_rate": 8.301590196447968e-06,
"loss": 1.2169,
"step": 2900
},
{
"epoch": 0.54,
"grad_norm": 2.4713288910964226,
"learning_rate": 8.296086627885022e-06,
"loss": 1.24,
"step": 2905
},
{
"epoch": 0.54,
"grad_norm": 2.57678209683771,
"learning_rate": 8.29057598760816e-06,
"loss": 1.2183,
"step": 2910
},
{
"epoch": 0.54,
"grad_norm": 2.388983411954069,
"learning_rate": 8.285058287440393e-06,
"loss": 1.2159,
"step": 2915
},
{
"epoch": 0.54,
"grad_norm": 2.5856899395384207,
"learning_rate": 8.279533539219881e-06,
"loss": 1.2423,
"step": 2920
},
{
"epoch": 0.55,
"grad_norm": 2.656085681106573,
"learning_rate": 8.274001754799902e-06,
"loss": 1.2425,
"step": 2925
},
{
"epoch": 0.55,
"grad_norm": 2.687407245664408,
"learning_rate": 8.268462946048839e-06,
"loss": 1.1885,
"step": 2930
},
{
"epoch": 0.55,
"grad_norm": 2.602375749001676,
"learning_rate": 8.262917124850132e-06,
"loss": 1.2539,
"step": 2935
},
{
"epoch": 0.55,
"grad_norm": 2.5032059088059926,
"learning_rate": 8.257364303102275e-06,
"loss": 1.2215,
"step": 2940
},
{
"epoch": 0.55,
"grad_norm": 2.1354807044438715,
"learning_rate": 8.251804492718781e-06,
"loss": 1.1698,
"step": 2945
},
{
"epoch": 0.55,
"grad_norm": 2.8821604335776976,
"learning_rate": 8.246237705628156e-06,
"loss": 1.2858,
"step": 2950
},
{
"epoch": 0.55,
"grad_norm": 2.808397596971899,
"learning_rate": 8.240663953773874e-06,
"loss": 1.2568,
"step": 2955
},
{
"epoch": 0.55,
"grad_norm": 2.5791493113582433,
"learning_rate": 8.23508324911435e-06,
"loss": 1.2067,
"step": 2960
},
{
"epoch": 0.55,
"grad_norm": 2.7351925838304054,
"learning_rate": 8.229495603622922e-06,
"loss": 1.2393,
"step": 2965
},
{
"epoch": 0.55,
"grad_norm": 2.937782581862869,
"learning_rate": 8.223901029287811e-06,
"loss": 1.219,
"step": 2970
},
{
"epoch": 0.55,
"grad_norm": 2.5782397077136476,
"learning_rate": 8.218299538112113e-06,
"loss": 1.2006,
"step": 2975
},
{
"epoch": 0.56,
"grad_norm": 2.6098996864666257,
"learning_rate": 8.212691142113759e-06,
"loss": 1.2895,
"step": 2980
},
{
"epoch": 0.56,
"grad_norm": 2.3838592097882865,
"learning_rate": 8.20707585332549e-06,
"loss": 1.2064,
"step": 2985
},
{
"epoch": 0.56,
"grad_norm": 2.9170817488913077,
"learning_rate": 8.201453683794844e-06,
"loss": 1.2256,
"step": 2990
},
{
"epoch": 0.56,
"grad_norm": 2.7415094542817506,
"learning_rate": 8.195824645584115e-06,
"loss": 1.2449,
"step": 2995
},
{
"epoch": 0.56,
"grad_norm": 3.020101156694181,
"learning_rate": 8.19018875077034e-06,
"loss": 1.196,
"step": 3000
},
{
"epoch": 0.56,
"grad_norm": 3.5256819382363416,
"learning_rate": 8.184546011445262e-06,
"loss": 1.2331,
"step": 3005
},
{
"epoch": 0.56,
"grad_norm": 2.489265808403552,
"learning_rate": 8.178896439715308e-06,
"loss": 1.2924,
"step": 3010
},
{
"epoch": 0.56,
"grad_norm": 2.4229165953443896,
"learning_rate": 8.173240047701566e-06,
"loss": 1.2201,
"step": 3015
},
{
"epoch": 0.56,
"grad_norm": 2.2691244384078706,
"learning_rate": 8.167576847539756e-06,
"loss": 1.2383,
"step": 3020
},
{
"epoch": 0.56,
"grad_norm": 2.811907668419909,
"learning_rate": 8.161906851380209e-06,
"loss": 1.2604,
"step": 3025
},
{
"epoch": 0.57,
"grad_norm": 2.259428143639486,
"learning_rate": 8.156230071387829e-06,
"loss": 1.2208,
"step": 3030
},
{
"epoch": 0.57,
"grad_norm": 2.8094443482735585,
"learning_rate": 8.15054651974208e-06,
"loss": 1.2036,
"step": 3035
},
{
"epoch": 0.57,
"grad_norm": 2.7986050897572725,
"learning_rate": 8.144856208636953e-06,
"loss": 1.1986,
"step": 3040
},
{
"epoch": 0.57,
"grad_norm": 2.3585431474571235,
"learning_rate": 8.139159150280942e-06,
"loss": 1.1997,
"step": 3045
},
{
"epoch": 0.57,
"grad_norm": 2.555014380109738,
"learning_rate": 8.133455356897017e-06,
"loss": 1.1899,
"step": 3050
},
{
"epoch": 0.57,
"grad_norm": 2.5930294103690983,
"learning_rate": 8.127744840722596e-06,
"loss": 1.2354,
"step": 3055
},
{
"epoch": 0.57,
"grad_norm": 2.4164140212051173,
"learning_rate": 8.122027614009525e-06,
"loss": 1.1931,
"step": 3060
},
{
"epoch": 0.57,
"grad_norm": 2.4544780551977095,
"learning_rate": 8.11630368902404e-06,
"loss": 1.2359,
"step": 3065
},
{
"epoch": 0.57,
"grad_norm": 2.6112887233948108,
"learning_rate": 8.11057307804676e-06,
"loss": 1.2432,
"step": 3070
},
{
"epoch": 0.57,
"grad_norm": 2.550549949346128,
"learning_rate": 8.104835793372634e-06,
"loss": 1.2374,
"step": 3075
},
{
"epoch": 0.57,
"grad_norm": 2.5394631647393333,
"learning_rate": 8.09909184731094e-06,
"loss": 1.2879,
"step": 3080
},
{
"epoch": 0.58,
"grad_norm": 2.448831569746075,
"learning_rate": 8.093341252185244e-06,
"loss": 1.2475,
"step": 3085
},
{
"epoch": 0.58,
"grad_norm": 2.262631971760282,
"learning_rate": 8.087584020333378e-06,
"loss": 1.2495,
"step": 3090
},
{
"epoch": 0.58,
"grad_norm": 2.3580312532472107,
"learning_rate": 8.081820164107417e-06,
"loss": 1.2066,
"step": 3095
},
{
"epoch": 0.58,
"grad_norm": 2.280906364318545,
"learning_rate": 8.076049695873637e-06,
"loss": 1.2622,
"step": 3100
},
{
"epoch": 0.58,
"grad_norm": 2.717321347834505,
"learning_rate": 8.070272628012517e-06,
"loss": 1.2626,
"step": 3105
},
{
"epoch": 0.58,
"grad_norm": 2.745888808242339,
"learning_rate": 8.064488972918678e-06,
"loss": 1.2383,
"step": 3110
},
{
"epoch": 0.58,
"grad_norm": 2.4430976141986522,
"learning_rate": 8.058698743000885e-06,
"loss": 1.2229,
"step": 3115
},
{
"epoch": 0.58,
"grad_norm": 2.2964872809414674,
"learning_rate": 8.052901950682007e-06,
"loss": 1.1775,
"step": 3120
},
{
"epoch": 0.58,
"grad_norm": 2.601409079456989,
"learning_rate": 8.047098608398991e-06,
"loss": 1.2243,
"step": 3125
},
{
"epoch": 0.58,
"grad_norm": 2.6030241980938924,
"learning_rate": 8.041288728602838e-06,
"loss": 1.2544,
"step": 3130
},
{
"epoch": 0.58,
"grad_norm": 2.4914596853603257,
"learning_rate": 8.035472323758572e-06,
"loss": 1.2529,
"step": 3135
},
{
"epoch": 0.59,
"grad_norm": 2.5770189045893295,
"learning_rate": 8.029649406345223e-06,
"loss": 1.1977,
"step": 3140
},
{
"epoch": 0.59,
"grad_norm": 2.631376359022823,
"learning_rate": 8.023819988855788e-06,
"loss": 1.1996,
"step": 3145
},
{
"epoch": 0.59,
"grad_norm": 2.1749751309010317,
"learning_rate": 8.01798408379721e-06,
"loss": 1.178,
"step": 3150
},
{
"epoch": 0.59,
"grad_norm": 2.7989592102512493,
"learning_rate": 8.012141703690356e-06,
"loss": 1.2529,
"step": 3155
},
{
"epoch": 0.59,
"grad_norm": 2.4294484853415645,
"learning_rate": 8.006292861069977e-06,
"loss": 1.2502,
"step": 3160
},
{
"epoch": 0.59,
"grad_norm": 2.5272009149226746,
"learning_rate": 8.0004375684847e-06,
"loss": 1.2616,
"step": 3165
},
{
"epoch": 0.59,
"grad_norm": 2.5718843710944066,
"learning_rate": 7.994575838496979e-06,
"loss": 1.2543,
"step": 3170
},
{
"epoch": 0.59,
"grad_norm": 2.4527756457949077,
"learning_rate": 7.988707683683088e-06,
"loss": 1.2024,
"step": 3175
},
{
"epoch": 0.59,
"grad_norm": 2.2354347723383285,
"learning_rate": 7.98283311663308e-06,
"loss": 1.2409,
"step": 3180
},
{
"epoch": 0.59,
"grad_norm": 2.526326729932712,
"learning_rate": 7.976952149950768e-06,
"loss": 1.2212,
"step": 3185
},
{
"epoch": 0.59,
"grad_norm": 2.372828024861733,
"learning_rate": 7.971064796253695e-06,
"loss": 1.2728,
"step": 3190
},
{
"epoch": 0.6,
"grad_norm": 2.487614641155206,
"learning_rate": 7.965171068173108e-06,
"loss": 1.2172,
"step": 3195
},
{
"epoch": 0.6,
"grad_norm": 2.567329023677344,
"learning_rate": 7.959270978353928e-06,
"loss": 1.1692,
"step": 3200
},
{
"epoch": 0.6,
"grad_norm": 2.157718620065115,
"learning_rate": 7.953364539454724e-06,
"loss": 1.2415,
"step": 3205
},
{
"epoch": 0.6,
"grad_norm": 2.8667488031824813,
"learning_rate": 7.947451764147694e-06,
"loss": 1.2713,
"step": 3210
},
{
"epoch": 0.6,
"grad_norm": 2.599094361368479,
"learning_rate": 7.941532665118621e-06,
"loss": 1.2505,
"step": 3215
},
{
"epoch": 0.6,
"grad_norm": 2.4541586898227084,
"learning_rate": 7.935607255066867e-06,
"loss": 1.2747,
"step": 3220
},
{
"epoch": 0.6,
"grad_norm": 2.3108705882965324,
"learning_rate": 7.929675546705318e-06,
"loss": 1.2838,
"step": 3225
},
{
"epoch": 0.6,
"grad_norm": 2.6954248407554617,
"learning_rate": 7.923737552760389e-06,
"loss": 1.2476,
"step": 3230
},
{
"epoch": 0.6,
"grad_norm": 2.5820318599293217,
"learning_rate": 7.917793285971972e-06,
"loss": 1.1464,
"step": 3235
},
{
"epoch": 0.6,
"grad_norm": 2.5389808963239773,
"learning_rate": 7.91184275909342e-06,
"loss": 1.2377,
"step": 3240
},
{
"epoch": 0.61,
"grad_norm": 2.4877167669233318,
"learning_rate": 7.905885984891514e-06,
"loss": 1.2184,
"step": 3245
},
{
"epoch": 0.61,
"grad_norm": 2.892266466172481,
"learning_rate": 7.899922976146446e-06,
"loss": 1.2468,
"step": 3250
},
{
"epoch": 0.61,
"grad_norm": 2.463904687153827,
"learning_rate": 7.893953745651774e-06,
"loss": 1.2562,
"step": 3255
},
{
"epoch": 0.61,
"grad_norm": 2.5775003087085153,
"learning_rate": 7.887978306214412e-06,
"loss": 1.2282,
"step": 3260
},
{
"epoch": 0.61,
"grad_norm": 2.5006547798451093,
"learning_rate": 7.881996670654593e-06,
"loss": 1.172,
"step": 3265
},
{
"epoch": 0.61,
"grad_norm": 2.5863204612482757,
"learning_rate": 7.876008851805841e-06,
"loss": 1.3017,
"step": 3270
},
{
"epoch": 0.61,
"grad_norm": 2.4859534188574264,
"learning_rate": 7.870014862514954e-06,
"loss": 1.2004,
"step": 3275
},
{
"epoch": 0.61,
"grad_norm": 2.773160463069392,
"learning_rate": 7.864014715641958e-06,
"loss": 1.3237,
"step": 3280
},
{
"epoch": 0.61,
"grad_norm": 2.787324849208085,
"learning_rate": 7.858008424060098e-06,
"loss": 1.2493,
"step": 3285
},
{
"epoch": 0.61,
"grad_norm": 2.5342367849847203,
"learning_rate": 7.8519960006558e-06,
"loss": 1.1987,
"step": 3290
},
{
"epoch": 0.61,
"grad_norm": 2.613544873848067,
"learning_rate": 7.845977458328641e-06,
"loss": 1.2352,
"step": 3295
},
{
"epoch": 0.62,
"grad_norm": 2.331618258739045,
"learning_rate": 7.83995280999134e-06,
"loss": 1.2313,
"step": 3300
},
{
"epoch": 0.62,
"grad_norm": 2.474266299250581,
"learning_rate": 7.833922068569699e-06,
"loss": 1.2009,
"step": 3305
},
{
"epoch": 0.62,
"grad_norm": 2.480106249358808,
"learning_rate": 7.827885247002604e-06,
"loss": 1.1864,
"step": 3310
},
{
"epoch": 0.62,
"grad_norm": 2.3278379337755157,
"learning_rate": 7.82184235824198e-06,
"loss": 1.1984,
"step": 3315
},
{
"epoch": 0.62,
"grad_norm": 2.61292859257405,
"learning_rate": 7.815793415252776e-06,
"loss": 1.1849,
"step": 3320
},
{
"epoch": 0.62,
"grad_norm": 2.292538348374208,
"learning_rate": 7.809738431012922e-06,
"loss": 1.2256,
"step": 3325
},
{
"epoch": 0.62,
"grad_norm": 2.481732800973384,
"learning_rate": 7.803677418513314e-06,
"loss": 1.221,
"step": 3330
},
{
"epoch": 0.62,
"grad_norm": 2.6567638646549647,
"learning_rate": 7.797610390757781e-06,
"loss": 1.2608,
"step": 3335
},
{
"epoch": 0.62,
"grad_norm": 2.7081940886811267,
"learning_rate": 7.79153736076306e-06,
"loss": 1.2218,
"step": 3340
},
{
"epoch": 0.62,
"grad_norm": 2.6863450330567775,
"learning_rate": 7.785458341558759e-06,
"loss": 1.271,
"step": 3345
},
{
"epoch": 0.62,
"grad_norm": 2.4418830984759747,
"learning_rate": 7.779373346187343e-06,
"loss": 1.2338,
"step": 3350
},
{
"epoch": 0.63,
"grad_norm": 2.5393616241756836,
"learning_rate": 7.773282387704095e-06,
"loss": 1.2133,
"step": 3355
},
{
"epoch": 0.63,
"grad_norm": 2.1693849381009347,
"learning_rate": 7.767185479177092e-06,
"loss": 1.2466,
"step": 3360
},
{
"epoch": 0.63,
"grad_norm": 2.3461814758289066,
"learning_rate": 7.76108263368718e-06,
"loss": 1.2427,
"step": 3365
},
{
"epoch": 0.63,
"grad_norm": 2.483322016918312,
"learning_rate": 7.754973864327934e-06,
"loss": 1.2986,
"step": 3370
},
{
"epoch": 0.63,
"grad_norm": 2.327833315357301,
"learning_rate": 7.748859184205653e-06,
"loss": 1.1808,
"step": 3375
},
{
"epoch": 0.63,
"grad_norm": 2.529809653474871,
"learning_rate": 7.742738606439304e-06,
"loss": 1.2513,
"step": 3380
},
{
"epoch": 0.63,
"grad_norm": 2.9085797113080685,
"learning_rate": 7.73661214416051e-06,
"loss": 1.2094,
"step": 3385
},
{
"epoch": 0.63,
"grad_norm": 2.3983677621797654,
"learning_rate": 7.730479810513527e-06,
"loss": 1.2382,
"step": 3390
},
{
"epoch": 0.63,
"grad_norm": 2.3326323104199544,
"learning_rate": 7.724341618655201e-06,
"loss": 1.2556,
"step": 3395
},
{
"epoch": 0.63,
"grad_norm": 2.564858311955747,
"learning_rate": 7.718197581754947e-06,
"loss": 1.2206,
"step": 3400
},
{
"epoch": 0.63,
"grad_norm": 2.5755523825131617,
"learning_rate": 7.712047712994722e-06,
"loss": 1.2795,
"step": 3405
},
{
"epoch": 0.64,
"grad_norm": 2.734228499086126,
"learning_rate": 7.705892025568996e-06,
"loss": 1.2526,
"step": 3410
},
{
"epoch": 0.64,
"grad_norm": 2.618647964492142,
"learning_rate": 7.69973053268472e-06,
"loss": 1.2267,
"step": 3415
},
{
"epoch": 0.64,
"grad_norm": 2.4211951168322323,
"learning_rate": 7.693563247561302e-06,
"loss": 1.1891,
"step": 3420
},
{
"epoch": 0.64,
"grad_norm": 2.956630395418801,
"learning_rate": 7.68739018343058e-06,
"loss": 1.2446,
"step": 3425
},
{
"epoch": 0.64,
"grad_norm": 2.3367616810534484,
"learning_rate": 7.681211353536787e-06,
"loss": 1.2206,
"step": 3430
},
{
"epoch": 0.64,
"grad_norm": 2.375902773902396,
"learning_rate": 7.675026771136528e-06,
"loss": 1.2322,
"step": 3435
},
{
"epoch": 0.64,
"grad_norm": 3.5013401997908384,
"learning_rate": 7.668836449498749e-06,
"loss": 1.196,
"step": 3440
},
{
"epoch": 0.64,
"grad_norm": 2.7544164936127697,
"learning_rate": 7.662640401904709e-06,
"loss": 1.2775,
"step": 3445
},
{
"epoch": 0.64,
"grad_norm": 2.33908987713843,
"learning_rate": 7.656438641647956e-06,
"loss": 1.2146,
"step": 3450
},
{
"epoch": 0.64,
"grad_norm": 2.427538531906867,
"learning_rate": 7.650231182034289e-06,
"loss": 1.2559,
"step": 3455
},
{
"epoch": 0.65,
"grad_norm": 2.850404299913776,
"learning_rate": 7.64401803638174e-06,
"loss": 1.1813,
"step": 3460
},
{
"epoch": 0.65,
"grad_norm": 2.3597140412277695,
"learning_rate": 7.637799218020533e-06,
"loss": 1.1822,
"step": 3465
},
{
"epoch": 0.65,
"grad_norm": 2.917103092601735,
"learning_rate": 7.631574740293073e-06,
"loss": 1.2315,
"step": 3470
},
{
"epoch": 0.65,
"grad_norm": 2.682869811949341,
"learning_rate": 7.625344616553897e-06,
"loss": 1.2308,
"step": 3475
},
{
"epoch": 0.65,
"grad_norm": 2.515419590465464,
"learning_rate": 7.6191088601696615e-06,
"loss": 1.2428,
"step": 3480
},
{
"epoch": 0.65,
"grad_norm": 2.233078474751225,
"learning_rate": 7.612867484519107e-06,
"loss": 1.2121,
"step": 3485
},
{
"epoch": 0.65,
"grad_norm": 2.23514466267859,
"learning_rate": 7.6066205029930294e-06,
"loss": 1.2103,
"step": 3490
},
{
"epoch": 0.65,
"grad_norm": 2.831634837573794,
"learning_rate": 7.600367928994252e-06,
"loss": 1.303,
"step": 3495
},
{
"epoch": 0.65,
"grad_norm": 2.7631860527584355,
"learning_rate": 7.594109775937595e-06,
"loss": 1.2173,
"step": 3500
},
{
"epoch": 0.65,
"grad_norm": 2.5870696417201717,
"learning_rate": 7.58784605724985e-06,
"loss": 1.2141,
"step": 3505
},
{
"epoch": 0.65,
"grad_norm": 2.34215008768121,
"learning_rate": 7.581576786369752e-06,
"loss": 1.2298,
"step": 3510
},
{
"epoch": 0.66,
"grad_norm": 2.1049228123123487,
"learning_rate": 7.575301976747944e-06,
"loss": 1.2195,
"step": 3515
},
{
"epoch": 0.66,
"grad_norm": 2.8693762742995124,
"learning_rate": 7.569021641846954e-06,
"loss": 1.2206,
"step": 3520
},
{
"epoch": 0.66,
"grad_norm": 2.551123234988086,
"learning_rate": 7.562735795141163e-06,
"loss": 1.185,
"step": 3525
},
{
"epoch": 0.66,
"grad_norm": 2.4250229123964004,
"learning_rate": 7.556444450116782e-06,
"loss": 1.1781,
"step": 3530
},
{
"epoch": 0.66,
"grad_norm": 2.4284380080427117,
"learning_rate": 7.550147620271812e-06,
"loss": 1.2299,
"step": 3535
},
{
"epoch": 0.66,
"grad_norm": 2.3780483099018777,
"learning_rate": 7.54384531911603e-06,
"loss": 1.2534,
"step": 3540
},
{
"epoch": 0.66,
"grad_norm": 2.7264232568865983,
"learning_rate": 7.537537560170941e-06,
"loss": 1.1865,
"step": 3545
},
{
"epoch": 0.66,
"grad_norm": 2.5294854459143767,
"learning_rate": 7.53122435696977e-06,
"loss": 1.2579,
"step": 3550
},
{
"epoch": 0.66,
"grad_norm": 2.2660161661863563,
"learning_rate": 7.5249057230574155e-06,
"loss": 1.2119,
"step": 3555
},
{
"epoch": 0.66,
"grad_norm": 3.2233305983948206,
"learning_rate": 7.51858167199043e-06,
"loss": 1.1961,
"step": 3560
},
{
"epoch": 0.66,
"grad_norm": 2.3319770150325603,
"learning_rate": 7.512252217336989e-06,
"loss": 1.2339,
"step": 3565
},
{
"epoch": 0.67,
"grad_norm": 3.071379491685929,
"learning_rate": 7.505917372676863e-06,
"loss": 1.2428,
"step": 3570
},
{
"epoch": 0.67,
"grad_norm": 2.5041481151057856,
"learning_rate": 7.4995771516013815e-06,
"loss": 1.2326,
"step": 3575
},
{
"epoch": 0.67,
"grad_norm": 2.4707186038211626,
"learning_rate": 7.4932315677134105e-06,
"loss": 1.2469,
"step": 3580
},
{
"epoch": 0.67,
"grad_norm": 2.4784712914427525,
"learning_rate": 7.486880634627328e-06,
"loss": 1.2581,
"step": 3585
},
{
"epoch": 0.67,
"grad_norm": 2.8602362997757402,
"learning_rate": 7.480524365968981e-06,
"loss": 1.2216,
"step": 3590
},
{
"epoch": 0.67,
"grad_norm": 2.5401997745842486,
"learning_rate": 7.474162775375667e-06,
"loss": 1.2347,
"step": 3595
},
{
"epoch": 0.67,
"grad_norm": 2.3472339553819674,
"learning_rate": 7.467795876496101e-06,
"loss": 1.2321,
"step": 3600
},
{
"epoch": 0.67,
"grad_norm": 2.7093269313095365,
"learning_rate": 7.4614236829903894e-06,
"loss": 1.2661,
"step": 3605
},
{
"epoch": 0.67,
"grad_norm": 2.7812103887052766,
"learning_rate": 7.455046208529996e-06,
"loss": 1.1655,
"step": 3610
},
{
"epoch": 0.67,
"grad_norm": 3.376729838029134,
"learning_rate": 7.448663466797713e-06,
"loss": 1.1841,
"step": 3615
},
{
"epoch": 0.68,
"grad_norm": 2.5747589195664333,
"learning_rate": 7.442275471487636e-06,
"loss": 1.2376,
"step": 3620
},
{
"epoch": 0.68,
"grad_norm": 2.451020523839669,
"learning_rate": 7.435882236305133e-06,
"loss": 1.206,
"step": 3625
},
{
"epoch": 0.68,
"grad_norm": 2.245081634477842,
"learning_rate": 7.4294837749668114e-06,
"loss": 1.1919,
"step": 3630
},
{
"epoch": 0.68,
"grad_norm": 2.4840846987847653,
"learning_rate": 7.4230801012004925e-06,
"loss": 1.2257,
"step": 3635
},
{
"epoch": 0.68,
"grad_norm": 2.278915066123373,
"learning_rate": 7.416671228745181e-06,
"loss": 1.1602,
"step": 3640
},
{
"epoch": 0.68,
"grad_norm": 2.4499028758210275,
"learning_rate": 7.410257171351034e-06,
"loss": 1.221,
"step": 3645
},
{
"epoch": 0.68,
"grad_norm": 3.6861762079284337,
"learning_rate": 7.403837942779337e-06,
"loss": 1.2513,
"step": 3650
},
{
"epoch": 0.68,
"grad_norm": 2.8481514636316234,
"learning_rate": 7.397413556802464e-06,
"loss": 1.1957,
"step": 3655
},
{
"epoch": 0.68,
"grad_norm": 2.2520010448465255,
"learning_rate": 7.390984027203858e-06,
"loss": 1.2156,
"step": 3660
},
{
"epoch": 0.68,
"grad_norm": 2.2144209219898223,
"learning_rate": 7.3845493677779975e-06,
"loss": 1.1903,
"step": 3665
},
{
"epoch": 0.68,
"grad_norm": 2.5534532157329597,
"learning_rate": 7.378109592330367e-06,
"loss": 1.2322,
"step": 3670
},
{
"epoch": 0.69,
"grad_norm": 2.3019311785840046,
"learning_rate": 7.371664714677424e-06,
"loss": 1.212,
"step": 3675
},
{
"epoch": 0.69,
"grad_norm": 2.635004055428559,
"learning_rate": 7.365214748646577e-06,
"loss": 1.2254,
"step": 3680
},
{
"epoch": 0.69,
"grad_norm": 4.04336948549475,
"learning_rate": 7.3587597080761505e-06,
"loss": 1.1767,
"step": 3685
},
{
"epoch": 0.69,
"grad_norm": 2.521375182264338,
"learning_rate": 7.352299606815356e-06,
"loss": 1.2269,
"step": 3690
},
{
"epoch": 0.69,
"grad_norm": 2.3893615350098143,
"learning_rate": 7.345834458724262e-06,
"loss": 1.2566,
"step": 3695
},
{
"epoch": 0.69,
"grad_norm": 3.0395949427272115,
"learning_rate": 7.3393642776737625e-06,
"loss": 1.2476,
"step": 3700
},
{
"epoch": 0.69,
"grad_norm": 2.739571762112554,
"learning_rate": 7.332889077545558e-06,
"loss": 1.1958,
"step": 3705
},
{
"epoch": 0.69,
"grad_norm": 2.65164618555974,
"learning_rate": 7.326408872232107e-06,
"loss": 1.26,
"step": 3710
},
{
"epoch": 0.69,
"grad_norm": 2.526691816795283,
"learning_rate": 7.319923675636615e-06,
"loss": 1.2448,
"step": 3715
},
{
"epoch": 0.69,
"grad_norm": 2.77656114577537,
"learning_rate": 7.313433501672991e-06,
"loss": 1.2472,
"step": 3720
},
{
"epoch": 0.69,
"grad_norm": 2.8625605756818553,
"learning_rate": 7.3069383642658245e-06,
"loss": 1.2494,
"step": 3725
},
{
"epoch": 0.7,
"grad_norm": 2.560276503666187,
"learning_rate": 7.300438277350354e-06,
"loss": 1.2785,
"step": 3730
},
{
"epoch": 0.7,
"grad_norm": 2.4990702759200243,
"learning_rate": 7.293933254872439e-06,
"loss": 1.2291,
"step": 3735
},
{
"epoch": 0.7,
"grad_norm": 3.0835335467596074,
"learning_rate": 7.287423310788524e-06,
"loss": 1.2341,
"step": 3740
},
{
"epoch": 0.7,
"grad_norm": 2.1231234217056913,
"learning_rate": 7.2809084590656174e-06,
"loss": 1.2133,
"step": 3745
},
{
"epoch": 0.7,
"grad_norm": 2.5855027096845027,
"learning_rate": 7.274388713681252e-06,
"loss": 1.2473,
"step": 3750
},
{
"epoch": 0.7,
"grad_norm": 2.3812570368674146,
"learning_rate": 7.267864088623465e-06,
"loss": 1.1988,
"step": 3755
},
{
"epoch": 0.7,
"grad_norm": 2.396924022356612,
"learning_rate": 7.261334597890759e-06,
"loss": 1.1919,
"step": 3760
},
{
"epoch": 0.7,
"grad_norm": 2.293490431649686,
"learning_rate": 7.254800255492078e-06,
"loss": 1.196,
"step": 3765
},
{
"epoch": 0.7,
"grad_norm": 3.116322647421966,
"learning_rate": 7.248261075446773e-06,
"loss": 1.2607,
"step": 3770
},
{
"epoch": 0.7,
"grad_norm": 2.511617024768492,
"learning_rate": 7.241717071784576e-06,
"loss": 1.2038,
"step": 3775
},
{
"epoch": 0.7,
"grad_norm": 2.4100645815926804,
"learning_rate": 7.235168258545569e-06,
"loss": 1.213,
"step": 3780
},
{
"epoch": 0.71,
"grad_norm": 2.699345731526474,
"learning_rate": 7.228614649780151e-06,
"loss": 1.2615,
"step": 3785
},
{
"epoch": 0.71,
"grad_norm": 2.8315172335251644,
"learning_rate": 7.222056259549009e-06,
"loss": 1.2952,
"step": 3790
},
{
"epoch": 0.71,
"grad_norm": 2.671653052341253,
"learning_rate": 7.21549310192309e-06,
"loss": 1.2385,
"step": 3795
},
{
"epoch": 0.71,
"grad_norm": 2.4881987224843463,
"learning_rate": 7.208925190983571e-06,
"loss": 1.1953,
"step": 3800
},
{
"epoch": 0.71,
"grad_norm": 5.397308968654964,
"learning_rate": 7.202352540821823e-06,
"loss": 1.188,
"step": 3805
},
{
"epoch": 0.71,
"grad_norm": 2.315595168564112,
"learning_rate": 7.195775165539389e-06,
"loss": 1.1864,
"step": 3810
},
{
"epoch": 0.71,
"grad_norm": 2.7725671060957984,
"learning_rate": 7.1891930792479466e-06,
"loss": 1.2379,
"step": 3815
},
{
"epoch": 0.71,
"grad_norm": 2.5781468482007175,
"learning_rate": 7.182606296069282e-06,
"loss": 1.2162,
"step": 3820
},
{
"epoch": 0.71,
"grad_norm": 2.3949068618768985,
"learning_rate": 7.176014830135262e-06,
"loss": 1.2382,
"step": 3825
},
{
"epoch": 0.71,
"grad_norm": 2.306974962429036,
"learning_rate": 7.169418695587791e-06,
"loss": 1.209,
"step": 3830
},
{
"epoch": 0.72,
"grad_norm": 2.4234118244741993,
"learning_rate": 7.1628179065788e-06,
"loss": 1.2339,
"step": 3835
},
{
"epoch": 0.72,
"grad_norm": 2.6049010934618946,
"learning_rate": 7.1562124772702014e-06,
"loss": 1.2113,
"step": 3840
},
{
"epoch": 0.72,
"grad_norm": 2.320735650008491,
"learning_rate": 7.1496024218338645e-06,
"loss": 1.2364,
"step": 3845
},
{
"epoch": 0.72,
"grad_norm": 2.483396009383728,
"learning_rate": 7.14298775445158e-06,
"loss": 1.2469,
"step": 3850
},
{
"epoch": 0.72,
"grad_norm": 2.518323908064992,
"learning_rate": 7.13636848931504e-06,
"loss": 1.2442,
"step": 3855
},
{
"epoch": 0.72,
"grad_norm": 3.17206675979895,
"learning_rate": 7.129744640625793e-06,
"loss": 1.2622,
"step": 3860
},
{
"epoch": 0.72,
"grad_norm": 2.9675291147438307,
"learning_rate": 7.123116222595233e-06,
"loss": 1.1703,
"step": 3865
},
{
"epoch": 0.72,
"grad_norm": 2.2495951739460156,
"learning_rate": 7.116483249444543e-06,
"loss": 1.2185,
"step": 3870
},
{
"epoch": 0.72,
"grad_norm": 2.4475476932021563,
"learning_rate": 7.109845735404693e-06,
"loss": 1.2649,
"step": 3875
},
{
"epoch": 0.72,
"grad_norm": 2.453807043549154,
"learning_rate": 7.103203694716387e-06,
"loss": 1.1565,
"step": 3880
},
{
"epoch": 0.72,
"grad_norm": 2.8207713989621284,
"learning_rate": 7.0965571416300394e-06,
"loss": 1.2525,
"step": 3885
},
{
"epoch": 0.73,
"grad_norm": 3.276570824599666,
"learning_rate": 7.089906090405754e-06,
"loss": 1.2386,
"step": 3890
},
{
"epoch": 0.73,
"grad_norm": 3.0267611954698816,
"learning_rate": 7.083250555313276e-06,
"loss": 1.1895,
"step": 3895
},
{
"epoch": 0.73,
"grad_norm": 2.4575239109756386,
"learning_rate": 7.076590550631979e-06,
"loss": 1.2019,
"step": 3900
},
{
"epoch": 0.73,
"grad_norm": 2.567815585392051,
"learning_rate": 7.069926090650821e-06,
"loss": 1.2415,
"step": 3905
},
{
"epoch": 0.73,
"grad_norm": 2.470581883290022,
"learning_rate": 7.063257189668321e-06,
"loss": 1.235,
"step": 3910
},
{
"epoch": 0.73,
"grad_norm": 2.786238810744643,
"learning_rate": 7.056583861992523e-06,
"loss": 1.2203,
"step": 3915
},
{
"epoch": 0.73,
"grad_norm": 2.1595794451729184,
"learning_rate": 7.049906121940974e-06,
"loss": 1.2336,
"step": 3920
},
{
"epoch": 0.73,
"grad_norm": 2.7786181362023474,
"learning_rate": 7.043223983840681e-06,
"loss": 1.2226,
"step": 3925
},
{
"epoch": 0.73,
"grad_norm": 2.2829956340398945,
"learning_rate": 7.036537462028093e-06,
"loss": 1.2451,
"step": 3930
},
{
"epoch": 0.73,
"grad_norm": 2.425946589100039,
"learning_rate": 7.029846570849059e-06,
"loss": 1.1942,
"step": 3935
},
{
"epoch": 0.73,
"grad_norm": 3.3968924559272717,
"learning_rate": 7.023151324658807e-06,
"loss": 1.2156,
"step": 3940
},
{
"epoch": 0.74,
"grad_norm": 2.162913532948155,
"learning_rate": 7.016451737821907e-06,
"loss": 1.2059,
"step": 3945
},
{
"epoch": 0.74,
"grad_norm": 2.0812857304137697,
"learning_rate": 7.009747824712238e-06,
"loss": 1.1948,
"step": 3950
},
{
"epoch": 0.74,
"grad_norm": 2.708378845612042,
"learning_rate": 7.003039599712968e-06,
"loss": 1.1979,
"step": 3955
},
{
"epoch": 0.74,
"grad_norm": 2.391502232822183,
"learning_rate": 6.996327077216509e-06,
"loss": 1.2545,
"step": 3960
},
{
"epoch": 0.74,
"grad_norm": 2.788548039051949,
"learning_rate": 6.989610271624501e-06,
"loss": 1.2301,
"step": 3965
},
{
"epoch": 0.74,
"grad_norm": 2.166345384579877,
"learning_rate": 6.982889197347764e-06,
"loss": 1.246,
"step": 3970
},
{
"epoch": 0.74,
"grad_norm": 2.3258307219501893,
"learning_rate": 6.9761638688062864e-06,
"loss": 1.1612,
"step": 3975
},
{
"epoch": 0.74,
"grad_norm": 2.430041825610154,
"learning_rate": 6.969434300429176e-06,
"loss": 1.2322,
"step": 3980
},
{
"epoch": 0.74,
"grad_norm": 2.4302309246907416,
"learning_rate": 6.962700506654641e-06,
"loss": 1.1925,
"step": 3985
},
{
"epoch": 0.74,
"grad_norm": 2.444079835086675,
"learning_rate": 6.9559625019299535e-06,
"loss": 1.2203,
"step": 3990
},
{
"epoch": 0.75,
"grad_norm": 2.7073060094127217,
"learning_rate": 6.949220300711422e-06,
"loss": 1.2291,
"step": 3995
},
{
"epoch": 0.75,
"grad_norm": 2.8700411507881056,
"learning_rate": 6.942473917464359e-06,
"loss": 1.2168,
"step": 4000
},
{
"epoch": 0.75,
"grad_norm": 2.719479119750876,
"learning_rate": 6.9357233666630445e-06,
"loss": 1.2352,
"step": 4005
},
{
"epoch": 0.75,
"grad_norm": 2.3178114215528183,
"learning_rate": 6.9289686627907045e-06,
"loss": 1.2019,
"step": 4010
},
{
"epoch": 0.75,
"grad_norm": 2.5183702507345362,
"learning_rate": 6.9222098203394764e-06,
"loss": 1.2316,
"step": 4015
},
{
"epoch": 0.75,
"grad_norm": 2.345595014646884,
"learning_rate": 6.915446853810373e-06,
"loss": 1.1935,
"step": 4020
},
{
"epoch": 0.75,
"grad_norm": 2.75761648382353,
"learning_rate": 6.908679777713256e-06,
"loss": 1.189,
"step": 4025
},
{
"epoch": 0.75,
"grad_norm": 2.4061222491813727,
"learning_rate": 6.90190860656681e-06,
"loss": 1.1825,
"step": 4030
},
{
"epoch": 0.75,
"grad_norm": 2.5927441284141244,
"learning_rate": 6.8951333548984935e-06,
"loss": 1.1896,
"step": 4035
},
{
"epoch": 0.75,
"grad_norm": 2.606025902673968,
"learning_rate": 6.888354037244533e-06,
"loss": 1.1845,
"step": 4040
},
{
"epoch": 0.75,
"grad_norm": 2.503320901473635,
"learning_rate": 6.881570668149868e-06,
"loss": 1.1908,
"step": 4045
},
{
"epoch": 0.76,
"grad_norm": 2.580339195095967,
"learning_rate": 6.874783262168138e-06,
"loss": 1.1778,
"step": 4050
},
{
"epoch": 0.76,
"grad_norm": 2.490610410423907,
"learning_rate": 6.867991833861638e-06,
"loss": 1.2201,
"step": 4055
},
{
"epoch": 0.76,
"grad_norm": 2.704097620844027,
"learning_rate": 6.861196397801297e-06,
"loss": 1.1642,
"step": 4060
},
{
"epoch": 0.76,
"grad_norm": 2.438950075594202,
"learning_rate": 6.854396968566639e-06,
"loss": 1.2412,
"step": 4065
},
{
"epoch": 0.76,
"grad_norm": 2.3645119920317783,
"learning_rate": 6.847593560745758e-06,
"loss": 1.2257,
"step": 4070
},
{
"epoch": 0.76,
"grad_norm": 2.5375149761967726,
"learning_rate": 6.840786188935282e-06,
"loss": 1.2149,
"step": 4075
},
{
"epoch": 0.76,
"grad_norm": 2.911196562829905,
"learning_rate": 6.833974867740347e-06,
"loss": 1.1989,
"step": 4080
},
{
"epoch": 0.76,
"grad_norm": 2.5478507571563096,
"learning_rate": 6.827159611774557e-06,
"loss": 1.2376,
"step": 4085
},
{
"epoch": 0.76,
"grad_norm": 2.5288802803835604,
"learning_rate": 6.820340435659962e-06,
"loss": 1.2257,
"step": 4090
},
{
"epoch": 0.76,
"grad_norm": 2.998542518592259,
"learning_rate": 6.813517354027025e-06,
"loss": 1.1785,
"step": 4095
},
{
"epoch": 0.76,
"grad_norm": 2.3206933604689834,
"learning_rate": 6.806690381514581e-06,
"loss": 1.2461,
"step": 4100
},
{
"epoch": 0.77,
"grad_norm": 2.5445568697444902,
"learning_rate": 6.799859532769818e-06,
"loss": 1.256,
"step": 4105
},
{
"epoch": 0.77,
"grad_norm": 2.608304464852376,
"learning_rate": 6.7930248224482385e-06,
"loss": 1.1701,
"step": 4110
},
{
"epoch": 0.77,
"grad_norm": 2.659114117064667,
"learning_rate": 6.786186265213633e-06,
"loss": 1.1906,
"step": 4115
},
{
"epoch": 0.77,
"grad_norm": 2.545472949009353,
"learning_rate": 6.779343875738042e-06,
"loss": 1.1745,
"step": 4120
},
{
"epoch": 0.77,
"grad_norm": 2.120427823947941,
"learning_rate": 6.772497668701727e-06,
"loss": 1.1989,
"step": 4125
},
{
"epoch": 0.77,
"grad_norm": 2.310856981052136,
"learning_rate": 6.7656476587931455e-06,
"loss": 1.2058,
"step": 4130
},
{
"epoch": 0.77,
"grad_norm": 2.454517067939305,
"learning_rate": 6.758793860708908e-06,
"loss": 1.2168,
"step": 4135
},
{
"epoch": 0.77,
"grad_norm": 2.2246318729776347,
"learning_rate": 6.751936289153759e-06,
"loss": 1.2466,
"step": 4140
},
{
"epoch": 0.77,
"grad_norm": 2.337677050118351,
"learning_rate": 6.7450749588405294e-06,
"loss": 1.2283,
"step": 4145
},
{
"epoch": 0.77,
"grad_norm": 2.6016517303715685,
"learning_rate": 6.7382098844901254e-06,
"loss": 1.2219,
"step": 4150
},
{
"epoch": 0.77,
"grad_norm": 2.39640045650889,
"learning_rate": 6.731341080831478e-06,
"loss": 1.1366,
"step": 4155
},
{
"epoch": 0.78,
"grad_norm": 2.332603496703322,
"learning_rate": 6.724468562601526e-06,
"loss": 1.2843,
"step": 4160
},
{
"epoch": 0.78,
"grad_norm": 2.599943943825198,
"learning_rate": 6.717592344545168e-06,
"loss": 1.242,
"step": 4165
},
{
"epoch": 0.78,
"grad_norm": 2.4915493854484976,
"learning_rate": 6.710712441415252e-06,
"loss": 1.1962,
"step": 4170
},
{
"epoch": 0.78,
"grad_norm": 2.1027225494150255,
"learning_rate": 6.7038288679725224e-06,
"loss": 1.2139,
"step": 4175
},
{
"epoch": 0.78,
"grad_norm": 2.356586756973305,
"learning_rate": 6.6969416389856065e-06,
"loss": 1.2431,
"step": 4180
},
{
"epoch": 0.78,
"grad_norm": 2.156520158907269,
"learning_rate": 6.690050769230968e-06,
"loss": 1.1878,
"step": 4185
},
{
"epoch": 0.78,
"grad_norm": 2.5483349132387425,
"learning_rate": 6.683156273492884e-06,
"loss": 1.1834,
"step": 4190
},
{
"epoch": 0.78,
"grad_norm": 2.283013595177182,
"learning_rate": 6.676258166563414e-06,
"loss": 1.259,
"step": 4195
},
{
"epoch": 0.78,
"grad_norm": 2.4877649502496584,
"learning_rate": 6.6693564632423626e-06,
"loss": 1.1508,
"step": 4200
},
{
"epoch": 0.78,
"grad_norm": 2.4403429188173584,
"learning_rate": 6.6624511783372494e-06,
"loss": 1.2324,
"step": 4205
},
{
"epoch": 0.79,
"grad_norm": 2.5999418377938555,
"learning_rate": 6.655542326663278e-06,
"loss": 1.2015,
"step": 4210
},
{
"epoch": 0.79,
"grad_norm": 2.3817244646066533,
"learning_rate": 6.648629923043308e-06,
"loss": 1.1421,
"step": 4215
},
{
"epoch": 0.79,
"grad_norm": 2.7861597820205075,
"learning_rate": 6.641713982307819e-06,
"loss": 1.2487,
"step": 4220
},
{
"epoch": 0.79,
"grad_norm": 2.604737390676733,
"learning_rate": 6.634794519294876e-06,
"loss": 1.1909,
"step": 4225
},
{
"epoch": 0.79,
"grad_norm": 2.1490298175211113,
"learning_rate": 6.627871548850101e-06,
"loss": 1.2128,
"step": 4230
},
{
"epoch": 0.79,
"grad_norm": 2.7478682969064128,
"learning_rate": 6.620945085826648e-06,
"loss": 1.2724,
"step": 4235
},
{
"epoch": 0.79,
"grad_norm": 2.4473416505881747,
"learning_rate": 6.614015145085157e-06,
"loss": 1.1413,
"step": 4240
},
{
"epoch": 0.79,
"grad_norm": 2.664153209181916,
"learning_rate": 6.607081741493731e-06,
"loss": 1.1926,
"step": 4245
},
{
"epoch": 0.79,
"grad_norm": 2.664772133407704,
"learning_rate": 6.6001448899279045e-06,
"loss": 1.2905,
"step": 4250
},
{
"epoch": 0.79,
"grad_norm": 2.463824076819037,
"learning_rate": 6.593204605270608e-06,
"loss": 1.1638,
"step": 4255
},
{
"epoch": 0.79,
"grad_norm": 2.754141316699185,
"learning_rate": 6.58626090241214e-06,
"loss": 1.2046,
"step": 4260
},
{
"epoch": 0.8,
"grad_norm": 2.2989864973374483,
"learning_rate": 6.579313796250127e-06,
"loss": 1.1798,
"step": 4265
},
{
"epoch": 0.8,
"grad_norm": 2.560534271429197,
"learning_rate": 6.572363301689504e-06,
"loss": 1.2023,
"step": 4270
},
{
"epoch": 0.8,
"grad_norm": 2.278717838505588,
"learning_rate": 6.565409433642471e-06,
"loss": 1.1969,
"step": 4275
},
{
"epoch": 0.8,
"grad_norm": 2.4833739605923917,
"learning_rate": 6.55845220702847e-06,
"loss": 1.1965,
"step": 4280
},
{
"epoch": 0.8,
"grad_norm": 2.430188408103723,
"learning_rate": 6.551491636774145e-06,
"loss": 1.2309,
"step": 4285
},
{
"epoch": 0.8,
"grad_norm": 2.6975259952464294,
"learning_rate": 6.544527737813315e-06,
"loss": 1.192,
"step": 4290
},
{
"epoch": 0.8,
"grad_norm": 2.4624516425401946,
"learning_rate": 6.5375605250869434e-06,
"loss": 1.2492,
"step": 4295
},
{
"epoch": 0.8,
"grad_norm": 2.562999083675579,
"learning_rate": 6.530590013543096e-06,
"loss": 1.1914,
"step": 4300
},
{
"epoch": 0.8,
"grad_norm": 2.1022978731614415,
"learning_rate": 6.523616218136924e-06,
"loss": 1.1834,
"step": 4305
},
{
"epoch": 0.8,
"grad_norm": 2.2686047736925516,
"learning_rate": 6.51663915383062e-06,
"loss": 1.2209,
"step": 4310
},
{
"epoch": 0.8,
"grad_norm": 2.5183595374337884,
"learning_rate": 6.509658835593391e-06,
"loss": 1.1786,
"step": 4315
},
{
"epoch": 0.81,
"grad_norm": 2.4391880597280107,
"learning_rate": 6.502675278401426e-06,
"loss": 1.1716,
"step": 4320
},
{
"epoch": 0.81,
"grad_norm": 2.484001041771429,
"learning_rate": 6.495688497237861e-06,
"loss": 1.2571,
"step": 4325
},
{
"epoch": 0.81,
"grad_norm": 2.2961665977309793,
"learning_rate": 6.48869850709275e-06,
"loss": 1.1841,
"step": 4330
},
{
"epoch": 0.81,
"grad_norm": 2.816933957342352,
"learning_rate": 6.481705322963035e-06,
"loss": 1.1983,
"step": 4335
},
{
"epoch": 0.81,
"grad_norm": 2.417764873329051,
"learning_rate": 6.474708959852504e-06,
"loss": 1.1827,
"step": 4340
},
{
"epoch": 0.81,
"grad_norm": 2.30617051197943,
"learning_rate": 6.467709432771772e-06,
"loss": 1.2341,
"step": 4345
},
{
"epoch": 0.81,
"grad_norm": 2.4526028616620983,
"learning_rate": 6.4607067567382385e-06,
"loss": 1.2723,
"step": 4350
},
{
"epoch": 0.81,
"grad_norm": 3.2024751951646775,
"learning_rate": 6.4537009467760605e-06,
"loss": 1.2418,
"step": 4355
},
{
"epoch": 0.81,
"grad_norm": 2.334451571090072,
"learning_rate": 6.446692017916118e-06,
"loss": 1.1688,
"step": 4360
},
{
"epoch": 0.81,
"grad_norm": 2.6791856005998147,
"learning_rate": 6.4396799851959815e-06,
"loss": 1.1688,
"step": 4365
},
{
"epoch": 0.81,
"grad_norm": 3.4087530077465655,
"learning_rate": 6.432664863659886e-06,
"loss": 1.229,
"step": 4370
},
{
"epoch": 0.82,
"grad_norm": 2.4259591534025837,
"learning_rate": 6.425646668358686e-06,
"loss": 1.2277,
"step": 4375
},
{
"epoch": 0.82,
"grad_norm": 2.495323027270627,
"learning_rate": 6.418625414349837e-06,
"loss": 1.2399,
"step": 4380
},
{
"epoch": 0.82,
"grad_norm": 2.896598795176137,
"learning_rate": 6.411601116697352e-06,
"loss": 1.2328,
"step": 4385
},
{
"epoch": 0.82,
"grad_norm": 2.458597929411267,
"learning_rate": 6.404573790471778e-06,
"loss": 1.197,
"step": 4390
},
{
"epoch": 0.82,
"grad_norm": 2.4345585639841105,
"learning_rate": 6.39754345075016e-06,
"loss": 1.268,
"step": 4395
},
{
"epoch": 0.82,
"grad_norm": 3.1164297658059166,
"learning_rate": 6.390510112616004e-06,
"loss": 1.2478,
"step": 4400
},
{
"epoch": 0.82,
"grad_norm": 2.3126599056846207,
"learning_rate": 6.383473791159252e-06,
"loss": 1.1894,
"step": 4405
},
{
"epoch": 0.82,
"grad_norm": 2.6416467416762117,
"learning_rate": 6.376434501476246e-06,
"loss": 1.2318,
"step": 4410
},
{
"epoch": 0.82,
"grad_norm": 2.5167282692335986,
"learning_rate": 6.3693922586697e-06,
"loss": 1.1838,
"step": 4415
},
{
"epoch": 0.82,
"grad_norm": 2.4469171349202745,
"learning_rate": 6.362347077848656e-06,
"loss": 1.1867,
"step": 4420
},
{
"epoch": 0.83,
"grad_norm": 2.7546951334155323,
"learning_rate": 6.355298974128467e-06,
"loss": 1.2411,
"step": 4425
},
{
"epoch": 0.83,
"grad_norm": 2.2263388765183314,
"learning_rate": 6.348247962630752e-06,
"loss": 1.2784,
"step": 4430
},
{
"epoch": 0.83,
"grad_norm": 2.4165973613827636,
"learning_rate": 6.341194058483375e-06,
"loss": 1.1628,
"step": 4435
},
{
"epoch": 0.83,
"grad_norm": 2.5473458867918954,
"learning_rate": 6.3341372768203965e-06,
"loss": 1.199,
"step": 4440
},
{
"epoch": 0.83,
"grad_norm": 2.5383914426106053,
"learning_rate": 6.3270776327820606e-06,
"loss": 1.1682,
"step": 4445
},
{
"epoch": 0.83,
"grad_norm": 2.4476722364188306,
"learning_rate": 6.320015141514744e-06,
"loss": 1.227,
"step": 4450
},
{
"epoch": 0.83,
"grad_norm": 2.5620633286885006,
"learning_rate": 6.31294981817094e-06,
"loss": 1.1892,
"step": 4455
},
{
"epoch": 0.83,
"grad_norm": 2.371789840507076,
"learning_rate": 6.305881677909211e-06,
"loss": 1.23,
"step": 4460
},
{
"epoch": 0.83,
"grad_norm": 2.316049180608952,
"learning_rate": 6.298810735894167e-06,
"loss": 1.2697,
"step": 4465
},
{
"epoch": 0.83,
"grad_norm": 2.6693599514255,
"learning_rate": 6.29173700729643e-06,
"loss": 1.2239,
"step": 4470
},
{
"epoch": 0.83,
"grad_norm": 2.5293838325304963,
"learning_rate": 6.284660507292599e-06,
"loss": 1.2408,
"step": 4475
},
{
"epoch": 0.84,
"grad_norm": 2.3739300942695007,
"learning_rate": 6.277581251065217e-06,
"loss": 1.2146,
"step": 4480
},
{
"epoch": 0.84,
"grad_norm": 2.1861207671387173,
"learning_rate": 6.270499253802742e-06,
"loss": 1.1965,
"step": 4485
},
{
"epoch": 0.84,
"grad_norm": 2.4626377264331505,
"learning_rate": 6.263414530699516e-06,
"loss": 1.2275,
"step": 4490
},
{
"epoch": 0.84,
"grad_norm": 2.4393045656196857,
"learning_rate": 6.256327096955726e-06,
"loss": 1.2562,
"step": 4495
},
{
"epoch": 0.84,
"grad_norm": 2.420788009586598,
"learning_rate": 6.249236967777373e-06,
"loss": 1.2532,
"step": 4500
},
{
"epoch": 0.84,
"grad_norm": 2.3682500570454206,
"learning_rate": 6.242144158376245e-06,
"loss": 1.2035,
"step": 4505
},
{
"epoch": 0.84,
"grad_norm": 2.617981119739426,
"learning_rate": 6.235048683969879e-06,
"loss": 1.2266,
"step": 4510
},
{
"epoch": 0.84,
"grad_norm": 2.6458940270866633,
"learning_rate": 6.227950559781528e-06,
"loss": 1.228,
"step": 4515
},
{
"epoch": 0.84,
"grad_norm": 2.589007218774909,
"learning_rate": 6.220849801040133e-06,
"loss": 1.2056,
"step": 4520
},
{
"epoch": 0.84,
"grad_norm": 2.877953150570889,
"learning_rate": 6.2137464229802835e-06,
"loss": 1.2288,
"step": 4525
},
{
"epoch": 0.84,
"grad_norm": 2.8622664846448544,
"learning_rate": 6.206640440842193e-06,
"loss": 1.2137,
"step": 4530
},
{
"epoch": 0.85,
"grad_norm": 2.633031891709981,
"learning_rate": 6.199531869871661e-06,
"loss": 1.2116,
"step": 4535
},
{
"epoch": 0.85,
"grad_norm": 2.524845529953313,
"learning_rate": 6.192420725320038e-06,
"loss": 1.1503,
"step": 4540
},
{
"epoch": 0.85,
"grad_norm": 2.1461215608130306,
"learning_rate": 6.185307022444198e-06,
"loss": 1.2836,
"step": 4545
},
{
"epoch": 0.85,
"grad_norm": 2.3233940868358087,
"learning_rate": 6.178190776506507e-06,
"loss": 1.2276,
"step": 4550
},
{
"epoch": 0.85,
"grad_norm": 2.571444746747503,
"learning_rate": 6.1710720027747835e-06,
"loss": 1.2365,
"step": 4555
},
{
"epoch": 0.85,
"grad_norm": 2.442325328268483,
"learning_rate": 6.163950716522268e-06,
"loss": 1.199,
"step": 4560
},
{
"epoch": 0.85,
"grad_norm": 2.8973344425831624,
"learning_rate": 6.1568269330275974e-06,
"loss": 1.2247,
"step": 4565
},
{
"epoch": 0.85,
"grad_norm": 2.265438975600916,
"learning_rate": 6.149700667574761e-06,
"loss": 1.2674,
"step": 4570
},
{
"epoch": 0.85,
"grad_norm": 2.427805906563997,
"learning_rate": 6.142571935453074e-06,
"loss": 1.2244,
"step": 4575
},
{
"epoch": 0.85,
"grad_norm": 2.71037186034558,
"learning_rate": 6.135440751957147e-06,
"loss": 1.2107,
"step": 4580
},
{
"epoch": 0.86,
"grad_norm": 2.47935744731514,
"learning_rate": 6.128307132386846e-06,
"loss": 1.2215,
"step": 4585
},
{
"epoch": 0.86,
"grad_norm": 2.4534851470879913,
"learning_rate": 6.121171092047267e-06,
"loss": 1.2313,
"step": 4590
},
{
"epoch": 0.86,
"grad_norm": 2.4955695058785343,
"learning_rate": 6.114032646248697e-06,
"loss": 1.1893,
"step": 4595
},
{
"epoch": 0.86,
"grad_norm": 2.7299946810100377,
"learning_rate": 6.106891810306584e-06,
"loss": 1.2553,
"step": 4600
},
{
"epoch": 0.86,
"grad_norm": 2.4519592083822843,
"learning_rate": 6.0997485995415065e-06,
"loss": 1.1517,
"step": 4605
},
{
"epoch": 0.86,
"grad_norm": 2.8237625939623467,
"learning_rate": 6.092603029279136e-06,
"loss": 1.2106,
"step": 4610
},
{
"epoch": 0.86,
"grad_norm": 2.2919033004960423,
"learning_rate": 6.0854551148502085e-06,
"loss": 1.2014,
"step": 4615
},
{
"epoch": 0.86,
"grad_norm": 2.353224396529788,
"learning_rate": 6.078304871590485e-06,
"loss": 1.2138,
"step": 4620
},
{
"epoch": 0.86,
"grad_norm": 2.5985774084094504,
"learning_rate": 6.071152314840726e-06,
"loss": 1.2418,
"step": 4625
},
{
"epoch": 0.86,
"grad_norm": 2.6678992574870835,
"learning_rate": 6.0639974599466585e-06,
"loss": 1.2188,
"step": 4630
},
{
"epoch": 0.86,
"grad_norm": 2.4496687988011967,
"learning_rate": 6.056840322258934e-06,
"loss": 1.1627,
"step": 4635
},
{
"epoch": 0.87,
"grad_norm": 2.4241155011469644,
"learning_rate": 6.049680917133106e-06,
"loss": 1.245,
"step": 4640
},
{
"epoch": 0.87,
"grad_norm": 2.381657027399795,
"learning_rate": 6.04251925992959e-06,
"loss": 1.188,
"step": 4645
},
{
"epoch": 0.87,
"grad_norm": 2.832348257790817,
"learning_rate": 6.035355366013638e-06,
"loss": 1.207,
"step": 4650
},
{
"epoch": 0.87,
"grad_norm": 2.5691320543151273,
"learning_rate": 6.028189250755293e-06,
"loss": 1.2605,
"step": 4655
},
{
"epoch": 0.87,
"grad_norm": 2.9545303881123983,
"learning_rate": 6.021020929529371e-06,
"loss": 1.1689,
"step": 4660
},
{
"epoch": 0.87,
"grad_norm": 2.651237945956081,
"learning_rate": 6.013850417715419e-06,
"loss": 1.24,
"step": 4665
},
{
"epoch": 0.87,
"grad_norm": 2.674374005358842,
"learning_rate": 6.006677730697682e-06,
"loss": 1.1851,
"step": 4670
},
{
"epoch": 0.87,
"grad_norm": 2.360244102364902,
"learning_rate": 5.999502883865073e-06,
"loss": 1.2477,
"step": 4675
},
{
"epoch": 0.87,
"grad_norm": 2.6191630744555634,
"learning_rate": 5.992325892611139e-06,
"loss": 1.2689,
"step": 4680
},
{
"epoch": 0.87,
"grad_norm": 2.342487040862795,
"learning_rate": 5.9851467723340295e-06,
"loss": 1.2413,
"step": 4685
},
{
"epoch": 0.87,
"grad_norm": 2.4245312125321803,
"learning_rate": 5.977965538436458e-06,
"loss": 1.2348,
"step": 4690
},
{
"epoch": 0.88,
"grad_norm": 2.2720239156004025,
"learning_rate": 5.9707822063256745e-06,
"loss": 1.1965,
"step": 4695
},
{
"epoch": 0.88,
"grad_norm": 2.5605489902324408,
"learning_rate": 5.9635967914134316e-06,
"loss": 1.2057,
"step": 4700
},
{
"epoch": 0.88,
"grad_norm": 2.55093673845062,
"learning_rate": 5.956409309115952e-06,
"loss": 1.2399,
"step": 4705
},
{
"epoch": 0.88,
"grad_norm": 2.482655438785984,
"learning_rate": 5.94921977485389e-06,
"loss": 1.2163,
"step": 4710
},
{
"epoch": 0.88,
"grad_norm": 2.92841769520834,
"learning_rate": 5.942028204052304e-06,
"loss": 1.1962,
"step": 4715
},
{
"epoch": 0.88,
"grad_norm": 2.403145984699388,
"learning_rate": 5.934834612140621e-06,
"loss": 1.2121,
"step": 4720
},
{
"epoch": 0.88,
"grad_norm": 2.5665557751164987,
"learning_rate": 5.927639014552608e-06,
"loss": 1.1943,
"step": 4725
},
{
"epoch": 0.88,
"grad_norm": 2.4345512205304614,
"learning_rate": 5.920441426726331e-06,
"loss": 1.2182,
"step": 4730
},
{
"epoch": 0.88,
"grad_norm": 2.4570381781809294,
"learning_rate": 5.9132418641041256e-06,
"loss": 1.1541,
"step": 4735
},
{
"epoch": 0.88,
"grad_norm": 2.42607286783389,
"learning_rate": 5.90604034213257e-06,
"loss": 1.2285,
"step": 4740
},
{
"epoch": 0.88,
"grad_norm": 2.1059221387409184,
"learning_rate": 5.898836876262439e-06,
"loss": 1.2695,
"step": 4745
},
{
"epoch": 0.89,
"grad_norm": 2.389763480309045,
"learning_rate": 5.891631481948684e-06,
"loss": 1.2433,
"step": 4750
},
{
"epoch": 0.89,
"grad_norm": 2.4520356682997058,
"learning_rate": 5.88442417465039e-06,
"loss": 1.2204,
"step": 4755
},
{
"epoch": 0.89,
"grad_norm": 2.466763311510685,
"learning_rate": 5.877214969830746e-06,
"loss": 1.1942,
"step": 4760
},
{
"epoch": 0.89,
"grad_norm": 2.456029969725007,
"learning_rate": 5.870003882957015e-06,
"loss": 1.224,
"step": 4765
},
{
"epoch": 0.89,
"grad_norm": 2.3061090049071393,
"learning_rate": 5.862790929500497e-06,
"loss": 1.1981,
"step": 4770
},
{
"epoch": 0.89,
"grad_norm": 2.4913818728157646,
"learning_rate": 5.855576124936496e-06,
"loss": 1.2408,
"step": 4775
},
{
"epoch": 0.89,
"grad_norm": 2.2335777978602813,
"learning_rate": 5.848359484744286e-06,
"loss": 1.1723,
"step": 4780
},
{
"epoch": 0.89,
"grad_norm": 2.347554354081274,
"learning_rate": 5.841141024407083e-06,
"loss": 1.2121,
"step": 4785
},
{
"epoch": 0.89,
"grad_norm": 2.3692356900549623,
"learning_rate": 5.833920759412006e-06,
"loss": 1.2553,
"step": 4790
},
{
"epoch": 0.89,
"grad_norm": 2.5382367232442076,
"learning_rate": 5.826698705250047e-06,
"loss": 1.2015,
"step": 4795
},
{
"epoch": 0.9,
"grad_norm": 2.2459980708439216,
"learning_rate": 5.819474877416031e-06,
"loss": 1.2305,
"step": 4800
},
{
"epoch": 0.9,
"grad_norm": 2.546675162324036,
"learning_rate": 5.8122492914086e-06,
"loss": 1.2121,
"step": 4805
},
{
"epoch": 0.9,
"grad_norm": 2.677752571152061,
"learning_rate": 5.805021962730155e-06,
"loss": 1.2224,
"step": 4810
},
{
"epoch": 0.9,
"grad_norm": 2.212925811044383,
"learning_rate": 5.797792906886847e-06,
"loss": 1.1743,
"step": 4815
},
{
"epoch": 0.9,
"grad_norm": 2.6409343572042823,
"learning_rate": 5.790562139388522e-06,
"loss": 1.1881,
"step": 4820
},
{
"epoch": 0.9,
"grad_norm": 2.3756770154413114,
"learning_rate": 5.783329675748711e-06,
"loss": 1.2045,
"step": 4825
},
{
"epoch": 0.9,
"grad_norm": 2.443951055189872,
"learning_rate": 5.776095531484574e-06,
"loss": 1.1974,
"step": 4830
},
{
"epoch": 0.9,
"grad_norm": 2.5850671647243453,
"learning_rate": 5.768859722116876e-06,
"loss": 1.2254,
"step": 4835
},
{
"epoch": 0.9,
"grad_norm": 2.9605499293438924,
"learning_rate": 5.761622263169964e-06,
"loss": 1.2438,
"step": 4840
},
{
"epoch": 0.9,
"grad_norm": 2.56592411129834,
"learning_rate": 5.754383170171716e-06,
"loss": 1.2011,
"step": 4845
},
{
"epoch": 0.9,
"grad_norm": 2.4821775141535216,
"learning_rate": 5.747142458653518e-06,
"loss": 1.2443,
"step": 4850
},
{
"epoch": 0.91,
"grad_norm": 2.4637437926975387,
"learning_rate": 5.7399001441502285e-06,
"loss": 1.2813,
"step": 4855
},
{
"epoch": 0.91,
"grad_norm": 2.7814434455793617,
"learning_rate": 5.732656242200148e-06,
"loss": 1.2608,
"step": 4860
},
{
"epoch": 0.91,
"grad_norm": 2.6231313656942885,
"learning_rate": 5.725410768344977e-06,
"loss": 1.1879,
"step": 4865
},
{
"epoch": 0.91,
"grad_norm": 2.455722926219591,
"learning_rate": 5.718163738129796e-06,
"loss": 1.191,
"step": 4870
},
{
"epoch": 0.91,
"grad_norm": 2.4432078663934735,
"learning_rate": 5.710915167103015e-06,
"loss": 1.2193,
"step": 4875
},
{
"epoch": 0.91,
"grad_norm": 2.424558271609936,
"learning_rate": 5.703665070816361e-06,
"loss": 1.19,
"step": 4880
},
{
"epoch": 0.91,
"grad_norm": 2.6356126763167746,
"learning_rate": 5.696413464824824e-06,
"loss": 1.2016,
"step": 4885
},
{
"epoch": 0.91,
"grad_norm": 2.449397519305924,
"learning_rate": 5.689160364686637e-06,
"loss": 1.1989,
"step": 4890
},
{
"epoch": 0.91,
"grad_norm": 2.459503260258635,
"learning_rate": 5.681905785963241e-06,
"loss": 1.1441,
"step": 4895
},
{
"epoch": 0.91,
"grad_norm": 2.3464241399602597,
"learning_rate": 5.6746497442192425e-06,
"loss": 1.2284,
"step": 4900
},
{
"epoch": 0.91,
"grad_norm": 3.1074072930659193,
"learning_rate": 5.667392255022396e-06,
"loss": 1.2021,
"step": 4905
},
{
"epoch": 0.92,
"grad_norm": 2.7749582202860004,
"learning_rate": 5.660133333943552e-06,
"loss": 1.1882,
"step": 4910
},
{
"epoch": 0.92,
"grad_norm": 2.9527764645138235,
"learning_rate": 5.652872996556643e-06,
"loss": 1.2202,
"step": 4915
},
{
"epoch": 0.92,
"grad_norm": 3.3891739848268343,
"learning_rate": 5.645611258438629e-06,
"loss": 1.2491,
"step": 4920
},
{
"epoch": 0.92,
"grad_norm": 2.7067471875098437,
"learning_rate": 5.638348135169486e-06,
"loss": 1.1965,
"step": 4925
},
{
"epoch": 0.92,
"grad_norm": 2.3502698720472175,
"learning_rate": 5.631083642332154e-06,
"loss": 1.1626,
"step": 4930
},
{
"epoch": 0.92,
"grad_norm": 2.6969350244271646,
"learning_rate": 5.623817795512515e-06,
"loss": 1.2453,
"step": 4935
},
{
"epoch": 0.92,
"grad_norm": 2.400025323822739,
"learning_rate": 5.616550610299355e-06,
"loss": 1.2092,
"step": 4940
},
{
"epoch": 0.92,
"grad_norm": 2.6363089823661365,
"learning_rate": 5.609282102284335e-06,
"loss": 1.1759,
"step": 4945
},
{
"epoch": 0.92,
"grad_norm": 2.4198122938368276,
"learning_rate": 5.602012287061945e-06,
"loss": 1.2302,
"step": 4950
},
{
"epoch": 0.92,
"grad_norm": 2.6128455829596833,
"learning_rate": 5.594741180229491e-06,
"loss": 1.1694,
"step": 4955
},
{
"epoch": 0.92,
"grad_norm": 2.4408008785514927,
"learning_rate": 5.58746879738704e-06,
"loss": 1.1767,
"step": 4960
},
{
"epoch": 0.93,
"grad_norm": 2.686534311568624,
"learning_rate": 5.580195154137405e-06,
"loss": 1.2308,
"step": 4965
},
{
"epoch": 0.93,
"grad_norm": 2.358550700692953,
"learning_rate": 5.5729202660860984e-06,
"loss": 1.2395,
"step": 4970
},
{
"epoch": 0.93,
"grad_norm": 2.4836834824616965,
"learning_rate": 5.565644148841303e-06,
"loss": 1.1877,
"step": 4975
},
{
"epoch": 0.93,
"grad_norm": 2.9208432064602214,
"learning_rate": 5.558366818013842e-06,
"loss": 1.157,
"step": 4980
},
{
"epoch": 0.93,
"grad_norm": 2.864971854169532,
"learning_rate": 5.551088289217139e-06,
"loss": 1.2356,
"step": 4985
},
{
"epoch": 0.93,
"grad_norm": 2.1570568875709517,
"learning_rate": 5.54380857806719e-06,
"loss": 1.1621,
"step": 4990
},
{
"epoch": 0.93,
"grad_norm": 2.825445528201766,
"learning_rate": 5.536527700182526e-06,
"loss": 1.2424,
"step": 4995
},
{
"epoch": 0.93,
"grad_norm": 2.541720844778179,
"learning_rate": 5.5292456711841845e-06,
"loss": 1.2044,
"step": 5000
},
{
"epoch": 0.93,
"grad_norm": 2.9093417603501726,
"learning_rate": 5.521962506695671e-06,
"loss": 1.2173,
"step": 5005
},
{
"epoch": 0.93,
"grad_norm": 2.385465021207973,
"learning_rate": 5.514678222342922e-06,
"loss": 1.1977,
"step": 5010
},
{
"epoch": 0.94,
"grad_norm": 2.647317844287573,
"learning_rate": 5.507392833754285e-06,
"loss": 1.1785,
"step": 5015
},
{
"epoch": 0.94,
"grad_norm": 2.473551842843763,
"learning_rate": 5.5001063565604705e-06,
"loss": 1.2015,
"step": 5020
},
{
"epoch": 0.94,
"grad_norm": 2.7241247663506782,
"learning_rate": 5.492818806394532e-06,
"loss": 1.2281,
"step": 5025
},
{
"epoch": 0.94,
"grad_norm": 2.518896282402336,
"learning_rate": 5.485530198891814e-06,
"loss": 1.2264,
"step": 5030
},
{
"epoch": 0.94,
"grad_norm": 2.4126046214585672,
"learning_rate": 5.47824054968994e-06,
"loss": 1.206,
"step": 5035
},
{
"epoch": 0.94,
"grad_norm": 2.724587523366145,
"learning_rate": 5.47094987442876e-06,
"loss": 1.1847,
"step": 5040
},
{
"epoch": 0.94,
"grad_norm": 2.1906266563183596,
"learning_rate": 5.463658188750333e-06,
"loss": 1.1167,
"step": 5045
},
{
"epoch": 0.94,
"grad_norm": 2.636133430468642,
"learning_rate": 5.456365508298882e-06,
"loss": 1.1843,
"step": 5050
},
{
"epoch": 0.94,
"grad_norm": 2.498460580189244,
"learning_rate": 5.449071848720762e-06,
"loss": 1.1861,
"step": 5055
},
{
"epoch": 0.94,
"grad_norm": 2.6456079430925814,
"learning_rate": 5.441777225664434e-06,
"loss": 1.2144,
"step": 5060
},
{
"epoch": 0.94,
"grad_norm": 2.635032475206284,
"learning_rate": 5.434481654780421e-06,
"loss": 1.1949,
"step": 5065
},
{
"epoch": 0.95,
"grad_norm": 2.414349835457233,
"learning_rate": 5.427185151721284e-06,
"loss": 1.1837,
"step": 5070
},
{
"epoch": 0.95,
"grad_norm": 2.517171242137346,
"learning_rate": 5.4198877321415785e-06,
"loss": 1.1937,
"step": 5075
},
{
"epoch": 0.95,
"grad_norm": 2.3322860590179593,
"learning_rate": 5.412589411697835e-06,
"loss": 1.221,
"step": 5080
},
{
"epoch": 0.95,
"grad_norm": 2.954235361194853,
"learning_rate": 5.405290206048508e-06,
"loss": 1.2411,
"step": 5085
},
{
"epoch": 0.95,
"grad_norm": 2.468404065206049,
"learning_rate": 5.397990130853955e-06,
"loss": 1.1789,
"step": 5090
},
{
"epoch": 0.95,
"grad_norm": 2.355727263917547,
"learning_rate": 5.390689201776399e-06,
"loss": 1.1613,
"step": 5095
},
{
"epoch": 0.95,
"grad_norm": 2.2426355580536663,
"learning_rate": 5.383387434479895e-06,
"loss": 1.208,
"step": 5100
},
{
"epoch": 0.95,
"grad_norm": 2.380680152812773,
"learning_rate": 5.376084844630298e-06,
"loss": 1.2055,
"step": 5105
},
{
"epoch": 0.95,
"grad_norm": 2.569418503583293,
"learning_rate": 5.3687814478952235e-06,
"loss": 1.184,
"step": 5110
},
{
"epoch": 0.95,
"grad_norm": 2.526665130298868,
"learning_rate": 5.361477259944022e-06,
"loss": 1.2064,
"step": 5115
},
{
"epoch": 0.95,
"grad_norm": 2.369294746436644,
"learning_rate": 5.35417229644774e-06,
"loss": 1.2072,
"step": 5120
},
{
"epoch": 0.96,
"grad_norm": 2.720314631994901,
"learning_rate": 5.346866573079088e-06,
"loss": 1.2119,
"step": 5125
},
{
"epoch": 0.96,
"grad_norm": 2.3314723731875633,
"learning_rate": 5.339560105512406e-06,
"loss": 1.1816,
"step": 5130
},
{
"epoch": 0.96,
"grad_norm": 2.26140189167418,
"learning_rate": 5.332252909423634e-06,
"loss": 1.1715,
"step": 5135
},
{
"epoch": 0.96,
"grad_norm": 2.7679372661959385,
"learning_rate": 5.324945000490271e-06,
"loss": 1.2743,
"step": 5140
},
{
"epoch": 0.96,
"grad_norm": 2.411111902010626,
"learning_rate": 5.317636394391346e-06,
"loss": 1.2343,
"step": 5145
},
{
"epoch": 0.96,
"grad_norm": 3.389608636178554,
"learning_rate": 5.310327106807387e-06,
"loss": 1.1983,
"step": 5150
},
{
"epoch": 0.96,
"grad_norm": 2.5290136942085546,
"learning_rate": 5.303017153420382e-06,
"loss": 1.2139,
"step": 5155
},
{
"epoch": 0.96,
"grad_norm": 2.5697341004928056,
"learning_rate": 5.295706549913746e-06,
"loss": 1.1951,
"step": 5160
},
{
"epoch": 0.96,
"grad_norm": 2.6841860442942402,
"learning_rate": 5.288395311972291e-06,
"loss": 1.21,
"step": 5165
},
{
"epoch": 0.96,
"grad_norm": 2.427048051792136,
"learning_rate": 5.281083455282187e-06,
"loss": 1.1969,
"step": 5170
},
{
"epoch": 0.97,
"grad_norm": 2.382376197299203,
"learning_rate": 5.273770995530938e-06,
"loss": 1.1831,
"step": 5175
},
{
"epoch": 0.97,
"grad_norm": 2.5132668182963886,
"learning_rate": 5.266457948407336e-06,
"loss": 1.2145,
"step": 5180
},
{
"epoch": 0.97,
"grad_norm": 2.3797244199180527,
"learning_rate": 5.259144329601434e-06,
"loss": 1.1751,
"step": 5185
},
{
"epoch": 0.97,
"grad_norm": 2.3611741368911194,
"learning_rate": 5.251830154804515e-06,
"loss": 1.2064,
"step": 5190
},
{
"epoch": 0.97,
"grad_norm": 2.3567756326205287,
"learning_rate": 5.244515439709048e-06,
"loss": 1.2077,
"step": 5195
},
{
"epoch": 0.97,
"grad_norm": 2.224791750500246,
"learning_rate": 5.23720020000867e-06,
"loss": 1.2123,
"step": 5200
},
{
"epoch": 0.97,
"grad_norm": 2.4136633237187897,
"learning_rate": 5.229884451398137e-06,
"loss": 1.2132,
"step": 5205
},
{
"epoch": 0.97,
"grad_norm": 2.304370257710667,
"learning_rate": 5.2225682095733e-06,
"loss": 1.2289,
"step": 5210
},
{
"epoch": 0.97,
"grad_norm": 2.5814542726053635,
"learning_rate": 5.215251490231066e-06,
"loss": 1.272,
"step": 5215
},
{
"epoch": 0.97,
"grad_norm": 2.3550014377964037,
"learning_rate": 5.207934309069369e-06,
"loss": 1.2217,
"step": 5220
},
{
"epoch": 0.97,
"grad_norm": 2.4860052514393702,
"learning_rate": 5.2006166817871304e-06,
"loss": 1.227,
"step": 5225
},
{
"epoch": 0.98,
"grad_norm": 2.463931792592068,
"learning_rate": 5.193298624084233e-06,
"loss": 1.2201,
"step": 5230
},
{
"epoch": 0.98,
"grad_norm": 2.6186162346655157,
"learning_rate": 5.185980151661479e-06,
"loss": 1.2194,
"step": 5235
},
{
"epoch": 0.98,
"grad_norm": 2.4191065869581587,
"learning_rate": 5.178661280220566e-06,
"loss": 1.1958,
"step": 5240
},
{
"epoch": 0.98,
"grad_norm": 2.5133518806542865,
"learning_rate": 5.1713420254640405e-06,
"loss": 1.1761,
"step": 5245
},
{
"epoch": 0.98,
"grad_norm": 2.325396160641914,
"learning_rate": 5.164022403095274e-06,
"loss": 1.178,
"step": 5250
},
{
"epoch": 0.98,
"grad_norm": 2.54773375649405,
"learning_rate": 5.156702428818431e-06,
"loss": 1.2196,
"step": 5255
},
{
"epoch": 0.98,
"grad_norm": 2.4900490503130928,
"learning_rate": 5.149382118338426e-06,
"loss": 1.2077,
"step": 5260
},
{
"epoch": 0.98,
"grad_norm": 2.252151812435276,
"learning_rate": 5.142061487360896e-06,
"loss": 1.23,
"step": 5265
},
{
"epoch": 0.98,
"grad_norm": 2.4776213894416377,
"learning_rate": 5.134740551592165e-06,
"loss": 1.2005,
"step": 5270
},
{
"epoch": 0.98,
"grad_norm": 2.4952786157824662,
"learning_rate": 5.127419326739216e-06,
"loss": 1.1988,
"step": 5275
},
{
"epoch": 0.98,
"grad_norm": 2.144004636814163,
"learning_rate": 5.120097828509645e-06,
"loss": 1.2339,
"step": 5280
},
{
"epoch": 0.99,
"grad_norm": 2.3998029555969547,
"learning_rate": 5.112776072611638e-06,
"loss": 1.1466,
"step": 5285
},
{
"epoch": 0.99,
"grad_norm": 2.37290798802621,
"learning_rate": 5.105454074753933e-06,
"loss": 1.1743,
"step": 5290
},
{
"epoch": 0.99,
"grad_norm": 2.6835621702851475,
"learning_rate": 5.09813185064579e-06,
"loss": 1.1698,
"step": 5295
},
{
"epoch": 0.99,
"grad_norm": 2.512506207520888,
"learning_rate": 5.09080941599695e-06,
"loss": 1.1823,
"step": 5300
},
{
"epoch": 0.99,
"grad_norm": 2.137533406863869,
"learning_rate": 5.083486786517608e-06,
"loss": 1.1922,
"step": 5305
},
{
"epoch": 0.99,
"grad_norm": 4.104624434210896,
"learning_rate": 5.076163977918379e-06,
"loss": 1.2323,
"step": 5310
},
{
"epoch": 0.99,
"grad_norm": 2.683564602708311,
"learning_rate": 5.068841005910256e-06,
"loss": 1.2063,
"step": 5315
},
{
"epoch": 0.99,
"grad_norm": 2.7608580905626847,
"learning_rate": 5.061517886204592e-06,
"loss": 1.2186,
"step": 5320
},
{
"epoch": 0.99,
"grad_norm": 2.4056499009625663,
"learning_rate": 5.054194634513047e-06,
"loss": 1.1975,
"step": 5325
},
{
"epoch": 0.99,
"grad_norm": 2.2975592932071107,
"learning_rate": 5.046871266547569e-06,
"loss": 1.2494,
"step": 5330
},
{
"epoch": 0.99,
"grad_norm": 2.741698022771611,
"learning_rate": 5.039547798020358e-06,
"loss": 1.235,
"step": 5335
},
{
"epoch": 1.0,
"grad_norm": 2.5059532665114173,
"learning_rate": 5.0322242446438265e-06,
"loss": 1.2242,
"step": 5340
},
{
"epoch": 1.0,
"grad_norm": 2.52273255127118,
"learning_rate": 5.0249006221305675e-06,
"loss": 1.2645,
"step": 5345
},
{
"epoch": 1.0,
"grad_norm": 2.513615769567221,
"learning_rate": 5.017576946193325e-06,
"loss": 1.1787,
"step": 5350
},
{
"epoch": 1.0,
"grad_norm": 2.3227647148793498,
"learning_rate": 5.01025323254496e-06,
"loss": 1.144,
"step": 5355
},
{
"epoch": 1.0,
"grad_norm": 2.3674913562659863,
"learning_rate": 5.002929496898407e-06,
"loss": 1.2229,
"step": 5360
},
{
"epoch": 1.0,
"grad_norm": 2.681463697074552,
"learning_rate": 4.995605754966652e-06,
"loss": 1.0945,
"step": 5365
},
{
"epoch": 1.0,
"grad_norm": 2.3815705086639753,
"learning_rate": 4.988282022462699e-06,
"loss": 1.0172,
"step": 5370
},
{
"epoch": 1.0,
"grad_norm": 2.419736712855131,
"learning_rate": 4.980958315099525e-06,
"loss": 1.0478,
"step": 5375
},
{
"epoch": 1.0,
"grad_norm": 2.4222079854149547,
"learning_rate": 4.973634648590055e-06,
"loss": 1.0388,
"step": 5380
},
{
"epoch": 1.0,
"grad_norm": 2.5561149360584605,
"learning_rate": 4.9663110386471245e-06,
"loss": 1.026,
"step": 5385
},
{
"epoch": 1.01,
"grad_norm": 2.197462219377321,
"learning_rate": 4.958987500983455e-06,
"loss": 1.0145,
"step": 5390
},
{
"epoch": 1.01,
"grad_norm": 2.5821217339392653,
"learning_rate": 4.951664051311604e-06,
"loss": 0.9705,
"step": 5395
},
{
"epoch": 1.01,
"grad_norm": 2.7055799291037737,
"learning_rate": 4.944340705343947e-06,
"loss": 1.0235,
"step": 5400
},
{
"epoch": 1.01,
"grad_norm": 2.3518618011207413,
"learning_rate": 4.9370174787926284e-06,
"loss": 0.9809,
"step": 5405
},
{
"epoch": 1.01,
"grad_norm": 2.41358216404054,
"learning_rate": 4.929694387369548e-06,
"loss": 1.0209,
"step": 5410
},
{
"epoch": 1.01,
"grad_norm": 2.1864823466664203,
"learning_rate": 4.922371446786308e-06,
"loss": 1.0187,
"step": 5415
},
{
"epoch": 1.01,
"grad_norm": 2.4422953264549574,
"learning_rate": 4.915048672754189e-06,
"loss": 1.0231,
"step": 5420
},
{
"epoch": 1.01,
"grad_norm": 2.4734873276063487,
"learning_rate": 4.907726080984112e-06,
"loss": 1.0291,
"step": 5425
},
{
"epoch": 1.01,
"grad_norm": 2.7353098849284017,
"learning_rate": 4.900403687186607e-06,
"loss": 1.0135,
"step": 5430
},
{
"epoch": 1.01,
"grad_norm": 2.2107790506597755,
"learning_rate": 4.893081507071788e-06,
"loss": 1.0397,
"step": 5435
},
{
"epoch": 1.01,
"grad_norm": 2.4179695684073854,
"learning_rate": 4.8857595563492996e-06,
"loss": 1.0093,
"step": 5440
},
{
"epoch": 1.02,
"grad_norm": 2.3001310602664566,
"learning_rate": 4.878437850728298e-06,
"loss": 1.0419,
"step": 5445
},
{
"epoch": 1.02,
"grad_norm": 2.1598933391908686,
"learning_rate": 4.871116405917413e-06,
"loss": 1.0052,
"step": 5450
},
{
"epoch": 1.02,
"grad_norm": 2.4913258463261476,
"learning_rate": 4.863795237624719e-06,
"loss": 1.0391,
"step": 5455
},
{
"epoch": 1.02,
"grad_norm": 2.5724129547310493,
"learning_rate": 4.856474361557692e-06,
"loss": 1.0534,
"step": 5460
},
{
"epoch": 1.02,
"grad_norm": 2.3554782257142155,
"learning_rate": 4.849153793423182e-06,
"loss": 1.0422,
"step": 5465
},
{
"epoch": 1.02,
"grad_norm": 2.2282999376603057,
"learning_rate": 4.841833548927379e-06,
"loss": 1.0285,
"step": 5470
},
{
"epoch": 1.02,
"grad_norm": 2.4711423811860342,
"learning_rate": 4.834513643775784e-06,
"loss": 0.9753,
"step": 5475
},
{
"epoch": 1.02,
"grad_norm": 2.669156173349385,
"learning_rate": 4.827194093673158e-06,
"loss": 1.0262,
"step": 5480
},
{
"epoch": 1.02,
"grad_norm": 2.6338258701202966,
"learning_rate": 4.81987491432351e-06,
"loss": 1.0041,
"step": 5485
},
{
"epoch": 1.02,
"grad_norm": 2.3801846206028876,
"learning_rate": 4.812556121430051e-06,
"loss": 1.0225,
"step": 5490
},
{
"epoch": 1.02,
"grad_norm": 2.6627376022318034,
"learning_rate": 4.805237730695161e-06,
"loss": 0.9694,
"step": 5495
},
{
"epoch": 1.03,
"grad_norm": 2.7291035336136678,
"learning_rate": 4.7979197578203606e-06,
"loss": 1.0077,
"step": 5500
},
{
"epoch": 1.03,
"grad_norm": 2.463838481125147,
"learning_rate": 4.790602218506271e-06,
"loss": 1.0493,
"step": 5505
},
{
"epoch": 1.03,
"grad_norm": 2.433765492278771,
"learning_rate": 4.783285128452584e-06,
"loss": 1.0118,
"step": 5510
},
{
"epoch": 1.03,
"grad_norm": 2.607628992304912,
"learning_rate": 4.775968503358026e-06,
"loss": 1.0695,
"step": 5515
},
{
"epoch": 1.03,
"grad_norm": 2.5295928086957846,
"learning_rate": 4.76865235892033e-06,
"loss": 1.0166,
"step": 5520
},
{
"epoch": 1.03,
"grad_norm": 2.203736263547884,
"learning_rate": 4.761336710836195e-06,
"loss": 0.9946,
"step": 5525
},
{
"epoch": 1.03,
"grad_norm": 2.869557388787276,
"learning_rate": 4.7540215748012545e-06,
"loss": 0.985,
"step": 5530
},
{
"epoch": 1.03,
"grad_norm": 2.531865966904552,
"learning_rate": 4.746706966510043e-06,
"loss": 0.9826,
"step": 5535
},
{
"epoch": 1.03,
"grad_norm": 2.7954245771234665,
"learning_rate": 4.739392901655966e-06,
"loss": 1.0198,
"step": 5540
},
{
"epoch": 1.03,
"grad_norm": 2.282486161494159,
"learning_rate": 4.732079395931261e-06,
"loss": 1.0174,
"step": 5545
},
{
"epoch": 1.04,
"grad_norm": 2.5815196933075706,
"learning_rate": 4.724766465026965e-06,
"loss": 1.0575,
"step": 5550
},
{
"epoch": 1.04,
"grad_norm": 2.5209180513093745,
"learning_rate": 4.717454124632883e-06,
"loss": 1.0617,
"step": 5555
},
{
"epoch": 1.04,
"grad_norm": 2.375427874388073,
"learning_rate": 4.710142390437551e-06,
"loss": 0.9906,
"step": 5560
},
{
"epoch": 1.04,
"grad_norm": 2.6600796567133838,
"learning_rate": 4.702831278128209e-06,
"loss": 1.03,
"step": 5565
},
{
"epoch": 1.04,
"grad_norm": 2.9237698514498924,
"learning_rate": 4.695520803390758e-06,
"loss": 1.0186,
"step": 5570
},
{
"epoch": 1.04,
"grad_norm": 2.759992337244157,
"learning_rate": 4.688210981909734e-06,
"loss": 0.9961,
"step": 5575
},
{
"epoch": 1.04,
"grad_norm": 2.536540198344336,
"learning_rate": 4.680901829368268e-06,
"loss": 1.075,
"step": 5580
},
{
"epoch": 1.04,
"grad_norm": 2.604142431895161,
"learning_rate": 4.673593361448062e-06,
"loss": 1.0397,
"step": 5585
},
{
"epoch": 1.04,
"grad_norm": 2.3833711820364543,
"learning_rate": 4.666285593829343e-06,
"loss": 0.9868,
"step": 5590
},
{
"epoch": 1.04,
"grad_norm": 2.6825074384656213,
"learning_rate": 4.658978542190839e-06,
"loss": 0.9972,
"step": 5595
},
{
"epoch": 1.04,
"grad_norm": 2.7437235400270397,
"learning_rate": 4.651672222209738e-06,
"loss": 1.0479,
"step": 5600
},
{
"epoch": 1.05,
"grad_norm": 2.140569839724797,
"learning_rate": 4.6443666495616646e-06,
"loss": 0.9903,
"step": 5605
},
{
"epoch": 1.05,
"grad_norm": 2.5891585356017837,
"learning_rate": 4.637061839920634e-06,
"loss": 1.0201,
"step": 5610
},
{
"epoch": 1.05,
"grad_norm": 2.3985815097913132,
"learning_rate": 4.62975780895903e-06,
"loss": 0.9745,
"step": 5615
},
{
"epoch": 1.05,
"grad_norm": 2.4716522327814774,
"learning_rate": 4.622454572347558e-06,
"loss": 1.0602,
"step": 5620
},
{
"epoch": 1.05,
"grad_norm": 2.4858109653750677,
"learning_rate": 4.615152145755224e-06,
"loss": 1.0142,
"step": 5625
},
{
"epoch": 1.05,
"grad_norm": 2.65100239334809,
"learning_rate": 4.6078505448492985e-06,
"loss": 0.9855,
"step": 5630
},
{
"epoch": 1.05,
"grad_norm": 2.622925290593546,
"learning_rate": 4.600549785295278e-06,
"loss": 0.9852,
"step": 5635
},
{
"epoch": 1.05,
"grad_norm": 2.3714256985145266,
"learning_rate": 4.5932498827568525e-06,
"loss": 0.9897,
"step": 5640
},
{
"epoch": 1.05,
"grad_norm": 2.7272449228012237,
"learning_rate": 4.585950852895872e-06,
"loss": 1.0032,
"step": 5645
},
{
"epoch": 1.05,
"grad_norm": 2.291278250033059,
"learning_rate": 4.578652711372321e-06,
"loss": 1.0166,
"step": 5650
},
{
"epoch": 1.05,
"grad_norm": 2.2818866479353472,
"learning_rate": 4.571355473844271e-06,
"loss": 1.0199,
"step": 5655
},
{
"epoch": 1.06,
"grad_norm": 2.81859930508086,
"learning_rate": 4.564059155967857e-06,
"loss": 1.0655,
"step": 5660
},
{
"epoch": 1.06,
"grad_norm": 2.4198976668946934,
"learning_rate": 4.556763773397242e-06,
"loss": 0.9985,
"step": 5665
},
{
"epoch": 1.06,
"grad_norm": 2.6075117504033107,
"learning_rate": 4.549469341784578e-06,
"loss": 1.0957,
"step": 5670
},
{
"epoch": 1.06,
"grad_norm": 2.394284879916922,
"learning_rate": 4.542175876779982e-06,
"loss": 1.0545,
"step": 5675
},
{
"epoch": 1.06,
"grad_norm": 2.4984055889771932,
"learning_rate": 4.5348833940314945e-06,
"loss": 1.0149,
"step": 5680
},
{
"epoch": 1.06,
"grad_norm": 2.2787926703782895,
"learning_rate": 4.527591909185049e-06,
"loss": 0.9748,
"step": 5685
},
{
"epoch": 1.06,
"grad_norm": 2.3430908487524498,
"learning_rate": 4.520301437884436e-06,
"loss": 1.0172,
"step": 5690
},
{
"epoch": 1.06,
"grad_norm": 2.555392414885646,
"learning_rate": 4.5130119957712746e-06,
"loss": 1.1048,
"step": 5695
},
{
"epoch": 1.06,
"grad_norm": 2.40399252374117,
"learning_rate": 4.5057235984849765e-06,
"loss": 0.9381,
"step": 5700
},
{
"epoch": 1.06,
"grad_norm": 2.507638615828214,
"learning_rate": 4.498436261662707e-06,
"loss": 1.0001,
"step": 5705
},
{
"epoch": 1.06,
"grad_norm": 2.955757388241879,
"learning_rate": 4.491150000939358e-06,
"loss": 1.0098,
"step": 5710
},
{
"epoch": 1.07,
"grad_norm": 2.3178496259731642,
"learning_rate": 4.483864831947518e-06,
"loss": 1.06,
"step": 5715
},
{
"epoch": 1.07,
"grad_norm": 2.861349825430097,
"learning_rate": 4.4765807703174256e-06,
"loss": 1.0323,
"step": 5720
},
{
"epoch": 1.07,
"grad_norm": 2.4567917565647823,
"learning_rate": 4.469297831676947e-06,
"loss": 1.0381,
"step": 5725
},
{
"epoch": 1.07,
"grad_norm": 2.3765296344065683,
"learning_rate": 4.4620160316515395e-06,
"loss": 1.001,
"step": 5730
},
{
"epoch": 1.07,
"grad_norm": 2.851117308902495,
"learning_rate": 4.454735385864215e-06,
"loss": 1.0239,
"step": 5735
},
{
"epoch": 1.07,
"grad_norm": 2.2475477463435363,
"learning_rate": 4.447455909935513e-06,
"loss": 1.006,
"step": 5740
},
{
"epoch": 1.07,
"grad_norm": 2.3691317072426736,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.9884,
"step": 5745
},
{
"epoch": 1.07,
"grad_norm": 2.7912385165600986,
"learning_rate": 4.432900530123543e-06,
"loss": 1.0358,
"step": 5750
},
{
"epoch": 1.07,
"grad_norm": 2.532523308925846,
"learning_rate": 4.425624657468662e-06,
"loss": 0.9825,
"step": 5755
},
{
"epoch": 1.07,
"grad_norm": 2.5901674352213346,
"learning_rate": 4.41835001712912e-06,
"loss": 0.9793,
"step": 5760
},
{
"epoch": 1.08,
"grad_norm": 2.526653070020505,
"learning_rate": 4.411076624712567e-06,
"loss": 1.0093,
"step": 5765
},
{
"epoch": 1.08,
"grad_norm": 2.5107555130617225,
"learning_rate": 4.403804495823979e-06,
"loss": 1.0016,
"step": 5770
},
{
"epoch": 1.08,
"grad_norm": 2.3247381049982936,
"learning_rate": 4.396533646065618e-06,
"loss": 1.0014,
"step": 5775
},
{
"epoch": 1.08,
"grad_norm": 2.750111283951606,
"learning_rate": 4.3892640910370085e-06,
"loss": 1.0422,
"step": 5780
},
{
"epoch": 1.08,
"grad_norm": 2.585259483582809,
"learning_rate": 4.381995846334891e-06,
"loss": 1.0446,
"step": 5785
},
{
"epoch": 1.08,
"grad_norm": 2.8484814547628146,
"learning_rate": 4.374728927553195e-06,
"loss": 1.0186,
"step": 5790
},
{
"epoch": 1.08,
"grad_norm": 2.5484297678172694,
"learning_rate": 4.367463350283008e-06,
"loss": 1.033,
"step": 5795
},
{
"epoch": 1.08,
"grad_norm": 3.2517629474206915,
"learning_rate": 4.3601991301125345e-06,
"loss": 1.0305,
"step": 5800
},
{
"epoch": 1.08,
"grad_norm": 2.317879031880154,
"learning_rate": 4.352936282627076e-06,
"loss": 1.009,
"step": 5805
},
{
"epoch": 1.08,
"grad_norm": 2.873136359968325,
"learning_rate": 4.345674823408983e-06,
"loss": 1.0396,
"step": 5810
},
{
"epoch": 1.08,
"grad_norm": 2.6184270537855627,
"learning_rate": 4.338414768037625e-06,
"loss": 1.0268,
"step": 5815
},
{
"epoch": 1.09,
"grad_norm": 2.5891458972953485,
"learning_rate": 4.3311561320893635e-06,
"loss": 1.0537,
"step": 5820
},
{
"epoch": 1.09,
"grad_norm": 2.3900157941023097,
"learning_rate": 4.3238989311375165e-06,
"loss": 0.9919,
"step": 5825
},
{
"epoch": 1.09,
"grad_norm": 2.188987069441718,
"learning_rate": 4.316643180752321e-06,
"loss": 0.9548,
"step": 5830
},
{
"epoch": 1.09,
"grad_norm": 2.7746815671456826,
"learning_rate": 4.309388896500899e-06,
"loss": 1.0482,
"step": 5835
},
{
"epoch": 1.09,
"grad_norm": 2.6451781748141037,
"learning_rate": 4.3021360939472296e-06,
"loss": 1.0075,
"step": 5840
},
{
"epoch": 1.09,
"grad_norm": 2.7182175432694016,
"learning_rate": 4.294884788652111e-06,
"loss": 1.0218,
"step": 5845
},
{
"epoch": 1.09,
"grad_norm": 2.7398738561495937,
"learning_rate": 4.2876349961731375e-06,
"loss": 1.031,
"step": 5850
},
{
"epoch": 1.09,
"grad_norm": 2.8602106485650522,
"learning_rate": 4.280386732064645e-06,
"loss": 0.9569,
"step": 5855
},
{
"epoch": 1.09,
"grad_norm": 2.5471621565131817,
"learning_rate": 4.273140011877698e-06,
"loss": 1.0346,
"step": 5860
},
{
"epoch": 1.09,
"grad_norm": 2.599772140063331,
"learning_rate": 4.265894851160047e-06,
"loss": 1.0129,
"step": 5865
},
{
"epoch": 1.09,
"grad_norm": 2.5023497756964166,
"learning_rate": 4.258651265456096e-06,
"loss": 0.9783,
"step": 5870
},
{
"epoch": 1.1,
"grad_norm": 2.3315837831174955,
"learning_rate": 4.251409270306872e-06,
"loss": 1.0385,
"step": 5875
},
{
"epoch": 1.1,
"grad_norm": 2.2339673161919738,
"learning_rate": 4.244168881249986e-06,
"loss": 0.9796,
"step": 5880
},
{
"epoch": 1.1,
"grad_norm": 2.9198639199308145,
"learning_rate": 4.236930113819605e-06,
"loss": 1.0418,
"step": 5885
},
{
"epoch": 1.1,
"grad_norm": 2.2996710537906098,
"learning_rate": 4.229692983546418e-06,
"loss": 0.992,
"step": 5890
},
{
"epoch": 1.1,
"grad_norm": 2.2304355168240186,
"learning_rate": 4.2224575059576e-06,
"loss": 1.0011,
"step": 5895
},
{
"epoch": 1.1,
"grad_norm": 2.5260445789366557,
"learning_rate": 4.215223696576781e-06,
"loss": 0.99,
"step": 5900
},
{
"epoch": 1.1,
"grad_norm": 2.465790598030769,
"learning_rate": 4.2079915709240095e-06,
"loss": 1.0134,
"step": 5905
},
{
"epoch": 1.1,
"grad_norm": 2.653095522617236,
"learning_rate": 4.200761144515724e-06,
"loss": 0.9633,
"step": 5910
},
{
"epoch": 1.1,
"grad_norm": 2.411929291599091,
"learning_rate": 4.193532432864718e-06,
"loss": 1.0315,
"step": 5915
},
{
"epoch": 1.1,
"grad_norm": 2.2292387604409547,
"learning_rate": 4.186305451480104e-06,
"loss": 1.0199,
"step": 5920
},
{
"epoch": 1.1,
"grad_norm": 2.7973446623576033,
"learning_rate": 4.179080215867282e-06,
"loss": 0.9596,
"step": 5925
},
{
"epoch": 1.11,
"grad_norm": 2.4745090701678882,
"learning_rate": 4.1718567415279085e-06,
"loss": 0.8889,
"step": 5930
},
{
"epoch": 1.11,
"grad_norm": 2.428997596511232,
"learning_rate": 4.164635043959861e-06,
"loss": 0.9836,
"step": 5935
},
{
"epoch": 1.11,
"grad_norm": 2.439369458922769,
"learning_rate": 4.157415138657203e-06,
"loss": 1.0524,
"step": 5940
},
{
"epoch": 1.11,
"grad_norm": 2.879991571996203,
"learning_rate": 4.150197041110154e-06,
"loss": 1.0329,
"step": 5945
},
{
"epoch": 1.11,
"grad_norm": 2.441256974326701,
"learning_rate": 4.142980766805055e-06,
"loss": 0.9827,
"step": 5950
},
{
"epoch": 1.11,
"grad_norm": 2.2621481444945166,
"learning_rate": 4.135766331224334e-06,
"loss": 1.0164,
"step": 5955
},
{
"epoch": 1.11,
"grad_norm": 2.4153632637114977,
"learning_rate": 4.128553749846477e-06,
"loss": 1.0018,
"step": 5960
},
{
"epoch": 1.11,
"grad_norm": 2.4291621713233456,
"learning_rate": 4.12134303814599e-06,
"loss": 1.0382,
"step": 5965
},
{
"epoch": 1.11,
"grad_norm": 2.4020903183200004,
"learning_rate": 4.114134211593366e-06,
"loss": 1.0146,
"step": 5970
},
{
"epoch": 1.11,
"grad_norm": 2.392997196335174,
"learning_rate": 4.106927285655052e-06,
"loss": 1.0068,
"step": 5975
},
{
"epoch": 1.12,
"grad_norm": 2.3982209610639917,
"learning_rate": 4.099722275793427e-06,
"loss": 0.9805,
"step": 5980
},
{
"epoch": 1.12,
"grad_norm": 2.4610808257505177,
"learning_rate": 4.0925191974667485e-06,
"loss": 0.9476,
"step": 5985
},
{
"epoch": 1.12,
"grad_norm": 2.410262471073752,
"learning_rate": 4.085318066129137e-06,
"loss": 0.9853,
"step": 5990
},
{
"epoch": 1.12,
"grad_norm": 2.520389162835538,
"learning_rate": 4.0781188972305245e-06,
"loss": 1.0039,
"step": 5995
},
{
"epoch": 1.12,
"grad_norm": 2.484693819518564,
"learning_rate": 4.07092170621665e-06,
"loss": 0.9892,
"step": 6000
},
{
"epoch": 1.12,
"grad_norm": 2.5500391170036405,
"learning_rate": 4.063726508528995e-06,
"loss": 1.0229,
"step": 6005
},
{
"epoch": 1.12,
"grad_norm": 2.620556442354383,
"learning_rate": 4.05653331960477e-06,
"loss": 1.0512,
"step": 6010
},
{
"epoch": 1.12,
"grad_norm": 2.8079283170008646,
"learning_rate": 4.049342154876871e-06,
"loss": 0.9743,
"step": 6015
},
{
"epoch": 1.12,
"grad_norm": 2.2753916987815677,
"learning_rate": 4.042153029773861e-06,
"loss": 0.9979,
"step": 6020
},
{
"epoch": 1.12,
"grad_norm": 3.294635777323174,
"learning_rate": 4.034965959719919e-06,
"loss": 0.9607,
"step": 6025
},
{
"epoch": 1.12,
"grad_norm": 2.6078387298892918,
"learning_rate": 4.027780960134813e-06,
"loss": 0.9941,
"step": 6030
},
{
"epoch": 1.13,
"grad_norm": 2.7700921852331244,
"learning_rate": 4.0205980464338765e-06,
"loss": 0.9948,
"step": 6035
},
{
"epoch": 1.13,
"grad_norm": 2.6699866883313113,
"learning_rate": 4.01341723402796e-06,
"loss": 0.9995,
"step": 6040
},
{
"epoch": 1.13,
"grad_norm": 2.194966684039299,
"learning_rate": 4.006238538323415e-06,
"loss": 1.005,
"step": 6045
},
{
"epoch": 1.13,
"grad_norm": 2.6389273378201152,
"learning_rate": 3.999061974722041e-06,
"loss": 1.0018,
"step": 6050
},
{
"epoch": 1.13,
"grad_norm": 2.288324978564677,
"learning_rate": 3.991887558621071e-06,
"loss": 1.0326,
"step": 6055
},
{
"epoch": 1.13,
"grad_norm": 2.382363146274824,
"learning_rate": 3.984715305413125e-06,
"loss": 0.9227,
"step": 6060
},
{
"epoch": 1.13,
"grad_norm": 2.545591262311767,
"learning_rate": 3.97754523048619e-06,
"loss": 0.9955,
"step": 6065
},
{
"epoch": 1.13,
"grad_norm": 2.696133141462619,
"learning_rate": 3.970377349223572e-06,
"loss": 1.0089,
"step": 6070
},
{
"epoch": 1.13,
"grad_norm": 2.5249857811313383,
"learning_rate": 3.963211677003872e-06,
"loss": 1.06,
"step": 6075
},
{
"epoch": 1.13,
"grad_norm": 2.6974517008498076,
"learning_rate": 3.956048229200956e-06,
"loss": 1.0025,
"step": 6080
},
{
"epoch": 1.13,
"grad_norm": 2.708829049326595,
"learning_rate": 3.948887021183911e-06,
"loss": 0.9934,
"step": 6085
},
{
"epoch": 1.14,
"grad_norm": 2.753223542699741,
"learning_rate": 3.941728068317026e-06,
"loss": 1.0309,
"step": 6090
},
{
"epoch": 1.14,
"grad_norm": 2.5571783646544533,
"learning_rate": 3.934571385959745e-06,
"loss": 1.0078,
"step": 6095
},
{
"epoch": 1.14,
"grad_norm": 2.9619368449595704,
"learning_rate": 3.927416989466644e-06,
"loss": 1.0263,
"step": 6100
},
{
"epoch": 1.14,
"grad_norm": 2.527413370302673,
"learning_rate": 3.920264894187393e-06,
"loss": 1.0192,
"step": 6105
},
{
"epoch": 1.14,
"grad_norm": 2.290679414980372,
"learning_rate": 3.913115115466728e-06,
"loss": 1.0123,
"step": 6110
},
{
"epoch": 1.14,
"grad_norm": 2.1791472017494864,
"learning_rate": 3.9059676686444095e-06,
"loss": 1.0294,
"step": 6115
},
{
"epoch": 1.14,
"grad_norm": 2.4477872352744097,
"learning_rate": 3.898822569055201e-06,
"loss": 0.9835,
"step": 6120
},
{
"epoch": 1.14,
"grad_norm": 2.4553291419944574,
"learning_rate": 3.8916798320288254e-06,
"loss": 0.9878,
"step": 6125
},
{
"epoch": 1.14,
"grad_norm": 2.5549208400564756,
"learning_rate": 3.884539472889938e-06,
"loss": 1.0245,
"step": 6130
},
{
"epoch": 1.14,
"grad_norm": 2.485934742620887,
"learning_rate": 3.877401506958095e-06,
"loss": 0.9945,
"step": 6135
},
{
"epoch": 1.15,
"grad_norm": 2.2578183339754023,
"learning_rate": 3.8702659495477144e-06,
"loss": 1.0209,
"step": 6140
},
{
"epoch": 1.15,
"grad_norm": 2.3465134445411304,
"learning_rate": 3.863132815968048e-06,
"loss": 1.0666,
"step": 6145
},
{
"epoch": 1.15,
"grad_norm": 2.29022378683888,
"learning_rate": 3.856002121523147e-06,
"loss": 1.0055,
"step": 6150
},
{
"epoch": 1.15,
"grad_norm": 2.660614384744464,
"learning_rate": 3.848873881511831e-06,
"loss": 1.0101,
"step": 6155
},
{
"epoch": 1.15,
"grad_norm": 2.615319592937916,
"learning_rate": 3.841748111227652e-06,
"loss": 0.9582,
"step": 6160
},
{
"epoch": 1.15,
"grad_norm": 2.817918853778393,
"learning_rate": 3.834624825958864e-06,
"loss": 0.9969,
"step": 6165
},
{
"epoch": 1.15,
"grad_norm": 2.4439689869979455,
"learning_rate": 3.827504040988388e-06,
"loss": 0.9914,
"step": 6170
},
{
"epoch": 1.15,
"grad_norm": 2.4582126352422438,
"learning_rate": 3.8203857715937845e-06,
"loss": 1.0226,
"step": 6175
},
{
"epoch": 1.15,
"grad_norm": 2.9601365316898867,
"learning_rate": 3.8132700330472124e-06,
"loss": 0.9979,
"step": 6180
},
{
"epoch": 1.15,
"grad_norm": 2.1744301885466704,
"learning_rate": 3.8061568406154035e-06,
"loss": 0.9677,
"step": 6185
},
{
"epoch": 1.15,
"grad_norm": 2.9345644472209083,
"learning_rate": 3.799046209559623e-06,
"loss": 0.9926,
"step": 6190
},
{
"epoch": 1.16,
"grad_norm": 2.487795059019343,
"learning_rate": 3.7919381551356478e-06,
"loss": 1.0199,
"step": 6195
},
{
"epoch": 1.16,
"grad_norm": 2.7762118273919603,
"learning_rate": 3.7848326925937207e-06,
"loss": 0.985,
"step": 6200
},
{
"epoch": 1.16,
"grad_norm": 2.385619014287349,
"learning_rate": 3.7777298371785257e-06,
"loss": 1.0049,
"step": 6205
},
{
"epoch": 1.16,
"grad_norm": 3.1858905013848773,
"learning_rate": 3.770629604129153e-06,
"loss": 1.0067,
"step": 6210
},
{
"epoch": 1.16,
"grad_norm": 2.646520353336217,
"learning_rate": 3.7635320086790635e-06,
"loss": 1.0203,
"step": 6215
},
{
"epoch": 1.16,
"grad_norm": 2.703518348813667,
"learning_rate": 3.7564370660560665e-06,
"loss": 0.9956,
"step": 6220
},
{
"epoch": 1.16,
"grad_norm": 2.6488264930358074,
"learning_rate": 3.7493447914822766e-06,
"loss": 1.0232,
"step": 6225
},
{
"epoch": 1.16,
"grad_norm": 3.0881970088316533,
"learning_rate": 3.7422552001740782e-06,
"loss": 0.9966,
"step": 6230
},
{
"epoch": 1.16,
"grad_norm": 2.6057449190922015,
"learning_rate": 3.7351683073421036e-06,
"loss": 1.0195,
"step": 6235
},
{
"epoch": 1.16,
"grad_norm": 2.442247586855112,
"learning_rate": 3.7280841281912007e-06,
"loss": 0.999,
"step": 6240
},
{
"epoch": 1.16,
"grad_norm": 2.6614307097799346,
"learning_rate": 3.7210026779203863e-06,
"loss": 1.0111,
"step": 6245
},
{
"epoch": 1.17,
"grad_norm": 2.883103898286734,
"learning_rate": 3.713923971722826e-06,
"loss": 0.989,
"step": 6250
},
{
"epoch": 1.17,
"grad_norm": 2.3731982209108873,
"learning_rate": 3.7068480247857975e-06,
"loss": 1.077,
"step": 6255
},
{
"epoch": 1.17,
"grad_norm": 2.5547604391314667,
"learning_rate": 3.6997748522906627e-06,
"loss": 1.0154,
"step": 6260
},
{
"epoch": 1.17,
"grad_norm": 2.6226187885824643,
"learning_rate": 3.6927044694128255e-06,
"loss": 1.0549,
"step": 6265
},
{
"epoch": 1.17,
"grad_norm": 3.0423886957703217,
"learning_rate": 3.685636891321706e-06,
"loss": 0.9967,
"step": 6270
},
{
"epoch": 1.17,
"grad_norm": 2.689767779138463,
"learning_rate": 3.678572133180708e-06,
"loss": 0.9555,
"step": 6275
},
{
"epoch": 1.17,
"grad_norm": 2.5322314585255947,
"learning_rate": 3.671510210147183e-06,
"loss": 0.9532,
"step": 6280
},
{
"epoch": 1.17,
"grad_norm": 2.5502386618866866,
"learning_rate": 3.6644511373724035e-06,
"loss": 1.0262,
"step": 6285
},
{
"epoch": 1.17,
"grad_norm": 2.9597133665829354,
"learning_rate": 3.657394930001524e-06,
"loss": 1.0865,
"step": 6290
},
{
"epoch": 1.17,
"grad_norm": 2.4548956067402226,
"learning_rate": 3.650341603173552e-06,
"loss": 1.0551,
"step": 6295
},
{
"epoch": 1.17,
"grad_norm": 2.415307407799765,
"learning_rate": 3.6432911720213127e-06,
"loss": 1.0286,
"step": 6300
},
{
"epoch": 1.18,
"grad_norm": 2.518703622236714,
"learning_rate": 3.636243651671424e-06,
"loss": 0.9453,
"step": 6305
},
{
"epoch": 1.18,
"grad_norm": 2.9236092035869587,
"learning_rate": 3.6291990572442527e-06,
"loss": 1.0284,
"step": 6310
},
{
"epoch": 1.18,
"grad_norm": 2.7290628973336393,
"learning_rate": 3.6221574038538926e-06,
"loss": 0.9858,
"step": 6315
},
{
"epoch": 1.18,
"grad_norm": 2.746796572631986,
"learning_rate": 3.615118706608125e-06,
"loss": 0.9979,
"step": 6320
},
{
"epoch": 1.18,
"grad_norm": 2.2102817767160086,
"learning_rate": 3.6080829806083885e-06,
"loss": 1.0017,
"step": 6325
},
{
"epoch": 1.18,
"grad_norm": 2.3433097259852578,
"learning_rate": 3.6010502409497493e-06,
"loss": 0.9991,
"step": 6330
},
{
"epoch": 1.18,
"grad_norm": 2.669305876106401,
"learning_rate": 3.594020502720865e-06,
"loss": 1.0773,
"step": 6335
},
{
"epoch": 1.18,
"grad_norm": 2.5602011856012203,
"learning_rate": 3.586993781003954e-06,
"loss": 1.003,
"step": 6340
},
{
"epoch": 1.18,
"grad_norm": 2.3053351276904617,
"learning_rate": 3.5799700908747607e-06,
"loss": 0.9884,
"step": 6345
},
{
"epoch": 1.18,
"grad_norm": 2.3995150259514064,
"learning_rate": 3.5729494474025296e-06,
"loss": 1.0127,
"step": 6350
},
{
"epoch": 1.19,
"grad_norm": 2.3227336825063083,
"learning_rate": 3.565931865649965e-06,
"loss": 1.0194,
"step": 6355
},
{
"epoch": 1.19,
"grad_norm": 2.5258564235269643,
"learning_rate": 3.5589173606732042e-06,
"loss": 1.0305,
"step": 6360
},
{
"epoch": 1.19,
"grad_norm": 2.5245519835366306,
"learning_rate": 3.551905947521781e-06,
"loss": 1.0075,
"step": 6365
},
{
"epoch": 1.19,
"grad_norm": 2.6202125604841493,
"learning_rate": 3.5448976412385994e-06,
"loss": 1.0026,
"step": 6370
},
{
"epoch": 1.19,
"grad_norm": 2.2985573829337143,
"learning_rate": 3.537892456859895e-06,
"loss": 1.0377,
"step": 6375
},
{
"epoch": 1.19,
"grad_norm": 2.5445427577379225,
"learning_rate": 3.5308904094152047e-06,
"loss": 1.0336,
"step": 6380
},
{
"epoch": 1.19,
"grad_norm": 2.587984616767967,
"learning_rate": 3.5238915139273387e-06,
"loss": 0.9534,
"step": 6385
},
{
"epoch": 1.19,
"grad_norm": 2.4587427920639082,
"learning_rate": 3.5168957854123386e-06,
"loss": 1.041,
"step": 6390
},
{
"epoch": 1.19,
"grad_norm": 2.5776496842048604,
"learning_rate": 3.5099032388794596e-06,
"loss": 1.0119,
"step": 6395
},
{
"epoch": 1.19,
"grad_norm": 2.9999317197733704,
"learning_rate": 3.5029138893311245e-06,
"loss": 1.0094,
"step": 6400
},
{
"epoch": 1.19,
"grad_norm": 2.370429876481651,
"learning_rate": 3.4959277517628953e-06,
"loss": 0.9934,
"step": 6405
},
{
"epoch": 1.2,
"grad_norm": 2.3397878191899757,
"learning_rate": 3.4889448411634465e-06,
"loss": 1.0449,
"step": 6410
},
{
"epoch": 1.2,
"grad_norm": 2.9633064271669465,
"learning_rate": 3.4819651725145303e-06,
"loss": 1.0328,
"step": 6415
},
{
"epoch": 1.2,
"grad_norm": 2.838930291488574,
"learning_rate": 3.4749887607909416e-06,
"loss": 0.9886,
"step": 6420
},
{
"epoch": 1.2,
"grad_norm": 2.662756964502569,
"learning_rate": 3.468015620960484e-06,
"loss": 0.9936,
"step": 6425
},
{
"epoch": 1.2,
"grad_norm": 2.5081288815673592,
"learning_rate": 3.4610457679839445e-06,
"loss": 1.0417,
"step": 6430
},
{
"epoch": 1.2,
"grad_norm": 2.1688593673892846,
"learning_rate": 3.4540792168150618e-06,
"loss": 0.988,
"step": 6435
},
{
"epoch": 1.2,
"grad_norm": 2.549976740705903,
"learning_rate": 3.447115982400485e-06,
"loss": 0.9867,
"step": 6440
},
{
"epoch": 1.2,
"grad_norm": 2.70924169526118,
"learning_rate": 3.440156079679749e-06,
"loss": 1.0214,
"step": 6445
},
{
"epoch": 1.2,
"grad_norm": 2.201236785717843,
"learning_rate": 3.4331995235852406e-06,
"loss": 0.996,
"step": 6450
},
{
"epoch": 1.2,
"grad_norm": 2.3856197789691813,
"learning_rate": 3.4262463290421654e-06,
"loss": 0.9932,
"step": 6455
},
{
"epoch": 1.2,
"grad_norm": 2.4317340460371484,
"learning_rate": 3.4192965109685215e-06,
"loss": 1.0217,
"step": 6460
},
{
"epoch": 1.21,
"grad_norm": 2.2660839453173676,
"learning_rate": 3.412350084275057e-06,
"loss": 0.995,
"step": 6465
},
{
"epoch": 1.21,
"grad_norm": 2.8062486187830142,
"learning_rate": 3.4054070638652458e-06,
"loss": 0.9988,
"step": 6470
},
{
"epoch": 1.21,
"grad_norm": 2.8224885243520745,
"learning_rate": 3.3984674646352543e-06,
"loss": 1.0084,
"step": 6475
},
{
"epoch": 1.21,
"grad_norm": 2.355107056771508,
"learning_rate": 3.39153130147391e-06,
"loss": 1.0043,
"step": 6480
},
{
"epoch": 1.21,
"grad_norm": 2.319540034779808,
"learning_rate": 3.3845985892626654e-06,
"loss": 0.9765,
"step": 6485
},
{
"epoch": 1.21,
"grad_norm": 2.4007112000801794,
"learning_rate": 3.3776693428755714e-06,
"loss": 1.0303,
"step": 6490
},
{
"epoch": 1.21,
"grad_norm": 2.7687023865947022,
"learning_rate": 3.3707435771792417e-06,
"loss": 1.0198,
"step": 6495
},
{
"epoch": 1.21,
"grad_norm": 2.2967682488576187,
"learning_rate": 3.3638213070328223e-06,
"loss": 1.0012,
"step": 6500
},
{
"epoch": 1.21,
"grad_norm": 3.00503160474256,
"learning_rate": 3.356902547287961e-06,
"loss": 1.0415,
"step": 6505
},
{
"epoch": 1.21,
"grad_norm": 2.581469591673954,
"learning_rate": 3.3499873127887726e-06,
"loss": 1.0051,
"step": 6510
},
{
"epoch": 1.21,
"grad_norm": 2.6784260311780703,
"learning_rate": 3.343075618371808e-06,
"loss": 0.968,
"step": 6515
},
{
"epoch": 1.22,
"grad_norm": 2.742914883406482,
"learning_rate": 3.336167478866024e-06,
"loss": 1.054,
"step": 6520
},
{
"epoch": 1.22,
"grad_norm": 2.5967700561103912,
"learning_rate": 3.329262909092752e-06,
"loss": 0.9289,
"step": 6525
},
{
"epoch": 1.22,
"grad_norm": 2.932847442630695,
"learning_rate": 3.322361923865661e-06,
"loss": 1.0103,
"step": 6530
},
{
"epoch": 1.22,
"grad_norm": 4.336241175871444,
"learning_rate": 3.3154645379907315e-06,
"loss": 0.9699,
"step": 6535
},
{
"epoch": 1.22,
"grad_norm": 2.2502099078873927,
"learning_rate": 3.3085707662662208e-06,
"loss": 1.0008,
"step": 6540
},
{
"epoch": 1.22,
"grad_norm": 2.3994266512691262,
"learning_rate": 3.3016806234826336e-06,
"loss": 0.9688,
"step": 6545
},
{
"epoch": 1.22,
"grad_norm": 2.6533685347490947,
"learning_rate": 3.294794124422688e-06,
"loss": 1.0142,
"step": 6550
},
{
"epoch": 1.22,
"grad_norm": 2.516906851667742,
"learning_rate": 3.2879112838612837e-06,
"loss": 0.9894,
"step": 6555
},
{
"epoch": 1.22,
"grad_norm": 2.372071290525371,
"learning_rate": 3.281032116565473e-06,
"loss": 1.0566,
"step": 6560
},
{
"epoch": 1.22,
"grad_norm": 2.6515335583185022,
"learning_rate": 3.274156637294421e-06,
"loss": 1.0129,
"step": 6565
},
{
"epoch": 1.23,
"grad_norm": 2.316592337770323,
"learning_rate": 3.2672848607993913e-06,
"loss": 0.9844,
"step": 6570
},
{
"epoch": 1.23,
"grad_norm": 2.7973489389722634,
"learning_rate": 3.260416801823694e-06,
"loss": 1.0009,
"step": 6575
},
{
"epoch": 1.23,
"grad_norm": 2.348401957786061,
"learning_rate": 3.253552475102668e-06,
"loss": 1.0305,
"step": 6580
},
{
"epoch": 1.23,
"grad_norm": 2.365305340735876,
"learning_rate": 3.2466918953636394e-06,
"loss": 0.9931,
"step": 6585
},
{
"epoch": 1.23,
"grad_norm": 2.536643326010143,
"learning_rate": 3.2398350773259035e-06,
"loss": 1.031,
"step": 6590
},
{
"epoch": 1.23,
"grad_norm": 2.6324859382475863,
"learning_rate": 3.232982035700678e-06,
"loss": 1.0175,
"step": 6595
},
{
"epoch": 1.23,
"grad_norm": 2.3943506218465296,
"learning_rate": 3.2261327851910827e-06,
"loss": 1.0144,
"step": 6600
},
{
"epoch": 1.23,
"grad_norm": 2.3642765166995945,
"learning_rate": 3.2192873404920966e-06,
"loss": 0.9929,
"step": 6605
},
{
"epoch": 1.23,
"grad_norm": 2.357721061823941,
"learning_rate": 3.212445716290543e-06,
"loss": 1.0205,
"step": 6610
},
{
"epoch": 1.23,
"grad_norm": 3.0330694342984508,
"learning_rate": 3.205607927265044e-06,
"loss": 1.031,
"step": 6615
},
{
"epoch": 1.23,
"grad_norm": 2.6767493180451973,
"learning_rate": 3.1987739880859902e-06,
"loss": 1.0337,
"step": 6620
},
{
"epoch": 1.24,
"grad_norm": 2.595396432430348,
"learning_rate": 3.1919439134155173e-06,
"loss": 1.0096,
"step": 6625
},
{
"epoch": 1.24,
"grad_norm": 2.4346297592857926,
"learning_rate": 3.1851177179074645e-06,
"loss": 1.0022,
"step": 6630
},
{
"epoch": 1.24,
"grad_norm": 2.7420307620549025,
"learning_rate": 3.1782954162073566e-06,
"loss": 1.0477,
"step": 6635
},
{
"epoch": 1.24,
"grad_norm": 3.026839985368973,
"learning_rate": 3.1714770229523563e-06,
"loss": 1.0093,
"step": 6640
},
{
"epoch": 1.24,
"grad_norm": 2.408559067936416,
"learning_rate": 3.1646625527712426e-06,
"loss": 0.963,
"step": 6645
},
{
"epoch": 1.24,
"grad_norm": 2.656230938630765,
"learning_rate": 3.1578520202843806e-06,
"loss": 1.0291,
"step": 6650
},
{
"epoch": 1.24,
"grad_norm": 3.1429634200338072,
"learning_rate": 3.151045440103685e-06,
"loss": 1.061,
"step": 6655
},
{
"epoch": 1.24,
"grad_norm": 3.0525857382979384,
"learning_rate": 3.1442428268325915e-06,
"loss": 1.0265,
"step": 6660
},
{
"epoch": 1.24,
"grad_norm": 2.5262738969825596,
"learning_rate": 3.137444195066023e-06,
"loss": 1.0584,
"step": 6665
},
{
"epoch": 1.24,
"grad_norm": 2.8656853508515683,
"learning_rate": 3.130649559390362e-06,
"loss": 1.0356,
"step": 6670
},
{
"epoch": 1.24,
"grad_norm": 2.678563207407218,
"learning_rate": 3.123858934383418e-06,
"loss": 1.0051,
"step": 6675
},
{
"epoch": 1.25,
"grad_norm": 2.6967742942369313,
"learning_rate": 3.1170723346143936e-06,
"loss": 0.9988,
"step": 6680
},
{
"epoch": 1.25,
"grad_norm": 2.6334943332730294,
"learning_rate": 3.110289774643857e-06,
"loss": 0.9862,
"step": 6685
},
{
"epoch": 1.25,
"grad_norm": 2.5356006656026158,
"learning_rate": 3.1035112690237086e-06,
"loss": 1.0316,
"step": 6690
},
{
"epoch": 1.25,
"grad_norm": 2.507316947200383,
"learning_rate": 3.096736832297148e-06,
"loss": 0.9558,
"step": 6695
},
{
"epoch": 1.25,
"grad_norm": 3.15448089703757,
"learning_rate": 3.08996647899865e-06,
"loss": 0.9614,
"step": 6700
},
{
"epoch": 1.25,
"grad_norm": 2.792658317811039,
"learning_rate": 3.0832002236539243e-06,
"loss": 0.9971,
"step": 6705
},
{
"epoch": 1.25,
"grad_norm": 2.680576302235998,
"learning_rate": 3.07643808077989e-06,
"loss": 1.0216,
"step": 6710
},
{
"epoch": 1.25,
"grad_norm": 2.5763702681537897,
"learning_rate": 3.069680064884641e-06,
"loss": 0.9948,
"step": 6715
},
{
"epoch": 1.25,
"grad_norm": 2.5684486308454457,
"learning_rate": 3.0629261904674206e-06,
"loss": 0.9914,
"step": 6720
},
{
"epoch": 1.25,
"grad_norm": 2.456176378605224,
"learning_rate": 3.056176472018584e-06,
"loss": 1.0467,
"step": 6725
},
{
"epoch": 1.26,
"grad_norm": 2.3265015455090543,
"learning_rate": 3.0494309240195706e-06,
"loss": 1.0284,
"step": 6730
},
{
"epoch": 1.26,
"grad_norm": 2.5090802418299103,
"learning_rate": 3.0426895609428713e-06,
"loss": 0.9892,
"step": 6735
},
{
"epoch": 1.26,
"grad_norm": 2.4337598655357797,
"learning_rate": 3.0359523972519976e-06,
"loss": 1.0318,
"step": 6740
},
{
"epoch": 1.26,
"grad_norm": 2.9534700024492158,
"learning_rate": 3.029219447401456e-06,
"loss": 0.9617,
"step": 6745
},
{
"epoch": 1.26,
"grad_norm": 2.5169294067657986,
"learning_rate": 3.0224907258367063e-06,
"loss": 1.0083,
"step": 6750
},
{
"epoch": 1.26,
"grad_norm": 2.5412138380242717,
"learning_rate": 3.0157662469941394e-06,
"loss": 1.0024,
"step": 6755
},
{
"epoch": 1.26,
"grad_norm": 2.383419470697286,
"learning_rate": 3.009046025301042e-06,
"loss": 1.0007,
"step": 6760
},
{
"epoch": 1.26,
"grad_norm": 2.346170762298227,
"learning_rate": 3.0023300751755713e-06,
"loss": 1.0101,
"step": 6765
},
{
"epoch": 1.26,
"grad_norm": 2.6426094982234685,
"learning_rate": 2.995618411026715e-06,
"loss": 1.0129,
"step": 6770
},
{
"epoch": 1.26,
"grad_norm": 2.6651384968183742,
"learning_rate": 2.988911047254268e-06,
"loss": 0.9953,
"step": 6775
},
{
"epoch": 1.26,
"grad_norm": 2.2908540342398376,
"learning_rate": 2.982207998248795e-06,
"loss": 0.9605,
"step": 6780
},
{
"epoch": 1.27,
"grad_norm": 2.605274696824992,
"learning_rate": 2.9755092783916106e-06,
"loss": 1.0554,
"step": 6785
},
{
"epoch": 1.27,
"grad_norm": 2.54625674691826,
"learning_rate": 2.968814902054735e-06,
"loss": 0.9603,
"step": 6790
},
{
"epoch": 1.27,
"grad_norm": 2.353611332509148,
"learning_rate": 2.962124883600874e-06,
"loss": 1.0114,
"step": 6795
},
{
"epoch": 1.27,
"grad_norm": 2.3484072442486625,
"learning_rate": 2.955439237383377e-06,
"loss": 0.9552,
"step": 6800
},
{
"epoch": 1.27,
"grad_norm": 2.538093611992046,
"learning_rate": 2.9487579777462165e-06,
"loss": 1.0189,
"step": 6805
},
{
"epoch": 1.27,
"grad_norm": 2.404444291570621,
"learning_rate": 2.94208111902396e-06,
"loss": 0.9687,
"step": 6810
},
{
"epoch": 1.27,
"grad_norm": 2.5355302453879744,
"learning_rate": 2.9354086755417226e-06,
"loss": 0.993,
"step": 6815
},
{
"epoch": 1.27,
"grad_norm": 2.63694242852118,
"learning_rate": 2.9287406616151513e-06,
"loss": 1.0192,
"step": 6820
},
{
"epoch": 1.27,
"grad_norm": 2.2105266714422642,
"learning_rate": 2.9220770915503884e-06,
"loss": 1.0121,
"step": 6825
},
{
"epoch": 1.27,
"grad_norm": 2.5519970083541343,
"learning_rate": 2.9154179796440463e-06,
"loss": 1.0066,
"step": 6830
},
{
"epoch": 1.27,
"grad_norm": 2.5838265435049017,
"learning_rate": 2.9087633401831654e-06,
"loss": 1.036,
"step": 6835
},
{
"epoch": 1.28,
"grad_norm": 2.5274982625597473,
"learning_rate": 2.9021131874451957e-06,
"loss": 1.0009,
"step": 6840
},
{
"epoch": 1.28,
"grad_norm": 2.594414889982177,
"learning_rate": 2.8954675356979566e-06,
"loss": 1.0087,
"step": 6845
},
{
"epoch": 1.28,
"grad_norm": 2.5135656552653973,
"learning_rate": 2.8888263991996172e-06,
"loss": 1.0323,
"step": 6850
},
{
"epoch": 1.28,
"grad_norm": 2.4202739645255775,
"learning_rate": 2.882189792198654e-06,
"loss": 0.9964,
"step": 6855
},
{
"epoch": 1.28,
"grad_norm": 2.5258407501718967,
"learning_rate": 2.8755577289338267e-06,
"loss": 0.9941,
"step": 6860
},
{
"epoch": 1.28,
"grad_norm": 2.8800933954985335,
"learning_rate": 2.868930223634149e-06,
"loss": 1.0738,
"step": 6865
},
{
"epoch": 1.28,
"grad_norm": 2.536918280979007,
"learning_rate": 2.862307290518846e-06,
"loss": 1.0208,
"step": 6870
},
{
"epoch": 1.28,
"grad_norm": 2.371938641774631,
"learning_rate": 2.855688943797348e-06,
"loss": 1.0219,
"step": 6875
},
{
"epoch": 1.28,
"grad_norm": 2.33992110293397,
"learning_rate": 2.8490751976692345e-06,
"loss": 0.9992,
"step": 6880
},
{
"epoch": 1.28,
"grad_norm": 2.635082343200966,
"learning_rate": 2.8424660663242178e-06,
"loss": 1.0006,
"step": 6885
},
{
"epoch": 1.28,
"grad_norm": 2.865808841306876,
"learning_rate": 2.835861563942107e-06,
"loss": 1.0357,
"step": 6890
},
{
"epoch": 1.29,
"grad_norm": 2.4160110151856866,
"learning_rate": 2.829261704692787e-06,
"loss": 0.9538,
"step": 6895
},
{
"epoch": 1.29,
"grad_norm": 2.544168845714422,
"learning_rate": 2.8226665027361753e-06,
"loss": 1.0393,
"step": 6900
},
{
"epoch": 1.29,
"grad_norm": 2.467913497990881,
"learning_rate": 2.8160759722221942e-06,
"loss": 1.0535,
"step": 6905
},
{
"epoch": 1.29,
"grad_norm": 2.9028252530293353,
"learning_rate": 2.809490127290746e-06,
"loss": 1.0147,
"step": 6910
},
{
"epoch": 1.29,
"grad_norm": 2.6008808280831097,
"learning_rate": 2.802908982071685e-06,
"loss": 1.047,
"step": 6915
},
{
"epoch": 1.29,
"grad_norm": 2.6288368611864863,
"learning_rate": 2.796332550684778e-06,
"loss": 1.0448,
"step": 6920
},
{
"epoch": 1.29,
"grad_norm": 2.448524250695845,
"learning_rate": 2.7897608472396776e-06,
"loss": 1.0377,
"step": 6925
},
{
"epoch": 1.29,
"grad_norm": 2.4677282962158085,
"learning_rate": 2.783193885835894e-06,
"loss": 0.9857,
"step": 6930
},
{
"epoch": 1.29,
"grad_norm": 2.7995182741898494,
"learning_rate": 2.7766316805627623e-06,
"loss": 1.0324,
"step": 6935
},
{
"epoch": 1.29,
"grad_norm": 2.6020088847197256,
"learning_rate": 2.7700742454994195e-06,
"loss": 0.9802,
"step": 6940
},
{
"epoch": 1.3,
"grad_norm": 2.584249913609,
"learning_rate": 2.7635215947147574e-06,
"loss": 1.0669,
"step": 6945
},
{
"epoch": 1.3,
"grad_norm": 2.6987668383796994,
"learning_rate": 2.7569737422674103e-06,
"loss": 1.0323,
"step": 6950
},
{
"epoch": 1.3,
"grad_norm": 2.784482570862148,
"learning_rate": 2.7504307022057152e-06,
"loss": 0.9833,
"step": 6955
},
{
"epoch": 1.3,
"grad_norm": 2.47770812516728,
"learning_rate": 2.7438924885676886e-06,
"loss": 0.9966,
"step": 6960
},
{
"epoch": 1.3,
"grad_norm": 2.73598006360337,
"learning_rate": 2.7373591153809864e-06,
"loss": 1.0233,
"step": 6965
},
{
"epoch": 1.3,
"grad_norm": 2.690482453892539,
"learning_rate": 2.7308305966628823e-06,
"loss": 0.9964,
"step": 6970
},
{
"epoch": 1.3,
"grad_norm": 2.4336117130705035,
"learning_rate": 2.724306946420234e-06,
"loss": 0.9767,
"step": 6975
},
{
"epoch": 1.3,
"grad_norm": 2.6576192606668587,
"learning_rate": 2.7177881786494538e-06,
"loss": 0.995,
"step": 6980
},
{
"epoch": 1.3,
"grad_norm": 2.142004404838875,
"learning_rate": 2.711274307336479e-06,
"loss": 0.9556,
"step": 6985
},
{
"epoch": 1.3,
"grad_norm": 2.533080687394775,
"learning_rate": 2.7047653464567416e-06,
"loss": 1.0542,
"step": 6990
},
{
"epoch": 1.3,
"grad_norm": 2.7053823200502127,
"learning_rate": 2.6982613099751376e-06,
"loss": 0.963,
"step": 6995
},
{
"epoch": 1.31,
"grad_norm": 2.597526805111531,
"learning_rate": 2.6917622118459975e-06,
"loss": 1.0127,
"step": 7000
},
{
"epoch": 1.31,
"grad_norm": 2.987091945981233,
"learning_rate": 2.6852680660130616e-06,
"loss": 1.0293,
"step": 7005
},
{
"epoch": 1.31,
"grad_norm": 2.3621389957330217,
"learning_rate": 2.678778886409438e-06,
"loss": 1.0367,
"step": 7010
},
{
"epoch": 1.31,
"grad_norm": 2.8142933423345666,
"learning_rate": 2.6722946869575837e-06,
"loss": 0.9871,
"step": 7015
},
{
"epoch": 1.31,
"grad_norm": 2.8818765360063967,
"learning_rate": 2.6658154815692693e-06,
"loss": 0.9571,
"step": 7020
},
{
"epoch": 1.31,
"grad_norm": 2.470646940527506,
"learning_rate": 2.659341284145553e-06,
"loss": 0.975,
"step": 7025
},
{
"epoch": 1.31,
"grad_norm": 3.5524754035548893,
"learning_rate": 2.6528721085767457e-06,
"loss": 1.0368,
"step": 7030
},
{
"epoch": 1.31,
"grad_norm": 2.612626897766346,
"learning_rate": 2.646407968742385e-06,
"loss": 0.9813,
"step": 7035
},
{
"epoch": 1.31,
"grad_norm": 2.4800999564613626,
"learning_rate": 2.6399488785112055e-06,
"loss": 1.0118,
"step": 7040
},
{
"epoch": 1.31,
"grad_norm": 2.5510537924141943,
"learning_rate": 2.633494851741104e-06,
"loss": 1.0442,
"step": 7045
},
{
"epoch": 1.31,
"grad_norm": 2.636748644407911,
"learning_rate": 2.6270459022791218e-06,
"loss": 1.0212,
"step": 7050
},
{
"epoch": 1.32,
"grad_norm": 2.7034201669908917,
"learning_rate": 2.6206020439613988e-06,
"loss": 1.0306,
"step": 7055
},
{
"epoch": 1.32,
"grad_norm": 2.337916153105764,
"learning_rate": 2.6141632906131578e-06,
"loss": 1.0701,
"step": 7060
},
{
"epoch": 1.32,
"grad_norm": 2.7302392511630447,
"learning_rate": 2.60772965604866e-06,
"loss": 1.0052,
"step": 7065
},
{
"epoch": 1.32,
"grad_norm": 2.622516557603338,
"learning_rate": 2.6013011540711954e-06,
"loss": 1.0121,
"step": 7070
},
{
"epoch": 1.32,
"grad_norm": 2.457751042922651,
"learning_rate": 2.5948777984730354e-06,
"loss": 0.9813,
"step": 7075
},
{
"epoch": 1.32,
"grad_norm": 2.5201979245792114,
"learning_rate": 2.5884596030354127e-06,
"loss": 1.0001,
"step": 7080
},
{
"epoch": 1.32,
"grad_norm": 3.100741302309318,
"learning_rate": 2.5820465815284845e-06,
"loss": 1.0238,
"step": 7085
},
{
"epoch": 1.32,
"grad_norm": 2.8057319534838077,
"learning_rate": 2.5756387477113154e-06,
"loss": 1.0933,
"step": 7090
},
{
"epoch": 1.32,
"grad_norm": 2.739482279660433,
"learning_rate": 2.5692361153318333e-06,
"loss": 1.053,
"step": 7095
},
{
"epoch": 1.32,
"grad_norm": 3.0396582626801707,
"learning_rate": 2.56283869812681e-06,
"loss": 1.0126,
"step": 7100
},
{
"epoch": 1.33,
"grad_norm": 2.8612314821867355,
"learning_rate": 2.5564465098218225e-06,
"loss": 1.0162,
"step": 7105
},
{
"epoch": 1.33,
"grad_norm": 2.540230694944602,
"learning_rate": 2.5500595641312344e-06,
"loss": 1.0183,
"step": 7110
},
{
"epoch": 1.33,
"grad_norm": 2.5484379608694723,
"learning_rate": 2.543677874758163e-06,
"loss": 1.0186,
"step": 7115
},
{
"epoch": 1.33,
"grad_norm": 2.7804517969115725,
"learning_rate": 2.5373014553944443e-06,
"loss": 1.0007,
"step": 7120
},
{
"epoch": 1.33,
"grad_norm": 2.4281339631896053,
"learning_rate": 2.530930319720608e-06,
"loss": 0.9656,
"step": 7125
},
{
"epoch": 1.33,
"grad_norm": 2.506088098134065,
"learning_rate": 2.5245644814058475e-06,
"loss": 1.0063,
"step": 7130
},
{
"epoch": 1.33,
"grad_norm": 2.8452273943404682,
"learning_rate": 2.5182039541079963e-06,
"loss": 0.9963,
"step": 7135
},
{
"epoch": 1.33,
"grad_norm": 2.717683642251381,
"learning_rate": 2.511848751473485e-06,
"loss": 0.9919,
"step": 7140
},
{
"epoch": 1.33,
"grad_norm": 2.3677700987643617,
"learning_rate": 2.5054988871373225e-06,
"loss": 1.0083,
"step": 7145
},
{
"epoch": 1.33,
"grad_norm": 2.5612910826475708,
"learning_rate": 2.499154374723069e-06,
"loss": 0.9992,
"step": 7150
},
{
"epoch": 1.33,
"grad_norm": 2.7925071554565806,
"learning_rate": 2.492815227842795e-06,
"loss": 1.0352,
"step": 7155
},
{
"epoch": 1.34,
"grad_norm": 2.6005784247999735,
"learning_rate": 2.486481460097068e-06,
"loss": 0.9978,
"step": 7160
},
{
"epoch": 1.34,
"grad_norm": 2.2973468599232247,
"learning_rate": 2.4801530850749086e-06,
"loss": 0.996,
"step": 7165
},
{
"epoch": 1.34,
"grad_norm": 2.831479201002048,
"learning_rate": 2.4738301163537675e-06,
"loss": 1.0229,
"step": 7170
},
{
"epoch": 1.34,
"grad_norm": 2.4063077604608827,
"learning_rate": 2.4675125674994983e-06,
"loss": 0.9992,
"step": 7175
},
{
"epoch": 1.34,
"grad_norm": 2.4806704995189772,
"learning_rate": 2.4612004520663246e-06,
"loss": 0.968,
"step": 7180
},
{
"epoch": 1.34,
"grad_norm": 2.6313623117831324,
"learning_rate": 2.4548937835968143e-06,
"loss": 0.988,
"step": 7185
},
{
"epoch": 1.34,
"grad_norm": 2.903837181362707,
"learning_rate": 2.448592575621849e-06,
"loss": 0.9885,
"step": 7190
},
{
"epoch": 1.34,
"grad_norm": 2.6711362094088305,
"learning_rate": 2.4422968416605897e-06,
"loss": 1.019,
"step": 7195
},
{
"epoch": 1.34,
"grad_norm": 3.1576982437768524,
"learning_rate": 2.4360065952204636e-06,
"loss": 1.033,
"step": 7200
},
{
"epoch": 1.34,
"grad_norm": 2.92877302700557,
"learning_rate": 2.429721849797115e-06,
"loss": 1.0043,
"step": 7205
},
{
"epoch": 1.34,
"grad_norm": 2.4781966294522877,
"learning_rate": 2.4234426188743893e-06,
"loss": 1.0097,
"step": 7210
},
{
"epoch": 1.35,
"grad_norm": 2.9640141348991427,
"learning_rate": 2.4171689159243005e-06,
"loss": 0.9569,
"step": 7215
},
{
"epoch": 1.35,
"grad_norm": 2.485092344715958,
"learning_rate": 2.4109007544070024e-06,
"loss": 1.0391,
"step": 7220
},
{
"epoch": 1.35,
"grad_norm": 2.700577108794825,
"learning_rate": 2.4046381477707587e-06,
"loss": 1.0008,
"step": 7225
},
{
"epoch": 1.35,
"grad_norm": 2.439655699190244,
"learning_rate": 2.3983811094519172e-06,
"loss": 0.9949,
"step": 7230
},
{
"epoch": 1.35,
"grad_norm": 2.2209150426109816,
"learning_rate": 2.3921296528748774e-06,
"loss": 1.0128,
"step": 7235
},
{
"epoch": 1.35,
"grad_norm": 2.3755750932435036,
"learning_rate": 2.385883791452061e-06,
"loss": 1.0717,
"step": 7240
},
{
"epoch": 1.35,
"grad_norm": 2.4671325785441445,
"learning_rate": 2.3796435385838926e-06,
"loss": 1.0212,
"step": 7245
},
{
"epoch": 1.35,
"grad_norm": 3.012659647333001,
"learning_rate": 2.3734089076587572e-06,
"loss": 1.0254,
"step": 7250
},
{
"epoch": 1.35,
"grad_norm": 2.512602473812372,
"learning_rate": 2.3671799120529793e-06,
"loss": 0.9908,
"step": 7255
},
{
"epoch": 1.35,
"grad_norm": 2.4269671022760564,
"learning_rate": 2.360956565130794e-06,
"loss": 0.9849,
"step": 7260
},
{
"epoch": 1.35,
"grad_norm": 2.4199392850875454,
"learning_rate": 2.3547388802443166e-06,
"loss": 0.9899,
"step": 7265
},
{
"epoch": 1.36,
"grad_norm": 2.483465002460721,
"learning_rate": 2.348526870733515e-06,
"loss": 0.9675,
"step": 7270
},
{
"epoch": 1.36,
"grad_norm": 2.7165707232570138,
"learning_rate": 2.3423205499261796e-06,
"loss": 1.0041,
"step": 7275
},
{
"epoch": 1.36,
"grad_norm": 2.507174897930294,
"learning_rate": 2.336119931137897e-06,
"loss": 1.0225,
"step": 7280
},
{
"epoch": 1.36,
"grad_norm": 2.486968006992097,
"learning_rate": 2.3299250276720175e-06,
"loss": 1.0073,
"step": 7285
},
{
"epoch": 1.36,
"grad_norm": 2.4942272969865287,
"learning_rate": 2.323735852819635e-06,
"loss": 1.0089,
"step": 7290
},
{
"epoch": 1.36,
"grad_norm": 2.6627456223879515,
"learning_rate": 2.31755241985955e-06,
"loss": 0.9922,
"step": 7295
},
{
"epoch": 1.36,
"grad_norm": 2.682610387682074,
"learning_rate": 2.3113747420582383e-06,
"loss": 1.0586,
"step": 7300
},
{
"epoch": 1.36,
"grad_norm": 2.3761170060074472,
"learning_rate": 2.3052028326698333e-06,
"loss": 0.9964,
"step": 7305
},
{
"epoch": 1.36,
"grad_norm": 2.6110018966383937,
"learning_rate": 2.299036704936095e-06,
"loss": 1.0244,
"step": 7310
},
{
"epoch": 1.36,
"grad_norm": 2.562086827792303,
"learning_rate": 2.292876372086375e-06,
"loss": 1.013,
"step": 7315
},
{
"epoch": 1.37,
"grad_norm": 2.6929334887746026,
"learning_rate": 2.286721847337591e-06,
"loss": 1.0022,
"step": 7320
},
{
"epoch": 1.37,
"grad_norm": 2.9940515986542255,
"learning_rate": 2.2805731438942016e-06,
"loss": 0.9905,
"step": 7325
},
{
"epoch": 1.37,
"grad_norm": 2.849568114317893,
"learning_rate": 2.274430274948179e-06,
"loss": 0.9861,
"step": 7330
},
{
"epoch": 1.37,
"grad_norm": 2.67404623335488,
"learning_rate": 2.2682932536789693e-06,
"loss": 0.9973,
"step": 7335
},
{
"epoch": 1.37,
"grad_norm": 2.5576579250667812,
"learning_rate": 2.26216209325348e-06,
"loss": 0.9966,
"step": 7340
},
{
"epoch": 1.37,
"grad_norm": 2.5678223637612834,
"learning_rate": 2.25603680682604e-06,
"loss": 1.0126,
"step": 7345
},
{
"epoch": 1.37,
"grad_norm": 2.24656998780328,
"learning_rate": 2.2499174075383762e-06,
"loss": 0.9919,
"step": 7350
},
{
"epoch": 1.37,
"grad_norm": 2.5754557845252344,
"learning_rate": 2.2438039085195897e-06,
"loss": 0.9964,
"step": 7355
},
{
"epoch": 1.37,
"grad_norm": 2.7952425186487737,
"learning_rate": 2.237696322886115e-06,
"loss": 1.0108,
"step": 7360
},
{
"epoch": 1.37,
"grad_norm": 2.3832345940042585,
"learning_rate": 2.2315946637417046e-06,
"loss": 0.9369,
"step": 7365
},
{
"epoch": 1.37,
"grad_norm": 2.582206718494254,
"learning_rate": 2.225498944177394e-06,
"loss": 0.9998,
"step": 7370
},
{
"epoch": 1.38,
"grad_norm": 3.1387040443518015,
"learning_rate": 2.2194091772714754e-06,
"loss": 0.9992,
"step": 7375
},
{
"epoch": 1.38,
"grad_norm": 2.6314196217784502,
"learning_rate": 2.2133253760894697e-06,
"loss": 0.9607,
"step": 7380
},
{
"epoch": 1.38,
"grad_norm": 2.8777705905638977,
"learning_rate": 2.207247553684099e-06,
"loss": 1.0068,
"step": 7385
},
{
"epoch": 1.38,
"grad_norm": 2.7533052839525145,
"learning_rate": 2.201175723095258e-06,
"loss": 1.0119,
"step": 7390
},
{
"epoch": 1.38,
"grad_norm": 2.6256471402176107,
"learning_rate": 2.1951098973499823e-06,
"loss": 0.9979,
"step": 7395
},
{
"epoch": 1.38,
"grad_norm": 2.3971531857526824,
"learning_rate": 2.189050089462433e-06,
"loss": 1.0022,
"step": 7400
},
{
"epoch": 1.38,
"grad_norm": 3.014527777899452,
"learning_rate": 2.1829963124338514e-06,
"loss": 1.0189,
"step": 7405
},
{
"epoch": 1.38,
"grad_norm": 2.4836659831352024,
"learning_rate": 2.1769485792525437e-06,
"loss": 0.9942,
"step": 7410
},
{
"epoch": 1.38,
"grad_norm": 2.49850328272449,
"learning_rate": 2.170906902893847e-06,
"loss": 0.9904,
"step": 7415
},
{
"epoch": 1.38,
"grad_norm": 2.4045192020286676,
"learning_rate": 2.164871296320106e-06,
"loss": 1.0055,
"step": 7420
},
{
"epoch": 1.38,
"grad_norm": 2.5940798839061148,
"learning_rate": 2.15884177248064e-06,
"loss": 0.9927,
"step": 7425
},
{
"epoch": 1.39,
"grad_norm": 3.2043613592102087,
"learning_rate": 2.152818344311721e-06,
"loss": 1.0198,
"step": 7430
},
{
"epoch": 1.39,
"grad_norm": 2.9549208216543086,
"learning_rate": 2.1468010247365373e-06,
"loss": 1.0524,
"step": 7435
},
{
"epoch": 1.39,
"grad_norm": 2.431983887961489,
"learning_rate": 2.1407898266651805e-06,
"loss": 1.0261,
"step": 7440
},
{
"epoch": 1.39,
"grad_norm": 2.8943587841934955,
"learning_rate": 2.134784762994601e-06,
"loss": 1.0147,
"step": 7445
},
{
"epoch": 1.39,
"grad_norm": 2.6469118849255504,
"learning_rate": 2.12878584660859e-06,
"loss": 1.011,
"step": 7450
},
{
"epoch": 1.39,
"grad_norm": 2.8503986008176567,
"learning_rate": 2.1227930903777517e-06,
"loss": 0.9992,
"step": 7455
},
{
"epoch": 1.39,
"grad_norm": 2.7444067798255296,
"learning_rate": 2.116806507159466e-06,
"loss": 1.0106,
"step": 7460
},
{
"epoch": 1.39,
"grad_norm": 2.8278801460868337,
"learning_rate": 2.1108261097978812e-06,
"loss": 1.007,
"step": 7465
},
{
"epoch": 1.39,
"grad_norm": 2.9725590428126054,
"learning_rate": 2.1048519111238647e-06,
"loss": 0.9747,
"step": 7470
},
{
"epoch": 1.39,
"grad_norm": 2.9983088444929225,
"learning_rate": 2.0988839239549875e-06,
"loss": 1.0049,
"step": 7475
},
{
"epoch": 1.39,
"grad_norm": 2.457640569736608,
"learning_rate": 2.092922161095492e-06,
"loss": 1.0196,
"step": 7480
},
{
"epoch": 1.4,
"grad_norm": 2.916737237944643,
"learning_rate": 2.0869666353362724e-06,
"loss": 1.0224,
"step": 7485
},
{
"epoch": 1.4,
"grad_norm": 2.749518033440132,
"learning_rate": 2.0810173594548356e-06,
"loss": 1.0098,
"step": 7490
},
{
"epoch": 1.4,
"grad_norm": 2.454945647019492,
"learning_rate": 2.0750743462152782e-06,
"loss": 0.9794,
"step": 7495
},
{
"epoch": 1.4,
"grad_norm": 2.5914266247015565,
"learning_rate": 2.0691376083682635e-06,
"loss": 1.0079,
"step": 7500
},
{
"epoch": 1.4,
"grad_norm": 2.6360537632556476,
"learning_rate": 2.063207158650994e-06,
"loss": 1.0306,
"step": 7505
},
{
"epoch": 1.4,
"grad_norm": 2.705751594055718,
"learning_rate": 2.0572830097871747e-06,
"loss": 0.9879,
"step": 7510
},
{
"epoch": 1.4,
"grad_norm": 2.417410823436373,
"learning_rate": 2.0513651744869966e-06,
"loss": 1.0156,
"step": 7515
},
{
"epoch": 1.4,
"grad_norm": 2.58329594764982,
"learning_rate": 2.0454536654471035e-06,
"loss": 0.9679,
"step": 7520
},
{
"epoch": 1.4,
"grad_norm": 2.5375397018150254,
"learning_rate": 2.039548495350564e-06,
"loss": 0.998,
"step": 7525
},
{
"epoch": 1.4,
"grad_norm": 2.5457670422000462,
"learning_rate": 2.0336496768668545e-06,
"loss": 1.0162,
"step": 7530
},
{
"epoch": 1.41,
"grad_norm": 2.5385173317223124,
"learning_rate": 2.0277572226518134e-06,
"loss": 0.9504,
"step": 7535
},
{
"epoch": 1.41,
"grad_norm": 2.6122071384819647,
"learning_rate": 2.021871145347632e-06,
"loss": 1.0261,
"step": 7540
},
{
"epoch": 1.41,
"grad_norm": 2.4559143987913004,
"learning_rate": 2.015991457582815e-06,
"loss": 1.0295,
"step": 7545
},
{
"epoch": 1.41,
"grad_norm": 2.4131900706265688,
"learning_rate": 2.010118171972165e-06,
"loss": 1.0106,
"step": 7550
},
{
"epoch": 1.41,
"grad_norm": 2.5830984989956187,
"learning_rate": 2.0042513011167447e-06,
"loss": 1.018,
"step": 7555
},
{
"epoch": 1.41,
"grad_norm": 2.5141342688412736,
"learning_rate": 1.998390857603853e-06,
"loss": 0.9796,
"step": 7560
},
{
"epoch": 1.41,
"grad_norm": 2.825530580323147,
"learning_rate": 1.9925368540070008e-06,
"loss": 1.0012,
"step": 7565
},
{
"epoch": 1.41,
"grad_norm": 2.570499501431887,
"learning_rate": 1.986689302885882e-06,
"loss": 0.9962,
"step": 7570
},
{
"epoch": 1.41,
"grad_norm": 2.5569510782263105,
"learning_rate": 1.980848216786346e-06,
"loss": 1.0039,
"step": 7575
},
{
"epoch": 1.41,
"grad_norm": 2.4037111815985623,
"learning_rate": 1.9750136082403735e-06,
"loss": 1.0027,
"step": 7580
},
{
"epoch": 1.41,
"grad_norm": 2.606403450673804,
"learning_rate": 1.9691854897660455e-06,
"loss": 1.0181,
"step": 7585
},
{
"epoch": 1.42,
"grad_norm": 2.6149611442247553,
"learning_rate": 1.9633638738675183e-06,
"loss": 0.9894,
"step": 7590
},
{
"epoch": 1.42,
"grad_norm": 2.731625908822058,
"learning_rate": 1.957548773035001e-06,
"loss": 0.9966,
"step": 7595
},
{
"epoch": 1.42,
"grad_norm": 2.6720427892224676,
"learning_rate": 1.9517401997447217e-06,
"loss": 0.9995,
"step": 7600
},
{
"epoch": 1.42,
"grad_norm": 2.2262976409569863,
"learning_rate": 1.945938166458902e-06,
"loss": 0.9857,
"step": 7605
},
{
"epoch": 1.42,
"grad_norm": 2.5234422281434123,
"learning_rate": 1.940142685625736e-06,
"loss": 0.9802,
"step": 7610
},
{
"epoch": 1.42,
"grad_norm": 2.904221387786334,
"learning_rate": 1.9343537696793555e-06,
"loss": 1.0409,
"step": 7615
},
{
"epoch": 1.42,
"grad_norm": 2.4909926104739384,
"learning_rate": 1.9285714310398117e-06,
"loss": 0.9962,
"step": 7620
},
{
"epoch": 1.42,
"grad_norm": 2.383606867119876,
"learning_rate": 1.9227956821130406e-06,
"loss": 1.0326,
"step": 7625
},
{
"epoch": 1.42,
"grad_norm": 2.8933537644372262,
"learning_rate": 1.9170265352908426e-06,
"loss": 0.9889,
"step": 7630
},
{
"epoch": 1.42,
"grad_norm": 2.5338221111488317,
"learning_rate": 1.911264002950849e-06,
"loss": 0.9956,
"step": 7635
},
{
"epoch": 1.42,
"grad_norm": 3.1011258166395628,
"learning_rate": 1.9055080974565076e-06,
"loss": 1.0198,
"step": 7640
},
{
"epoch": 1.43,
"grad_norm": 2.7342133503101147,
"learning_rate": 1.8997588311570425e-06,
"loss": 0.9802,
"step": 7645
},
{
"epoch": 1.43,
"grad_norm": 2.8091842542893977,
"learning_rate": 1.894016216387436e-06,
"loss": 0.9592,
"step": 7650
},
{
"epoch": 1.43,
"grad_norm": 3.407053889176312,
"learning_rate": 1.888280265468393e-06,
"loss": 1.0095,
"step": 7655
},
{
"epoch": 1.43,
"grad_norm": 2.693200133443283,
"learning_rate": 1.8825509907063328e-06,
"loss": 1.0003,
"step": 7660
},
{
"epoch": 1.43,
"grad_norm": 2.246216665996247,
"learning_rate": 1.8768284043933427e-06,
"loss": 0.9846,
"step": 7665
},
{
"epoch": 1.43,
"grad_norm": 2.346758280901885,
"learning_rate": 1.8711125188071621e-06,
"loss": 1.0293,
"step": 7670
},
{
"epoch": 1.43,
"grad_norm": 2.787357287294682,
"learning_rate": 1.865403346211153e-06,
"loss": 1.0274,
"step": 7675
},
{
"epoch": 1.43,
"grad_norm": 3.7658119105298917,
"learning_rate": 1.8597008988542791e-06,
"loss": 0.9855,
"step": 7680
},
{
"epoch": 1.43,
"grad_norm": 2.5022564517891235,
"learning_rate": 1.854005188971071e-06,
"loss": 1.0077,
"step": 7685
},
{
"epoch": 1.43,
"grad_norm": 2.934534992469462,
"learning_rate": 1.848316228781606e-06,
"loss": 1.0042,
"step": 7690
},
{
"epoch": 1.44,
"grad_norm": 2.8338775537793035,
"learning_rate": 1.8426340304914764e-06,
"loss": 0.9496,
"step": 7695
},
{
"epoch": 1.44,
"grad_norm": 2.4873584490481133,
"learning_rate": 1.8369586062917693e-06,
"loss": 1.0171,
"step": 7700
},
{
"epoch": 1.44,
"grad_norm": 3.0088916973384765,
"learning_rate": 1.8312899683590424e-06,
"loss": 0.9972,
"step": 7705
},
{
"epoch": 1.44,
"grad_norm": 2.656464767123038,
"learning_rate": 1.8256281288552885e-06,
"loss": 1.0099,
"step": 7710
},
{
"epoch": 1.44,
"grad_norm": 3.0511739775979114,
"learning_rate": 1.8199730999279147e-06,
"loss": 1.0329,
"step": 7715
},
{
"epoch": 1.44,
"grad_norm": 2.586345592514905,
"learning_rate": 1.8143248937097153e-06,
"loss": 1.0266,
"step": 7720
},
{
"epoch": 1.44,
"grad_norm": 2.5776112571536216,
"learning_rate": 1.8086835223188542e-06,
"loss": 1.0502,
"step": 7725
},
{
"epoch": 1.44,
"grad_norm": 2.5687232321997833,
"learning_rate": 1.8030489978588195e-06,
"loss": 1.0314,
"step": 7730
},
{
"epoch": 1.44,
"grad_norm": 2.589975182340955,
"learning_rate": 1.7974213324184176e-06,
"loss": 1.0744,
"step": 7735
},
{
"epoch": 1.44,
"grad_norm": 2.4216590891248093,
"learning_rate": 1.7918005380717341e-06,
"loss": 1.0085,
"step": 7740
},
{
"epoch": 1.44,
"grad_norm": 2.6874915737338756,
"learning_rate": 1.7861866268781186e-06,
"loss": 0.9891,
"step": 7745
},
{
"epoch": 1.45,
"grad_norm": 2.730115388264179,
"learning_rate": 1.7805796108821483e-06,
"loss": 1.0821,
"step": 7750
},
{
"epoch": 1.45,
"grad_norm": 2.58907575983276,
"learning_rate": 1.7749795021136074e-06,
"loss": 1.0293,
"step": 7755
},
{
"epoch": 1.45,
"grad_norm": 2.400032501236323,
"learning_rate": 1.769386312587462e-06,
"loss": 0.9911,
"step": 7760
},
{
"epoch": 1.45,
"grad_norm": 2.5616088617013593,
"learning_rate": 1.7638000543038326e-06,
"loss": 1.022,
"step": 7765
},
{
"epoch": 1.45,
"grad_norm": 3.011699779238265,
"learning_rate": 1.7582207392479684e-06,
"loss": 0.9719,
"step": 7770
},
{
"epoch": 1.45,
"grad_norm": 2.7120556534956206,
"learning_rate": 1.7526483793902215e-06,
"loss": 1.0045,
"step": 7775
},
{
"epoch": 1.45,
"grad_norm": 2.41683174537852,
"learning_rate": 1.747082986686024e-06,
"loss": 1.0072,
"step": 7780
},
{
"epoch": 1.45,
"grad_norm": 2.32504667984806,
"learning_rate": 1.741524573075855e-06,
"loss": 0.9902,
"step": 7785
},
{
"epoch": 1.45,
"grad_norm": 2.6375990189427148,
"learning_rate": 1.7359731504852283e-06,
"loss": 0.9992,
"step": 7790
},
{
"epoch": 1.45,
"grad_norm": 2.404224852309051,
"learning_rate": 1.7304287308246514e-06,
"loss": 0.9802,
"step": 7795
},
{
"epoch": 1.45,
"grad_norm": 2.7029433370239087,
"learning_rate": 1.72489132598961e-06,
"loss": 1.0207,
"step": 7800
},
{
"epoch": 1.46,
"grad_norm": 2.3249820631357827,
"learning_rate": 1.7193609478605388e-06,
"loss": 1.0503,
"step": 7805
},
{
"epoch": 1.46,
"grad_norm": 2.634946032303947,
"learning_rate": 1.713837608302797e-06,
"loss": 1.0101,
"step": 7810
},
{
"epoch": 1.46,
"grad_norm": 2.3929583667727585,
"learning_rate": 1.7083213191666425e-06,
"loss": 0.9947,
"step": 7815
},
{
"epoch": 1.46,
"grad_norm": 2.71027976931746,
"learning_rate": 1.7028120922872071e-06,
"loss": 1.017,
"step": 7820
},
{
"epoch": 1.46,
"grad_norm": 2.668532399284522,
"learning_rate": 1.69730993948447e-06,
"loss": 1.0405,
"step": 7825
},
{
"epoch": 1.46,
"grad_norm": 2.6708797038580636,
"learning_rate": 1.6918148725632322e-06,
"loss": 0.9941,
"step": 7830
},
{
"epoch": 1.46,
"grad_norm": 2.5676328647086586,
"learning_rate": 1.6863269033130963e-06,
"loss": 1.0316,
"step": 7835
},
{
"epoch": 1.46,
"grad_norm": 2.9899945946380626,
"learning_rate": 1.6808460435084316e-06,
"loss": 0.9909,
"step": 7840
},
{
"epoch": 1.46,
"grad_norm": 3.054882339496845,
"learning_rate": 1.6753723049083575e-06,
"loss": 1.0078,
"step": 7845
},
{
"epoch": 1.46,
"grad_norm": 2.609112302401605,
"learning_rate": 1.6699056992567132e-06,
"loss": 0.989,
"step": 7850
},
{
"epoch": 1.46,
"grad_norm": 2.430195809966236,
"learning_rate": 1.6644462382820348e-06,
"loss": 1.0013,
"step": 7855
},
{
"epoch": 1.47,
"grad_norm": 3.2390129601989823,
"learning_rate": 1.6589939336975298e-06,
"loss": 0.9794,
"step": 7860
},
{
"epoch": 1.47,
"grad_norm": 2.4217885170245452,
"learning_rate": 1.6535487972010511e-06,
"loss": 0.9815,
"step": 7865
},
{
"epoch": 1.47,
"grad_norm": 2.929203958696014,
"learning_rate": 1.648110840475074e-06,
"loss": 0.9624,
"step": 7870
},
{
"epoch": 1.47,
"grad_norm": 2.5488052903322385,
"learning_rate": 1.6426800751866662e-06,
"loss": 1.0104,
"step": 7875
},
{
"epoch": 1.47,
"grad_norm": 2.8700612666358607,
"learning_rate": 1.637256512987473e-06,
"loss": 1.0434,
"step": 7880
},
{
"epoch": 1.47,
"grad_norm": 2.8596393228875967,
"learning_rate": 1.6318401655136807e-06,
"loss": 0.9891,
"step": 7885
},
{
"epoch": 1.47,
"grad_norm": 3.0486393042996323,
"learning_rate": 1.6264310443859937e-06,
"loss": 1.0076,
"step": 7890
},
{
"epoch": 1.47,
"grad_norm": 2.3273538904602655,
"learning_rate": 1.6210291612096168e-06,
"loss": 0.9968,
"step": 7895
},
{
"epoch": 1.47,
"grad_norm": 2.914791726331403,
"learning_rate": 1.6156345275742274e-06,
"loss": 1.0513,
"step": 7900
},
{
"epoch": 1.47,
"grad_norm": 3.1399450509554034,
"learning_rate": 1.610247155053945e-06,
"loss": 1.0086,
"step": 7905
},
{
"epoch": 1.48,
"grad_norm": 2.7207488493952297,
"learning_rate": 1.6048670552073125e-06,
"loss": 0.9984,
"step": 7910
},
{
"epoch": 1.48,
"grad_norm": 2.6048326574991654,
"learning_rate": 1.5994942395772668e-06,
"loss": 1.0135,
"step": 7915
},
{
"epoch": 1.48,
"grad_norm": 2.6367561560889774,
"learning_rate": 1.5941287196911236e-06,
"loss": 0.9907,
"step": 7920
},
{
"epoch": 1.48,
"grad_norm": 2.4769120488306227,
"learning_rate": 1.5887705070605364e-06,
"loss": 1.0166,
"step": 7925
},
{
"epoch": 1.48,
"grad_norm": 2.5189491566743256,
"learning_rate": 1.5834196131814872e-06,
"loss": 1.0052,
"step": 7930
},
{
"epoch": 1.48,
"grad_norm": 2.4493928951618664,
"learning_rate": 1.578076049534253e-06,
"loss": 1.0553,
"step": 7935
},
{
"epoch": 1.48,
"grad_norm": 2.5351068015113625,
"learning_rate": 1.572739827583384e-06,
"loss": 1.0467,
"step": 7940
},
{
"epoch": 1.48,
"grad_norm": 2.655233690562782,
"learning_rate": 1.5674109587776825e-06,
"loss": 0.9715,
"step": 7945
},
{
"epoch": 1.48,
"grad_norm": 2.5534781581480117,
"learning_rate": 1.5620894545501709e-06,
"loss": 0.9374,
"step": 7950
},
{
"epoch": 1.48,
"grad_norm": 2.7528561246346124,
"learning_rate": 1.5567753263180712e-06,
"loss": 0.9979,
"step": 7955
},
{
"epoch": 1.48,
"grad_norm": 3.0595711702661093,
"learning_rate": 1.551468585482782e-06,
"loss": 0.9997,
"step": 7960
},
{
"epoch": 1.49,
"grad_norm": 2.6100795976750435,
"learning_rate": 1.5461692434298515e-06,
"loss": 1.0002,
"step": 7965
},
{
"epoch": 1.49,
"grad_norm": 2.873387156051466,
"learning_rate": 1.5408773115289527e-06,
"loss": 1.0216,
"step": 7970
},
{
"epoch": 1.49,
"grad_norm": 2.6690217461993444,
"learning_rate": 1.5355928011338628e-06,
"loss": 0.9939,
"step": 7975
},
{
"epoch": 1.49,
"grad_norm": 2.506394501418365,
"learning_rate": 1.5303157235824323e-06,
"loss": 1.0374,
"step": 7980
},
{
"epoch": 1.49,
"grad_norm": 2.26834419318478,
"learning_rate": 1.5250460901965702e-06,
"loss": 0.9837,
"step": 7985
},
{
"epoch": 1.49,
"grad_norm": 2.626617307266259,
"learning_rate": 1.51978391228221e-06,
"loss": 1.0259,
"step": 7990
},
{
"epoch": 1.49,
"grad_norm": 2.514926149191691,
"learning_rate": 1.5145292011292905e-06,
"loss": 0.9788,
"step": 7995
},
{
"epoch": 1.49,
"grad_norm": 2.566211120378057,
"learning_rate": 1.509281968011731e-06,
"loss": 0.9221,
"step": 8000
},
{
"epoch": 1.49,
"grad_norm": 3.090252271917728,
"learning_rate": 1.5040422241874058e-06,
"loss": 1.027,
"step": 8005
},
{
"epoch": 1.49,
"grad_norm": 2.7794131086955,
"learning_rate": 1.4988099808981227e-06,
"loss": 0.9859,
"step": 8010
},
{
"epoch": 1.49,
"grad_norm": 2.4233202674174334,
"learning_rate": 1.4935852493695962e-06,
"loss": 1.0112,
"step": 8015
},
{
"epoch": 1.5,
"grad_norm": 2.6253735306012476,
"learning_rate": 1.4883680408114238e-06,
"loss": 0.9951,
"step": 8020
},
{
"epoch": 1.5,
"grad_norm": 2.7054085641131747,
"learning_rate": 1.4831583664170619e-06,
"loss": 0.9924,
"step": 8025
},
{
"epoch": 1.5,
"grad_norm": 2.66331786719659,
"learning_rate": 1.4779562373638073e-06,
"loss": 1.0151,
"step": 8030
},
{
"epoch": 1.5,
"grad_norm": 2.4687092490748976,
"learning_rate": 1.472761664812763e-06,
"loss": 1.0439,
"step": 8035
},
{
"epoch": 1.5,
"grad_norm": 2.2381123844793023,
"learning_rate": 1.4675746599088208e-06,
"loss": 0.992,
"step": 8040
},
{
"epoch": 1.5,
"grad_norm": 2.7820004676047483,
"learning_rate": 1.4623952337806397e-06,
"loss": 0.9766,
"step": 8045
},
{
"epoch": 1.5,
"grad_norm": 2.843933021496936,
"learning_rate": 1.4572233975406102e-06,
"loss": 0.9946,
"step": 8050
},
{
"epoch": 1.5,
"grad_norm": 2.847159497897028,
"learning_rate": 1.4520591622848495e-06,
"loss": 1.01,
"step": 8055
},
{
"epoch": 1.5,
"grad_norm": 2.5912260687609376,
"learning_rate": 1.4469025390931595e-06,
"loss": 1.0257,
"step": 8060
},
{
"epoch": 1.5,
"grad_norm": 2.499004311563004,
"learning_rate": 1.4417535390290139e-06,
"loss": 1.0954,
"step": 8065
},
{
"epoch": 1.5,
"grad_norm": 2.6280528719952874,
"learning_rate": 1.4366121731395266e-06,
"loss": 1.0619,
"step": 8070
},
{
"epoch": 1.51,
"grad_norm": 2.5818235310207545,
"learning_rate": 1.4314784524554405e-06,
"loss": 1.0654,
"step": 8075
},
{
"epoch": 1.51,
"grad_norm": 2.857141504036705,
"learning_rate": 1.4263523879910902e-06,
"loss": 0.9869,
"step": 8080
},
{
"epoch": 1.51,
"grad_norm": 2.4989029878764413,
"learning_rate": 1.4212339907443822e-06,
"loss": 1.0482,
"step": 8085
},
{
"epoch": 1.51,
"grad_norm": 2.6923172446783954,
"learning_rate": 1.416123271696776e-06,
"loss": 0.9947,
"step": 8090
},
{
"epoch": 1.51,
"grad_norm": 2.5223697256777005,
"learning_rate": 1.4110202418132596e-06,
"loss": 1.0431,
"step": 8095
},
{
"epoch": 1.51,
"grad_norm": 2.5206572271590697,
"learning_rate": 1.4059249120423213e-06,
"loss": 1.0203,
"step": 8100
},
{
"epoch": 1.51,
"grad_norm": 2.274549996984812,
"learning_rate": 1.4008372933159287e-06,
"loss": 0.9754,
"step": 8105
},
{
"epoch": 1.51,
"grad_norm": 2.463111888028792,
"learning_rate": 1.3957573965495064e-06,
"loss": 1.026,
"step": 8110
},
{
"epoch": 1.51,
"grad_norm": 2.3299161035359477,
"learning_rate": 1.3906852326419096e-06,
"loss": 1.0099,
"step": 8115
},
{
"epoch": 1.51,
"grad_norm": 2.3383819268062713,
"learning_rate": 1.385620812475409e-06,
"loss": 0.9251,
"step": 8120
},
{
"epoch": 1.52,
"grad_norm": 2.780420753115608,
"learning_rate": 1.3805641469156523e-06,
"loss": 0.9935,
"step": 8125
},
{
"epoch": 1.52,
"grad_norm": 2.5750186026284276,
"learning_rate": 1.3755152468116562e-06,
"loss": 1.0058,
"step": 8130
},
{
"epoch": 1.52,
"grad_norm": 2.935898032597156,
"learning_rate": 1.3704741229957723e-06,
"loss": 0.9991,
"step": 8135
},
{
"epoch": 1.52,
"grad_norm": 2.5933993665579242,
"learning_rate": 1.365440786283674e-06,
"loss": 1.0002,
"step": 8140
},
{
"epoch": 1.52,
"grad_norm": 2.7639651828531577,
"learning_rate": 1.3604152474743233e-06,
"loss": 0.9922,
"step": 8145
},
{
"epoch": 1.52,
"grad_norm": 3.2441453180376048,
"learning_rate": 1.3553975173499516e-06,
"loss": 0.9989,
"step": 8150
},
{
"epoch": 1.52,
"grad_norm": 2.4970579337537373,
"learning_rate": 1.350387606676039e-06,
"loss": 1.0367,
"step": 8155
},
{
"epoch": 1.52,
"grad_norm": 2.5016675644006856,
"learning_rate": 1.3453855262012877e-06,
"loss": 0.9925,
"step": 8160
},
{
"epoch": 1.52,
"grad_norm": 2.712422105166325,
"learning_rate": 1.3403912866576013e-06,
"loss": 0.9872,
"step": 8165
},
{
"epoch": 1.52,
"grad_norm": 2.6277080069881475,
"learning_rate": 1.3354048987600597e-06,
"loss": 0.9896,
"step": 8170
},
{
"epoch": 1.52,
"grad_norm": 2.2742594923450294,
"learning_rate": 1.3304263732068978e-06,
"loss": 1.0459,
"step": 8175
},
{
"epoch": 1.53,
"grad_norm": 3.198036891220997,
"learning_rate": 1.3254557206794805e-06,
"loss": 1.0019,
"step": 8180
},
{
"epoch": 1.53,
"grad_norm": 2.8910090926748206,
"learning_rate": 1.3204929518422854e-06,
"loss": 1.0001,
"step": 8185
},
{
"epoch": 1.53,
"grad_norm": 2.588419180976511,
"learning_rate": 1.3155380773428716e-06,
"loss": 1.0491,
"step": 8190
},
{
"epoch": 1.53,
"grad_norm": 2.9831121608734863,
"learning_rate": 1.3105911078118623e-06,
"loss": 1.0183,
"step": 8195
},
{
"epoch": 1.53,
"grad_norm": 2.7492226038465577,
"learning_rate": 1.3056520538629202e-06,
"loss": 0.9878,
"step": 8200
},
{
"epoch": 1.53,
"grad_norm": 2.3860445073262215,
"learning_rate": 1.3007209260927262e-06,
"loss": 0.9998,
"step": 8205
},
{
"epoch": 1.53,
"grad_norm": 2.7245221333282394,
"learning_rate": 1.2957977350809548e-06,
"loss": 1.0239,
"step": 8210
},
{
"epoch": 1.53,
"grad_norm": 2.436510051916248,
"learning_rate": 1.290882491390253e-06,
"loss": 1.0715,
"step": 8215
},
{
"epoch": 1.53,
"grad_norm": 2.278573195904812,
"learning_rate": 1.2859752055662144e-06,
"loss": 0.9442,
"step": 8220
},
{
"epoch": 1.53,
"grad_norm": 2.4519680767900773,
"learning_rate": 1.2810758881373653e-06,
"loss": 0.9701,
"step": 8225
},
{
"epoch": 1.53,
"grad_norm": 2.8358651175843423,
"learning_rate": 1.2761845496151299e-06,
"loss": 0.9333,
"step": 8230
},
{
"epoch": 1.54,
"grad_norm": 2.5083663458407113,
"learning_rate": 1.2713012004938157e-06,
"loss": 1.0233,
"step": 8235
},
{
"epoch": 1.54,
"grad_norm": 2.33214694353536,
"learning_rate": 1.26642585125059e-06,
"loss": 0.9518,
"step": 8240
},
{
"epoch": 1.54,
"grad_norm": 2.833861135741682,
"learning_rate": 1.2615585123454522e-06,
"loss": 0.9277,
"step": 8245
},
{
"epoch": 1.54,
"grad_norm": 2.67836542381953,
"learning_rate": 1.2566991942212226e-06,
"loss": 1.008,
"step": 8250
},
{
"epoch": 1.54,
"grad_norm": 2.3412731570156002,
"learning_rate": 1.2518479073035078e-06,
"loss": 0.9933,
"step": 8255
},
{
"epoch": 1.54,
"grad_norm": 2.637129686981733,
"learning_rate": 1.247004662000686e-06,
"loss": 0.9752,
"step": 8260
},
{
"epoch": 1.54,
"grad_norm": 2.503883482079726,
"learning_rate": 1.2421694687038782e-06,
"loss": 0.9784,
"step": 8265
},
{
"epoch": 1.54,
"grad_norm": 3.93407055926674,
"learning_rate": 1.2373423377869376e-06,
"loss": 1.036,
"step": 8270
},
{
"epoch": 1.54,
"grad_norm": 2.454241790604054,
"learning_rate": 1.232523279606413e-06,
"loss": 1.0169,
"step": 8275
},
{
"epoch": 1.54,
"grad_norm": 2.53575212107835,
"learning_rate": 1.227712304501536e-06,
"loss": 0.9813,
"step": 8280
},
{
"epoch": 1.55,
"grad_norm": 2.3951270616987146,
"learning_rate": 1.222909422794194e-06,
"loss": 1.0607,
"step": 8285
},
{
"epoch": 1.55,
"grad_norm": 2.49607239228924,
"learning_rate": 1.2181146447889102e-06,
"loss": 0.9752,
"step": 8290
},
{
"epoch": 1.55,
"grad_norm": 2.942783317969311,
"learning_rate": 1.213327980772826e-06,
"loss": 0.9916,
"step": 8295
},
{
"epoch": 1.55,
"grad_norm": 2.655950651887915,
"learning_rate": 1.2085494410156702e-06,
"loss": 0.983,
"step": 8300
},
{
"epoch": 1.55,
"grad_norm": 2.254472695034608,
"learning_rate": 1.203779035769741e-06,
"loss": 1.0008,
"step": 8305
},
{
"epoch": 1.55,
"grad_norm": 2.5598190175983193,
"learning_rate": 1.1990167752698844e-06,
"loss": 0.9649,
"step": 8310
},
{
"epoch": 1.55,
"grad_norm": 2.8832350083558667,
"learning_rate": 1.1942626697334765e-06,
"loss": 0.978,
"step": 8315
},
{
"epoch": 1.55,
"grad_norm": 2.2769055391490576,
"learning_rate": 1.1895167293603882e-06,
"loss": 1.01,
"step": 8320
},
{
"epoch": 1.55,
"grad_norm": 2.868510340527335,
"learning_rate": 1.1847789643329799e-06,
"loss": 0.9954,
"step": 8325
},
{
"epoch": 1.55,
"grad_norm": 2.975783073583449,
"learning_rate": 1.1800493848160666e-06,
"loss": 0.9835,
"step": 8330
},
{
"epoch": 1.55,
"grad_norm": 2.839117035143341,
"learning_rate": 1.1753280009569068e-06,
"loss": 0.9374,
"step": 8335
},
{
"epoch": 1.56,
"grad_norm": 2.9242643444387855,
"learning_rate": 1.170614822885171e-06,
"loss": 1.0051,
"step": 8340
},
{
"epoch": 1.56,
"grad_norm": 2.4750110312055495,
"learning_rate": 1.165909860712926e-06,
"loss": 0.9987,
"step": 8345
},
{
"epoch": 1.56,
"grad_norm": 2.8365937492878324,
"learning_rate": 1.161213124534611e-06,
"loss": 1.0153,
"step": 8350
},
{
"epoch": 1.56,
"grad_norm": 2.8947472669930394,
"learning_rate": 1.1565246244270162e-06,
"loss": 0.974,
"step": 8355
},
{
"epoch": 1.56,
"grad_norm": 2.7331996001322367,
"learning_rate": 1.1518443704492616e-06,
"loss": 0.9935,
"step": 8360
},
{
"epoch": 1.56,
"grad_norm": 2.630249509479508,
"learning_rate": 1.1471723726427764e-06,
"loss": 1.0287,
"step": 8365
},
{
"epoch": 1.56,
"grad_norm": 2.591959586184885,
"learning_rate": 1.1425086410312741e-06,
"loss": 1.0038,
"step": 8370
},
{
"epoch": 1.56,
"grad_norm": 2.6408943576305535,
"learning_rate": 1.1378531856207337e-06,
"loss": 0.9827,
"step": 8375
},
{
"epoch": 1.56,
"grad_norm": 2.55792605523133,
"learning_rate": 1.1332060163993807e-06,
"loss": 0.9853,
"step": 8380
},
{
"epoch": 1.56,
"grad_norm": 3.0398087791861568,
"learning_rate": 1.1285671433376593e-06,
"loss": 0.9964,
"step": 8385
},
{
"epoch": 1.56,
"grad_norm": 2.6670733962637074,
"learning_rate": 1.1239365763882154e-06,
"loss": 0.9925,
"step": 8390
},
{
"epoch": 1.57,
"grad_norm": 3.4420907762244157,
"learning_rate": 1.119314325485874e-06,
"loss": 1.0508,
"step": 8395
},
{
"epoch": 1.57,
"grad_norm": 2.4398172452504143,
"learning_rate": 1.1147004005476192e-06,
"loss": 0.9965,
"step": 8400
},
{
"epoch": 1.57,
"grad_norm": 2.837429151462753,
"learning_rate": 1.1100948114725695e-06,
"loss": 1.022,
"step": 8405
},
{
"epoch": 1.57,
"grad_norm": 2.50970813890442,
"learning_rate": 1.1054975681419622e-06,
"loss": 0.9722,
"step": 8410
},
{
"epoch": 1.57,
"grad_norm": 2.6008952386375968,
"learning_rate": 1.100908680419126e-06,
"loss": 1.0042,
"step": 8415
},
{
"epoch": 1.57,
"grad_norm": 2.397601424695823,
"learning_rate": 1.0963281581494628e-06,
"loss": 1.009,
"step": 8420
},
{
"epoch": 1.57,
"grad_norm": 2.341096455070129,
"learning_rate": 1.0917560111604302e-06,
"loss": 0.9997,
"step": 8425
},
{
"epoch": 1.57,
"grad_norm": 2.72165027387936,
"learning_rate": 1.0871922492615133e-06,
"loss": 1.0503,
"step": 8430
},
{
"epoch": 1.57,
"grad_norm": 2.3765933075053876,
"learning_rate": 1.0826368822442078e-06,
"loss": 0.9567,
"step": 8435
},
{
"epoch": 1.57,
"grad_norm": 2.7711690893945997,
"learning_rate": 1.0780899198819983e-06,
"loss": 1.0002,
"step": 8440
},
{
"epoch": 1.57,
"grad_norm": 2.5883255453415086,
"learning_rate": 1.073551371930337e-06,
"loss": 1.0195,
"step": 8445
},
{
"epoch": 1.58,
"grad_norm": 2.984516847385917,
"learning_rate": 1.0690212481266243e-06,
"loss": 0.9722,
"step": 8450
},
{
"epoch": 1.58,
"grad_norm": 2.4493087022991276,
"learning_rate": 1.0644995581901856e-06,
"loss": 0.9661,
"step": 8455
},
{
"epoch": 1.58,
"grad_norm": 2.4118000504228387,
"learning_rate": 1.0599863118222503e-06,
"loss": 1.0067,
"step": 8460
},
{
"epoch": 1.58,
"grad_norm": 2.776031639098257,
"learning_rate": 1.055481518705937e-06,
"loss": 0.9601,
"step": 8465
},
{
"epoch": 1.58,
"grad_norm": 2.8333438440892866,
"learning_rate": 1.0509851885062227e-06,
"loss": 1.0002,
"step": 8470
},
{
"epoch": 1.58,
"grad_norm": 2.6776557473356597,
"learning_rate": 1.0464973308699311e-06,
"loss": 0.9865,
"step": 8475
},
{
"epoch": 1.58,
"grad_norm": 2.784349412463351,
"learning_rate": 1.0420179554257037e-06,
"loss": 1.0139,
"step": 8480
},
{
"epoch": 1.58,
"grad_norm": 2.862349613553008,
"learning_rate": 1.0375470717839864e-06,
"loss": 0.9772,
"step": 8485
},
{
"epoch": 1.58,
"grad_norm": 2.6005297452393377,
"learning_rate": 1.0330846895370078e-06,
"loss": 0.9669,
"step": 8490
},
{
"epoch": 1.58,
"grad_norm": 2.2580387526234844,
"learning_rate": 1.0286308182587546e-06,
"loss": 1.0336,
"step": 8495
},
{
"epoch": 1.59,
"grad_norm": 2.715174129691373,
"learning_rate": 1.0241854675049534e-06,
"loss": 1.0184,
"step": 8500
},
{
"epoch": 1.59,
"grad_norm": 2.2875458045237487,
"learning_rate": 1.0197486468130485e-06,
"loss": 0.9621,
"step": 8505
},
{
"epoch": 1.59,
"grad_norm": 3.1661384763089004,
"learning_rate": 1.0153203657021894e-06,
"loss": 1.0024,
"step": 8510
},
{
"epoch": 1.59,
"grad_norm": 2.3304188336777094,
"learning_rate": 1.0109006336731947e-06,
"loss": 1.0389,
"step": 8515
},
{
"epoch": 1.59,
"grad_norm": 2.525969437894621,
"learning_rate": 1.006489460208549e-06,
"loss": 0.9968,
"step": 8520
},
{
"epoch": 1.59,
"grad_norm": 2.387820313938779,
"learning_rate": 1.0020868547723694e-06,
"loss": 0.9955,
"step": 8525
},
{
"epoch": 1.59,
"grad_norm": 2.9191333287504664,
"learning_rate": 9.97692826810393e-07,
"loss": 0.9721,
"step": 8530
},
{
"epoch": 1.59,
"grad_norm": 2.8134674262578456,
"learning_rate": 9.933073857499554e-07,
"loss": 0.953,
"step": 8535
},
{
"epoch": 1.59,
"grad_norm": 2.716709444476881,
"learning_rate": 9.889305409999656e-07,
"loss": 1.0351,
"step": 8540
},
{
"epoch": 1.59,
"grad_norm": 2.5286335152563497,
"learning_rate": 9.845623019508915e-07,
"loss": 0.959,
"step": 8545
},
{
"epoch": 1.59,
"grad_norm": 2.870530482870655,
"learning_rate": 9.80202677974737e-07,
"loss": 1.01,
"step": 8550
},
{
"epoch": 1.6,
"grad_norm": 2.6542379957905853,
"learning_rate": 9.758516784250222e-07,
"loss": 0.9648,
"step": 8555
},
{
"epoch": 1.6,
"grad_norm": 2.4225005154609076,
"learning_rate": 9.715093126367643e-07,
"loss": 0.9523,
"step": 8560
},
{
"epoch": 1.6,
"grad_norm": 2.488467696082567,
"learning_rate": 9.671755899264562e-07,
"loss": 0.9919,
"step": 8565
},
{
"epoch": 1.6,
"grad_norm": 2.732784552303335,
"learning_rate": 9.628505195920463e-07,
"loss": 0.9653,
"step": 8570
},
{
"epoch": 1.6,
"grad_norm": 2.58854203634735,
"learning_rate": 9.585341109129227e-07,
"loss": 0.9737,
"step": 8575
},
{
"epoch": 1.6,
"grad_norm": 3.0065804562154033,
"learning_rate": 9.542263731498875e-07,
"loss": 1.0292,
"step": 8580
},
{
"epoch": 1.6,
"grad_norm": 3.1293104736581174,
"learning_rate": 9.499273155451388e-07,
"loss": 0.9769,
"step": 8585
},
{
"epoch": 1.6,
"grad_norm": 2.871745660395818,
"learning_rate": 9.456369473222538e-07,
"loss": 0.9853,
"step": 8590
},
{
"epoch": 1.6,
"grad_norm": 2.588651633798483,
"learning_rate": 9.413552776861645e-07,
"loss": 0.9886,
"step": 8595
},
{
"epoch": 1.6,
"grad_norm": 2.821289962013644,
"learning_rate": 9.370823158231418e-07,
"loss": 1.0294,
"step": 8600
},
{
"epoch": 1.6,
"grad_norm": 2.6198307828799154,
"learning_rate": 9.328180709007734e-07,
"loss": 0.9639,
"step": 8605
},
{
"epoch": 1.61,
"grad_norm": 2.589055089856369,
"learning_rate": 9.285625520679447e-07,
"loss": 1.0009,
"step": 8610
},
{
"epoch": 1.61,
"grad_norm": 2.6817091221550564,
"learning_rate": 9.243157684548199e-07,
"loss": 0.9768,
"step": 8615
},
{
"epoch": 1.61,
"grad_norm": 2.446688278010068,
"learning_rate": 9.200777291728224e-07,
"loss": 1.0152,
"step": 8620
},
{
"epoch": 1.61,
"grad_norm": 2.6819493855737067,
"learning_rate": 9.158484433146142e-07,
"loss": 0.9845,
"step": 8625
},
{
"epoch": 1.61,
"grad_norm": 2.4531681967551897,
"learning_rate": 9.116279199540768e-07,
"loss": 1.0151,
"step": 8630
},
{
"epoch": 1.61,
"grad_norm": 2.518328043463977,
"learning_rate": 9.074161681462906e-07,
"loss": 0.9412,
"step": 8635
},
{
"epoch": 1.61,
"grad_norm": 2.6161600853086715,
"learning_rate": 9.032131969275198e-07,
"loss": 0.9948,
"step": 8640
},
{
"epoch": 1.61,
"grad_norm": 2.622763413959486,
"learning_rate": 8.990190153151874e-07,
"loss": 0.9626,
"step": 8645
},
{
"epoch": 1.61,
"grad_norm": 2.574191956837669,
"learning_rate": 8.948336323078593e-07,
"loss": 1.0239,
"step": 8650
},
{
"epoch": 1.61,
"grad_norm": 2.72808187078771,
"learning_rate": 8.90657056885224e-07,
"loss": 0.9803,
"step": 8655
},
{
"epoch": 1.61,
"grad_norm": 2.305170921371496,
"learning_rate": 8.864892980080719e-07,
"loss": 0.9985,
"step": 8660
},
{
"epoch": 1.62,
"grad_norm": 2.5223234611611147,
"learning_rate": 8.823303646182823e-07,
"loss": 0.982,
"step": 8665
},
{
"epoch": 1.62,
"grad_norm": 2.6678684960116246,
"learning_rate": 8.781802656387955e-07,
"loss": 0.9834,
"step": 8670
},
{
"epoch": 1.62,
"grad_norm": 2.443915108308519,
"learning_rate": 8.740390099735973e-07,
"loss": 0.9941,
"step": 8675
},
{
"epoch": 1.62,
"grad_norm": 3.0455073502536436,
"learning_rate": 8.699066065077005e-07,
"loss": 0.9633,
"step": 8680
},
{
"epoch": 1.62,
"grad_norm": 2.8874510477383653,
"learning_rate": 8.657830641071297e-07,
"loss": 0.9885,
"step": 8685
},
{
"epoch": 1.62,
"grad_norm": 2.501539958965497,
"learning_rate": 8.616683916188939e-07,
"loss": 1.027,
"step": 8690
},
{
"epoch": 1.62,
"grad_norm": 2.739630521686995,
"learning_rate": 8.575625978709723e-07,
"loss": 1.0207,
"step": 8695
},
{
"epoch": 1.62,
"grad_norm": 2.800804703467251,
"learning_rate": 8.534656916722966e-07,
"loss": 1.0162,
"step": 8700
},
{
"epoch": 1.62,
"grad_norm": 2.6458421698177403,
"learning_rate": 8.493776818127275e-07,
"loss": 0.9937,
"step": 8705
},
{
"epoch": 1.62,
"grad_norm": 2.4116964460512316,
"learning_rate": 8.452985770630451e-07,
"loss": 0.9697,
"step": 8710
},
{
"epoch": 1.63,
"grad_norm": 2.5314410018943403,
"learning_rate": 8.412283861749149e-07,
"loss": 0.9986,
"step": 8715
},
{
"epoch": 1.63,
"grad_norm": 2.5639593529419744,
"learning_rate": 8.371671178808832e-07,
"loss": 0.957,
"step": 8720
},
{
"epoch": 1.63,
"grad_norm": 2.5695870932919584,
"learning_rate": 8.331147808943507e-07,
"loss": 1.0161,
"step": 8725
},
{
"epoch": 1.63,
"grad_norm": 2.603768679055716,
"learning_rate": 8.290713839095605e-07,
"loss": 1.0269,
"step": 8730
},
{
"epoch": 1.63,
"grad_norm": 2.447043867628572,
"learning_rate": 8.250369356015698e-07,
"loss": 0.9981,
"step": 8735
},
{
"epoch": 1.63,
"grad_norm": 2.3880441714776466,
"learning_rate": 8.210114446262391e-07,
"loss": 1.0256,
"step": 8740
},
{
"epoch": 1.63,
"grad_norm": 2.7658758501345826,
"learning_rate": 8.169949196202099e-07,
"loss": 1.0113,
"step": 8745
},
{
"epoch": 1.63,
"grad_norm": 3.1004638384283023,
"learning_rate": 8.129873692008883e-07,
"loss": 0.9592,
"step": 8750
},
{
"epoch": 1.63,
"grad_norm": 2.5833091864252467,
"learning_rate": 8.089888019664255e-07,
"loss": 0.964,
"step": 8755
},
{
"epoch": 1.63,
"grad_norm": 2.6515268577779207,
"learning_rate": 8.049992264956985e-07,
"loss": 0.9743,
"step": 8760
},
{
"epoch": 1.63,
"grad_norm": 2.8611413957572727,
"learning_rate": 8.010186513482931e-07,
"loss": 0.9987,
"step": 8765
},
{
"epoch": 1.64,
"grad_norm": 2.6172758310161157,
"learning_rate": 7.970470850644841e-07,
"loss": 0.9748,
"step": 8770
},
{
"epoch": 1.64,
"grad_norm": 3.1282851318839215,
"learning_rate": 7.930845361652212e-07,
"loss": 0.9794,
"step": 8775
},
{
"epoch": 1.64,
"grad_norm": 3.0573290970246303,
"learning_rate": 7.891310131521041e-07,
"loss": 0.9787,
"step": 8780
},
{
"epoch": 1.64,
"grad_norm": 2.1506066941256243,
"learning_rate": 7.851865245073675e-07,
"loss": 1.0085,
"step": 8785
},
{
"epoch": 1.64,
"grad_norm": 2.5240271641211263,
"learning_rate": 7.812510786938654e-07,
"loss": 1.0368,
"step": 8790
},
{
"epoch": 1.64,
"grad_norm": 2.423560041926841,
"learning_rate": 7.77324684155048e-07,
"loss": 0.9355,
"step": 8795
},
{
"epoch": 1.64,
"grad_norm": 2.96046939164647,
"learning_rate": 7.73407349314948e-07,
"loss": 1.0029,
"step": 8800
},
{
"epoch": 1.64,
"grad_norm": 2.8005942518059963,
"learning_rate": 7.69499082578159e-07,
"loss": 0.9935,
"step": 8805
},
{
"epoch": 1.64,
"grad_norm": 3.2244592316095817,
"learning_rate": 7.655998923298197e-07,
"loss": 0.9929,
"step": 8810
},
{
"epoch": 1.64,
"grad_norm": 2.70767496399375,
"learning_rate": 7.617097869355972e-07,
"loss": 1.0294,
"step": 8815
},
{
"epoch": 1.64,
"grad_norm": 2.6297837313438106,
"learning_rate": 7.57828774741664e-07,
"loss": 0.9631,
"step": 8820
},
{
"epoch": 1.65,
"grad_norm": 2.4742233021313593,
"learning_rate": 7.539568640746859e-07,
"loss": 1.019,
"step": 8825
},
{
"epoch": 1.65,
"grad_norm": 2.776492788189207,
"learning_rate": 7.500940632418003e-07,
"loss": 1.012,
"step": 8830
},
{
"epoch": 1.65,
"grad_norm": 2.4002889591019696,
"learning_rate": 7.462403805305968e-07,
"loss": 1.0021,
"step": 8835
},
{
"epoch": 1.65,
"grad_norm": 2.6289778414523037,
"learning_rate": 7.423958242091078e-07,
"loss": 0.998,
"step": 8840
},
{
"epoch": 1.65,
"grad_norm": 2.7338119825458262,
"learning_rate": 7.38560402525782e-07,
"loss": 0.9947,
"step": 8845
},
{
"epoch": 1.65,
"grad_norm": 2.414954611873808,
"learning_rate": 7.347341237094697e-07,
"loss": 1.0184,
"step": 8850
},
{
"epoch": 1.65,
"grad_norm": 3.019980775613549,
"learning_rate": 7.30916995969404e-07,
"loss": 1.0255,
"step": 8855
},
{
"epoch": 1.65,
"grad_norm": 2.6991242420719725,
"learning_rate": 7.271090274951898e-07,
"loss": 0.9912,
"step": 8860
},
{
"epoch": 1.65,
"grad_norm": 2.4453775704614746,
"learning_rate": 7.233102264567749e-07,
"loss": 0.9877,
"step": 8865
},
{
"epoch": 1.65,
"grad_norm": 2.3079920807165593,
"learning_rate": 7.195206010044425e-07,
"loss": 1.0195,
"step": 8870
},
{
"epoch": 1.66,
"grad_norm": 2.7013123383670576,
"learning_rate": 7.157401592687851e-07,
"loss": 1.043,
"step": 8875
},
{
"epoch": 1.66,
"grad_norm": 2.717114035440022,
"learning_rate": 7.119689093606974e-07,
"loss": 0.9513,
"step": 8880
},
{
"epoch": 1.66,
"grad_norm": 2.4488620724748977,
"learning_rate": 7.082068593713498e-07,
"loss": 0.9934,
"step": 8885
},
{
"epoch": 1.66,
"grad_norm": 2.533131093439108,
"learning_rate": 7.044540173721743e-07,
"loss": 0.942,
"step": 8890
},
{
"epoch": 1.66,
"grad_norm": 2.739407015115311,
"learning_rate": 7.007103914148483e-07,
"loss": 0.9678,
"step": 8895
},
{
"epoch": 1.66,
"grad_norm": 2.9594049097283412,
"learning_rate": 6.969759895312749e-07,
"loss": 1.0177,
"step": 8900
},
{
"epoch": 1.66,
"grad_norm": 2.5416612175743785,
"learning_rate": 6.932508197335719e-07,
"loss": 0.9669,
"step": 8905
},
{
"epoch": 1.66,
"grad_norm": 2.741704966229588,
"learning_rate": 6.895348900140414e-07,
"loss": 0.9719,
"step": 8910
},
{
"epoch": 1.66,
"grad_norm": 2.650765378296894,
"learning_rate": 6.858282083451673e-07,
"loss": 1.0236,
"step": 8915
},
{
"epoch": 1.66,
"grad_norm": 2.6247728708420914,
"learning_rate": 6.821307826795897e-07,
"loss": 0.9939,
"step": 8920
},
{
"epoch": 1.66,
"grad_norm": 2.8682972388443977,
"learning_rate": 6.784426209500916e-07,
"loss": 1.0111,
"step": 8925
},
{
"epoch": 1.67,
"grad_norm": 2.532873552583512,
"learning_rate": 6.747637310695792e-07,
"loss": 0.9711,
"step": 8930
},
{
"epoch": 1.67,
"grad_norm": 2.76208256828962,
"learning_rate": 6.710941209310656e-07,
"loss": 0.962,
"step": 8935
},
{
"epoch": 1.67,
"grad_norm": 2.7328359763817804,
"learning_rate": 6.674337984076551e-07,
"loss": 1.0309,
"step": 8940
},
{
"epoch": 1.67,
"grad_norm": 2.6713338056299203,
"learning_rate": 6.637827713525252e-07,
"loss": 1.0181,
"step": 8945
},
{
"epoch": 1.67,
"grad_norm": 2.299256870197894,
"learning_rate": 6.601410475989106e-07,
"loss": 0.9893,
"step": 8950
},
{
"epoch": 1.67,
"grad_norm": 2.581073111238717,
"learning_rate": 6.565086349600847e-07,
"loss": 1.0204,
"step": 8955
},
{
"epoch": 1.67,
"grad_norm": 2.6116790427110574,
"learning_rate": 6.52885541229345e-07,
"loss": 0.9743,
"step": 8960
},
{
"epoch": 1.67,
"grad_norm": 2.7309921989649677,
"learning_rate": 6.492717741799948e-07,
"loss": 0.9694,
"step": 8965
},
{
"epoch": 1.67,
"grad_norm": 2.586747283406679,
"learning_rate": 6.456673415653292e-07,
"loss": 1.0078,
"step": 8970
},
{
"epoch": 1.67,
"grad_norm": 2.479290657154231,
"learning_rate": 6.420722511186134e-07,
"loss": 0.9972,
"step": 8975
},
{
"epoch": 1.67,
"grad_norm": 2.4648199168537412,
"learning_rate": 6.384865105530708e-07,
"loss": 1.0074,
"step": 8980
},
{
"epoch": 1.68,
"grad_norm": 2.2138902601386574,
"learning_rate": 6.349101275618641e-07,
"loss": 0.9987,
"step": 8985
},
{
"epoch": 1.68,
"grad_norm": 2.2720662118151544,
"learning_rate": 6.3134310981808e-07,
"loss": 1.0317,
"step": 8990
},
{
"epoch": 1.68,
"grad_norm": 2.3287005891438817,
"learning_rate": 6.277854649747111e-07,
"loss": 1.0005,
"step": 8995
},
{
"epoch": 1.68,
"grad_norm": 2.4627011404225,
"learning_rate": 6.242372006646419e-07,
"loss": 1.0005,
"step": 9000
},
{
"epoch": 1.68,
"grad_norm": 2.511510221864267,
"learning_rate": 6.206983245006304e-07,
"loss": 1.0113,
"step": 9005
},
{
"epoch": 1.68,
"grad_norm": 2.5333977257602363,
"learning_rate": 6.171688440752915e-07,
"loss": 0.9587,
"step": 9010
},
{
"epoch": 1.68,
"grad_norm": 2.2279459443378635,
"learning_rate": 6.136487669610841e-07,
"loss": 1.0043,
"step": 9015
},
{
"epoch": 1.68,
"grad_norm": 3.028251278959122,
"learning_rate": 6.101381007102914e-07,
"loss": 0.9891,
"step": 9020
},
{
"epoch": 1.68,
"grad_norm": 2.7570318740615942,
"learning_rate": 6.066368528550037e-07,
"loss": 1.0377,
"step": 9025
},
{
"epoch": 1.68,
"grad_norm": 2.6810202304359287,
"learning_rate": 6.031450309071068e-07,
"loss": 1.0084,
"step": 9030
},
{
"epoch": 1.68,
"grad_norm": 2.4733864750094394,
"learning_rate": 5.996626423582624e-07,
"loss": 1.0223,
"step": 9035
},
{
"epoch": 1.69,
"grad_norm": 2.399472156625489,
"learning_rate": 5.961896946798923e-07,
"loss": 1.0306,
"step": 9040
},
{
"epoch": 1.69,
"grad_norm": 3.010499926572972,
"learning_rate": 5.927261953231639e-07,
"loss": 0.9157,
"step": 9045
},
{
"epoch": 1.69,
"grad_norm": 3.138764769840231,
"learning_rate": 5.892721517189726e-07,
"loss": 0.9091,
"step": 9050
},
{
"epoch": 1.69,
"grad_norm": 3.052701911605609,
"learning_rate": 5.858275712779288e-07,
"loss": 0.9896,
"step": 9055
},
{
"epoch": 1.69,
"grad_norm": 2.6461404601065834,
"learning_rate": 5.823924613903365e-07,
"loss": 0.9955,
"step": 9060
},
{
"epoch": 1.69,
"grad_norm": 2.6462011414933206,
"learning_rate": 5.789668294261841e-07,
"loss": 0.9824,
"step": 9065
},
{
"epoch": 1.69,
"grad_norm": 2.4145981121391333,
"learning_rate": 5.755506827351214e-07,
"loss": 1.024,
"step": 9070
},
{
"epoch": 1.69,
"grad_norm": 2.726076929054993,
"learning_rate": 5.721440286464491e-07,
"loss": 1.0127,
"step": 9075
},
{
"epoch": 1.69,
"grad_norm": 2.832791332944865,
"learning_rate": 5.687468744691049e-07,
"loss": 1.0005,
"step": 9080
},
{
"epoch": 1.69,
"grad_norm": 2.873484163008957,
"learning_rate": 5.653592274916408e-07,
"loss": 1.0268,
"step": 9085
},
{
"epoch": 1.7,
"grad_norm": 2.4948123943547125,
"learning_rate": 5.619810949822124e-07,
"loss": 1.017,
"step": 9090
},
{
"epoch": 1.7,
"grad_norm": 2.500406538069392,
"learning_rate": 5.586124841885604e-07,
"loss": 1.0411,
"step": 9095
},
{
"epoch": 1.7,
"grad_norm": 2.850558812351205,
"learning_rate": 5.552534023380024e-07,
"loss": 0.9598,
"step": 9100
},
{
"epoch": 1.7,
"grad_norm": 2.9195318309343645,
"learning_rate": 5.519038566374041e-07,
"loss": 0.986,
"step": 9105
},
{
"epoch": 1.7,
"grad_norm": 2.8803650261654665,
"learning_rate": 5.485638542731764e-07,
"loss": 1.0123,
"step": 9110
},
{
"epoch": 1.7,
"grad_norm": 2.5711087382307873,
"learning_rate": 5.452334024112533e-07,
"loss": 0.9667,
"step": 9115
},
{
"epoch": 1.7,
"grad_norm": 2.600124755629203,
"learning_rate": 5.419125081970805e-07,
"loss": 1.0223,
"step": 9120
},
{
"epoch": 1.7,
"grad_norm": 2.692133520812537,
"learning_rate": 5.386011787555951e-07,
"loss": 1.0128,
"step": 9125
},
{
"epoch": 1.7,
"grad_norm": 2.841992509801276,
"learning_rate": 5.352994211912144e-07,
"loss": 0.977,
"step": 9130
},
{
"epoch": 1.7,
"grad_norm": 2.4308516211470703,
"learning_rate": 5.320072425878198e-07,
"loss": 0.9886,
"step": 9135
},
{
"epoch": 1.7,
"grad_norm": 2.528221958734492,
"learning_rate": 5.287246500087401e-07,
"loss": 0.9715,
"step": 9140
},
{
"epoch": 1.71,
"grad_norm": 2.6059596924040513,
"learning_rate": 5.254516504967377e-07,
"loss": 1.0196,
"step": 9145
},
{
"epoch": 1.71,
"grad_norm": 3.0973405688579327,
"learning_rate": 5.221882510739939e-07,
"loss": 0.9772,
"step": 9150
},
{
"epoch": 1.71,
"grad_norm": 2.334645224986343,
"learning_rate": 5.189344587420924e-07,
"loss": 1.0351,
"step": 9155
},
{
"epoch": 1.71,
"grad_norm": 2.7613541361820593,
"learning_rate": 5.156902804820041e-07,
"loss": 0.9751,
"step": 9160
},
{
"epoch": 1.71,
"grad_norm": 2.536194202059706,
"learning_rate": 5.124557232540761e-07,
"loss": 1.0051,
"step": 9165
},
{
"epoch": 1.71,
"grad_norm": 2.583988794361422,
"learning_rate": 5.092307939980112e-07,
"loss": 0.9559,
"step": 9170
},
{
"epoch": 1.71,
"grad_norm": 2.6904789600878045,
"learning_rate": 5.060154996328554e-07,
"loss": 0.9627,
"step": 9175
},
{
"epoch": 1.71,
"grad_norm": 2.9369318056690186,
"learning_rate": 5.028098470569843e-07,
"loss": 1.0039,
"step": 9180
},
{
"epoch": 1.71,
"grad_norm": 2.842613702289357,
"learning_rate": 4.996138431480862e-07,
"loss": 1.0297,
"step": 9185
},
{
"epoch": 1.71,
"grad_norm": 2.538587614782097,
"learning_rate": 4.964274947631492e-07,
"loss": 1.049,
"step": 9190
},
{
"epoch": 1.71,
"grad_norm": 2.3897911615036356,
"learning_rate": 4.932508087384452e-07,
"loss": 1.0171,
"step": 9195
},
{
"epoch": 1.72,
"grad_norm": 2.608616242680837,
"learning_rate": 4.900837918895152e-07,
"loss": 1.0567,
"step": 9200
},
{
"epoch": 1.72,
"grad_norm": 2.7563694990221403,
"learning_rate": 4.869264510111543e-07,
"loss": 1.0163,
"step": 9205
},
{
"epoch": 1.72,
"grad_norm": 2.8931596725585047,
"learning_rate": 4.837787928774013e-07,
"loss": 1.0264,
"step": 9210
},
{
"epoch": 1.72,
"grad_norm": 2.56930829905871,
"learning_rate": 4.806408242415184e-07,
"loss": 1.0019,
"step": 9215
},
{
"epoch": 1.72,
"grad_norm": 2.5389017003142,
"learning_rate": 4.77512551835978e-07,
"loss": 0.9896,
"step": 9220
},
{
"epoch": 1.72,
"grad_norm": 2.4572664571372367,
"learning_rate": 4.743939823724508e-07,
"loss": 0.9452,
"step": 9225
},
{
"epoch": 1.72,
"grad_norm": 2.597573031717897,
"learning_rate": 4.7128512254179083e-07,
"loss": 0.9848,
"step": 9230
},
{
"epoch": 1.72,
"grad_norm": 2.9459449902717694,
"learning_rate": 4.6818597901401864e-07,
"loss": 0.9703,
"step": 9235
},
{
"epoch": 1.72,
"grad_norm": 2.672050813178242,
"learning_rate": 4.6509655843830827e-07,
"loss": 0.9637,
"step": 9240
},
{
"epoch": 1.72,
"grad_norm": 2.5874000556032435,
"learning_rate": 4.620168674429754e-07,
"loss": 1.0507,
"step": 9245
},
{
"epoch": 1.73,
"grad_norm": 2.65520217601556,
"learning_rate": 4.5894691263545845e-07,
"loss": 1.0023,
"step": 9250
},
{
"epoch": 1.73,
"grad_norm": 2.969014031244479,
"learning_rate": 4.5588670060230933e-07,
"loss": 1.0131,
"step": 9255
},
{
"epoch": 1.73,
"grad_norm": 2.7090713448074206,
"learning_rate": 4.528362379091766e-07,
"loss": 0.9715,
"step": 9260
},
{
"epoch": 1.73,
"grad_norm": 2.6279641402257043,
"learning_rate": 4.4979553110078957e-07,
"loss": 0.9686,
"step": 9265
},
{
"epoch": 1.73,
"grad_norm": 2.5834971940560294,
"learning_rate": 4.467645867009468e-07,
"loss": 0.9521,
"step": 9270
},
{
"epoch": 1.73,
"grad_norm": 3.0675999534568468,
"learning_rate": 4.4374341121250496e-07,
"loss": 0.9839,
"step": 9275
},
{
"epoch": 1.73,
"grad_norm": 3.2037422202426504,
"learning_rate": 4.407320111173591e-07,
"loss": 0.967,
"step": 9280
},
{
"epoch": 1.73,
"grad_norm": 3.020695568265947,
"learning_rate": 4.3773039287643095e-07,
"loss": 0.9436,
"step": 9285
},
{
"epoch": 1.73,
"grad_norm": 2.7269030107954655,
"learning_rate": 4.347385629296552e-07,
"loss": 1.0461,
"step": 9290
},
{
"epoch": 1.73,
"grad_norm": 2.5327864316035207,
"learning_rate": 4.3175652769597007e-07,
"loss": 0.9584,
"step": 9295
},
{
"epoch": 1.73,
"grad_norm": 2.519235011867918,
"learning_rate": 4.2878429357329277e-07,
"loss": 0.9824,
"step": 9300
},
{
"epoch": 1.74,
"grad_norm": 2.310799267177094,
"learning_rate": 4.2582186693851666e-07,
"loss": 1.0077,
"step": 9305
},
{
"epoch": 1.74,
"grad_norm": 2.4863765476136823,
"learning_rate": 4.228692541474921e-07,
"loss": 1.0093,
"step": 9310
},
{
"epoch": 1.74,
"grad_norm": 2.6063805598016225,
"learning_rate": 4.199264615350135e-07,
"loss": 1.0481,
"step": 9315
},
{
"epoch": 1.74,
"grad_norm": 2.985739971909663,
"learning_rate": 4.16993495414808e-07,
"loss": 0.9816,
"step": 9320
},
{
"epoch": 1.74,
"grad_norm": 2.466317302363233,
"learning_rate": 4.1407036207951833e-07,
"loss": 0.9676,
"step": 9325
},
{
"epoch": 1.74,
"grad_norm": 2.7455631489194876,
"learning_rate": 4.1115706780069155e-07,
"loss": 0.9819,
"step": 9330
},
{
"epoch": 1.74,
"grad_norm": 2.908316238615218,
"learning_rate": 4.08253618828765e-07,
"loss": 1.0758,
"step": 9335
},
{
"epoch": 1.74,
"grad_norm": 2.8169316671572573,
"learning_rate": 4.053600213930536e-07,
"loss": 0.9638,
"step": 9340
},
{
"epoch": 1.74,
"grad_norm": 2.381506348033514,
"learning_rate": 4.0247628170173524e-07,
"loss": 1.0118,
"step": 9345
},
{
"epoch": 1.74,
"grad_norm": 2.755468426326752,
"learning_rate": 3.9960240594183906e-07,
"loss": 0.9953,
"step": 9350
},
{
"epoch": 1.74,
"grad_norm": 2.605028560077565,
"learning_rate": 3.9673840027923014e-07,
"loss": 1.0409,
"step": 9355
},
{
"epoch": 1.75,
"grad_norm": 3.3235218437167777,
"learning_rate": 3.938842708585977e-07,
"loss": 0.969,
"step": 9360
},
{
"epoch": 1.75,
"grad_norm": 2.5416660450574007,
"learning_rate": 3.910400238034423e-07,
"loss": 1.0235,
"step": 9365
},
{
"epoch": 1.75,
"grad_norm": 2.990917512675436,
"learning_rate": 3.8820566521606207e-07,
"loss": 0.9659,
"step": 9370
},
{
"epoch": 1.75,
"grad_norm": 2.605011528586985,
"learning_rate": 3.853812011775382e-07,
"loss": 1.0126,
"step": 9375
},
{
"epoch": 1.75,
"grad_norm": 2.7008321338964403,
"learning_rate": 3.8256663774772383e-07,
"loss": 0.9629,
"step": 9380
},
{
"epoch": 1.75,
"grad_norm": 2.528423641712258,
"learning_rate": 3.797619809652314e-07,
"loss": 0.9808,
"step": 9385
},
{
"epoch": 1.75,
"grad_norm": 2.31165739813913,
"learning_rate": 3.769672368474181e-07,
"loss": 0.9899,
"step": 9390
},
{
"epoch": 1.75,
"grad_norm": 2.7287545120118013,
"learning_rate": 3.741824113903725e-07,
"loss": 0.9959,
"step": 9395
},
{
"epoch": 1.75,
"grad_norm": 2.5020104494413253,
"learning_rate": 3.714075105689041e-07,
"loss": 1.0553,
"step": 9400
},
{
"epoch": 1.75,
"grad_norm": 2.604353933763939,
"learning_rate": 3.686425403365307e-07,
"loss": 0.9977,
"step": 9405
},
{
"epoch": 1.75,
"grad_norm": 3.1903595327659082,
"learning_rate": 3.658875066254608e-07,
"loss": 1.0267,
"step": 9410
},
{
"epoch": 1.76,
"grad_norm": 2.5140954157235638,
"learning_rate": 3.631424153465862e-07,
"loss": 1.0241,
"step": 9415
},
{
"epoch": 1.76,
"grad_norm": 3.1645831576006813,
"learning_rate": 3.6040727238946807e-07,
"loss": 1.0294,
"step": 9420
},
{
"epoch": 1.76,
"grad_norm": 2.1964802552754197,
"learning_rate": 3.576820836223194e-07,
"loss": 0.9772,
"step": 9425
},
{
"epoch": 1.76,
"grad_norm": 2.668666804034786,
"learning_rate": 3.5496685489200265e-07,
"loss": 1.0015,
"step": 9430
},
{
"epoch": 1.76,
"grad_norm": 2.733206900757186,
"learning_rate": 3.5226159202400747e-07,
"loss": 0.9567,
"step": 9435
},
{
"epoch": 1.76,
"grad_norm": 3.2939125262903644,
"learning_rate": 3.4956630082244204e-07,
"loss": 0.9883,
"step": 9440
},
{
"epoch": 1.76,
"grad_norm": 2.8479567257890976,
"learning_rate": 3.468809870700207e-07,
"loss": 0.9406,
"step": 9445
},
{
"epoch": 1.76,
"grad_norm": 2.731564303566274,
"learning_rate": 3.4420565652805215e-07,
"loss": 0.986,
"step": 9450
},
{
"epoch": 1.76,
"grad_norm": 2.8052153951482466,
"learning_rate": 3.4154031493642716e-07,
"loss": 0.9648,
"step": 9455
},
{
"epoch": 1.76,
"grad_norm": 2.3411545722993226,
"learning_rate": 3.388849680136008e-07,
"loss": 1.0088,
"step": 9460
},
{
"epoch": 1.77,
"grad_norm": 2.5499289340441837,
"learning_rate": 3.362396214565888e-07,
"loss": 1.0102,
"step": 9465
},
{
"epoch": 1.77,
"grad_norm": 2.30165208282686,
"learning_rate": 3.336042809409512e-07,
"loss": 0.9971,
"step": 9470
},
{
"epoch": 1.77,
"grad_norm": 2.6602655807503988,
"learning_rate": 3.3097895212077845e-07,
"loss": 0.996,
"step": 9475
},
{
"epoch": 1.77,
"grad_norm": 2.578917739046354,
"learning_rate": 3.2836364062868165e-07,
"loss": 0.9575,
"step": 9480
},
{
"epoch": 1.77,
"grad_norm": 2.7011601669086116,
"learning_rate": 3.2575835207577957e-07,
"loss": 1.0345,
"step": 9485
},
{
"epoch": 1.77,
"grad_norm": 3.187787275788192,
"learning_rate": 3.2316309205168705e-07,
"loss": 0.9859,
"step": 9490
},
{
"epoch": 1.77,
"grad_norm": 2.6122898621423216,
"learning_rate": 3.205778661245051e-07,
"loss": 1.0141,
"step": 9495
},
{
"epoch": 1.77,
"grad_norm": 2.443420355234737,
"learning_rate": 3.180026798408026e-07,
"loss": 1.0211,
"step": 9500
},
{
"epoch": 1.77,
"grad_norm": 2.540209666631077,
"learning_rate": 3.1543753872561087e-07,
"loss": 1.0208,
"step": 9505
},
{
"epoch": 1.77,
"grad_norm": 2.6964911580456654,
"learning_rate": 3.128824482824083e-07,
"loss": 0.9891,
"step": 9510
},
{
"epoch": 1.77,
"grad_norm": 2.4469653887447746,
"learning_rate": 3.1033741399311204e-07,
"loss": 1.0305,
"step": 9515
},
{
"epoch": 1.78,
"grad_norm": 2.291299962290693,
"learning_rate": 3.0780244131806193e-07,
"loss": 1.0168,
"step": 9520
},
{
"epoch": 1.78,
"grad_norm": 2.522210460018589,
"learning_rate": 3.0527753569601083e-07,
"loss": 0.961,
"step": 9525
},
{
"epoch": 1.78,
"grad_norm": 2.5659714010554526,
"learning_rate": 3.027627025441132e-07,
"loss": 0.9547,
"step": 9530
},
{
"epoch": 1.78,
"grad_norm": 2.7052846580426855,
"learning_rate": 3.0025794725791315e-07,
"loss": 0.9943,
"step": 9535
},
{
"epoch": 1.78,
"grad_norm": 2.484865366051049,
"learning_rate": 2.977632752113324e-07,
"loss": 0.9765,
"step": 9540
},
{
"epoch": 1.78,
"grad_norm": 2.352793754428888,
"learning_rate": 2.9527869175666025e-07,
"loss": 0.9883,
"step": 9545
},
{
"epoch": 1.78,
"grad_norm": 2.830152940630546,
"learning_rate": 2.928042022245398e-07,
"loss": 0.9651,
"step": 9550
},
{
"epoch": 1.78,
"grad_norm": 2.5427539631035554,
"learning_rate": 2.903398119239581e-07,
"loss": 0.9826,
"step": 9555
},
{
"epoch": 1.78,
"grad_norm": 2.9400021212833787,
"learning_rate": 2.878855261422359e-07,
"loss": 1.0063,
"step": 9560
},
{
"epoch": 1.78,
"grad_norm": 2.3427477540958725,
"learning_rate": 2.8544135014501264e-07,
"loss": 1.0179,
"step": 9565
},
{
"epoch": 1.78,
"grad_norm": 2.7757453717128824,
"learning_rate": 2.8300728917623866e-07,
"loss": 0.9788,
"step": 9570
},
{
"epoch": 1.79,
"grad_norm": 2.4936847575193193,
"learning_rate": 2.8058334845816214e-07,
"loss": 1.0008,
"step": 9575
},
{
"epoch": 1.79,
"grad_norm": 2.579956273305064,
"learning_rate": 2.7816953319131865e-07,
"loss": 1.0594,
"step": 9580
},
{
"epoch": 1.79,
"grad_norm": 2.6638295637588993,
"learning_rate": 2.757658485545195e-07,
"loss": 0.9649,
"step": 9585
},
{
"epoch": 1.79,
"grad_norm": 2.9718905778082707,
"learning_rate": 2.7337229970484104e-07,
"loss": 1.0001,
"step": 9590
},
{
"epoch": 1.79,
"grad_norm": 2.470369084422218,
"learning_rate": 2.709888917776132e-07,
"loss": 1.0225,
"step": 9595
},
{
"epoch": 1.79,
"grad_norm": 2.9125743810651463,
"learning_rate": 2.6861562988640786e-07,
"loss": 1.0356,
"step": 9600
},
{
"epoch": 1.79,
"grad_norm": 2.282995672450367,
"learning_rate": 2.662525191230314e-07,
"loss": 0.9894,
"step": 9605
},
{
"epoch": 1.79,
"grad_norm": 2.45891203140545,
"learning_rate": 2.638995645575082e-07,
"loss": 1.0093,
"step": 9610
},
{
"epoch": 1.79,
"grad_norm": 3.1540839609729803,
"learning_rate": 2.615567712380751e-07,
"loss": 1.0343,
"step": 9615
},
{
"epoch": 1.79,
"grad_norm": 2.4897966159719283,
"learning_rate": 2.592241441911641e-07,
"loss": 1.0053,
"step": 9620
},
{
"epoch": 1.79,
"grad_norm": 2.6750189260740074,
"learning_rate": 2.5690168842140097e-07,
"loss": 0.9827,
"step": 9625
},
{
"epoch": 1.8,
"grad_norm": 2.4749333331718457,
"learning_rate": 2.5458940891158525e-07,
"loss": 1.028,
"step": 9630
},
{
"epoch": 1.8,
"grad_norm": 2.690092812167263,
"learning_rate": 2.5228731062268476e-07,
"loss": 0.9687,
"step": 9635
},
{
"epoch": 1.8,
"grad_norm": 2.521697310830935,
"learning_rate": 2.4999539849382304e-07,
"loss": 0.9968,
"step": 9640
},
{
"epoch": 1.8,
"grad_norm": 2.598292223752983,
"learning_rate": 2.477136774422706e-07,
"loss": 1.0522,
"step": 9645
},
{
"epoch": 1.8,
"grad_norm": 2.449309049508482,
"learning_rate": 2.4544215236343206e-07,
"loss": 1.0047,
"step": 9650
},
{
"epoch": 1.8,
"grad_norm": 2.6542058491455385,
"learning_rate": 2.4318082813083734e-07,
"loss": 1.006,
"step": 9655
},
{
"epoch": 1.8,
"grad_norm": 2.5090872392658103,
"learning_rate": 2.4092970959612885e-07,
"loss": 0.9827,
"step": 9660
},
{
"epoch": 1.8,
"grad_norm": 2.5844303394951806,
"learning_rate": 2.3868880158905426e-07,
"loss": 1.0087,
"step": 9665
},
{
"epoch": 1.8,
"grad_norm": 2.3442031349160604,
"learning_rate": 2.3645810891745556e-07,
"loss": 0.997,
"step": 9670
},
{
"epoch": 1.8,
"grad_norm": 2.654874365932498,
"learning_rate": 2.3423763636725717e-07,
"loss": 0.9667,
"step": 9675
},
{
"epoch": 1.81,
"grad_norm": 2.5095988006665046,
"learning_rate": 2.320273887024549e-07,
"loss": 0.9777,
"step": 9680
},
{
"epoch": 1.81,
"grad_norm": 2.715845469741196,
"learning_rate": 2.298273706651083e-07,
"loss": 0.9743,
"step": 9685
},
{
"epoch": 1.81,
"grad_norm": 2.675360652874929,
"learning_rate": 2.2763758697533112e-07,
"loss": 0.9735,
"step": 9690
},
{
"epoch": 1.81,
"grad_norm": 2.9739364378003312,
"learning_rate": 2.2545804233127578e-07,
"loss": 1.0247,
"step": 9695
},
{
"epoch": 1.81,
"grad_norm": 3.1442927146127273,
"learning_rate": 2.2328874140912948e-07,
"loss": 1.0427,
"step": 9700
},
{
"epoch": 1.81,
"grad_norm": 2.586761975300717,
"learning_rate": 2.2112968886310037e-07,
"loss": 1.0241,
"step": 9705
},
{
"epoch": 1.81,
"grad_norm": 2.521670484167634,
"learning_rate": 2.1898088932541139e-07,
"loss": 1.0188,
"step": 9710
},
{
"epoch": 1.81,
"grad_norm": 2.803812342944837,
"learning_rate": 2.1684234740628474e-07,
"loss": 1.03,
"step": 9715
},
{
"epoch": 1.81,
"grad_norm": 2.748460050126293,
"learning_rate": 2.1471406769393633e-07,
"loss": 0.9911,
"step": 9720
},
{
"epoch": 1.81,
"grad_norm": 3.1096781734009644,
"learning_rate": 2.1259605475456468e-07,
"loss": 1.0387,
"step": 9725
},
{
"epoch": 1.81,
"grad_norm": 2.730958677130239,
"learning_rate": 2.1048831313234042e-07,
"loss": 0.9288,
"step": 9730
},
{
"epoch": 1.82,
"grad_norm": 2.5620713671460504,
"learning_rate": 2.0839084734939784e-07,
"loss": 0.9943,
"step": 9735
},
{
"epoch": 1.82,
"grad_norm": 3.296656696817953,
"learning_rate": 2.063036619058245e-07,
"loss": 1.0116,
"step": 9740
},
{
"epoch": 1.82,
"grad_norm": 2.460667225354087,
"learning_rate": 2.0422676127965169e-07,
"loss": 1.0158,
"step": 9745
},
{
"epoch": 1.82,
"grad_norm": 2.4612694765541314,
"learning_rate": 2.021601499268433e-07,
"loss": 1.0556,
"step": 9750
},
{
"epoch": 1.82,
"grad_norm": 2.3591424193825685,
"learning_rate": 2.0010383228129038e-07,
"loss": 1.0146,
"step": 9755
},
{
"epoch": 1.82,
"grad_norm": 2.8113416401044202,
"learning_rate": 1.980578127547972e-07,
"loss": 0.9576,
"step": 9760
},
{
"epoch": 1.82,
"grad_norm": 2.706080390163917,
"learning_rate": 1.9602209573707397e-07,
"loss": 0.9713,
"step": 9765
},
{
"epoch": 1.82,
"grad_norm": 2.845030926145939,
"learning_rate": 1.9399668559572593e-07,
"loss": 1.0307,
"step": 9770
},
{
"epoch": 1.82,
"grad_norm": 2.674120579688675,
"learning_rate": 1.9198158667624756e-07,
"loss": 1.022,
"step": 9775
},
{
"epoch": 1.82,
"grad_norm": 2.1385275862773003,
"learning_rate": 1.8997680330200775e-07,
"loss": 0.978,
"step": 9780
},
{
"epoch": 1.82,
"grad_norm": 2.843210914091474,
"learning_rate": 1.879823397742453e-07,
"loss": 1.0131,
"step": 9785
},
{
"epoch": 1.83,
"grad_norm": 3.0167556280462704,
"learning_rate": 1.8599820037205785e-07,
"loss": 0.9705,
"step": 9790
},
{
"epoch": 1.83,
"grad_norm": 2.7679496204235217,
"learning_rate": 1.8402438935239187e-07,
"loss": 0.9938,
"step": 9795
},
{
"epoch": 1.83,
"grad_norm": 2.8044277031578955,
"learning_rate": 1.8206091095003543e-07,
"loss": 1.034,
"step": 9800
},
{
"epoch": 1.83,
"grad_norm": 2.6766193581300706,
"learning_rate": 1.801077693776082e-07,
"loss": 0.9989,
"step": 9805
},
{
"epoch": 1.83,
"grad_norm": 2.4536021788135027,
"learning_rate": 1.78164968825551e-07,
"loss": 0.9419,
"step": 9810
},
{
"epoch": 1.83,
"grad_norm": 2.7894618880044963,
"learning_rate": 1.7623251346211902e-07,
"loss": 0.9509,
"step": 9815
},
{
"epoch": 1.83,
"grad_norm": 2.682428541717693,
"learning_rate": 1.743104074333718e-07,
"loss": 1.0204,
"step": 9820
},
{
"epoch": 1.83,
"grad_norm": 2.8164877468179195,
"learning_rate": 1.7239865486316508e-07,
"loss": 0.9888,
"step": 9825
},
{
"epoch": 1.83,
"grad_norm": 2.7847626207224816,
"learning_rate": 1.7049725985313959e-07,
"loss": 0.9771,
"step": 9830
},
{
"epoch": 1.83,
"grad_norm": 3.3563234753041744,
"learning_rate": 1.6860622648271653e-07,
"loss": 1.0058,
"step": 9835
},
{
"epoch": 1.84,
"grad_norm": 3.3102502358004706,
"learning_rate": 1.667255588090838e-07,
"loss": 1.0054,
"step": 9840
},
{
"epoch": 1.84,
"grad_norm": 2.7169721452197635,
"learning_rate": 1.6485526086719217e-07,
"loss": 0.9396,
"step": 9845
},
{
"epoch": 1.84,
"grad_norm": 2.6443166952288104,
"learning_rate": 1.62995336669744e-07,
"loss": 1.0504,
"step": 9850
},
{
"epoch": 1.84,
"grad_norm": 2.6784711600031637,
"learning_rate": 1.6114579020718223e-07,
"loss": 0.9967,
"step": 9855
},
{
"epoch": 1.84,
"grad_norm": 2.7021265214230525,
"learning_rate": 1.593066254476866e-07,
"loss": 1.0112,
"step": 9860
},
{
"epoch": 1.84,
"grad_norm": 2.4415845222324783,
"learning_rate": 1.57477846337164e-07,
"loss": 0.9537,
"step": 9865
},
{
"epoch": 1.84,
"grad_norm": 2.71926218122397,
"learning_rate": 1.5565945679923699e-07,
"loss": 1.0178,
"step": 9870
},
{
"epoch": 1.84,
"grad_norm": 2.704907322234658,
"learning_rate": 1.5385146073523926e-07,
"loss": 1.0128,
"step": 9875
},
{
"epoch": 1.84,
"grad_norm": 2.4769145265711363,
"learning_rate": 1.5205386202420235e-07,
"loss": 1.0083,
"step": 9880
},
{
"epoch": 1.84,
"grad_norm": 2.6895279389899067,
"learning_rate": 1.5026666452285565e-07,
"loss": 1.0007,
"step": 9885
},
{
"epoch": 1.84,
"grad_norm": 2.627092854331598,
"learning_rate": 1.4848987206560695e-07,
"loss": 1.0009,
"step": 9890
},
{
"epoch": 1.85,
"grad_norm": 2.60277939316083,
"learning_rate": 1.4672348846454408e-07,
"loss": 1.0241,
"step": 9895
},
{
"epoch": 1.85,
"grad_norm": 2.7442165962309515,
"learning_rate": 1.4496751750942173e-07,
"loss": 1.0583,
"step": 9900
},
{
"epoch": 1.85,
"grad_norm": 2.7088131815525895,
"learning_rate": 1.4322196296765346e-07,
"loss": 1.0226,
"step": 9905
},
{
"epoch": 1.85,
"grad_norm": 2.4976027705254693,
"learning_rate": 1.414868285843063e-07,
"loss": 1.0096,
"step": 9910
},
{
"epoch": 1.85,
"grad_norm": 3.6650855699844134,
"learning_rate": 1.3976211808209018e-07,
"loss": 0.9702,
"step": 9915
},
{
"epoch": 1.85,
"grad_norm": 2.8355048109869494,
"learning_rate": 1.3804783516135012e-07,
"loss": 1.0196,
"step": 9920
},
{
"epoch": 1.85,
"grad_norm": 2.805735805208835,
"learning_rate": 1.3634398350005907e-07,
"loss": 1.0183,
"step": 9925
},
{
"epoch": 1.85,
"grad_norm": 2.2774993379472583,
"learning_rate": 1.3465056675381117e-07,
"loss": 0.9935,
"step": 9930
},
{
"epoch": 1.85,
"grad_norm": 2.3630526568932333,
"learning_rate": 1.3296758855581015e-07,
"loss": 1.0068,
"step": 9935
},
{
"epoch": 1.85,
"grad_norm": 2.8721281507955574,
"learning_rate": 1.3129505251686603e-07,
"loss": 1.0101,
"step": 9940
},
{
"epoch": 1.85,
"grad_norm": 2.6647193040662813,
"learning_rate": 1.2963296222538446e-07,
"loss": 1.0117,
"step": 9945
},
{
"epoch": 1.86,
"grad_norm": 2.3667246578642827,
"learning_rate": 1.2798132124736074e-07,
"loss": 0.991,
"step": 9950
},
{
"epoch": 1.86,
"grad_norm": 2.8919200983344076,
"learning_rate": 1.2634013312636973e-07,
"loss": 0.9733,
"step": 9955
},
{
"epoch": 1.86,
"grad_norm": 2.36008020599268,
"learning_rate": 1.247094013835609e-07,
"loss": 0.994,
"step": 9960
},
{
"epoch": 1.86,
"grad_norm": 2.467275486733601,
"learning_rate": 1.230891295176495e-07,
"loss": 0.9721,
"step": 9965
},
{
"epoch": 1.86,
"grad_norm": 2.684510172486263,
"learning_rate": 1.2147932100490866e-07,
"loss": 1.0237,
"step": 9970
},
{
"epoch": 1.86,
"grad_norm": 2.970515110780605,
"learning_rate": 1.1987997929916395e-07,
"loss": 0.9639,
"step": 9975
},
{
"epoch": 1.86,
"grad_norm": 2.517212726647846,
"learning_rate": 1.1829110783178332e-07,
"loss": 0.9784,
"step": 9980
},
{
"epoch": 1.86,
"grad_norm": 2.6269061771187623,
"learning_rate": 1.1671271001167106e-07,
"loss": 0.993,
"step": 9985
},
{
"epoch": 1.86,
"grad_norm": 2.4276654463627274,
"learning_rate": 1.1514478922525996e-07,
"loss": 1.0237,
"step": 9990
},
{
"epoch": 1.86,
"grad_norm": 2.628761439871769,
"learning_rate": 1.1358734883650635e-07,
"loss": 0.9542,
"step": 9995
},
{
"epoch": 1.86,
"grad_norm": 2.8787281265293165,
"learning_rate": 1.1204039218687901e-07,
"loss": 1.0165,
"step": 10000
},
{
"epoch": 1.87,
"grad_norm": 2.9024745064799125,
"learning_rate": 1.105039225953547e-07,
"loss": 0.9447,
"step": 10005
},
{
"epoch": 1.87,
"grad_norm": 2.781988132946109,
"learning_rate": 1.0897794335841094e-07,
"loss": 0.9774,
"step": 10010
},
{
"epoch": 1.87,
"grad_norm": 2.732110290110213,
"learning_rate": 1.0746245775001607e-07,
"loss": 0.9785,
"step": 10015
},
{
"epoch": 1.87,
"grad_norm": 2.475660794796942,
"learning_rate": 1.0595746902162751e-07,
"loss": 1.0015,
"step": 10020
},
{
"epoch": 1.87,
"grad_norm": 2.549244214480255,
"learning_rate": 1.0446298040218017e-07,
"loss": 0.9978,
"step": 10025
},
{
"epoch": 1.87,
"grad_norm": 2.7669451232914866,
"learning_rate": 1.0297899509808085e-07,
"loss": 0.9641,
"step": 10030
},
{
"epoch": 1.87,
"grad_norm": 2.7884785856697345,
"learning_rate": 1.0150551629320215e-07,
"loss": 1.0045,
"step": 10035
},
{
"epoch": 1.87,
"grad_norm": 2.822600979742313,
"learning_rate": 1.000425471488753e-07,
"loss": 0.9792,
"step": 10040
},
{
"epoch": 1.87,
"grad_norm": 2.34346287354799,
"learning_rate": 9.859009080388338e-08,
"loss": 1.0256,
"step": 10045
},
{
"epoch": 1.87,
"grad_norm": 3.0091617199806744,
"learning_rate": 9.714815037445202e-08,
"loss": 0.9503,
"step": 10050
},
{
"epoch": 1.88,
"grad_norm": 2.454168024542919,
"learning_rate": 9.571672895424822e-08,
"loss": 0.9699,
"step": 10055
},
{
"epoch": 1.88,
"grad_norm": 2.577866899775397,
"learning_rate": 9.429582961436978e-08,
"loss": 1.0006,
"step": 10060
},
{
"epoch": 1.88,
"grad_norm": 2.4283231702938752,
"learning_rate": 9.28854554033376e-08,
"loss": 1.0203,
"step": 10065
},
{
"epoch": 1.88,
"grad_norm": 2.4012436890293922,
"learning_rate": 9.148560934709395e-08,
"loss": 0.9573,
"step": 10070
},
{
"epoch": 1.88,
"grad_norm": 2.6430208059626805,
"learning_rate": 9.009629444899081e-08,
"loss": 1.0115,
"step": 10075
},
{
"epoch": 1.88,
"grad_norm": 2.3785722525752853,
"learning_rate": 8.871751368978554e-08,
"loss": 1.0351,
"step": 10080
},
{
"epoch": 1.88,
"grad_norm": 2.5618218348538813,
"learning_rate": 8.734927002763793e-08,
"loss": 1.0202,
"step": 10085
},
{
"epoch": 1.88,
"grad_norm": 2.3559694341229704,
"learning_rate": 8.599156639809591e-08,
"loss": 0.9665,
"step": 10090
},
{
"epoch": 1.88,
"grad_norm": 2.498377208318561,
"learning_rate": 8.46444057140977e-08,
"loss": 1.0156,
"step": 10095
},
{
"epoch": 1.88,
"grad_norm": 2.8502573752110387,
"learning_rate": 8.330779086595852e-08,
"loss": 0.9877,
"step": 10100
},
{
"epoch": 1.88,
"grad_norm": 2.730502176083712,
"learning_rate": 8.198172472137111e-08,
"loss": 0.9944,
"step": 10105
},
{
"epoch": 1.89,
"grad_norm": 2.5665957223980884,
"learning_rate": 8.0666210125393e-08,
"loss": 0.9516,
"step": 10110
},
{
"epoch": 1.89,
"grad_norm": 2.599168705119983,
"learning_rate": 7.936124990044591e-08,
"loss": 1.0021,
"step": 10115
},
{
"epoch": 1.89,
"grad_norm": 2.509541356997699,
"learning_rate": 7.806684684630528e-08,
"loss": 0.9792,
"step": 10120
},
{
"epoch": 1.89,
"grad_norm": 2.5079499486537,
"learning_rate": 7.678300374009795e-08,
"loss": 1.0283,
"step": 10125
},
{
"epoch": 1.89,
"grad_norm": 2.7116455936430652,
"learning_rate": 7.550972333629336e-08,
"loss": 1.0421,
"step": 10130
},
{
"epoch": 1.89,
"grad_norm": 2.4374078572896356,
"learning_rate": 7.424700836669851e-08,
"loss": 1.0299,
"step": 10135
},
{
"epoch": 1.89,
"grad_norm": 2.6644682159026503,
"learning_rate": 7.299486154045465e-08,
"loss": 0.9972,
"step": 10140
},
{
"epoch": 1.89,
"grad_norm": 2.533516269766712,
"learning_rate": 7.175328554402561e-08,
"loss": 0.9605,
"step": 10145
},
{
"epoch": 1.89,
"grad_norm": 2.422491510120683,
"learning_rate": 7.052228304119945e-08,
"loss": 1.0066,
"step": 10150
},
{
"epoch": 1.89,
"grad_norm": 2.701319935117471,
"learning_rate": 6.930185667307632e-08,
"loss": 0.9592,
"step": 10155
},
{
"epoch": 1.89,
"grad_norm": 2.4325140710805977,
"learning_rate": 6.809200905806612e-08,
"loss": 0.9774,
"step": 10160
},
{
"epoch": 1.9,
"grad_norm": 3.3040592558385087,
"learning_rate": 6.689274279188252e-08,
"loss": 1.0472,
"step": 10165
},
{
"epoch": 1.9,
"grad_norm": 2.9349662595476644,
"learning_rate": 6.570406044753675e-08,
"loss": 0.9517,
"step": 10170
},
{
"epoch": 1.9,
"grad_norm": 2.9432283064145315,
"learning_rate": 6.452596457533211e-08,
"loss": 0.9954,
"step": 10175
},
{
"epoch": 1.9,
"grad_norm": 2.446315385822753,
"learning_rate": 6.335845770285897e-08,
"loss": 0.9763,
"step": 10180
},
{
"epoch": 1.9,
"grad_norm": 2.4541488493898207,
"learning_rate": 6.220154233498976e-08,
"loss": 0.9846,
"step": 10185
},
{
"epoch": 1.9,
"grad_norm": 2.740747449768843,
"learning_rate": 6.105522095387173e-08,
"loss": 1.0473,
"step": 10190
},
{
"epoch": 1.9,
"grad_norm": 2.412788490791504,
"learning_rate": 5.991949601892366e-08,
"loss": 1.0206,
"step": 10195
},
{
"epoch": 1.9,
"grad_norm": 2.2280455313309404,
"learning_rate": 5.87943699668303e-08,
"loss": 1.0077,
"step": 10200
},
{
"epoch": 1.9,
"grad_norm": 2.5793911483638077,
"learning_rate": 5.767984521153569e-08,
"loss": 0.995,
"step": 10205
},
{
"epoch": 1.9,
"grad_norm": 2.729204350644033,
"learning_rate": 5.657592414423818e-08,
"loss": 0.9881,
"step": 10210
},
{
"epoch": 1.9,
"grad_norm": 2.6928681239961385,
"learning_rate": 5.548260913338766e-08,
"loss": 0.9753,
"step": 10215
},
{
"epoch": 1.91,
"grad_norm": 2.7934332324069606,
"learning_rate": 5.439990252467886e-08,
"loss": 1.0208,
"step": 10220
},
{
"epoch": 1.91,
"grad_norm": 2.4601360704368305,
"learning_rate": 5.332780664104531e-08,
"loss": 1.0288,
"step": 10225
},
{
"epoch": 1.91,
"grad_norm": 2.6515972842534627,
"learning_rate": 5.226632378265484e-08,
"loss": 1.0382,
"step": 10230
},
{
"epoch": 1.91,
"grad_norm": 2.6197630708573176,
"learning_rate": 5.121545622690738e-08,
"loss": 1.0731,
"step": 10235
},
{
"epoch": 1.91,
"grad_norm": 2.844455313015032,
"learning_rate": 5.017520622842609e-08,
"loss": 0.9925,
"step": 10240
},
{
"epoch": 1.91,
"grad_norm": 2.3475689287172785,
"learning_rate": 4.914557601905456e-08,
"loss": 0.9502,
"step": 10245
},
{
"epoch": 1.91,
"grad_norm": 2.429152996391864,
"learning_rate": 4.812656780785185e-08,
"loss": 0.961,
"step": 10250
},
{
"epoch": 1.91,
"grad_norm": 2.530617111574067,
"learning_rate": 4.711818378108801e-08,
"loss": 0.9803,
"step": 10255
},
{
"epoch": 1.91,
"grad_norm": 2.3730172494335977,
"learning_rate": 4.6120426102238547e-08,
"loss": 0.9925,
"step": 10260
},
{
"epoch": 1.91,
"grad_norm": 3.0046001348839546,
"learning_rate": 4.513329691198054e-08,
"loss": 1.0298,
"step": 10265
},
{
"epoch": 1.92,
"grad_norm": 2.5230634308534228,
"learning_rate": 4.4156798328187645e-08,
"loss": 0.9695,
"step": 10270
},
{
"epoch": 1.92,
"grad_norm": 2.6072373901168846,
"learning_rate": 4.319093244592565e-08,
"loss": 0.9855,
"step": 10275
},
{
"epoch": 1.92,
"grad_norm": 2.7290151915323144,
"learning_rate": 4.2235701337448585e-08,
"loss": 1.0341,
"step": 10280
},
{
"epoch": 1.92,
"grad_norm": 2.5839402508390577,
"learning_rate": 4.129110705219319e-08,
"loss": 1.0088,
"step": 10285
},
{
"epoch": 1.92,
"grad_norm": 2.734863021668995,
"learning_rate": 4.035715161677389e-08,
"loss": 1.0405,
"step": 10290
},
{
"epoch": 1.92,
"grad_norm": 2.537803008246443,
"learning_rate": 3.943383703498171e-08,
"loss": 0.9874,
"step": 10295
},
{
"epoch": 1.92,
"grad_norm": 2.604401291284094,
"learning_rate": 3.8521165287777605e-08,
"loss": 0.9927,
"step": 10300
},
{
"epoch": 1.92,
"grad_norm": 2.5104863117844776,
"learning_rate": 3.761913833328634e-08,
"loss": 0.9748,
"step": 10305
},
{
"epoch": 1.92,
"grad_norm": 2.9857002776001,
"learning_rate": 3.672775810679652e-08,
"loss": 0.9856,
"step": 10310
},
{
"epoch": 1.92,
"grad_norm": 3.487877028129344,
"learning_rate": 3.584702652075333e-08,
"loss": 0.9325,
"step": 10315
},
{
"epoch": 1.92,
"grad_norm": 2.462719813368756,
"learning_rate": 3.4976945464756385e-08,
"loss": 0.9738,
"step": 10320
},
{
"epoch": 1.93,
"grad_norm": 2.6647821276664305,
"learning_rate": 3.4117516805552976e-08,
"loss": 0.9519,
"step": 10325
},
{
"epoch": 1.93,
"grad_norm": 2.7581010174320104,
"learning_rate": 3.326874238703814e-08,
"loss": 0.9946,
"step": 10330
},
{
"epoch": 1.93,
"grad_norm": 2.8054155893587867,
"learning_rate": 3.2430624030245773e-08,
"loss": 1.0044,
"step": 10335
},
{
"epoch": 1.93,
"grad_norm": 2.887227101606384,
"learning_rate": 3.1603163533349136e-08,
"loss": 1.0235,
"step": 10340
},
{
"epoch": 1.93,
"grad_norm": 2.557612920941727,
"learning_rate": 3.0786362671655356e-08,
"loss": 1.0399,
"step": 10345
},
{
"epoch": 1.93,
"grad_norm": 2.317041647221803,
"learning_rate": 2.998022319760041e-08,
"loss": 1.0367,
"step": 10350
},
{
"epoch": 1.93,
"grad_norm": 3.3628041239154904,
"learning_rate": 2.9184746840746346e-08,
"loss": 1.0432,
"step": 10355
},
{
"epoch": 1.93,
"grad_norm": 2.45529949397882,
"learning_rate": 2.8399935307778516e-08,
"loss": 0.9676,
"step": 10360
},
{
"epoch": 1.93,
"grad_norm": 2.601408268617963,
"learning_rate": 2.7625790282500564e-08,
"loss": 1.0007,
"step": 10365
},
{
"epoch": 1.93,
"grad_norm": 2.6735204157162165,
"learning_rate": 2.6862313425830567e-08,
"loss": 1.0625,
"step": 10370
},
{
"epoch": 1.93,
"grad_norm": 2.359922200975906,
"learning_rate": 2.6109506375799896e-08,
"loss": 1.0015,
"step": 10375
},
{
"epoch": 1.94,
"grad_norm": 2.3056241737305734,
"learning_rate": 2.5367370747546026e-08,
"loss": 0.9744,
"step": 10380
},
{
"epoch": 1.94,
"grad_norm": 2.637133489900889,
"learning_rate": 2.4635908133312514e-08,
"loss": 0.9974,
"step": 10385
},
{
"epoch": 1.94,
"grad_norm": 2.4683171286123557,
"learning_rate": 2.3915120102443458e-08,
"loss": 0.9822,
"step": 10390
},
{
"epoch": 1.94,
"grad_norm": 2.503440993401632,
"learning_rate": 2.3205008201380718e-08,
"loss": 1.0011,
"step": 10395
},
{
"epoch": 1.94,
"grad_norm": 2.8108246729696376,
"learning_rate": 2.2505573953661153e-08,
"loss": 1.0272,
"step": 10400
},
{
"epoch": 1.94,
"grad_norm": 2.664659762392884,
"learning_rate": 2.181681885991216e-08,
"loss": 0.9898,
"step": 10405
},
{
"epoch": 1.94,
"grad_norm": 2.585990693063831,
"learning_rate": 2.113874439784891e-08,
"loss": 1.0153,
"step": 10410
},
{
"epoch": 1.94,
"grad_norm": 2.8947291462365046,
"learning_rate": 2.0471352022272683e-08,
"loss": 0.9818,
"step": 10415
},
{
"epoch": 1.94,
"grad_norm": 2.3630315438689227,
"learning_rate": 1.981464316506587e-08,
"loss": 0.9976,
"step": 10420
},
{
"epoch": 1.94,
"grad_norm": 2.9333155999303084,
"learning_rate": 1.9168619235188646e-08,
"loss": 0.9797,
"step": 10425
},
{
"epoch": 1.95,
"grad_norm": 2.7761353274100236,
"learning_rate": 1.8533281618677846e-08,
"loss": 0.9809,
"step": 10430
},
{
"epoch": 1.95,
"grad_norm": 2.926917528544316,
"learning_rate": 1.790863167864254e-08,
"loss": 0.9823,
"step": 10435
},
{
"epoch": 1.95,
"grad_norm": 2.4755499922436677,
"learning_rate": 1.7294670755262365e-08,
"loss": 0.9795,
"step": 10440
},
{
"epoch": 1.95,
"grad_norm": 2.9766426008228724,
"learning_rate": 1.669140016578197e-08,
"loss": 0.9984,
"step": 10445
},
{
"epoch": 1.95,
"grad_norm": 2.7200469233636877,
"learning_rate": 1.6098821204511562e-08,
"loss": 0.9475,
"step": 10450
},
{
"epoch": 1.95,
"grad_norm": 2.762204423209276,
"learning_rate": 1.5516935142821933e-08,
"loss": 0.9846,
"step": 10455
},
{
"epoch": 1.95,
"grad_norm": 2.5807977939478377,
"learning_rate": 1.4945743229142772e-08,
"loss": 1.0107,
"step": 10460
},
{
"epoch": 1.95,
"grad_norm": 2.7922748699778244,
"learning_rate": 1.4385246688959909e-08,
"loss": 0.9823,
"step": 10465
},
{
"epoch": 1.95,
"grad_norm": 2.4732334084056387,
"learning_rate": 1.3835446724810852e-08,
"loss": 1.0322,
"step": 10470
},
{
"epoch": 1.95,
"grad_norm": 2.918632979618341,
"learning_rate": 1.3296344516285364e-08,
"loss": 1.0167,
"step": 10475
},
{
"epoch": 1.95,
"grad_norm": 2.5363575883520846,
"learning_rate": 1.2767941220020453e-08,
"loss": 0.9996,
"step": 10480
},
{
"epoch": 1.96,
"grad_norm": 2.7942753065364663,
"learning_rate": 1.2250237969699263e-08,
"loss": 1.0262,
"step": 10485
},
{
"epoch": 1.96,
"grad_norm": 2.6713751330423516,
"learning_rate": 1.1743235876047753e-08,
"loss": 1.0308,
"step": 10490
},
{
"epoch": 1.96,
"grad_norm": 2.6195251533724946,
"learning_rate": 1.1246936026832466e-08,
"loss": 0.9306,
"step": 10495
},
{
"epoch": 1.96,
"grad_norm": 2.673566984697063,
"learning_rate": 1.0761339486859424e-08,
"loss": 1.0086,
"step": 10500
},
{
"epoch": 1.96,
"grad_norm": 2.6680706352937285,
"learning_rate": 1.0286447297969682e-08,
"loss": 1.0331,
"step": 10505
},
{
"epoch": 1.96,
"grad_norm": 2.7859246917263087,
"learning_rate": 9.822260479038226e-09,
"loss": 0.992,
"step": 10510
},
{
"epoch": 1.96,
"grad_norm": 2.3574649028330765,
"learning_rate": 9.368780025972856e-09,
"loss": 1.0498,
"step": 10515
},
{
"epoch": 1.96,
"grad_norm": 2.2777885112050225,
"learning_rate": 8.926006911710305e-09,
"loss": 0.9413,
"step": 10520
},
{
"epoch": 1.96,
"grad_norm": 2.5295332699440527,
"learning_rate": 8.493942086214568e-09,
"loss": 1.0083,
"step": 10525
},
{
"epoch": 1.96,
"grad_norm": 2.7853874196558737,
"learning_rate": 8.072586476475797e-09,
"loss": 1.0256,
"step": 10530
},
{
"epoch": 1.96,
"grad_norm": 2.6333818412429935,
"learning_rate": 7.661940986507522e-09,
"loss": 1.0239,
"step": 10535
},
{
"epoch": 1.97,
"grad_norm": 2.549575059800705,
"learning_rate": 7.262006497344431e-09,
"loss": 1.0001,
"step": 10540
},
{
"epoch": 1.97,
"grad_norm": 2.2772256153873895,
"learning_rate": 6.872783867041266e-09,
"loss": 0.958,
"step": 10545
},
{
"epoch": 1.97,
"grad_norm": 2.6683332998055147,
"learning_rate": 6.494273930670036e-09,
"loss": 1.0027,
"step": 10550
},
{
"epoch": 1.97,
"grad_norm": 2.5321362849934115,
"learning_rate": 6.126477500319472e-09,
"loss": 1.025,
"step": 10555
},
{
"epoch": 1.97,
"grad_norm": 2.602531237642126,
"learning_rate": 5.7693953650928e-09,
"loss": 1.063,
"step": 10560
},
{
"epoch": 1.97,
"grad_norm": 2.305373620166385,
"learning_rate": 5.42302829110497e-09,
"loss": 1.0504,
"step": 10565
},
{
"epoch": 1.97,
"grad_norm": 2.6405599677279197,
"learning_rate": 5.087377021482099e-09,
"loss": 1.0107,
"step": 10570
},
{
"epoch": 1.97,
"grad_norm": 2.4886321109441547,
"learning_rate": 4.762442276360357e-09,
"loss": 0.9846,
"step": 10575
},
{
"epoch": 1.97,
"grad_norm": 2.404716196110838,
"learning_rate": 4.448224752883201e-09,
"loss": 1.0061,
"step": 10580
},
{
"epoch": 1.97,
"grad_norm": 2.6094386667769336,
"learning_rate": 4.144725125200255e-09,
"loss": 0.9906,
"step": 10585
},
{
"epoch": 1.97,
"grad_norm": 3.0107792922490466,
"learning_rate": 3.85194404446676e-09,
"loss": 1.0194,
"step": 10590
},
{
"epoch": 1.98,
"grad_norm": 2.849093851045047,
"learning_rate": 3.5698821388407966e-09,
"loss": 0.9846,
"step": 10595
},
{
"epoch": 1.98,
"grad_norm": 2.625766042793939,
"learning_rate": 3.2985400134827318e-09,
"loss": 0.9563,
"step": 10600
},
{
"epoch": 1.98,
"grad_norm": 2.6570589229153803,
"learning_rate": 3.037918250553551e-09,
"loss": 1.0326,
"step": 10605
},
{
"epoch": 1.98,
"grad_norm": 2.3590551907855484,
"learning_rate": 2.7880174092148603e-09,
"loss": 0.9736,
"step": 10610
},
{
"epoch": 1.98,
"grad_norm": 2.3928499657909925,
"learning_rate": 2.548838025625e-09,
"loss": 0.9729,
"step": 10615
},
{
"epoch": 1.98,
"grad_norm": 2.6017921989296315,
"learning_rate": 2.3203806129407092e-09,
"loss": 0.9762,
"step": 10620
},
{
"epoch": 1.98,
"grad_norm": 2.5268978088365888,
"learning_rate": 2.1026456613143508e-09,
"loss": 1.0171,
"step": 10625
},
{
"epoch": 1.98,
"grad_norm": 2.647145075223155,
"learning_rate": 1.8956336378944673e-09,
"loss": 0.9712,
"step": 10630
},
{
"epoch": 1.98,
"grad_norm": 3.2187105524163186,
"learning_rate": 1.6993449868218937e-09,
"loss": 0.9911,
"step": 10635
},
{
"epoch": 1.98,
"grad_norm": 2.6908790885372715,
"learning_rate": 1.5137801292325338e-09,
"loss": 0.9815,
"step": 10640
},
{
"epoch": 1.99,
"grad_norm": 2.5243341858010324,
"learning_rate": 1.3389394632523644e-09,
"loss": 1.0148,
"step": 10645
},
{
"epoch": 1.99,
"grad_norm": 3.157707192953073,
"learning_rate": 1.1748233640007655e-09,
"loss": 1.0329,
"step": 10650
},
{
"epoch": 1.99,
"grad_norm": 2.789228515124869,
"learning_rate": 1.02143218358719e-09,
"loss": 1.0507,
"step": 10655
},
{
"epoch": 1.99,
"grad_norm": 2.494956425208091,
"learning_rate": 8.787662511094975e-10,
"loss": 1.009,
"step": 10660
},
{
"epoch": 1.99,
"grad_norm": 2.3012094212140335,
"learning_rate": 7.468258726561761e-10,
"loss": 1.0183,
"step": 10665
},
{
"epoch": 1.99,
"grad_norm": 3.027904709311952,
"learning_rate": 6.256113313035661e-10,
"loss": 0.9953,
"step": 10670
},
{
"epoch": 1.99,
"grad_norm": 2.6485213093531996,
"learning_rate": 5.151228871164149e-10,
"loss": 1.0034,
"step": 10675
},
{
"epoch": 1.99,
"grad_norm": 2.341845531967239,
"learning_rate": 4.1536077714621206e-10,
"loss": 0.9533,
"step": 10680
},
{
"epoch": 1.99,
"grad_norm": 2.866112027395985,
"learning_rate": 3.2632521543118913e-10,
"loss": 0.9769,
"step": 10685
},
{
"epoch": 1.99,
"grad_norm": 2.3197040560602704,
"learning_rate": 2.4801639299576465e-10,
"loss": 0.9946,
"step": 10690
},
{
"epoch": 1.99,
"grad_norm": 2.961026582811728,
"learning_rate": 1.8043447785109914e-10,
"loss": 0.9354,
"step": 10695
},
{
"epoch": 2.0,
"grad_norm": 2.7427152602676603,
"learning_rate": 1.2357961499342985e-10,
"loss": 0.9992,
"step": 10700
},
{
"epoch": 2.0,
"grad_norm": 2.674107444859358,
"learning_rate": 7.745192640351562e-11,
"loss": 1.0313,
"step": 10705
},
{
"epoch": 2.0,
"grad_norm": 2.591490853219579,
"learning_rate": 4.2051511048302184e-11,
"loss": 0.9762,
"step": 10710
},
{
"epoch": 2.0,
"grad_norm": 2.318869364141943,
"learning_rate": 1.7378444879256885e-11,
"loss": 0.9667,
"step": 10715
},
{
"epoch": 2.0,
"grad_norm": 2.6035429249427002,
"learning_rate": 3.4327808312584197e-12,
"loss": 1.0605,
"step": 10720
},
{
"epoch": 2.0,
"step": 10724,
"total_flos": 1.675657501170729e+17,
"train_loss": 1.1225166902173949,
"train_runtime": 37990.1661,
"train_samples_per_second": 36.134,
"train_steps_per_second": 0.282
}
],
"logging_steps": 5,
"max_steps": 10724,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 1000,
"total_flos": 1.675657501170729e+17,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}