atsuki-yamaguchi's picture
Upload folder using huggingface_hub
dc39d5e verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.0473597903539947,
"eval_steps": 500,
"global_step": 6000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 32.95980453491211,
"learning_rate": 5e-06,
"loss": 10.6174,
"step": 5
},
{
"epoch": 0.0,
"grad_norm": 23.08175277709961,
"learning_rate": 1e-05,
"loss": 10.1269,
"step": 10
},
{
"epoch": 0.0,
"grad_norm": 19.50652503967285,
"learning_rate": 1.5e-05,
"loss": 9.1906,
"step": 15
},
{
"epoch": 0.0,
"grad_norm": 13.88753890991211,
"learning_rate": 2e-05,
"loss": 8.2921,
"step": 20
},
{
"epoch": 0.0,
"grad_norm": 7.801230430603027,
"learning_rate": 2.5e-05,
"loss": 7.6554,
"step": 25
},
{
"epoch": 0.0,
"grad_norm": 4.9045634269714355,
"learning_rate": 3e-05,
"loss": 7.4797,
"step": 30
},
{
"epoch": 0.0,
"grad_norm": 4.769446849822998,
"learning_rate": 3.5e-05,
"loss": 7.3007,
"step": 35
},
{
"epoch": 0.0,
"grad_norm": 3.60318922996521,
"learning_rate": 4e-05,
"loss": 7.1262,
"step": 40
},
{
"epoch": 0.0,
"grad_norm": 4.022258281707764,
"learning_rate": 4.5e-05,
"loss": 6.9873,
"step": 45
},
{
"epoch": 0.0,
"grad_norm": 3.828768730163574,
"learning_rate": 5e-05,
"loss": 6.6907,
"step": 50
},
{
"epoch": 0.0,
"grad_norm": 3.7684431076049805,
"learning_rate": 5.500000000000001e-05,
"loss": 6.3761,
"step": 55
},
{
"epoch": 0.0,
"grad_norm": 3.8739750385284424,
"learning_rate": 6e-05,
"loss": 6.2971,
"step": 60
},
{
"epoch": 0.0,
"grad_norm": 22.622005462646484,
"learning_rate": 6.500000000000001e-05,
"loss": 6.0035,
"step": 65
},
{
"epoch": 0.0,
"grad_norm": 4.665385723114014,
"learning_rate": 7e-05,
"loss": 5.9092,
"step": 70
},
{
"epoch": 0.0,
"grad_norm": 3.8950252532958984,
"learning_rate": 7.500000000000001e-05,
"loss": 5.6709,
"step": 75
},
{
"epoch": 0.0,
"grad_norm": 4.0300421714782715,
"learning_rate": 8e-05,
"loss": 5.5943,
"step": 80
},
{
"epoch": 0.0,
"grad_norm": 3.3713412284851074,
"learning_rate": 8.5e-05,
"loss": 5.2952,
"step": 85
},
{
"epoch": 0.0,
"grad_norm": 2.9832828044891357,
"learning_rate": 9e-05,
"loss": 5.4213,
"step": 90
},
{
"epoch": 0.0,
"grad_norm": 3.131589651107788,
"learning_rate": 9.5e-05,
"loss": 5.1793,
"step": 95
},
{
"epoch": 0.0,
"grad_norm": 3.034811019897461,
"learning_rate": 0.0001,
"loss": 5.099,
"step": 100
},
{
"epoch": 0.0,
"grad_norm": 3.0053954124450684,
"learning_rate": 9.999999961506464e-05,
"loss": 4.9262,
"step": 105
},
{
"epoch": 0.0,
"grad_norm": 2.5153939723968506,
"learning_rate": 9.999999846025851e-05,
"loss": 4.9408,
"step": 110
},
{
"epoch": 0.0,
"grad_norm": 2.802779197692871,
"learning_rate": 9.999999653558166e-05,
"loss": 4.8329,
"step": 115
},
{
"epoch": 0.0,
"grad_norm": 3.006596803665161,
"learning_rate": 9.999999384103412e-05,
"loss": 4.6361,
"step": 120
},
{
"epoch": 0.0,
"grad_norm": 2.791879653930664,
"learning_rate": 9.999999037661592e-05,
"loss": 4.5578,
"step": 125
},
{
"epoch": 0.0,
"grad_norm": 2.5133559703826904,
"learning_rate": 9.999998614232713e-05,
"loss": 4.5047,
"step": 130
},
{
"epoch": 0.0,
"grad_norm": 2.5018651485443115,
"learning_rate": 9.99999811381678e-05,
"loss": 4.3281,
"step": 135
},
{
"epoch": 0.0,
"grad_norm": 2.6733438968658447,
"learning_rate": 9.9999975364138e-05,
"loss": 4.3313,
"step": 140
},
{
"epoch": 0.0,
"grad_norm": 2.1609246730804443,
"learning_rate": 9.999996882023784e-05,
"loss": 4.127,
"step": 145
},
{
"epoch": 0.0,
"grad_norm": 2.0396323204040527,
"learning_rate": 9.99999615064674e-05,
"loss": 4.0959,
"step": 150
},
{
"epoch": 0.0,
"grad_norm": 2.0834591388702393,
"learning_rate": 9.99999534228268e-05,
"loss": 4.0395,
"step": 155
},
{
"epoch": 0.0,
"grad_norm": 2.3303310871124268,
"learning_rate": 9.999994456931619e-05,
"loss": 4.0032,
"step": 160
},
{
"epoch": 0.0,
"grad_norm": 2.1243906021118164,
"learning_rate": 9.999993494593566e-05,
"loss": 3.9607,
"step": 165
},
{
"epoch": 0.0,
"grad_norm": 2.3404269218444824,
"learning_rate": 9.999992455268539e-05,
"loss": 3.8629,
"step": 170
},
{
"epoch": 0.0,
"grad_norm": 2.0201656818389893,
"learning_rate": 9.999991338956554e-05,
"loss": 3.9762,
"step": 175
},
{
"epoch": 0.0,
"grad_norm": 2.18098521232605,
"learning_rate": 9.999990145657626e-05,
"loss": 3.762,
"step": 180
},
{
"epoch": 0.0,
"grad_norm": 1.9303104877471924,
"learning_rate": 9.999988875371776e-05,
"loss": 3.7669,
"step": 185
},
{
"epoch": 0.0,
"grad_norm": 1.970390796661377,
"learning_rate": 9.999987528099021e-05,
"loss": 3.6189,
"step": 190
},
{
"epoch": 0.0,
"grad_norm": 2.543450117111206,
"learning_rate": 9.999986103839384e-05,
"loss": 3.5614,
"step": 195
},
{
"epoch": 0.0,
"grad_norm": 2.151980400085449,
"learning_rate": 9.999984602592884e-05,
"loss": 3.5602,
"step": 200
},
{
"epoch": 0.0,
"grad_norm": 2.0567076206207275,
"learning_rate": 9.999983024359549e-05,
"loss": 3.5707,
"step": 205
},
{
"epoch": 0.0,
"grad_norm": 1.9052320718765259,
"learning_rate": 9.9999813691394e-05,
"loss": 3.4812,
"step": 210
},
{
"epoch": 0.0,
"grad_norm": 1.8563047647476196,
"learning_rate": 9.999979636932461e-05,
"loss": 3.3568,
"step": 215
},
{
"epoch": 0.0,
"grad_norm": 1.658285140991211,
"learning_rate": 9.999977827738762e-05,
"loss": 3.4196,
"step": 220
},
{
"epoch": 0.0,
"grad_norm": 1.7364400625228882,
"learning_rate": 9.999975941558328e-05,
"loss": 3.4028,
"step": 225
},
{
"epoch": 0.0,
"grad_norm": 1.9363423585891724,
"learning_rate": 9.999973978391191e-05,
"loss": 3.2965,
"step": 230
},
{
"epoch": 0.0,
"grad_norm": 1.875470519065857,
"learning_rate": 9.99997193823738e-05,
"loss": 3.4251,
"step": 235
},
{
"epoch": 0.0,
"grad_norm": 1.824401617050171,
"learning_rate": 9.999969821096924e-05,
"loss": 3.3772,
"step": 240
},
{
"epoch": 0.0,
"grad_norm": 1.6702059507369995,
"learning_rate": 9.99996762696986e-05,
"loss": 3.2796,
"step": 245
},
{
"epoch": 0.0,
"grad_norm": 1.5805175304412842,
"learning_rate": 9.999965355856218e-05,
"loss": 3.2094,
"step": 250
},
{
"epoch": 0.0,
"grad_norm": 1.5987428426742554,
"learning_rate": 9.999963007756035e-05,
"loss": 3.2734,
"step": 255
},
{
"epoch": 0.0,
"grad_norm": 1.6558455228805542,
"learning_rate": 9.999960582669346e-05,
"loss": 3.2781,
"step": 260
},
{
"epoch": 0.0,
"grad_norm": 1.8360902070999146,
"learning_rate": 9.99995808059619e-05,
"loss": 3.1587,
"step": 265
},
{
"epoch": 0.0,
"grad_norm": 1.484212040901184,
"learning_rate": 9.999955501536603e-05,
"loss": 3.1896,
"step": 270
},
{
"epoch": 0.0,
"grad_norm": 1.5533018112182617,
"learning_rate": 9.999952845490626e-05,
"loss": 3.2841,
"step": 275
},
{
"epoch": 0.0,
"grad_norm": 1.5394784212112427,
"learning_rate": 9.999950112458302e-05,
"loss": 3.0576,
"step": 280
},
{
"epoch": 0.0,
"grad_norm": 1.705702781677246,
"learning_rate": 9.99994730243967e-05,
"loss": 3.2049,
"step": 285
},
{
"epoch": 0.0,
"grad_norm": 2.4536681175231934,
"learning_rate": 9.999944415434774e-05,
"loss": 3.0977,
"step": 290
},
{
"epoch": 0.0,
"grad_norm": 2.2552566528320312,
"learning_rate": 9.99994145144366e-05,
"loss": 3.0715,
"step": 295
},
{
"epoch": 0.0,
"grad_norm": 1.7731832265853882,
"learning_rate": 9.999938410466371e-05,
"loss": 3.1072,
"step": 300
},
{
"epoch": 0.0,
"grad_norm": 1.4365235567092896,
"learning_rate": 9.999935292502958e-05,
"loss": 3.0695,
"step": 305
},
{
"epoch": 0.0,
"grad_norm": 1.5288578271865845,
"learning_rate": 9.999932097553463e-05,
"loss": 3.0721,
"step": 310
},
{
"epoch": 0.0,
"grad_norm": 1.8409128189086914,
"learning_rate": 9.999928825617941e-05,
"loss": 3.0173,
"step": 315
},
{
"epoch": 0.0,
"grad_norm": 1.3911570310592651,
"learning_rate": 9.999925476696439e-05,
"loss": 2.9696,
"step": 320
},
{
"epoch": 0.0,
"grad_norm": 1.4508342742919922,
"learning_rate": 9.99992205078901e-05,
"loss": 2.9962,
"step": 325
},
{
"epoch": 0.0,
"grad_norm": 1.5994690656661987,
"learning_rate": 9.999918547895705e-05,
"loss": 3.014,
"step": 330
},
{
"epoch": 0.0,
"grad_norm": 1.4494816064834595,
"learning_rate": 9.999914968016579e-05,
"loss": 3.0387,
"step": 335
},
{
"epoch": 0.0,
"grad_norm": 1.5155937671661377,
"learning_rate": 9.999911311151689e-05,
"loss": 2.9023,
"step": 340
},
{
"epoch": 0.0,
"grad_norm": 1.4000442028045654,
"learning_rate": 9.999907577301089e-05,
"loss": 3.0064,
"step": 345
},
{
"epoch": 0.0,
"grad_norm": 1.396157145500183,
"learning_rate": 9.999903766464837e-05,
"loss": 2.9525,
"step": 350
},
{
"epoch": 0.0,
"grad_norm": 1.3950470685958862,
"learning_rate": 9.99989987864299e-05,
"loss": 2.9472,
"step": 355
},
{
"epoch": 0.0,
"grad_norm": 1.5075054168701172,
"learning_rate": 9.999895913835613e-05,
"loss": 2.9132,
"step": 360
},
{
"epoch": 0.0,
"grad_norm": 1.3249679803848267,
"learning_rate": 9.999891872042762e-05,
"loss": 3.0044,
"step": 365
},
{
"epoch": 0.0,
"grad_norm": 1.390251636505127,
"learning_rate": 9.9998877532645e-05,
"loss": 2.9182,
"step": 370
},
{
"epoch": 0.0,
"grad_norm": 1.3941525220870972,
"learning_rate": 9.999883557500892e-05,
"loss": 2.8691,
"step": 375
},
{
"epoch": 0.0,
"grad_norm": 1.3029892444610596,
"learning_rate": 9.999879284752003e-05,
"loss": 2.906,
"step": 380
},
{
"epoch": 0.0,
"grad_norm": 1.2555627822875977,
"learning_rate": 9.999874935017897e-05,
"loss": 2.8948,
"step": 385
},
{
"epoch": 0.0,
"grad_norm": 1.2675596475601196,
"learning_rate": 9.999870508298641e-05,
"loss": 2.9366,
"step": 390
},
{
"epoch": 0.0,
"grad_norm": 1.3020423650741577,
"learning_rate": 9.999866004594303e-05,
"loss": 2.9958,
"step": 395
},
{
"epoch": 0.0,
"grad_norm": 1.2635390758514404,
"learning_rate": 9.999861423904954e-05,
"loss": 2.7824,
"step": 400
},
{
"epoch": 0.0,
"grad_norm": 1.3098176717758179,
"learning_rate": 9.999856766230665e-05,
"loss": 2.9318,
"step": 405
},
{
"epoch": 0.0,
"grad_norm": 1.2534728050231934,
"learning_rate": 9.999852031571506e-05,
"loss": 2.8804,
"step": 410
},
{
"epoch": 0.0,
"grad_norm": 1.2704709768295288,
"learning_rate": 9.999847219927549e-05,
"loss": 2.8769,
"step": 415
},
{
"epoch": 0.0,
"grad_norm": 1.3780895471572876,
"learning_rate": 9.999842331298872e-05,
"loss": 2.8193,
"step": 420
},
{
"epoch": 0.0,
"grad_norm": 1.3754754066467285,
"learning_rate": 9.999837365685544e-05,
"loss": 2.8855,
"step": 425
},
{
"epoch": 0.0,
"grad_norm": 4.035409927368164,
"learning_rate": 9.999832323087649e-05,
"loss": 2.8485,
"step": 430
},
{
"epoch": 0.0,
"grad_norm": 1.3403300046920776,
"learning_rate": 9.999827203505259e-05,
"loss": 2.8813,
"step": 435
},
{
"epoch": 0.0,
"grad_norm": 1.3428093194961548,
"learning_rate": 9.999822006938455e-05,
"loss": 2.7208,
"step": 440
},
{
"epoch": 0.0,
"grad_norm": 1.2568292617797852,
"learning_rate": 9.999816733387316e-05,
"loss": 2.8367,
"step": 445
},
{
"epoch": 0.0,
"grad_norm": 1.1782286167144775,
"learning_rate": 9.999811382851924e-05,
"loss": 2.8307,
"step": 450
},
{
"epoch": 0.0,
"grad_norm": 1.2177588939666748,
"learning_rate": 9.999805955332362e-05,
"loss": 2.732,
"step": 455
},
{
"epoch": 0.0,
"grad_norm": 1.187841773033142,
"learning_rate": 9.999800450828711e-05,
"loss": 2.8451,
"step": 460
},
{
"epoch": 0.0,
"grad_norm": 1.336360216140747,
"learning_rate": 9.99979486934106e-05,
"loss": 2.8259,
"step": 465
},
{
"epoch": 0.0,
"grad_norm": 1.3569526672363281,
"learning_rate": 9.99978921086949e-05,
"loss": 2.8134,
"step": 470
},
{
"epoch": 0.0,
"grad_norm": 1.176292896270752,
"learning_rate": 9.999783475414093e-05,
"loss": 2.8574,
"step": 475
},
{
"epoch": 0.0,
"grad_norm": 1.1667404174804688,
"learning_rate": 9.999777662974952e-05,
"loss": 2.8403,
"step": 480
},
{
"epoch": 0.0,
"grad_norm": 1.2586623430252075,
"learning_rate": 9.999771773552162e-05,
"loss": 2.8177,
"step": 485
},
{
"epoch": 0.0,
"grad_norm": 1.1664625406265259,
"learning_rate": 9.999765807145811e-05,
"loss": 2.7642,
"step": 490
},
{
"epoch": 0.0,
"grad_norm": 1.1835579872131348,
"learning_rate": 9.99975976375599e-05,
"loss": 2.6916,
"step": 495
},
{
"epoch": 0.0,
"grad_norm": 1.2055819034576416,
"learning_rate": 9.999753643382793e-05,
"loss": 2.7263,
"step": 500
},
{
"epoch": 0.0,
"grad_norm": 1.1588612794876099,
"learning_rate": 9.999747446026315e-05,
"loss": 2.7711,
"step": 505
},
{
"epoch": 0.0,
"grad_norm": 1.1667373180389404,
"learning_rate": 9.999741171686649e-05,
"loss": 2.7734,
"step": 510
},
{
"epoch": 0.0,
"grad_norm": 1.2584174871444702,
"learning_rate": 9.999734820363895e-05,
"loss": 2.7403,
"step": 515
},
{
"epoch": 0.0,
"grad_norm": 1.156795620918274,
"learning_rate": 9.99972839205815e-05,
"loss": 2.6839,
"step": 520
},
{
"epoch": 0.0,
"grad_norm": 2.5467302799224854,
"learning_rate": 9.999721886769511e-05,
"loss": 2.7323,
"step": 525
},
{
"epoch": 0.0,
"grad_norm": 1.2152918577194214,
"learning_rate": 9.999715304498078e-05,
"loss": 2.8581,
"step": 530
},
{
"epoch": 0.0,
"grad_norm": 1.1377496719360352,
"learning_rate": 9.999708645243955e-05,
"loss": 2.7058,
"step": 535
},
{
"epoch": 0.0,
"grad_norm": 1.141626000404358,
"learning_rate": 9.999701909007242e-05,
"loss": 2.7032,
"step": 540
},
{
"epoch": 0.0,
"grad_norm": 1.169816255569458,
"learning_rate": 9.999695095788044e-05,
"loss": 2.6926,
"step": 545
},
{
"epoch": 0.0,
"grad_norm": 1.2900584936141968,
"learning_rate": 9.999688205586468e-05,
"loss": 2.8123,
"step": 550
},
{
"epoch": 0.0,
"grad_norm": 1.1548844575881958,
"learning_rate": 9.999681238402615e-05,
"loss": 2.7452,
"step": 555
},
{
"epoch": 0.0,
"grad_norm": 1.1868460178375244,
"learning_rate": 9.999674194236597e-05,
"loss": 2.5828,
"step": 560
},
{
"epoch": 0.0,
"grad_norm": 1.2031657695770264,
"learning_rate": 9.999667073088519e-05,
"loss": 2.8189,
"step": 565
},
{
"epoch": 0.0,
"grad_norm": 1.1584038734436035,
"learning_rate": 9.999659874958493e-05,
"loss": 2.5919,
"step": 570
},
{
"epoch": 0.0,
"grad_norm": 1.1079555749893188,
"learning_rate": 9.99965259984663e-05,
"loss": 2.6586,
"step": 575
},
{
"epoch": 0.0,
"grad_norm": 1.1593067646026611,
"learning_rate": 9.999645247753039e-05,
"loss": 2.6506,
"step": 580
},
{
"epoch": 0.0,
"grad_norm": 1.1107124090194702,
"learning_rate": 9.999637818677836e-05,
"loss": 2.7492,
"step": 585
},
{
"epoch": 0.0,
"grad_norm": 1.5372860431671143,
"learning_rate": 9.999630312621136e-05,
"loss": 2.6704,
"step": 590
},
{
"epoch": 0.0,
"grad_norm": 1.162360429763794,
"learning_rate": 9.999622729583051e-05,
"loss": 2.7192,
"step": 595
},
{
"epoch": 0.0,
"grad_norm": 1.131689429283142,
"learning_rate": 9.999615069563701e-05,
"loss": 2.6704,
"step": 600
},
{
"epoch": 0.0,
"grad_norm": 1.2108142375946045,
"learning_rate": 9.999607332563205e-05,
"loss": 2.6746,
"step": 605
},
{
"epoch": 0.0,
"grad_norm": 1.1424001455307007,
"learning_rate": 9.999599518581676e-05,
"loss": 2.6269,
"step": 610
},
{
"epoch": 0.0,
"grad_norm": 1.127795934677124,
"learning_rate": 9.999591627619242e-05,
"loss": 2.6739,
"step": 615
},
{
"epoch": 0.0,
"grad_norm": 1.1234729290008545,
"learning_rate": 9.999583659676019e-05,
"loss": 2.6392,
"step": 620
},
{
"epoch": 0.0,
"grad_norm": 1.108341932296753,
"learning_rate": 9.999575614752133e-05,
"loss": 2.6916,
"step": 625
},
{
"epoch": 0.0,
"grad_norm": 1.201235294342041,
"learning_rate": 9.999567492847706e-05,
"loss": 2.5429,
"step": 630
},
{
"epoch": 0.01,
"grad_norm": 1.0808579921722412,
"learning_rate": 9.999559293962864e-05,
"loss": 2.665,
"step": 635
},
{
"epoch": 0.01,
"grad_norm": 1.175994873046875,
"learning_rate": 9.99955101809773e-05,
"loss": 2.7239,
"step": 640
},
{
"epoch": 0.01,
"grad_norm": 1.1204140186309814,
"learning_rate": 9.999542665252437e-05,
"loss": 2.6005,
"step": 645
},
{
"epoch": 0.01,
"grad_norm": 1.466207504272461,
"learning_rate": 9.999534235427109e-05,
"loss": 2.6616,
"step": 650
},
{
"epoch": 0.01,
"grad_norm": 1.1603835821151733,
"learning_rate": 9.999525728621879e-05,
"loss": 2.6176,
"step": 655
},
{
"epoch": 0.01,
"grad_norm": 1.0453072786331177,
"learning_rate": 9.999517144836876e-05,
"loss": 2.6543,
"step": 660
},
{
"epoch": 0.01,
"grad_norm": 1.0758476257324219,
"learning_rate": 9.999508484072233e-05,
"loss": 2.679,
"step": 665
},
{
"epoch": 0.01,
"grad_norm": 1.1041420698165894,
"learning_rate": 9.999499746328083e-05,
"loss": 2.7054,
"step": 670
},
{
"epoch": 0.01,
"grad_norm": 1.1205840110778809,
"learning_rate": 9.99949093160456e-05,
"loss": 2.6226,
"step": 675
},
{
"epoch": 0.01,
"grad_norm": 1.0702414512634277,
"learning_rate": 9.9994820399018e-05,
"loss": 2.5689,
"step": 680
},
{
"epoch": 0.01,
"grad_norm": 1.0805797576904297,
"learning_rate": 9.999473071219941e-05,
"loss": 2.6365,
"step": 685
},
{
"epoch": 0.01,
"grad_norm": 1.1297557353973389,
"learning_rate": 9.999464025559121e-05,
"loss": 2.6007,
"step": 690
},
{
"epoch": 0.01,
"grad_norm": 1.0443118810653687,
"learning_rate": 9.999454902919478e-05,
"loss": 2.6287,
"step": 695
},
{
"epoch": 0.01,
"grad_norm": 1.0508939027786255,
"learning_rate": 9.999445703301154e-05,
"loss": 2.7079,
"step": 700
},
{
"epoch": 0.01,
"grad_norm": 1.096401333808899,
"learning_rate": 9.999436426704287e-05,
"loss": 2.6502,
"step": 705
},
{
"epoch": 0.01,
"grad_norm": 1.0624600648880005,
"learning_rate": 9.999427073129026e-05,
"loss": 2.6078,
"step": 710
},
{
"epoch": 0.01,
"grad_norm": 1.2751412391662598,
"learning_rate": 9.999417642575509e-05,
"loss": 2.597,
"step": 715
},
{
"epoch": 0.01,
"grad_norm": 1.0477944612503052,
"learning_rate": 9.999408135043886e-05,
"loss": 2.5954,
"step": 720
},
{
"epoch": 0.01,
"grad_norm": 1.0045342445373535,
"learning_rate": 9.999398550534298e-05,
"loss": 2.5694,
"step": 725
},
{
"epoch": 0.01,
"grad_norm": 1.0026309490203857,
"learning_rate": 9.999388889046899e-05,
"loss": 2.5706,
"step": 730
},
{
"epoch": 0.01,
"grad_norm": 0.9893562197685242,
"learning_rate": 9.999379150581833e-05,
"loss": 2.641,
"step": 735
},
{
"epoch": 0.01,
"grad_norm": 1.0654165744781494,
"learning_rate": 9.999369335139251e-05,
"loss": 2.6301,
"step": 740
},
{
"epoch": 0.01,
"grad_norm": 1.0744380950927734,
"learning_rate": 9.999359442719306e-05,
"loss": 2.6489,
"step": 745
},
{
"epoch": 0.01,
"grad_norm": 1.009368658065796,
"learning_rate": 9.999349473322149e-05,
"loss": 2.6512,
"step": 750
},
{
"epoch": 0.01,
"grad_norm": 1.046723484992981,
"learning_rate": 9.999339426947932e-05,
"loss": 2.5826,
"step": 755
},
{
"epoch": 0.01,
"grad_norm": 1.0702534914016724,
"learning_rate": 9.999329303596812e-05,
"loss": 2.5603,
"step": 760
},
{
"epoch": 0.01,
"grad_norm": 0.9764877557754517,
"learning_rate": 9.999319103268945e-05,
"loss": 2.6447,
"step": 765
},
{
"epoch": 0.01,
"grad_norm": 0.9929907917976379,
"learning_rate": 9.999308825964485e-05,
"loss": 2.6057,
"step": 770
},
{
"epoch": 0.01,
"grad_norm": 1.1112033128738403,
"learning_rate": 9.999298471683593e-05,
"loss": 2.6257,
"step": 775
},
{
"epoch": 0.01,
"grad_norm": 1.1541080474853516,
"learning_rate": 9.999288040426429e-05,
"loss": 2.5506,
"step": 780
},
{
"epoch": 0.01,
"grad_norm": 1.0221292972564697,
"learning_rate": 9.999277532193151e-05,
"loss": 2.6348,
"step": 785
},
{
"epoch": 0.01,
"grad_norm": 1.0220556259155273,
"learning_rate": 9.999266946983923e-05,
"loss": 2.6159,
"step": 790
},
{
"epoch": 0.01,
"grad_norm": 1.0188376903533936,
"learning_rate": 9.999256284798906e-05,
"loss": 2.5773,
"step": 795
},
{
"epoch": 0.01,
"grad_norm": 1.0471076965332031,
"learning_rate": 9.999245545638267e-05,
"loss": 2.51,
"step": 800
},
{
"epoch": 0.01,
"grad_norm": 1.0504218339920044,
"learning_rate": 9.999234729502167e-05,
"loss": 2.6157,
"step": 805
},
{
"epoch": 0.01,
"grad_norm": 1.148425579071045,
"learning_rate": 9.999223836390777e-05,
"loss": 2.5515,
"step": 810
},
{
"epoch": 0.01,
"grad_norm": 1.1203477382659912,
"learning_rate": 9.999212866304262e-05,
"loss": 2.6196,
"step": 815
},
{
"epoch": 0.01,
"grad_norm": 1.0279690027236938,
"learning_rate": 9.999201819242794e-05,
"loss": 2.5179,
"step": 820
},
{
"epoch": 0.01,
"grad_norm": 1.049363136291504,
"learning_rate": 9.99919069520654e-05,
"loss": 2.6311,
"step": 825
},
{
"epoch": 0.01,
"grad_norm": 1.0632222890853882,
"learning_rate": 9.99917949419567e-05,
"loss": 2.5137,
"step": 830
},
{
"epoch": 0.01,
"grad_norm": 1.4006502628326416,
"learning_rate": 9.999168216210362e-05,
"loss": 2.5498,
"step": 835
},
{
"epoch": 0.01,
"grad_norm": 1.0094521045684814,
"learning_rate": 9.999156861250784e-05,
"loss": 2.5208,
"step": 840
},
{
"epoch": 0.01,
"grad_norm": 1.0634709596633911,
"learning_rate": 9.999145429317114e-05,
"loss": 2.5981,
"step": 845
},
{
"epoch": 0.01,
"grad_norm": 1.0341681241989136,
"learning_rate": 9.999133920409527e-05,
"loss": 2.4564,
"step": 850
},
{
"epoch": 0.01,
"grad_norm": 1.0016040802001953,
"learning_rate": 9.999122334528199e-05,
"loss": 2.6095,
"step": 855
},
{
"epoch": 0.01,
"grad_norm": 1.0196380615234375,
"learning_rate": 9.999110671673313e-05,
"loss": 2.5775,
"step": 860
},
{
"epoch": 0.01,
"grad_norm": 0.9892910122871399,
"learning_rate": 9.999098931845044e-05,
"loss": 2.656,
"step": 865
},
{
"epoch": 0.01,
"grad_norm": 1.0621706247329712,
"learning_rate": 9.999087115043573e-05,
"loss": 2.5605,
"step": 870
},
{
"epoch": 0.01,
"grad_norm": 1.0043002367019653,
"learning_rate": 9.999075221269083e-05,
"loss": 2.5334,
"step": 875
},
{
"epoch": 0.01,
"grad_norm": 1.033072829246521,
"learning_rate": 9.999063250521759e-05,
"loss": 2.5778,
"step": 880
},
{
"epoch": 0.01,
"grad_norm": 1.0772734880447388,
"learning_rate": 9.999051202801781e-05,
"loss": 2.5504,
"step": 885
},
{
"epoch": 0.01,
"grad_norm": 0.9885766506195068,
"learning_rate": 9.999039078109341e-05,
"loss": 2.538,
"step": 890
},
{
"epoch": 0.01,
"grad_norm": 1.198051929473877,
"learning_rate": 9.999026876444617e-05,
"loss": 2.5959,
"step": 895
},
{
"epoch": 0.01,
"grad_norm": 1.007222056388855,
"learning_rate": 9.999014597807805e-05,
"loss": 2.533,
"step": 900
},
{
"epoch": 0.01,
"grad_norm": 0.9983527660369873,
"learning_rate": 9.99900224219909e-05,
"loss": 2.484,
"step": 905
},
{
"epoch": 0.01,
"grad_norm": 1.1200263500213623,
"learning_rate": 9.998989809618662e-05,
"loss": 2.5374,
"step": 910
},
{
"epoch": 0.01,
"grad_norm": 1.0435030460357666,
"learning_rate": 9.998977300066715e-05,
"loss": 2.5753,
"step": 915
},
{
"epoch": 0.01,
"grad_norm": 1.1321920156478882,
"learning_rate": 9.998964713543437e-05,
"loss": 2.6016,
"step": 920
},
{
"epoch": 0.01,
"grad_norm": 1.1007788181304932,
"learning_rate": 9.998952050049026e-05,
"loss": 2.5412,
"step": 925
},
{
"epoch": 0.01,
"grad_norm": 0.9599917531013489,
"learning_rate": 9.998939309583677e-05,
"loss": 2.5929,
"step": 930
},
{
"epoch": 0.01,
"grad_norm": 0.9470053911209106,
"learning_rate": 9.998926492147585e-05,
"loss": 2.5021,
"step": 935
},
{
"epoch": 0.01,
"grad_norm": 0.9383249282836914,
"learning_rate": 9.998913597740945e-05,
"loss": 2.5888,
"step": 940
},
{
"epoch": 0.01,
"grad_norm": 0.9767156839370728,
"learning_rate": 9.99890062636396e-05,
"loss": 2.4786,
"step": 945
},
{
"epoch": 0.01,
"grad_norm": 0.9423871636390686,
"learning_rate": 9.998887578016827e-05,
"loss": 2.4054,
"step": 950
},
{
"epoch": 0.01,
"grad_norm": 0.9604974389076233,
"learning_rate": 9.998874452699747e-05,
"loss": 2.4476,
"step": 955
},
{
"epoch": 0.01,
"grad_norm": 0.9511391520500183,
"learning_rate": 9.998861250412922e-05,
"loss": 2.5004,
"step": 960
},
{
"epoch": 0.01,
"grad_norm": 1.0014891624450684,
"learning_rate": 9.998847971156556e-05,
"loss": 2.4607,
"step": 965
},
{
"epoch": 0.01,
"grad_norm": 0.9947559833526611,
"learning_rate": 9.998834614930855e-05,
"loss": 2.5944,
"step": 970
},
{
"epoch": 0.01,
"grad_norm": 1.0023629665374756,
"learning_rate": 9.99882118173602e-05,
"loss": 2.501,
"step": 975
},
{
"epoch": 0.01,
"grad_norm": 0.9713792204856873,
"learning_rate": 9.998807671572263e-05,
"loss": 2.5254,
"step": 980
},
{
"epoch": 0.01,
"grad_norm": 0.9444621801376343,
"learning_rate": 9.998794084439789e-05,
"loss": 2.4623,
"step": 985
},
{
"epoch": 0.01,
"grad_norm": 1.0840632915496826,
"learning_rate": 9.998780420338809e-05,
"loss": 2.5279,
"step": 990
},
{
"epoch": 0.01,
"grad_norm": 0.9299571514129639,
"learning_rate": 9.998766679269531e-05,
"loss": 2.611,
"step": 995
},
{
"epoch": 0.01,
"grad_norm": 0.9813534617424011,
"learning_rate": 9.998752861232168e-05,
"loss": 2.5523,
"step": 1000
},
{
"epoch": 0.01,
"grad_norm": 1.064868688583374,
"learning_rate": 9.998738966226933e-05,
"loss": 2.4938,
"step": 1005
},
{
"epoch": 0.01,
"grad_norm": 0.9773799777030945,
"learning_rate": 9.998724994254041e-05,
"loss": 2.4786,
"step": 1010
},
{
"epoch": 0.01,
"grad_norm": 0.9777352213859558,
"learning_rate": 9.998710945313705e-05,
"loss": 2.4999,
"step": 1015
},
{
"epoch": 0.01,
"grad_norm": 0.9362644553184509,
"learning_rate": 9.998696819406141e-05,
"loss": 2.4366,
"step": 1020
},
{
"epoch": 0.01,
"grad_norm": 0.9722654819488525,
"learning_rate": 9.99868261653157e-05,
"loss": 2.4774,
"step": 1025
},
{
"epoch": 0.01,
"grad_norm": 0.9268703460693359,
"learning_rate": 9.998668336690206e-05,
"loss": 2.586,
"step": 1030
},
{
"epoch": 0.01,
"grad_norm": 0.9543352127075195,
"learning_rate": 9.998653979882271e-05,
"loss": 2.513,
"step": 1035
},
{
"epoch": 0.01,
"grad_norm": 0.9115052223205566,
"learning_rate": 9.998639546107989e-05,
"loss": 2.559,
"step": 1040
},
{
"epoch": 0.01,
"grad_norm": 0.9557726383209229,
"learning_rate": 9.998625035367577e-05,
"loss": 2.5197,
"step": 1045
},
{
"epoch": 0.01,
"grad_norm": 0.9358950257301331,
"learning_rate": 9.998610447661261e-05,
"loss": 2.5415,
"step": 1050
},
{
"epoch": 0.01,
"grad_norm": 0.9870069026947021,
"learning_rate": 9.998595782989266e-05,
"loss": 2.3759,
"step": 1055
},
{
"epoch": 0.01,
"grad_norm": 0.9148921370506287,
"learning_rate": 9.998581041351818e-05,
"loss": 2.4227,
"step": 1060
},
{
"epoch": 0.01,
"grad_norm": 0.9112080335617065,
"learning_rate": 9.998566222749142e-05,
"loss": 2.5108,
"step": 1065
},
{
"epoch": 0.01,
"grad_norm": 0.9223835468292236,
"learning_rate": 9.998551327181468e-05,
"loss": 2.4249,
"step": 1070
},
{
"epoch": 0.01,
"grad_norm": 0.9529135823249817,
"learning_rate": 9.998536354649026e-05,
"loss": 2.4926,
"step": 1075
},
{
"epoch": 0.01,
"grad_norm": 0.9169237613677979,
"learning_rate": 9.998521305152045e-05,
"loss": 2.5502,
"step": 1080
},
{
"epoch": 0.01,
"grad_norm": 0.9463605880737305,
"learning_rate": 9.998506178690755e-05,
"loss": 2.5636,
"step": 1085
},
{
"epoch": 0.01,
"grad_norm": 0.966305673122406,
"learning_rate": 9.998490975265393e-05,
"loss": 2.4698,
"step": 1090
},
{
"epoch": 0.01,
"grad_norm": 0.9776231050491333,
"learning_rate": 9.99847569487619e-05,
"loss": 2.5655,
"step": 1095
},
{
"epoch": 0.01,
"grad_norm": 0.9171642065048218,
"learning_rate": 9.998460337523383e-05,
"loss": 2.4615,
"step": 1100
},
{
"epoch": 0.01,
"grad_norm": 0.9231149554252625,
"learning_rate": 9.998444903207207e-05,
"loss": 2.4601,
"step": 1105
},
{
"epoch": 0.01,
"grad_norm": 0.946007490158081,
"learning_rate": 9.998429391927902e-05,
"loss": 2.5345,
"step": 1110
},
{
"epoch": 0.01,
"grad_norm": 0.9144755601882935,
"learning_rate": 9.998413803685705e-05,
"loss": 2.4527,
"step": 1115
},
{
"epoch": 0.01,
"grad_norm": 0.9138039350509644,
"learning_rate": 9.998398138480855e-05,
"loss": 2.4387,
"step": 1120
},
{
"epoch": 0.01,
"grad_norm": 0.9384303689002991,
"learning_rate": 9.998382396313597e-05,
"loss": 2.4052,
"step": 1125
},
{
"epoch": 0.01,
"grad_norm": 0.9103301167488098,
"learning_rate": 9.998366577184168e-05,
"loss": 2.4573,
"step": 1130
},
{
"epoch": 0.01,
"grad_norm": 0.9118425250053406,
"learning_rate": 9.998350681092817e-05,
"loss": 2.5117,
"step": 1135
},
{
"epoch": 0.01,
"grad_norm": 0.9298896789550781,
"learning_rate": 9.998334708039785e-05,
"loss": 2.4959,
"step": 1140
},
{
"epoch": 0.01,
"grad_norm": 1.3021560907363892,
"learning_rate": 9.998318658025318e-05,
"loss": 2.4709,
"step": 1145
},
{
"epoch": 0.01,
"grad_norm": 0.9491179585456848,
"learning_rate": 9.998302531049667e-05,
"loss": 2.5453,
"step": 1150
},
{
"epoch": 0.01,
"grad_norm": 0.9604113698005676,
"learning_rate": 9.998286327113076e-05,
"loss": 2.5037,
"step": 1155
},
{
"epoch": 0.01,
"grad_norm": 0.9123629927635193,
"learning_rate": 9.998270046215797e-05,
"loss": 2.4952,
"step": 1160
},
{
"epoch": 0.01,
"grad_norm": 0.9616356492042542,
"learning_rate": 9.998253688358078e-05,
"loss": 2.5024,
"step": 1165
},
{
"epoch": 0.01,
"grad_norm": 0.9099243879318237,
"learning_rate": 9.998237253540174e-05,
"loss": 2.4577,
"step": 1170
},
{
"epoch": 0.01,
"grad_norm": 0.8995581269264221,
"learning_rate": 9.998220741762338e-05,
"loss": 2.5199,
"step": 1175
},
{
"epoch": 0.01,
"grad_norm": 0.9029030203819275,
"learning_rate": 9.998204153024821e-05,
"loss": 2.4,
"step": 1180
},
{
"epoch": 0.01,
"grad_norm": 0.9215284585952759,
"learning_rate": 9.998187487327882e-05,
"loss": 2.4054,
"step": 1185
},
{
"epoch": 0.01,
"grad_norm": 0.8998051881790161,
"learning_rate": 9.998170744671775e-05,
"loss": 2.3941,
"step": 1190
},
{
"epoch": 0.01,
"grad_norm": 0.9433353543281555,
"learning_rate": 9.998153925056758e-05,
"loss": 2.3906,
"step": 1195
},
{
"epoch": 0.01,
"grad_norm": 0.8889665007591248,
"learning_rate": 9.998137028483092e-05,
"loss": 2.4342,
"step": 1200
},
{
"epoch": 0.01,
"grad_norm": 0.9126563668251038,
"learning_rate": 9.998120054951035e-05,
"loss": 2.423,
"step": 1205
},
{
"epoch": 0.01,
"grad_norm": 0.9330198764801025,
"learning_rate": 9.99810300446085e-05,
"loss": 2.4753,
"step": 1210
},
{
"epoch": 0.01,
"grad_norm": 0.885710597038269,
"learning_rate": 9.9980858770128e-05,
"loss": 2.448,
"step": 1215
},
{
"epoch": 0.01,
"grad_norm": 0.955104649066925,
"learning_rate": 9.998068672607146e-05,
"loss": 2.5193,
"step": 1220
},
{
"epoch": 0.01,
"grad_norm": 0.9296510219573975,
"learning_rate": 9.998051391244155e-05,
"loss": 2.432,
"step": 1225
},
{
"epoch": 0.01,
"grad_norm": 0.915410578250885,
"learning_rate": 9.998034032924092e-05,
"loss": 2.4303,
"step": 1230
},
{
"epoch": 0.01,
"grad_norm": 0.9149826765060425,
"learning_rate": 9.998016597647225e-05,
"loss": 2.4296,
"step": 1235
},
{
"epoch": 0.01,
"grad_norm": 0.9270094633102417,
"learning_rate": 9.997999085413823e-05,
"loss": 2.4242,
"step": 1240
},
{
"epoch": 0.01,
"grad_norm": 0.9151135087013245,
"learning_rate": 9.997981496224153e-05,
"loss": 2.4867,
"step": 1245
},
{
"epoch": 0.01,
"grad_norm": 0.9048399329185486,
"learning_rate": 9.99796383007849e-05,
"loss": 2.3713,
"step": 1250
},
{
"epoch": 0.01,
"grad_norm": 0.9345071911811829,
"learning_rate": 9.997946086977103e-05,
"loss": 2.4998,
"step": 1255
},
{
"epoch": 0.01,
"grad_norm": 0.9002808928489685,
"learning_rate": 9.997928266920266e-05,
"loss": 2.5592,
"step": 1260
},
{
"epoch": 0.01,
"grad_norm": 0.9870762825012207,
"learning_rate": 9.997910369908254e-05,
"loss": 2.4635,
"step": 1265
},
{
"epoch": 0.01,
"grad_norm": 0.9023085236549377,
"learning_rate": 9.99789239594134e-05,
"loss": 2.4521,
"step": 1270
},
{
"epoch": 0.01,
"grad_norm": 0.8951256275177002,
"learning_rate": 9.997874345019805e-05,
"loss": 2.413,
"step": 1275
},
{
"epoch": 0.01,
"grad_norm": 0.8877105712890625,
"learning_rate": 9.997856217143921e-05,
"loss": 2.5026,
"step": 1280
},
{
"epoch": 0.01,
"grad_norm": 0.8981291055679321,
"learning_rate": 9.997838012313974e-05,
"loss": 2.4024,
"step": 1285
},
{
"epoch": 0.01,
"grad_norm": 0.9028763771057129,
"learning_rate": 9.99781973053024e-05,
"loss": 2.4518,
"step": 1290
},
{
"epoch": 0.01,
"grad_norm": 0.8955935835838318,
"learning_rate": 9.997801371793001e-05,
"loss": 2.3482,
"step": 1295
},
{
"epoch": 0.01,
"grad_norm": 0.9060973525047302,
"learning_rate": 9.997782936102542e-05,
"loss": 2.3617,
"step": 1300
},
{
"epoch": 0.01,
"grad_norm": 0.8658581972122192,
"learning_rate": 9.997764423459144e-05,
"loss": 2.5117,
"step": 1305
},
{
"epoch": 0.01,
"grad_norm": 0.8440374135971069,
"learning_rate": 9.997745833863092e-05,
"loss": 2.4486,
"step": 1310
},
{
"epoch": 0.01,
"grad_norm": 0.8818914294242859,
"learning_rate": 9.997727167314676e-05,
"loss": 2.479,
"step": 1315
},
{
"epoch": 0.01,
"grad_norm": 0.8601047396659851,
"learning_rate": 9.997708423814178e-05,
"loss": 2.4193,
"step": 1320
},
{
"epoch": 0.01,
"grad_norm": 0.8747096657752991,
"learning_rate": 9.99768960336189e-05,
"loss": 2.4259,
"step": 1325
},
{
"epoch": 0.01,
"grad_norm": 0.913184642791748,
"learning_rate": 9.997670705958102e-05,
"loss": 2.4655,
"step": 1330
},
{
"epoch": 0.01,
"grad_norm": 0.8921045660972595,
"learning_rate": 9.997651731603104e-05,
"loss": 2.4242,
"step": 1335
},
{
"epoch": 0.01,
"grad_norm": 1.1094772815704346,
"learning_rate": 9.997632680297188e-05,
"loss": 2.4102,
"step": 1340
},
{
"epoch": 0.01,
"grad_norm": 0.8828011155128479,
"learning_rate": 9.997613552040648e-05,
"loss": 2.4166,
"step": 1345
},
{
"epoch": 0.01,
"grad_norm": 0.8802856206893921,
"learning_rate": 9.997594346833779e-05,
"loss": 2.4426,
"step": 1350
},
{
"epoch": 0.01,
"grad_norm": 0.9135480523109436,
"learning_rate": 9.997575064676875e-05,
"loss": 2.4824,
"step": 1355
},
{
"epoch": 0.01,
"grad_norm": 0.9367683529853821,
"learning_rate": 9.997555705570232e-05,
"loss": 2.4444,
"step": 1360
},
{
"epoch": 0.01,
"grad_norm": 0.8855845332145691,
"learning_rate": 9.997536269514152e-05,
"loss": 2.4998,
"step": 1365
},
{
"epoch": 0.01,
"grad_norm": 0.8944286704063416,
"learning_rate": 9.99751675650893e-05,
"loss": 2.4665,
"step": 1370
},
{
"epoch": 0.01,
"grad_norm": 0.9414265155792236,
"learning_rate": 9.99749716655487e-05,
"loss": 2.3831,
"step": 1375
},
{
"epoch": 0.01,
"grad_norm": 0.9087872505187988,
"learning_rate": 9.997477499652273e-05,
"loss": 2.5116,
"step": 1380
},
{
"epoch": 0.01,
"grad_norm": 0.8749736547470093,
"learning_rate": 9.99745775580144e-05,
"loss": 2.3583,
"step": 1385
},
{
"epoch": 0.01,
"grad_norm": 0.8725540637969971,
"learning_rate": 9.997437935002676e-05,
"loss": 2.4517,
"step": 1390
},
{
"epoch": 0.01,
"grad_norm": 0.9350659847259521,
"learning_rate": 9.997418037256285e-05,
"loss": 2.4592,
"step": 1395
},
{
"epoch": 0.01,
"grad_norm": 0.9635329246520996,
"learning_rate": 9.997398062562575e-05,
"loss": 2.4431,
"step": 1400
},
{
"epoch": 0.01,
"grad_norm": 0.8694053888320923,
"learning_rate": 9.997378010921856e-05,
"loss": 2.3769,
"step": 1405
},
{
"epoch": 0.01,
"grad_norm": 0.8675988912582397,
"learning_rate": 9.99735788233443e-05,
"loss": 2.4562,
"step": 1410
},
{
"epoch": 0.01,
"grad_norm": 0.9171274900436401,
"learning_rate": 9.997337676800613e-05,
"loss": 2.4224,
"step": 1415
},
{
"epoch": 0.01,
"grad_norm": 0.9519357681274414,
"learning_rate": 9.997317394320713e-05,
"loss": 2.4209,
"step": 1420
},
{
"epoch": 0.01,
"grad_norm": 0.8641920685768127,
"learning_rate": 9.997297034895044e-05,
"loss": 2.3827,
"step": 1425
},
{
"epoch": 0.01,
"grad_norm": 0.8366517424583435,
"learning_rate": 9.99727659852392e-05,
"loss": 2.3705,
"step": 1430
},
{
"epoch": 0.01,
"grad_norm": 0.8750792741775513,
"learning_rate": 9.997256085207652e-05,
"loss": 2.455,
"step": 1435
},
{
"epoch": 0.01,
"grad_norm": 0.8980041146278381,
"learning_rate": 9.997235494946557e-05,
"loss": 2.3978,
"step": 1440
},
{
"epoch": 0.01,
"grad_norm": 0.822719395160675,
"learning_rate": 9.997214827740955e-05,
"loss": 2.4085,
"step": 1445
},
{
"epoch": 0.01,
"grad_norm": 0.8546679019927979,
"learning_rate": 9.997194083591164e-05,
"loss": 2.3923,
"step": 1450
},
{
"epoch": 0.01,
"grad_norm": 0.8284981846809387,
"learning_rate": 9.9971732624975e-05,
"loss": 2.4288,
"step": 1455
},
{
"epoch": 0.01,
"grad_norm": 0.8693990111351013,
"learning_rate": 9.997152364460287e-05,
"loss": 2.4196,
"step": 1460
},
{
"epoch": 0.01,
"grad_norm": 0.8502383232116699,
"learning_rate": 9.997131389479844e-05,
"loss": 2.3379,
"step": 1465
},
{
"epoch": 0.01,
"grad_norm": 0.8509696125984192,
"learning_rate": 9.997110337556494e-05,
"loss": 2.4017,
"step": 1470
},
{
"epoch": 0.01,
"grad_norm": 0.8833335041999817,
"learning_rate": 9.997089208690564e-05,
"loss": 2.4779,
"step": 1475
},
{
"epoch": 0.01,
"grad_norm": 0.89153653383255,
"learning_rate": 9.997068002882377e-05,
"loss": 2.3608,
"step": 1480
},
{
"epoch": 0.01,
"grad_norm": 0.8419952988624573,
"learning_rate": 9.997046720132261e-05,
"loss": 2.4207,
"step": 1485
},
{
"epoch": 0.01,
"grad_norm": 0.881877064704895,
"learning_rate": 9.997025360440541e-05,
"loss": 2.4667,
"step": 1490
},
{
"epoch": 0.01,
"grad_norm": 0.8301999568939209,
"learning_rate": 9.997003923807551e-05,
"loss": 2.3211,
"step": 1495
},
{
"epoch": 0.01,
"grad_norm": 0.8392807841300964,
"learning_rate": 9.996982410233615e-05,
"loss": 2.4645,
"step": 1500
},
{
"epoch": 0.01,
"grad_norm": 0.853757381439209,
"learning_rate": 9.996960819719068e-05,
"loss": 2.4622,
"step": 1505
},
{
"epoch": 0.01,
"grad_norm": 0.8599218726158142,
"learning_rate": 9.996939152264242e-05,
"loss": 2.3768,
"step": 1510
},
{
"epoch": 0.01,
"grad_norm": 0.9592317938804626,
"learning_rate": 9.99691740786947e-05,
"loss": 2.4149,
"step": 1515
},
{
"epoch": 0.01,
"grad_norm": 0.8576536774635315,
"learning_rate": 9.996895586535086e-05,
"loss": 2.4424,
"step": 1520
},
{
"epoch": 0.01,
"grad_norm": 0.8099011182785034,
"learning_rate": 9.996873688261428e-05,
"loss": 2.4376,
"step": 1525
},
{
"epoch": 0.01,
"grad_norm": 0.9041072726249695,
"learning_rate": 9.996851713048831e-05,
"loss": 2.3744,
"step": 1530
},
{
"epoch": 0.01,
"grad_norm": 0.8630540370941162,
"learning_rate": 9.996829660897637e-05,
"loss": 2.4249,
"step": 1535
},
{
"epoch": 0.01,
"grad_norm": 0.8749147057533264,
"learning_rate": 9.996807531808181e-05,
"loss": 2.4476,
"step": 1540
},
{
"epoch": 0.01,
"grad_norm": 0.8820419907569885,
"learning_rate": 9.996785325780806e-05,
"loss": 2.5427,
"step": 1545
},
{
"epoch": 0.01,
"grad_norm": 0.8545085787773132,
"learning_rate": 9.996763042815853e-05,
"loss": 2.3553,
"step": 1550
},
{
"epoch": 0.01,
"grad_norm": 0.8746365308761597,
"learning_rate": 9.996740682913668e-05,
"loss": 2.3407,
"step": 1555
},
{
"epoch": 0.01,
"grad_norm": 0.977430522441864,
"learning_rate": 9.996718246074591e-05,
"loss": 2.3651,
"step": 1560
},
{
"epoch": 0.01,
"grad_norm": 0.839228093624115,
"learning_rate": 9.99669573229897e-05,
"loss": 2.3566,
"step": 1565
},
{
"epoch": 0.01,
"grad_norm": 0.930707573890686,
"learning_rate": 9.996673141587151e-05,
"loss": 2.47,
"step": 1570
},
{
"epoch": 0.01,
"grad_norm": 0.8668906092643738,
"learning_rate": 9.996650473939482e-05,
"loss": 2.4418,
"step": 1575
},
{
"epoch": 0.01,
"grad_norm": 0.8575398921966553,
"learning_rate": 9.996627729356314e-05,
"loss": 2.4174,
"step": 1580
},
{
"epoch": 0.01,
"grad_norm": 0.8358970880508423,
"learning_rate": 9.996604907837993e-05,
"loss": 2.4093,
"step": 1585
},
{
"epoch": 0.01,
"grad_norm": 0.8708803653717041,
"learning_rate": 9.996582009384874e-05,
"loss": 2.4766,
"step": 1590
},
{
"epoch": 0.01,
"grad_norm": 0.8505133986473083,
"learning_rate": 9.996559033997307e-05,
"loss": 2.4108,
"step": 1595
},
{
"epoch": 0.01,
"grad_norm": 0.8573861122131348,
"learning_rate": 9.996535981675648e-05,
"loss": 2.4504,
"step": 1600
},
{
"epoch": 0.01,
"grad_norm": 0.8563638925552368,
"learning_rate": 9.996512852420251e-05,
"loss": 2.3746,
"step": 1605
},
{
"epoch": 0.01,
"grad_norm": 0.8990402817726135,
"learning_rate": 9.996489646231471e-05,
"loss": 2.4193,
"step": 1610
},
{
"epoch": 0.01,
"grad_norm": 0.8627756834030151,
"learning_rate": 9.996466363109667e-05,
"loss": 2.3711,
"step": 1615
},
{
"epoch": 0.01,
"grad_norm": 0.8257574439048767,
"learning_rate": 9.996443003055196e-05,
"loss": 2.3728,
"step": 1620
},
{
"epoch": 0.01,
"grad_norm": 0.8396132588386536,
"learning_rate": 9.99641956606842e-05,
"loss": 2.3601,
"step": 1625
},
{
"epoch": 0.01,
"grad_norm": 0.8632432818412781,
"learning_rate": 9.996396052149698e-05,
"loss": 2.3399,
"step": 1630
},
{
"epoch": 0.01,
"grad_norm": 0.8119171261787415,
"learning_rate": 9.996372461299392e-05,
"loss": 2.3117,
"step": 1635
},
{
"epoch": 0.01,
"grad_norm": 0.8263778686523438,
"learning_rate": 9.996348793517866e-05,
"loss": 2.3892,
"step": 1640
},
{
"epoch": 0.01,
"grad_norm": 0.8197963237762451,
"learning_rate": 9.996325048805484e-05,
"loss": 2.3052,
"step": 1645
},
{
"epoch": 0.01,
"grad_norm": 0.8172884583473206,
"learning_rate": 9.996301227162612e-05,
"loss": 2.3004,
"step": 1650
},
{
"epoch": 0.01,
"grad_norm": 0.8418741822242737,
"learning_rate": 9.996277328589618e-05,
"loss": 2.3195,
"step": 1655
},
{
"epoch": 0.01,
"grad_norm": 0.830545961856842,
"learning_rate": 9.996253353086866e-05,
"loss": 2.3688,
"step": 1660
},
{
"epoch": 0.01,
"grad_norm": 0.7909248471260071,
"learning_rate": 9.996229300654729e-05,
"loss": 2.3493,
"step": 1665
},
{
"epoch": 0.01,
"grad_norm": 0.923248827457428,
"learning_rate": 9.996205171293576e-05,
"loss": 2.432,
"step": 1670
},
{
"epoch": 0.01,
"grad_norm": 0.8337894678115845,
"learning_rate": 9.996180965003778e-05,
"loss": 2.4622,
"step": 1675
},
{
"epoch": 0.01,
"grad_norm": 0.9023234844207764,
"learning_rate": 9.996156681785708e-05,
"loss": 2.4802,
"step": 1680
},
{
"epoch": 0.01,
"grad_norm": 0.8517597317695618,
"learning_rate": 9.996132321639741e-05,
"loss": 2.4099,
"step": 1685
},
{
"epoch": 0.01,
"grad_norm": 0.8151902556419373,
"learning_rate": 9.996107884566252e-05,
"loss": 2.3,
"step": 1690
},
{
"epoch": 0.01,
"grad_norm": 0.8822351694107056,
"learning_rate": 9.996083370565616e-05,
"loss": 2.3961,
"step": 1695
},
{
"epoch": 0.01,
"grad_norm": 0.8180464506149292,
"learning_rate": 9.99605877963821e-05,
"loss": 2.4271,
"step": 1700
},
{
"epoch": 0.01,
"grad_norm": 0.8311895728111267,
"learning_rate": 9.996034111784413e-05,
"loss": 2.3834,
"step": 1705
},
{
"epoch": 0.01,
"grad_norm": 0.849658727645874,
"learning_rate": 9.996009367004607e-05,
"loss": 2.3185,
"step": 1710
},
{
"epoch": 0.01,
"grad_norm": 0.8437408804893494,
"learning_rate": 9.995984545299171e-05,
"loss": 2.4068,
"step": 1715
},
{
"epoch": 0.01,
"grad_norm": 0.8224195241928101,
"learning_rate": 9.995959646668488e-05,
"loss": 2.3665,
"step": 1720
},
{
"epoch": 0.01,
"grad_norm": 0.8556229472160339,
"learning_rate": 9.995934671112941e-05,
"loss": 2.3838,
"step": 1725
},
{
"epoch": 0.01,
"grad_norm": 0.8254923224449158,
"learning_rate": 9.995909618632915e-05,
"loss": 2.3534,
"step": 1730
},
{
"epoch": 0.01,
"grad_norm": 0.8680881857872009,
"learning_rate": 9.995884489228794e-05,
"loss": 2.3599,
"step": 1735
},
{
"epoch": 0.01,
"grad_norm": 0.8698033690452576,
"learning_rate": 9.995859282900967e-05,
"loss": 2.3728,
"step": 1740
},
{
"epoch": 0.01,
"grad_norm": 0.8474555611610413,
"learning_rate": 9.995833999649821e-05,
"loss": 2.3038,
"step": 1745
},
{
"epoch": 0.01,
"grad_norm": 0.8396309614181519,
"learning_rate": 9.995808639475746e-05,
"loss": 2.3818,
"step": 1750
},
{
"epoch": 0.01,
"grad_norm": 0.9070517420768738,
"learning_rate": 9.995783202379131e-05,
"loss": 2.4223,
"step": 1755
},
{
"epoch": 0.01,
"grad_norm": 0.867617130279541,
"learning_rate": 9.99575768836037e-05,
"loss": 2.4105,
"step": 1760
},
{
"epoch": 0.01,
"grad_norm": 0.8572276830673218,
"learning_rate": 9.995732097419855e-05,
"loss": 2.3065,
"step": 1765
},
{
"epoch": 0.01,
"grad_norm": 0.8537548184394836,
"learning_rate": 9.995706429557978e-05,
"loss": 2.4077,
"step": 1770
},
{
"epoch": 0.01,
"grad_norm": 0.8255508542060852,
"learning_rate": 9.995680684775137e-05,
"loss": 2.4587,
"step": 1775
},
{
"epoch": 0.01,
"grad_norm": 0.870722770690918,
"learning_rate": 9.995654863071729e-05,
"loss": 2.2974,
"step": 1780
},
{
"epoch": 0.01,
"grad_norm": 0.8325221538543701,
"learning_rate": 9.995628964448147e-05,
"loss": 2.2874,
"step": 1785
},
{
"epoch": 0.01,
"grad_norm": 0.8294020891189575,
"learning_rate": 9.995602988904793e-05,
"loss": 2.4584,
"step": 1790
},
{
"epoch": 0.01,
"grad_norm": 0.8577289581298828,
"learning_rate": 9.995576936442068e-05,
"loss": 2.3433,
"step": 1795
},
{
"epoch": 0.01,
"grad_norm": 0.8739635348320007,
"learning_rate": 9.995550807060371e-05,
"loss": 2.3547,
"step": 1800
},
{
"epoch": 0.01,
"grad_norm": 0.8251529335975647,
"learning_rate": 9.995524600760105e-05,
"loss": 2.354,
"step": 1805
},
{
"epoch": 0.01,
"grad_norm": 0.8267414569854736,
"learning_rate": 9.995498317541674e-05,
"loss": 2.3868,
"step": 1810
},
{
"epoch": 0.01,
"grad_norm": 0.8029406070709229,
"learning_rate": 9.995471957405483e-05,
"loss": 2.3988,
"step": 1815
},
{
"epoch": 0.01,
"grad_norm": 0.8370579481124878,
"learning_rate": 9.995445520351935e-05,
"loss": 2.316,
"step": 1820
},
{
"epoch": 0.01,
"grad_norm": 0.8098414540290833,
"learning_rate": 9.995419006381442e-05,
"loss": 2.3879,
"step": 1825
},
{
"epoch": 0.01,
"grad_norm": 0.7726863026618958,
"learning_rate": 9.995392415494407e-05,
"loss": 2.4128,
"step": 1830
},
{
"epoch": 0.01,
"grad_norm": 0.8195619583129883,
"learning_rate": 9.995365747691243e-05,
"loss": 2.3588,
"step": 1835
},
{
"epoch": 0.01,
"grad_norm": 0.8154809474945068,
"learning_rate": 9.995339002972359e-05,
"loss": 2.3515,
"step": 1840
},
{
"epoch": 0.01,
"grad_norm": 0.8370392322540283,
"learning_rate": 9.995312181338168e-05,
"loss": 2.3805,
"step": 1845
},
{
"epoch": 0.01,
"grad_norm": 0.8787909150123596,
"learning_rate": 9.995285282789082e-05,
"loss": 2.3736,
"step": 1850
},
{
"epoch": 0.01,
"grad_norm": 0.8334288597106934,
"learning_rate": 9.995258307325515e-05,
"loss": 2.3737,
"step": 1855
},
{
"epoch": 0.01,
"grad_norm": 0.8454689979553223,
"learning_rate": 9.995231254947883e-05,
"loss": 2.366,
"step": 1860
},
{
"epoch": 0.01,
"grad_norm": 0.852279007434845,
"learning_rate": 9.995204125656603e-05,
"loss": 2.4119,
"step": 1865
},
{
"epoch": 0.01,
"grad_norm": 0.838905930519104,
"learning_rate": 9.995176919452092e-05,
"loss": 2.2984,
"step": 1870
},
{
"epoch": 0.01,
"grad_norm": 0.8040510416030884,
"learning_rate": 9.995149636334769e-05,
"loss": 2.411,
"step": 1875
},
{
"epoch": 0.01,
"grad_norm": 0.8262283802032471,
"learning_rate": 9.995122276305054e-05,
"loss": 2.4327,
"step": 1880
},
{
"epoch": 0.01,
"grad_norm": 0.7874197363853455,
"learning_rate": 9.995094839363368e-05,
"loss": 2.3131,
"step": 1885
},
{
"epoch": 0.01,
"grad_norm": 0.8189416527748108,
"learning_rate": 9.995067325510132e-05,
"loss": 2.3871,
"step": 1890
},
{
"epoch": 0.01,
"grad_norm": 0.8729049563407898,
"learning_rate": 9.995039734745774e-05,
"loss": 2.4426,
"step": 1895
},
{
"epoch": 0.01,
"grad_norm": 0.8288578987121582,
"learning_rate": 9.995012067070715e-05,
"loss": 2.3394,
"step": 1900
},
{
"epoch": 0.02,
"grad_norm": 0.8030306696891785,
"learning_rate": 9.994984322485383e-05,
"loss": 2.3682,
"step": 1905
},
{
"epoch": 0.02,
"grad_norm": 0.8330934047698975,
"learning_rate": 9.994956500990205e-05,
"loss": 2.2872,
"step": 1910
},
{
"epoch": 0.02,
"grad_norm": 0.8806706666946411,
"learning_rate": 9.994928602585608e-05,
"loss": 2.3558,
"step": 1915
},
{
"epoch": 0.02,
"grad_norm": 0.8309563994407654,
"learning_rate": 9.994900627272023e-05,
"loss": 2.4081,
"step": 1920
},
{
"epoch": 0.02,
"grad_norm": 0.791577160358429,
"learning_rate": 9.99487257504988e-05,
"loss": 2.3026,
"step": 1925
},
{
"epoch": 0.02,
"grad_norm": 0.8165927529335022,
"learning_rate": 9.994844445919612e-05,
"loss": 2.3803,
"step": 1930
},
{
"epoch": 0.02,
"grad_norm": 0.8097928762435913,
"learning_rate": 9.994816239881648e-05,
"loss": 2.3904,
"step": 1935
},
{
"epoch": 0.02,
"grad_norm": 0.8042891621589661,
"learning_rate": 9.99478795693643e-05,
"loss": 2.3482,
"step": 1940
},
{
"epoch": 0.02,
"grad_norm": 0.807779848575592,
"learning_rate": 9.994759597084384e-05,
"loss": 2.4438,
"step": 1945
},
{
"epoch": 0.02,
"grad_norm": 0.8150476813316345,
"learning_rate": 9.994731160325955e-05,
"loss": 2.3489,
"step": 1950
},
{
"epoch": 0.02,
"grad_norm": 0.8188815116882324,
"learning_rate": 9.994702646661577e-05,
"loss": 2.3754,
"step": 1955
},
{
"epoch": 0.02,
"grad_norm": 0.841265082359314,
"learning_rate": 9.994674056091689e-05,
"loss": 2.301,
"step": 1960
},
{
"epoch": 0.02,
"grad_norm": 0.8434075713157654,
"learning_rate": 9.994645388616732e-05,
"loss": 2.3622,
"step": 1965
},
{
"epoch": 0.02,
"grad_norm": 0.7885850667953491,
"learning_rate": 9.994616644237148e-05,
"loss": 2.34,
"step": 1970
},
{
"epoch": 0.02,
"grad_norm": 0.8558570742607117,
"learning_rate": 9.994587822953378e-05,
"loss": 2.4269,
"step": 1975
},
{
"epoch": 0.02,
"grad_norm": 0.8093646168708801,
"learning_rate": 9.994558924765866e-05,
"loss": 2.3819,
"step": 1980
},
{
"epoch": 0.02,
"grad_norm": 0.7693427801132202,
"learning_rate": 9.99452994967506e-05,
"loss": 2.3401,
"step": 1985
},
{
"epoch": 0.02,
"grad_norm": 0.8213228583335876,
"learning_rate": 9.994500897681401e-05,
"loss": 2.2366,
"step": 1990
},
{
"epoch": 0.02,
"grad_norm": 0.7763933539390564,
"learning_rate": 9.994471768785339e-05,
"loss": 2.3407,
"step": 1995
},
{
"epoch": 0.02,
"grad_norm": 0.8061359524726868,
"learning_rate": 9.994442562987324e-05,
"loss": 2.3409,
"step": 2000
},
{
"epoch": 0.02,
"grad_norm": 0.8034436106681824,
"learning_rate": 9.994413280287803e-05,
"loss": 2.3132,
"step": 2005
},
{
"epoch": 0.02,
"grad_norm": 0.8211275339126587,
"learning_rate": 9.994383920687229e-05,
"loss": 2.4276,
"step": 2010
},
{
"epoch": 0.02,
"grad_norm": 0.8341994881629944,
"learning_rate": 9.994354484186053e-05,
"loss": 2.4719,
"step": 2015
},
{
"epoch": 0.02,
"grad_norm": 0.7997161149978638,
"learning_rate": 9.994324970784727e-05,
"loss": 2.3593,
"step": 2020
},
{
"epoch": 0.02,
"grad_norm": 0.8052873015403748,
"learning_rate": 9.994295380483709e-05,
"loss": 2.396,
"step": 2025
},
{
"epoch": 0.02,
"grad_norm": 0.7901674509048462,
"learning_rate": 9.99426571328345e-05,
"loss": 2.3404,
"step": 2030
},
{
"epoch": 0.02,
"grad_norm": 0.8188238739967346,
"learning_rate": 9.994235969184411e-05,
"loss": 2.3232,
"step": 2035
},
{
"epoch": 0.02,
"grad_norm": 0.8371001482009888,
"learning_rate": 9.994206148187046e-05,
"loss": 2.3469,
"step": 2040
},
{
"epoch": 0.02,
"grad_norm": 0.8383309245109558,
"learning_rate": 9.994176250291818e-05,
"loss": 2.3914,
"step": 2045
},
{
"epoch": 0.02,
"grad_norm": 0.8136764764785767,
"learning_rate": 9.994146275499186e-05,
"loss": 2.4021,
"step": 2050
},
{
"epoch": 0.02,
"grad_norm": 0.8402037620544434,
"learning_rate": 9.994116223809611e-05,
"loss": 2.3279,
"step": 2055
},
{
"epoch": 0.02,
"grad_norm": 0.8816250562667847,
"learning_rate": 9.994086095223557e-05,
"loss": 2.3664,
"step": 2060
},
{
"epoch": 0.02,
"grad_norm": 0.864163339138031,
"learning_rate": 9.994055889741485e-05,
"loss": 2.3721,
"step": 2065
},
{
"epoch": 0.02,
"grad_norm": 0.8118492960929871,
"learning_rate": 9.994025607363862e-05,
"loss": 2.2617,
"step": 2070
},
{
"epoch": 0.02,
"grad_norm": 0.8354922533035278,
"learning_rate": 9.993995248091155e-05,
"loss": 2.2772,
"step": 2075
},
{
"epoch": 0.02,
"grad_norm": 0.8222642540931702,
"learning_rate": 9.993964811923831e-05,
"loss": 2.2767,
"step": 2080
},
{
"epoch": 0.02,
"grad_norm": 0.8496009111404419,
"learning_rate": 9.993934298862358e-05,
"loss": 2.4048,
"step": 2085
},
{
"epoch": 0.02,
"grad_norm": 0.8409602046012878,
"learning_rate": 9.993903708907205e-05,
"loss": 2.3102,
"step": 2090
},
{
"epoch": 0.02,
"grad_norm": 0.7944759726524353,
"learning_rate": 9.993873042058845e-05,
"loss": 2.3983,
"step": 2095
},
{
"epoch": 0.02,
"grad_norm": 0.7996658682823181,
"learning_rate": 9.993842298317749e-05,
"loss": 2.3538,
"step": 2100
},
{
"epoch": 0.02,
"grad_norm": 0.8359138369560242,
"learning_rate": 9.993811477684392e-05,
"loss": 2.4723,
"step": 2105
},
{
"epoch": 0.02,
"grad_norm": 0.8169530034065247,
"learning_rate": 9.993780580159246e-05,
"loss": 2.3283,
"step": 2110
},
{
"epoch": 0.02,
"grad_norm": 0.8531467318534851,
"learning_rate": 9.993749605742788e-05,
"loss": 2.3247,
"step": 2115
},
{
"epoch": 0.02,
"grad_norm": 0.8172628283500671,
"learning_rate": 9.993718554435495e-05,
"loss": 2.3176,
"step": 2120
},
{
"epoch": 0.02,
"grad_norm": 0.800510585308075,
"learning_rate": 9.993687426237846e-05,
"loss": 2.3249,
"step": 2125
},
{
"epoch": 0.02,
"grad_norm": 0.8383053541183472,
"learning_rate": 9.993656221150318e-05,
"loss": 2.3961,
"step": 2130
},
{
"epoch": 0.02,
"grad_norm": 0.8201522827148438,
"learning_rate": 9.993624939173394e-05,
"loss": 2.4808,
"step": 2135
},
{
"epoch": 0.02,
"grad_norm": 0.7858189940452576,
"learning_rate": 9.993593580307553e-05,
"loss": 2.2732,
"step": 2140
},
{
"epoch": 0.02,
"grad_norm": 0.8177255988121033,
"learning_rate": 9.99356214455328e-05,
"loss": 2.3332,
"step": 2145
},
{
"epoch": 0.02,
"grad_norm": 0.8163393139839172,
"learning_rate": 9.993530631911059e-05,
"loss": 2.3076,
"step": 2150
},
{
"epoch": 0.02,
"grad_norm": 0.7773647308349609,
"learning_rate": 9.993499042381373e-05,
"loss": 2.336,
"step": 2155
},
{
"epoch": 0.02,
"grad_norm": 0.7859750390052795,
"learning_rate": 9.993467375964711e-05,
"loss": 2.3515,
"step": 2160
},
{
"epoch": 0.02,
"grad_norm": 0.81662917137146,
"learning_rate": 9.99343563266156e-05,
"loss": 2.2917,
"step": 2165
},
{
"epoch": 0.02,
"grad_norm": 0.771237850189209,
"learning_rate": 9.993403812472408e-05,
"loss": 2.3551,
"step": 2170
},
{
"epoch": 0.02,
"grad_norm": 0.7982082366943359,
"learning_rate": 9.993371915397745e-05,
"loss": 2.3385,
"step": 2175
},
{
"epoch": 0.02,
"grad_norm": 0.8058829307556152,
"learning_rate": 9.993339941438062e-05,
"loss": 2.2757,
"step": 2180
},
{
"epoch": 0.02,
"grad_norm": 0.799950122833252,
"learning_rate": 9.993307890593853e-05,
"loss": 2.328,
"step": 2185
},
{
"epoch": 0.02,
"grad_norm": 0.782180666923523,
"learning_rate": 9.993275762865609e-05,
"loss": 2.2733,
"step": 2190
},
{
"epoch": 0.02,
"grad_norm": 0.7389876842498779,
"learning_rate": 9.993243558253826e-05,
"loss": 2.3271,
"step": 2195
},
{
"epoch": 0.02,
"grad_norm": 0.7932037711143494,
"learning_rate": 9.993211276759e-05,
"loss": 2.4068,
"step": 2200
},
{
"epoch": 0.02,
"grad_norm": 0.8344466090202332,
"learning_rate": 9.993178918381628e-05,
"loss": 2.4415,
"step": 2205
},
{
"epoch": 0.02,
"grad_norm": 0.9178872108459473,
"learning_rate": 9.993146483122207e-05,
"loss": 2.3783,
"step": 2210
},
{
"epoch": 0.02,
"grad_norm": 0.8039920926094055,
"learning_rate": 9.993113970981238e-05,
"loss": 2.3165,
"step": 2215
},
{
"epoch": 0.02,
"grad_norm": 0.8377980589866638,
"learning_rate": 9.993081381959221e-05,
"loss": 2.3917,
"step": 2220
},
{
"epoch": 0.02,
"grad_norm": 0.8306506276130676,
"learning_rate": 9.99304871605666e-05,
"loss": 2.3285,
"step": 2225
},
{
"epoch": 0.02,
"grad_norm": 0.8235587477684021,
"learning_rate": 9.993015973274053e-05,
"loss": 2.4039,
"step": 2230
},
{
"epoch": 0.02,
"grad_norm": 0.8542500138282776,
"learning_rate": 9.992983153611908e-05,
"loss": 2.3191,
"step": 2235
},
{
"epoch": 0.02,
"grad_norm": 0.833791971206665,
"learning_rate": 9.99295025707073e-05,
"loss": 2.3775,
"step": 2240
},
{
"epoch": 0.02,
"grad_norm": 0.8063716292381287,
"learning_rate": 9.992917283651024e-05,
"loss": 2.2759,
"step": 2245
},
{
"epoch": 0.02,
"grad_norm": 0.8285427093505859,
"learning_rate": 9.992884233353298e-05,
"loss": 2.2706,
"step": 2250
},
{
"epoch": 0.02,
"grad_norm": 0.7727683782577515,
"learning_rate": 9.992851106178063e-05,
"loss": 2.3248,
"step": 2255
},
{
"epoch": 0.02,
"grad_norm": 0.7583116292953491,
"learning_rate": 9.992817902125826e-05,
"loss": 2.2826,
"step": 2260
},
{
"epoch": 0.02,
"grad_norm": 0.7820054292678833,
"learning_rate": 9.9927846211971e-05,
"loss": 2.3142,
"step": 2265
},
{
"epoch": 0.02,
"grad_norm": 0.814536452293396,
"learning_rate": 9.992751263392397e-05,
"loss": 2.3259,
"step": 2270
},
{
"epoch": 0.02,
"grad_norm": 0.8349671363830566,
"learning_rate": 9.992717828712232e-05,
"loss": 2.2862,
"step": 2275
},
{
"epoch": 0.02,
"grad_norm": 0.8213352560997009,
"learning_rate": 9.992684317157117e-05,
"loss": 2.3648,
"step": 2280
},
{
"epoch": 0.02,
"grad_norm": 0.8106248378753662,
"learning_rate": 9.992650728727571e-05,
"loss": 2.3491,
"step": 2285
},
{
"epoch": 0.02,
"grad_norm": 0.8015328645706177,
"learning_rate": 9.99261706342411e-05,
"loss": 2.457,
"step": 2290
},
{
"epoch": 0.02,
"grad_norm": 0.8064042329788208,
"learning_rate": 9.99258332124725e-05,
"loss": 2.3696,
"step": 2295
},
{
"epoch": 0.02,
"grad_norm": 0.7859455347061157,
"learning_rate": 9.992549502197514e-05,
"loss": 2.2891,
"step": 2300
},
{
"epoch": 0.02,
"grad_norm": 0.7974657416343689,
"learning_rate": 9.992515606275423e-05,
"loss": 2.3534,
"step": 2305
},
{
"epoch": 0.02,
"grad_norm": 0.8065370321273804,
"learning_rate": 9.992481633481495e-05,
"loss": 2.3398,
"step": 2310
},
{
"epoch": 0.02,
"grad_norm": 0.801386296749115,
"learning_rate": 9.992447583816257e-05,
"loss": 2.3518,
"step": 2315
},
{
"epoch": 0.02,
"grad_norm": 0.8371037244796753,
"learning_rate": 9.992413457280231e-05,
"loss": 2.3655,
"step": 2320
},
{
"epoch": 0.02,
"grad_norm": 0.7933724522590637,
"learning_rate": 9.992379253873943e-05,
"loss": 2.2849,
"step": 2325
},
{
"epoch": 0.02,
"grad_norm": 0.8001721501350403,
"learning_rate": 9.99234497359792e-05,
"loss": 2.309,
"step": 2330
},
{
"epoch": 0.02,
"grad_norm": 0.8331447839736938,
"learning_rate": 9.992310616452689e-05,
"loss": 2.3225,
"step": 2335
},
{
"epoch": 0.02,
"grad_norm": 0.8447306752204895,
"learning_rate": 9.99227618243878e-05,
"loss": 2.4171,
"step": 2340
},
{
"epoch": 0.02,
"grad_norm": 0.806386411190033,
"learning_rate": 9.992241671556724e-05,
"loss": 2.3152,
"step": 2345
},
{
"epoch": 0.02,
"grad_norm": 0.815006673336029,
"learning_rate": 9.99220708380705e-05,
"loss": 2.2347,
"step": 2350
},
{
"epoch": 0.02,
"grad_norm": 0.8030174970626831,
"learning_rate": 9.99217241919029e-05,
"loss": 2.3719,
"step": 2355
},
{
"epoch": 0.02,
"grad_norm": 0.830703616142273,
"learning_rate": 9.992137677706982e-05,
"loss": 2.2908,
"step": 2360
},
{
"epoch": 0.02,
"grad_norm": 0.7835960388183594,
"learning_rate": 9.99210285935766e-05,
"loss": 2.364,
"step": 2365
},
{
"epoch": 0.02,
"grad_norm": 0.787173330783844,
"learning_rate": 9.992067964142856e-05,
"loss": 2.3161,
"step": 2370
},
{
"epoch": 0.02,
"grad_norm": 0.8124563097953796,
"learning_rate": 9.992032992063109e-05,
"loss": 2.2959,
"step": 2375
},
{
"epoch": 0.02,
"grad_norm": 0.8323636651039124,
"learning_rate": 9.99199794311896e-05,
"loss": 2.3037,
"step": 2380
},
{
"epoch": 0.02,
"grad_norm": 0.804873526096344,
"learning_rate": 9.991962817310947e-05,
"loss": 2.3198,
"step": 2385
},
{
"epoch": 0.02,
"grad_norm": 0.7786262631416321,
"learning_rate": 9.991927614639612e-05,
"loss": 2.2223,
"step": 2390
},
{
"epoch": 0.02,
"grad_norm": 0.8154786229133606,
"learning_rate": 9.991892335105493e-05,
"loss": 2.3512,
"step": 2395
},
{
"epoch": 0.02,
"grad_norm": 0.7992445826530457,
"learning_rate": 9.991856978709137e-05,
"loss": 2.3167,
"step": 2400
},
{
"epoch": 0.02,
"grad_norm": 0.7551295757293701,
"learning_rate": 9.99182154545109e-05,
"loss": 2.2938,
"step": 2405
},
{
"epoch": 0.02,
"grad_norm": 0.8158405423164368,
"learning_rate": 9.991786035331893e-05,
"loss": 2.303,
"step": 2410
},
{
"epoch": 0.02,
"grad_norm": 0.8352602124214172,
"learning_rate": 9.991750448352096e-05,
"loss": 2.321,
"step": 2415
},
{
"epoch": 0.02,
"grad_norm": 0.77967369556427,
"learning_rate": 9.991714784512243e-05,
"loss": 2.2544,
"step": 2420
},
{
"epoch": 0.02,
"grad_norm": 0.7667365074157715,
"learning_rate": 9.991679043812887e-05,
"loss": 2.2496,
"step": 2425
},
{
"epoch": 0.02,
"grad_norm": 0.7527759075164795,
"learning_rate": 9.99164322625458e-05,
"loss": 2.2589,
"step": 2430
},
{
"epoch": 0.02,
"grad_norm": 0.826384961605072,
"learning_rate": 9.991607331837867e-05,
"loss": 2.2631,
"step": 2435
},
{
"epoch": 0.02,
"grad_norm": 0.7508112192153931,
"learning_rate": 9.991571360563306e-05,
"loss": 2.3196,
"step": 2440
},
{
"epoch": 0.02,
"grad_norm": 0.7706888914108276,
"learning_rate": 9.99153531243145e-05,
"loss": 2.287,
"step": 2445
},
{
"epoch": 0.02,
"grad_norm": 0.77959805727005,
"learning_rate": 9.991499187442852e-05,
"loss": 2.2557,
"step": 2450
},
{
"epoch": 0.02,
"grad_norm": 0.7784958481788635,
"learning_rate": 9.99146298559807e-05,
"loss": 2.3505,
"step": 2455
},
{
"epoch": 0.02,
"grad_norm": 0.7773137092590332,
"learning_rate": 9.991426706897661e-05,
"loss": 2.2434,
"step": 2460
},
{
"epoch": 0.02,
"grad_norm": 0.810344398021698,
"learning_rate": 9.991390351342185e-05,
"loss": 2.3382,
"step": 2465
},
{
"epoch": 0.02,
"grad_norm": 0.7817177772521973,
"learning_rate": 9.991353918932198e-05,
"loss": 2.2682,
"step": 2470
},
{
"epoch": 0.02,
"grad_norm": 0.752217173576355,
"learning_rate": 9.991317409668264e-05,
"loss": 2.1936,
"step": 2475
},
{
"epoch": 0.02,
"grad_norm": 0.8046365976333618,
"learning_rate": 9.991280823550945e-05,
"loss": 2.3546,
"step": 2480
},
{
"epoch": 0.02,
"grad_norm": 0.8171698451042175,
"learning_rate": 9.991244160580803e-05,
"loss": 2.3222,
"step": 2485
},
{
"epoch": 0.02,
"grad_norm": 0.7982907295227051,
"learning_rate": 9.991207420758404e-05,
"loss": 2.3071,
"step": 2490
},
{
"epoch": 0.02,
"grad_norm": 0.848210334777832,
"learning_rate": 9.991170604084314e-05,
"loss": 2.3529,
"step": 2495
},
{
"epoch": 0.02,
"grad_norm": 0.8181712031364441,
"learning_rate": 9.991133710559097e-05,
"loss": 2.3902,
"step": 2500
},
{
"epoch": 0.02,
"grad_norm": 0.8075858950614929,
"learning_rate": 9.991096740183324e-05,
"loss": 2.3688,
"step": 2505
},
{
"epoch": 0.02,
"grad_norm": 0.7965590953826904,
"learning_rate": 9.991059692957563e-05,
"loss": 2.3312,
"step": 2510
},
{
"epoch": 0.02,
"grad_norm": 0.7993034720420837,
"learning_rate": 9.991022568882384e-05,
"loss": 2.3073,
"step": 2515
},
{
"epoch": 0.02,
"grad_norm": 0.7413946390151978,
"learning_rate": 9.99098536795836e-05,
"loss": 2.2938,
"step": 2520
},
{
"epoch": 0.02,
"grad_norm": 0.7868495583534241,
"learning_rate": 9.990948090186063e-05,
"loss": 2.3277,
"step": 2525
},
{
"epoch": 0.02,
"grad_norm": 0.8130128383636475,
"learning_rate": 9.990910735566065e-05,
"loss": 2.35,
"step": 2530
},
{
"epoch": 0.02,
"grad_norm": 0.844701886177063,
"learning_rate": 9.990873304098946e-05,
"loss": 2.3395,
"step": 2535
},
{
"epoch": 0.02,
"grad_norm": 0.8003630638122559,
"learning_rate": 9.990835795785278e-05,
"loss": 2.3277,
"step": 2540
},
{
"epoch": 0.02,
"grad_norm": 0.825343668460846,
"learning_rate": 9.990798210625642e-05,
"loss": 2.3812,
"step": 2545
},
{
"epoch": 0.02,
"grad_norm": 0.9025837182998657,
"learning_rate": 9.990760548620613e-05,
"loss": 2.3598,
"step": 2550
},
{
"epoch": 0.02,
"grad_norm": 1.0060158967971802,
"learning_rate": 9.990722809770773e-05,
"loss": 2.3781,
"step": 2555
},
{
"epoch": 0.02,
"grad_norm": 0.7704659700393677,
"learning_rate": 9.990684994076702e-05,
"loss": 2.2909,
"step": 2560
},
{
"epoch": 0.02,
"grad_norm": 0.7716596722602844,
"learning_rate": 9.990647101538983e-05,
"loss": 2.2676,
"step": 2565
},
{
"epoch": 0.02,
"grad_norm": 0.778334379196167,
"learning_rate": 9.990609132158201e-05,
"loss": 2.373,
"step": 2570
},
{
"epoch": 0.02,
"grad_norm": 0.7821041941642761,
"learning_rate": 9.990571085934938e-05,
"loss": 2.3273,
"step": 2575
},
{
"epoch": 0.02,
"grad_norm": 0.8228597640991211,
"learning_rate": 9.990532962869783e-05,
"loss": 2.24,
"step": 2580
},
{
"epoch": 0.02,
"grad_norm": 0.7717438340187073,
"learning_rate": 9.990494762963318e-05,
"loss": 2.2995,
"step": 2585
},
{
"epoch": 0.02,
"grad_norm": 0.8166317939758301,
"learning_rate": 9.990456486216136e-05,
"loss": 2.3476,
"step": 2590
},
{
"epoch": 0.02,
"grad_norm": 0.7950439453125,
"learning_rate": 9.990418132628824e-05,
"loss": 2.3631,
"step": 2595
},
{
"epoch": 0.02,
"grad_norm": 0.7446428537368774,
"learning_rate": 9.990379702201974e-05,
"loss": 2.2814,
"step": 2600
},
{
"epoch": 0.02,
"grad_norm": 0.7886162400245667,
"learning_rate": 9.990341194936173e-05,
"loss": 2.2883,
"step": 2605
},
{
"epoch": 0.02,
"grad_norm": 0.7807297110557556,
"learning_rate": 9.990302610832021e-05,
"loss": 2.2614,
"step": 2610
},
{
"epoch": 0.02,
"grad_norm": 1.103926420211792,
"learning_rate": 9.990263949890108e-05,
"loss": 2.3401,
"step": 2615
},
{
"epoch": 0.02,
"grad_norm": 0.7823458909988403,
"learning_rate": 9.990225212111029e-05,
"loss": 2.2377,
"step": 2620
},
{
"epoch": 0.02,
"grad_norm": 0.7610365748405457,
"learning_rate": 9.990186397495383e-05,
"loss": 2.347,
"step": 2625
},
{
"epoch": 0.02,
"grad_norm": 0.8271576166152954,
"learning_rate": 9.990147506043766e-05,
"loss": 2.3774,
"step": 2630
},
{
"epoch": 0.02,
"grad_norm": 0.8160341382026672,
"learning_rate": 9.990108537756775e-05,
"loss": 2.3621,
"step": 2635
},
{
"epoch": 0.02,
"grad_norm": 0.804689347743988,
"learning_rate": 9.990069492635015e-05,
"loss": 2.3145,
"step": 2640
},
{
"epoch": 0.02,
"grad_norm": 0.7709464430809021,
"learning_rate": 9.990030370679082e-05,
"loss": 2.2497,
"step": 2645
},
{
"epoch": 0.02,
"grad_norm": 0.824783444404602,
"learning_rate": 9.989991171889581e-05,
"loss": 2.3339,
"step": 2650
},
{
"epoch": 0.02,
"grad_norm": 0.7638097405433655,
"learning_rate": 9.989951896267115e-05,
"loss": 2.2861,
"step": 2655
},
{
"epoch": 0.02,
"grad_norm": 0.7840123176574707,
"learning_rate": 9.989912543812289e-05,
"loss": 2.3274,
"step": 2660
},
{
"epoch": 0.02,
"grad_norm": 0.811965823173523,
"learning_rate": 9.989873114525709e-05,
"loss": 2.2741,
"step": 2665
},
{
"epoch": 0.02,
"grad_norm": 0.7927914261817932,
"learning_rate": 9.989833608407981e-05,
"loss": 2.3365,
"step": 2670
},
{
"epoch": 0.02,
"grad_norm": 0.7686588764190674,
"learning_rate": 9.989794025459715e-05,
"loss": 2.2705,
"step": 2675
},
{
"epoch": 0.02,
"grad_norm": 0.7874404788017273,
"learning_rate": 9.989754365681519e-05,
"loss": 2.3382,
"step": 2680
},
{
"epoch": 0.02,
"grad_norm": 0.7749989628791809,
"learning_rate": 9.989714629074006e-05,
"loss": 2.2996,
"step": 2685
},
{
"epoch": 0.02,
"grad_norm": 0.7693790197372437,
"learning_rate": 9.989674815637784e-05,
"loss": 2.3346,
"step": 2690
},
{
"epoch": 0.02,
"grad_norm": 0.7790564894676208,
"learning_rate": 9.989634925373468e-05,
"loss": 2.2063,
"step": 2695
},
{
"epoch": 0.02,
"grad_norm": 0.7817680239677429,
"learning_rate": 9.989594958281673e-05,
"loss": 2.3381,
"step": 2700
},
{
"epoch": 0.02,
"grad_norm": 0.7555989623069763,
"learning_rate": 9.989554914363015e-05,
"loss": 2.2745,
"step": 2705
},
{
"epoch": 0.02,
"grad_norm": 0.7444673776626587,
"learning_rate": 9.989514793618109e-05,
"loss": 2.2174,
"step": 2710
},
{
"epoch": 0.02,
"grad_norm": 0.7561413049697876,
"learning_rate": 9.989474596047572e-05,
"loss": 2.3708,
"step": 2715
},
{
"epoch": 0.02,
"grad_norm": 0.8025218844413757,
"learning_rate": 9.989434321652024e-05,
"loss": 2.2371,
"step": 2720
},
{
"epoch": 0.02,
"grad_norm": 0.8423010110855103,
"learning_rate": 9.989393970432085e-05,
"loss": 2.2464,
"step": 2725
},
{
"epoch": 0.02,
"grad_norm": 0.8212866187095642,
"learning_rate": 9.989353542388378e-05,
"loss": 2.2834,
"step": 2730
},
{
"epoch": 0.02,
"grad_norm": 0.8092983961105347,
"learning_rate": 9.989313037521524e-05,
"loss": 2.3657,
"step": 2735
},
{
"epoch": 0.02,
"grad_norm": 0.7843234539031982,
"learning_rate": 9.989272455832145e-05,
"loss": 2.3437,
"step": 2740
},
{
"epoch": 0.02,
"grad_norm": 0.7900530099868774,
"learning_rate": 9.989231797320869e-05,
"loss": 2.2444,
"step": 2745
},
{
"epoch": 0.02,
"grad_norm": 0.8011190891265869,
"learning_rate": 9.989191061988319e-05,
"loss": 2.2646,
"step": 2750
},
{
"epoch": 0.02,
"grad_norm": 0.7694355249404907,
"learning_rate": 9.989150249835125e-05,
"loss": 2.3337,
"step": 2755
},
{
"epoch": 0.02,
"grad_norm": 0.7959652543067932,
"learning_rate": 9.989109360861913e-05,
"loss": 2.3659,
"step": 2760
},
{
"epoch": 0.02,
"grad_norm": 0.7657417058944702,
"learning_rate": 9.989068395069313e-05,
"loss": 2.3654,
"step": 2765
},
{
"epoch": 0.02,
"grad_norm": 0.8033024072647095,
"learning_rate": 9.989027352457958e-05,
"loss": 2.2972,
"step": 2770
},
{
"epoch": 0.02,
"grad_norm": 0.7785532474517822,
"learning_rate": 9.988986233028478e-05,
"loss": 2.2482,
"step": 2775
},
{
"epoch": 0.02,
"grad_norm": 0.7383233308792114,
"learning_rate": 9.988945036781507e-05,
"loss": 2.2911,
"step": 2780
},
{
"epoch": 0.02,
"grad_norm": 0.7808538675308228,
"learning_rate": 9.988903763717679e-05,
"loss": 2.3362,
"step": 2785
},
{
"epoch": 0.02,
"grad_norm": 0.7831281423568726,
"learning_rate": 9.988862413837628e-05,
"loss": 2.3955,
"step": 2790
},
{
"epoch": 0.02,
"grad_norm": 0.7887628674507141,
"learning_rate": 9.988820987141993e-05,
"loss": 2.2985,
"step": 2795
},
{
"epoch": 0.02,
"grad_norm": 0.8031251430511475,
"learning_rate": 9.988779483631409e-05,
"loss": 2.2062,
"step": 2800
},
{
"epoch": 0.02,
"grad_norm": 0.7751171588897705,
"learning_rate": 9.98873790330652e-05,
"loss": 2.3602,
"step": 2805
},
{
"epoch": 0.02,
"grad_norm": 0.9482797384262085,
"learning_rate": 9.988696246167961e-05,
"loss": 2.2772,
"step": 2810
},
{
"epoch": 0.02,
"grad_norm": 0.749188244342804,
"learning_rate": 9.988654512216378e-05,
"loss": 2.2857,
"step": 2815
},
{
"epoch": 0.02,
"grad_norm": 0.8156470656394958,
"learning_rate": 9.988612701452409e-05,
"loss": 2.3276,
"step": 2820
},
{
"epoch": 0.02,
"grad_norm": 0.8180411458015442,
"learning_rate": 9.988570813876701e-05,
"loss": 2.2386,
"step": 2825
},
{
"epoch": 0.02,
"grad_norm": 0.7540989518165588,
"learning_rate": 9.988528849489897e-05,
"loss": 2.4144,
"step": 2830
},
{
"epoch": 0.02,
"grad_norm": 0.7952907085418701,
"learning_rate": 9.988486808292646e-05,
"loss": 2.3341,
"step": 2835
},
{
"epoch": 0.02,
"grad_norm": 0.7904815673828125,
"learning_rate": 9.988444690285592e-05,
"loss": 2.2426,
"step": 2840
},
{
"epoch": 0.02,
"grad_norm": 0.7669328451156616,
"learning_rate": 9.988402495469385e-05,
"loss": 2.3253,
"step": 2845
},
{
"epoch": 0.02,
"grad_norm": 0.7687118053436279,
"learning_rate": 9.988360223844673e-05,
"loss": 2.292,
"step": 2850
},
{
"epoch": 0.02,
"grad_norm": 0.7611505389213562,
"learning_rate": 9.98831787541211e-05,
"loss": 2.2126,
"step": 2855
},
{
"epoch": 0.02,
"grad_norm": 0.7588258981704712,
"learning_rate": 9.988275450172346e-05,
"loss": 2.3161,
"step": 2860
},
{
"epoch": 0.02,
"grad_norm": 0.7292211651802063,
"learning_rate": 9.988232948126037e-05,
"loss": 2.2365,
"step": 2865
},
{
"epoch": 0.02,
"grad_norm": 0.7894567251205444,
"learning_rate": 9.988190369273833e-05,
"loss": 2.2709,
"step": 2870
},
{
"epoch": 0.02,
"grad_norm": 0.7431198954582214,
"learning_rate": 9.988147713616391e-05,
"loss": 2.2799,
"step": 2875
},
{
"epoch": 0.02,
"grad_norm": 0.766562283039093,
"learning_rate": 9.988104981154371e-05,
"loss": 2.3821,
"step": 2880
},
{
"epoch": 0.02,
"grad_norm": 0.726231038570404,
"learning_rate": 9.988062171888429e-05,
"loss": 2.2011,
"step": 2885
},
{
"epoch": 0.02,
"grad_norm": 0.7627571225166321,
"learning_rate": 9.988019285819221e-05,
"loss": 2.3042,
"step": 2890
},
{
"epoch": 0.02,
"grad_norm": 0.7949878573417664,
"learning_rate": 9.987976322947411e-05,
"loss": 2.2765,
"step": 2895
},
{
"epoch": 0.02,
"grad_norm": 0.9218003749847412,
"learning_rate": 9.98793328327366e-05,
"loss": 2.2742,
"step": 2900
},
{
"epoch": 0.02,
"grad_norm": 0.7685514092445374,
"learning_rate": 9.98789016679863e-05,
"loss": 2.295,
"step": 2905
},
{
"epoch": 0.02,
"grad_norm": 0.7606549263000488,
"learning_rate": 9.987846973522985e-05,
"loss": 2.3258,
"step": 2910
},
{
"epoch": 0.02,
"grad_norm": 0.7988168001174927,
"learning_rate": 9.98780370344739e-05,
"loss": 2.2588,
"step": 2915
},
{
"epoch": 0.02,
"grad_norm": 0.7924713492393494,
"learning_rate": 9.987760356572512e-05,
"loss": 2.2535,
"step": 2920
},
{
"epoch": 0.02,
"grad_norm": 0.7237165570259094,
"learning_rate": 9.987716932899018e-05,
"loss": 2.2497,
"step": 2925
},
{
"epoch": 0.02,
"grad_norm": 0.7524308562278748,
"learning_rate": 9.987673432427576e-05,
"loss": 2.262,
"step": 2930
},
{
"epoch": 0.02,
"grad_norm": 0.8643974661827087,
"learning_rate": 9.987629855158855e-05,
"loss": 2.2258,
"step": 2935
},
{
"epoch": 0.02,
"grad_norm": 0.7577559351921082,
"learning_rate": 9.98758620109353e-05,
"loss": 2.2539,
"step": 2940
},
{
"epoch": 0.02,
"grad_norm": 0.7918326258659363,
"learning_rate": 9.98754247023227e-05,
"loss": 2.2085,
"step": 2945
},
{
"epoch": 0.02,
"grad_norm": 0.7608367800712585,
"learning_rate": 9.987498662575747e-05,
"loss": 2.222,
"step": 2950
},
{
"epoch": 0.02,
"grad_norm": 0.8014614582061768,
"learning_rate": 9.987454778124639e-05,
"loss": 2.3716,
"step": 2955
},
{
"epoch": 0.02,
"grad_norm": 0.8235347270965576,
"learning_rate": 9.987410816879619e-05,
"loss": 2.2309,
"step": 2960
},
{
"epoch": 0.02,
"grad_norm": 0.7728661894798279,
"learning_rate": 9.987366778841365e-05,
"loss": 2.2386,
"step": 2965
},
{
"epoch": 0.02,
"grad_norm": 0.8065659999847412,
"learning_rate": 9.987322664010554e-05,
"loss": 2.3138,
"step": 2970
},
{
"epoch": 0.02,
"grad_norm": 0.846693754196167,
"learning_rate": 9.987278472387867e-05,
"loss": 2.2257,
"step": 2975
},
{
"epoch": 0.02,
"grad_norm": 1.541448712348938,
"learning_rate": 9.987234203973986e-05,
"loss": 2.2435,
"step": 2980
},
{
"epoch": 0.02,
"grad_norm": 0.7528800368309021,
"learning_rate": 9.987189858769586e-05,
"loss": 2.2612,
"step": 2985
},
{
"epoch": 0.02,
"grad_norm": 0.7742182612419128,
"learning_rate": 9.987145436775357e-05,
"loss": 2.2236,
"step": 2990
},
{
"epoch": 0.02,
"grad_norm": 0.7810326218605042,
"learning_rate": 9.98710093799198e-05,
"loss": 2.2376,
"step": 2995
},
{
"epoch": 0.02,
"grad_norm": 0.7470199465751648,
"learning_rate": 9.987056362420138e-05,
"loss": 2.2864,
"step": 3000
},
{
"epoch": 0.02,
"grad_norm": 0.7890694737434387,
"learning_rate": 9.98701171006052e-05,
"loss": 2.2939,
"step": 3005
},
{
"epoch": 0.02,
"grad_norm": 0.7770738005638123,
"learning_rate": 9.986966980913817e-05,
"loss": 2.2442,
"step": 3010
},
{
"epoch": 0.02,
"grad_norm": 0.7154539823532104,
"learning_rate": 9.986922174980709e-05,
"loss": 2.2553,
"step": 3015
},
{
"epoch": 0.02,
"grad_norm": 0.8018057346343994,
"learning_rate": 9.986877292261893e-05,
"loss": 2.2843,
"step": 3020
},
{
"epoch": 0.02,
"grad_norm": 0.7703126072883606,
"learning_rate": 9.986832332758058e-05,
"loss": 2.2368,
"step": 3025
},
{
"epoch": 0.02,
"grad_norm": 0.7295985817909241,
"learning_rate": 9.986787296469895e-05,
"loss": 2.315,
"step": 3030
},
{
"epoch": 0.02,
"grad_norm": 0.7249435782432556,
"learning_rate": 9.986742183398099e-05,
"loss": 2.3282,
"step": 3035
},
{
"epoch": 0.02,
"grad_norm": 0.8233409523963928,
"learning_rate": 9.986696993543366e-05,
"loss": 2.3645,
"step": 3040
},
{
"epoch": 0.02,
"grad_norm": 0.7459914088249207,
"learning_rate": 9.986651726906387e-05,
"loss": 2.2303,
"step": 3045
},
{
"epoch": 0.02,
"grad_norm": 0.758183479309082,
"learning_rate": 9.986606383487862e-05,
"loss": 2.2497,
"step": 3050
},
{
"epoch": 0.02,
"grad_norm": 0.733689546585083,
"learning_rate": 9.98656096328849e-05,
"loss": 2.2874,
"step": 3055
},
{
"epoch": 0.02,
"grad_norm": 0.7754899859428406,
"learning_rate": 9.98651546630897e-05,
"loss": 2.2735,
"step": 3060
},
{
"epoch": 0.02,
"grad_norm": 1.1103088855743408,
"learning_rate": 9.986469892550001e-05,
"loss": 2.3101,
"step": 3065
},
{
"epoch": 0.02,
"grad_norm": 0.7256807088851929,
"learning_rate": 9.986424242012287e-05,
"loss": 2.3247,
"step": 3070
},
{
"epoch": 0.02,
"grad_norm": 0.7757099270820618,
"learning_rate": 9.986378514696528e-05,
"loss": 2.306,
"step": 3075
},
{
"epoch": 0.02,
"grad_norm": 0.7849761843681335,
"learning_rate": 9.98633271060343e-05,
"loss": 2.2742,
"step": 3080
},
{
"epoch": 0.02,
"grad_norm": 0.7703383564949036,
"learning_rate": 9.9862868297337e-05,
"loss": 2.3667,
"step": 3085
},
{
"epoch": 0.02,
"grad_norm": 0.7254393100738525,
"learning_rate": 9.98624087208804e-05,
"loss": 2.2762,
"step": 3090
},
{
"epoch": 0.02,
"grad_norm": 0.7525180578231812,
"learning_rate": 9.98619483766716e-05,
"loss": 2.2141,
"step": 3095
},
{
"epoch": 0.02,
"grad_norm": 0.7308431267738342,
"learning_rate": 9.98614872647177e-05,
"loss": 2.3251,
"step": 3100
},
{
"epoch": 0.02,
"grad_norm": 0.7849510312080383,
"learning_rate": 9.986102538502579e-05,
"loss": 2.2762,
"step": 3105
},
{
"epoch": 0.02,
"grad_norm": 0.7393447756767273,
"learning_rate": 9.986056273760297e-05,
"loss": 2.2472,
"step": 3110
},
{
"epoch": 0.02,
"grad_norm": 0.7799934148788452,
"learning_rate": 9.986009932245637e-05,
"loss": 2.2718,
"step": 3115
},
{
"epoch": 0.02,
"grad_norm": 0.7899098992347717,
"learning_rate": 9.985963513959313e-05,
"loss": 2.2794,
"step": 3120
},
{
"epoch": 0.02,
"grad_norm": 0.786720335483551,
"learning_rate": 9.985917018902041e-05,
"loss": 2.3342,
"step": 3125
},
{
"epoch": 0.02,
"grad_norm": 0.7941493988037109,
"learning_rate": 9.985870447074535e-05,
"loss": 2.3042,
"step": 3130
},
{
"epoch": 0.02,
"grad_norm": 0.7838422656059265,
"learning_rate": 9.985823798477512e-05,
"loss": 2.184,
"step": 3135
},
{
"epoch": 0.02,
"grad_norm": 0.750948965549469,
"learning_rate": 9.985777073111691e-05,
"loss": 2.3011,
"step": 3140
},
{
"epoch": 0.02,
"grad_norm": 0.735377311706543,
"learning_rate": 9.985730270977791e-05,
"loss": 2.2743,
"step": 3145
},
{
"epoch": 0.02,
"grad_norm": 0.7587282657623291,
"learning_rate": 9.985683392076534e-05,
"loss": 2.2879,
"step": 3150
},
{
"epoch": 0.02,
"grad_norm": 0.7545079588890076,
"learning_rate": 9.98563643640864e-05,
"loss": 2.3096,
"step": 3155
},
{
"epoch": 0.02,
"grad_norm": 0.7551508545875549,
"learning_rate": 9.985589403974833e-05,
"loss": 2.2702,
"step": 3160
},
{
"epoch": 0.02,
"grad_norm": 0.7722675204277039,
"learning_rate": 9.985542294775838e-05,
"loss": 2.2509,
"step": 3165
},
{
"epoch": 0.03,
"grad_norm": 0.7377732396125793,
"learning_rate": 9.985495108812377e-05,
"loss": 2.2203,
"step": 3170
},
{
"epoch": 0.03,
"grad_norm": 0.737814724445343,
"learning_rate": 9.985447846085181e-05,
"loss": 2.2663,
"step": 3175
},
{
"epoch": 0.03,
"grad_norm": 0.7906785607337952,
"learning_rate": 9.985400506594974e-05,
"loss": 2.327,
"step": 3180
},
{
"epoch": 0.03,
"grad_norm": 0.7829717993736267,
"learning_rate": 9.985353090342487e-05,
"loss": 2.3065,
"step": 3185
},
{
"epoch": 0.03,
"grad_norm": 0.7877334356307983,
"learning_rate": 9.985305597328449e-05,
"loss": 2.306,
"step": 3190
},
{
"epoch": 0.03,
"grad_norm": 0.7634186148643494,
"learning_rate": 9.985258027553593e-05,
"loss": 2.1856,
"step": 3195
},
{
"epoch": 0.03,
"grad_norm": 0.7351861000061035,
"learning_rate": 9.98521038101865e-05,
"loss": 2.2821,
"step": 3200
},
{
"epoch": 0.03,
"grad_norm": 0.8012837171554565,
"learning_rate": 9.985162657724354e-05,
"loss": 2.2903,
"step": 3205
},
{
"epoch": 0.03,
"grad_norm": 0.8202405571937561,
"learning_rate": 9.98511485767144e-05,
"loss": 2.3465,
"step": 3210
},
{
"epoch": 0.03,
"grad_norm": 0.7695194482803345,
"learning_rate": 9.985066980860642e-05,
"loss": 2.3357,
"step": 3215
},
{
"epoch": 0.03,
"grad_norm": 0.7718874216079712,
"learning_rate": 9.9850190272927e-05,
"loss": 2.3274,
"step": 3220
},
{
"epoch": 0.03,
"grad_norm": 0.7420881390571594,
"learning_rate": 9.984970996968352e-05,
"loss": 2.2773,
"step": 3225
},
{
"epoch": 0.03,
"grad_norm": 0.7779717445373535,
"learning_rate": 9.984922889888335e-05,
"loss": 2.3143,
"step": 3230
},
{
"epoch": 0.03,
"grad_norm": 0.7514321804046631,
"learning_rate": 9.984874706053393e-05,
"loss": 2.2958,
"step": 3235
},
{
"epoch": 0.03,
"grad_norm": 0.7656455636024475,
"learning_rate": 9.984826445464266e-05,
"loss": 2.2702,
"step": 3240
},
{
"epoch": 0.03,
"grad_norm": 0.7379190325737,
"learning_rate": 9.984778108121698e-05,
"loss": 2.3099,
"step": 3245
},
{
"epoch": 0.03,
"grad_norm": 0.7915068864822388,
"learning_rate": 9.984729694026431e-05,
"loss": 2.3127,
"step": 3250
},
{
"epoch": 0.03,
"grad_norm": 0.8166426420211792,
"learning_rate": 9.984681203179214e-05,
"loss": 2.3164,
"step": 3255
},
{
"epoch": 0.03,
"grad_norm": 0.7281964421272278,
"learning_rate": 9.984632635580792e-05,
"loss": 2.3097,
"step": 3260
},
{
"epoch": 0.03,
"grad_norm": 0.7576284408569336,
"learning_rate": 9.984583991231913e-05,
"loss": 2.2812,
"step": 3265
},
{
"epoch": 0.03,
"grad_norm": 0.8970535397529602,
"learning_rate": 9.984535270133324e-05,
"loss": 2.2866,
"step": 3270
},
{
"epoch": 0.03,
"grad_norm": 0.7504796981811523,
"learning_rate": 9.984486472285777e-05,
"loss": 2.3273,
"step": 3275
},
{
"epoch": 0.03,
"grad_norm": 0.7535706162452698,
"learning_rate": 9.984437597690024e-05,
"loss": 2.1453,
"step": 3280
},
{
"epoch": 0.03,
"grad_norm": 0.7760560512542725,
"learning_rate": 9.984388646346817e-05,
"loss": 2.2972,
"step": 3285
},
{
"epoch": 0.03,
"grad_norm": 0.7605026960372925,
"learning_rate": 9.984339618256909e-05,
"loss": 2.2884,
"step": 3290
},
{
"epoch": 0.03,
"grad_norm": 0.7673270106315613,
"learning_rate": 9.984290513421057e-05,
"loss": 2.3473,
"step": 3295
},
{
"epoch": 0.03,
"grad_norm": 0.7707405686378479,
"learning_rate": 9.984241331840015e-05,
"loss": 2.2803,
"step": 3300
},
{
"epoch": 0.03,
"grad_norm": 0.7661117911338806,
"learning_rate": 9.984192073514538e-05,
"loss": 2.2792,
"step": 3305
},
{
"epoch": 0.03,
"grad_norm": 0.7346450090408325,
"learning_rate": 9.98414273844539e-05,
"loss": 2.2393,
"step": 3310
},
{
"epoch": 0.03,
"grad_norm": 0.7762331962585449,
"learning_rate": 9.984093326633325e-05,
"loss": 2.2804,
"step": 3315
},
{
"epoch": 0.03,
"grad_norm": 0.8068987131118774,
"learning_rate": 9.98404383807911e-05,
"loss": 2.2583,
"step": 3320
},
{
"epoch": 0.03,
"grad_norm": 0.7443338632583618,
"learning_rate": 9.983994272783502e-05,
"loss": 2.267,
"step": 3325
},
{
"epoch": 0.03,
"grad_norm": 0.747916042804718,
"learning_rate": 9.983944630747266e-05,
"loss": 2.3469,
"step": 3330
},
{
"epoch": 0.03,
"grad_norm": 0.7459864616394043,
"learning_rate": 9.983894911971167e-05,
"loss": 2.1693,
"step": 3335
},
{
"epoch": 0.03,
"grad_norm": 0.7667864561080933,
"learning_rate": 9.983845116455969e-05,
"loss": 2.1407,
"step": 3340
},
{
"epoch": 0.03,
"grad_norm": 0.7649180889129639,
"learning_rate": 9.983795244202439e-05,
"loss": 2.217,
"step": 3345
},
{
"epoch": 0.03,
"grad_norm": 0.7473801970481873,
"learning_rate": 9.983745295211347e-05,
"loss": 2.3101,
"step": 3350
},
{
"epoch": 0.03,
"grad_norm": 0.749112069606781,
"learning_rate": 9.983695269483459e-05,
"loss": 2.2523,
"step": 3355
},
{
"epoch": 0.03,
"grad_norm": 0.7355244159698486,
"learning_rate": 9.983645167019546e-05,
"loss": 2.2606,
"step": 3360
},
{
"epoch": 0.03,
"grad_norm": 0.7610791921615601,
"learning_rate": 9.983594987820381e-05,
"loss": 2.2804,
"step": 3365
},
{
"epoch": 0.03,
"grad_norm": 0.7378023862838745,
"learning_rate": 9.983544731886735e-05,
"loss": 2.2753,
"step": 3370
},
{
"epoch": 0.03,
"grad_norm": 0.7339537739753723,
"learning_rate": 9.983494399219386e-05,
"loss": 2.2287,
"step": 3375
},
{
"epoch": 0.03,
"grad_norm": 0.7985178828239441,
"learning_rate": 9.983443989819103e-05,
"loss": 2.2588,
"step": 3380
},
{
"epoch": 0.03,
"grad_norm": 0.7475832104682922,
"learning_rate": 9.983393503686665e-05,
"loss": 2.1719,
"step": 3385
},
{
"epoch": 0.03,
"grad_norm": 0.7476961016654968,
"learning_rate": 9.983342940822849e-05,
"loss": 2.2788,
"step": 3390
},
{
"epoch": 0.03,
"grad_norm": 0.7462399005889893,
"learning_rate": 9.983292301228432e-05,
"loss": 2.2509,
"step": 3395
},
{
"epoch": 0.03,
"grad_norm": 1.0284730195999146,
"learning_rate": 9.983241584904197e-05,
"loss": 2.3372,
"step": 3400
},
{
"epoch": 0.03,
"grad_norm": 0.7454382181167603,
"learning_rate": 9.983190791850925e-05,
"loss": 2.2196,
"step": 3405
},
{
"epoch": 0.03,
"grad_norm": 0.7688352465629578,
"learning_rate": 9.983139922069394e-05,
"loss": 2.2973,
"step": 3410
},
{
"epoch": 0.03,
"grad_norm": 0.7922349572181702,
"learning_rate": 9.983088975560391e-05,
"loss": 2.2343,
"step": 3415
},
{
"epoch": 0.03,
"grad_norm": 0.7532771229743958,
"learning_rate": 9.983037952324699e-05,
"loss": 2.2585,
"step": 3420
},
{
"epoch": 0.03,
"grad_norm": 0.7738056778907776,
"learning_rate": 9.982986852363102e-05,
"loss": 2.2314,
"step": 3425
},
{
"epoch": 0.03,
"grad_norm": 0.7581791281700134,
"learning_rate": 9.98293567567639e-05,
"loss": 2.3659,
"step": 3430
},
{
"epoch": 0.03,
"grad_norm": 0.7392639517784119,
"learning_rate": 9.982884422265349e-05,
"loss": 2.2552,
"step": 3435
},
{
"epoch": 0.03,
"grad_norm": 0.7250624299049377,
"learning_rate": 9.982833092130768e-05,
"loss": 2.2775,
"step": 3440
},
{
"epoch": 0.03,
"grad_norm": 0.7871769666671753,
"learning_rate": 9.982781685273438e-05,
"loss": 2.3651,
"step": 3445
},
{
"epoch": 0.03,
"grad_norm": 0.7372907996177673,
"learning_rate": 9.982730201694153e-05,
"loss": 2.3051,
"step": 3450
},
{
"epoch": 0.03,
"grad_norm": 0.7637687921524048,
"learning_rate": 9.982678641393701e-05,
"loss": 2.2698,
"step": 3455
},
{
"epoch": 0.03,
"grad_norm": 0.7562505602836609,
"learning_rate": 9.98262700437288e-05,
"loss": 2.2445,
"step": 3460
},
{
"epoch": 0.03,
"grad_norm": 0.7814033031463623,
"learning_rate": 9.982575290632481e-05,
"loss": 2.3265,
"step": 3465
},
{
"epoch": 0.03,
"grad_norm": 0.8336643576622009,
"learning_rate": 9.982523500173304e-05,
"loss": 2.2118,
"step": 3470
},
{
"epoch": 0.03,
"grad_norm": 0.7395555973052979,
"learning_rate": 9.982471632996144e-05,
"loss": 2.2737,
"step": 3475
},
{
"epoch": 0.03,
"grad_norm": 0.7723784446716309,
"learning_rate": 9.9824196891018e-05,
"loss": 2.2991,
"step": 3480
},
{
"epoch": 0.03,
"grad_norm": 0.7597246766090393,
"learning_rate": 9.982367668491075e-05,
"loss": 2.2585,
"step": 3485
},
{
"epoch": 0.03,
"grad_norm": 0.7042340040206909,
"learning_rate": 9.982315571164765e-05,
"loss": 2.3494,
"step": 3490
},
{
"epoch": 0.03,
"grad_norm": 0.7320727705955505,
"learning_rate": 9.982263397123675e-05,
"loss": 2.2109,
"step": 3495
},
{
"epoch": 0.03,
"grad_norm": 0.7801787257194519,
"learning_rate": 9.982211146368608e-05,
"loss": 2.3336,
"step": 3500
},
{
"epoch": 0.03,
"grad_norm": 0.7796906232833862,
"learning_rate": 9.982158818900369e-05,
"loss": 2.3184,
"step": 3505
},
{
"epoch": 0.03,
"grad_norm": 0.7338558435440063,
"learning_rate": 9.982106414719762e-05,
"loss": 2.2779,
"step": 3510
},
{
"epoch": 0.03,
"grad_norm": 0.7506589889526367,
"learning_rate": 9.982053933827595e-05,
"loss": 2.298,
"step": 3515
},
{
"epoch": 0.03,
"grad_norm": 0.7650732398033142,
"learning_rate": 9.982001376224677e-05,
"loss": 2.285,
"step": 3520
},
{
"epoch": 0.03,
"grad_norm": 0.7226093411445618,
"learning_rate": 9.981948741911816e-05,
"loss": 2.2696,
"step": 3525
},
{
"epoch": 0.03,
"grad_norm": 0.7864460945129395,
"learning_rate": 9.981896030889821e-05,
"loss": 2.2255,
"step": 3530
},
{
"epoch": 0.03,
"grad_norm": 0.7817670702934265,
"learning_rate": 9.981843243159507e-05,
"loss": 2.331,
"step": 3535
},
{
"epoch": 0.03,
"grad_norm": 0.780349612236023,
"learning_rate": 9.981790378721684e-05,
"loss": 2.3259,
"step": 3540
},
{
"epoch": 0.03,
"grad_norm": 0.7508283853530884,
"learning_rate": 9.98173743757717e-05,
"loss": 2.2353,
"step": 3545
},
{
"epoch": 0.03,
"grad_norm": 0.7514976263046265,
"learning_rate": 9.981684419726775e-05,
"loss": 2.2264,
"step": 3550
},
{
"epoch": 0.03,
"grad_norm": 0.7356547117233276,
"learning_rate": 9.981631325171318e-05,
"loss": 2.2888,
"step": 3555
},
{
"epoch": 0.03,
"grad_norm": 0.8579287528991699,
"learning_rate": 9.981578153911616e-05,
"loss": 2.3032,
"step": 3560
},
{
"epoch": 0.03,
"grad_norm": 0.8231850862503052,
"learning_rate": 9.981524905948487e-05,
"loss": 2.2611,
"step": 3565
},
{
"epoch": 0.03,
"grad_norm": 0.753423810005188,
"learning_rate": 9.981471581282753e-05,
"loss": 2.2111,
"step": 3570
},
{
"epoch": 0.03,
"grad_norm": 0.7484989166259766,
"learning_rate": 9.981418179915234e-05,
"loss": 2.2122,
"step": 3575
},
{
"epoch": 0.03,
"grad_norm": 0.7779796719551086,
"learning_rate": 9.981364701846752e-05,
"loss": 2.3053,
"step": 3580
},
{
"epoch": 0.03,
"grad_norm": 0.7838788628578186,
"learning_rate": 9.981311147078128e-05,
"loss": 2.2961,
"step": 3585
},
{
"epoch": 0.03,
"grad_norm": 0.7868286967277527,
"learning_rate": 9.981257515610192e-05,
"loss": 2.3102,
"step": 3590
},
{
"epoch": 0.03,
"grad_norm": 0.7895642518997192,
"learning_rate": 9.981203807443764e-05,
"loss": 2.2845,
"step": 3595
},
{
"epoch": 0.03,
"grad_norm": 0.7732324004173279,
"learning_rate": 9.981150022579676e-05,
"loss": 2.2642,
"step": 3600
},
{
"epoch": 0.03,
"grad_norm": 0.7438845038414001,
"learning_rate": 9.981096161018754e-05,
"loss": 2.3276,
"step": 3605
},
{
"epoch": 0.03,
"grad_norm": 0.7430480122566223,
"learning_rate": 9.981042222761826e-05,
"loss": 2.3,
"step": 3610
},
{
"epoch": 0.03,
"grad_norm": 0.7510012984275818,
"learning_rate": 9.980988207809726e-05,
"loss": 2.2584,
"step": 3615
},
{
"epoch": 0.03,
"grad_norm": 0.7372361421585083,
"learning_rate": 9.98093411616328e-05,
"loss": 2.2706,
"step": 3620
},
{
"epoch": 0.03,
"grad_norm": 0.7566350698471069,
"learning_rate": 9.980879947823327e-05,
"loss": 2.2053,
"step": 3625
},
{
"epoch": 0.03,
"grad_norm": 0.7359904646873474,
"learning_rate": 9.980825702790696e-05,
"loss": 2.3195,
"step": 3630
},
{
"epoch": 0.03,
"grad_norm": 0.766076385974884,
"learning_rate": 9.980771381066227e-05,
"loss": 2.2409,
"step": 3635
},
{
"epoch": 0.03,
"grad_norm": 0.7602216601371765,
"learning_rate": 9.980716982650753e-05,
"loss": 2.3081,
"step": 3640
},
{
"epoch": 0.03,
"grad_norm": 0.7077452540397644,
"learning_rate": 9.980662507545113e-05,
"loss": 2.251,
"step": 3645
},
{
"epoch": 0.03,
"grad_norm": 0.7858209013938904,
"learning_rate": 9.980607955750146e-05,
"loss": 2.312,
"step": 3650
},
{
"epoch": 0.03,
"grad_norm": 0.7604818344116211,
"learning_rate": 9.98055332726669e-05,
"loss": 2.1794,
"step": 3655
},
{
"epoch": 0.03,
"grad_norm": 0.7449327707290649,
"learning_rate": 9.980498622095588e-05,
"loss": 2.2553,
"step": 3660
},
{
"epoch": 0.03,
"grad_norm": 0.7738369107246399,
"learning_rate": 9.980443840237681e-05,
"loss": 2.2817,
"step": 3665
},
{
"epoch": 0.03,
"grad_norm": 0.7516607642173767,
"learning_rate": 9.980388981693816e-05,
"loss": 2.1983,
"step": 3670
},
{
"epoch": 0.03,
"grad_norm": 0.7499752640724182,
"learning_rate": 9.980334046464832e-05,
"loss": 2.33,
"step": 3675
},
{
"epoch": 0.03,
"grad_norm": 0.7446860074996948,
"learning_rate": 9.980279034551579e-05,
"loss": 2.2585,
"step": 3680
},
{
"epoch": 0.03,
"grad_norm": 0.7675132155418396,
"learning_rate": 9.980223945954902e-05,
"loss": 2.2751,
"step": 3685
},
{
"epoch": 0.03,
"grad_norm": 0.7864475846290588,
"learning_rate": 9.980168780675651e-05,
"loss": 2.3084,
"step": 3690
},
{
"epoch": 0.03,
"grad_norm": 0.7555897831916809,
"learning_rate": 9.980113538714675e-05,
"loss": 2.2484,
"step": 3695
},
{
"epoch": 0.03,
"grad_norm": 0.7966665625572205,
"learning_rate": 9.980058220072825e-05,
"loss": 2.2549,
"step": 3700
},
{
"epoch": 0.03,
"grad_norm": 0.7549835443496704,
"learning_rate": 9.98000282475095e-05,
"loss": 2.2778,
"step": 3705
},
{
"epoch": 0.03,
"grad_norm": 0.7632366418838501,
"learning_rate": 9.979947352749906e-05,
"loss": 2.2579,
"step": 3710
},
{
"epoch": 0.03,
"grad_norm": 0.8138110637664795,
"learning_rate": 9.979891804070545e-05,
"loss": 2.3114,
"step": 3715
},
{
"epoch": 0.03,
"grad_norm": 0.7309870719909668,
"learning_rate": 9.979836178713724e-05,
"loss": 2.3154,
"step": 3720
},
{
"epoch": 0.03,
"grad_norm": 0.7168034315109253,
"learning_rate": 9.979780476680298e-05,
"loss": 2.2157,
"step": 3725
},
{
"epoch": 0.03,
"grad_norm": 0.7405194640159607,
"learning_rate": 9.979724697971125e-05,
"loss": 2.2982,
"step": 3730
},
{
"epoch": 0.03,
"grad_norm": 0.7334524989128113,
"learning_rate": 9.979668842587065e-05,
"loss": 2.2174,
"step": 3735
},
{
"epoch": 0.03,
"grad_norm": 0.7381799817085266,
"learning_rate": 9.979612910528976e-05,
"loss": 2.287,
"step": 3740
},
{
"epoch": 0.03,
"grad_norm": 0.7213761210441589,
"learning_rate": 9.979556901797723e-05,
"loss": 2.2034,
"step": 3745
},
{
"epoch": 0.03,
"grad_norm": 0.7332202792167664,
"learning_rate": 9.979500816394165e-05,
"loss": 2.2527,
"step": 3750
},
{
"epoch": 0.03,
"grad_norm": 0.7268756628036499,
"learning_rate": 9.979444654319165e-05,
"loss": 2.2862,
"step": 3755
},
{
"epoch": 0.03,
"grad_norm": 0.8504984378814697,
"learning_rate": 9.979388415573591e-05,
"loss": 2.2032,
"step": 3760
},
{
"epoch": 0.03,
"grad_norm": 0.7761270999908447,
"learning_rate": 9.979332100158307e-05,
"loss": 2.2328,
"step": 3765
},
{
"epoch": 0.03,
"grad_norm": 0.7488829493522644,
"learning_rate": 9.979275708074179e-05,
"loss": 2.2358,
"step": 3770
},
{
"epoch": 0.03,
"grad_norm": 0.7451363205909729,
"learning_rate": 9.979219239322078e-05,
"loss": 2.2302,
"step": 3775
},
{
"epoch": 0.03,
"grad_norm": 0.7336910367012024,
"learning_rate": 9.979162693902871e-05,
"loss": 2.1444,
"step": 3780
},
{
"epoch": 0.03,
"grad_norm": 0.7972361445426941,
"learning_rate": 9.97910607181743e-05,
"loss": 2.2402,
"step": 3785
},
{
"epoch": 0.03,
"grad_norm": 0.7630717158317566,
"learning_rate": 9.979049373066626e-05,
"loss": 2.3184,
"step": 3790
},
{
"epoch": 0.03,
"grad_norm": 0.7924901843070984,
"learning_rate": 9.978992597651334e-05,
"loss": 2.2307,
"step": 3795
},
{
"epoch": 0.03,
"grad_norm": 0.7356935739517212,
"learning_rate": 9.978935745572424e-05,
"loss": 2.2528,
"step": 3800
},
{
"epoch": 0.03,
"grad_norm": 0.7206233739852905,
"learning_rate": 9.978878816830776e-05,
"loss": 2.2149,
"step": 3805
},
{
"epoch": 0.03,
"grad_norm": 0.754388689994812,
"learning_rate": 9.978821811427263e-05,
"loss": 2.12,
"step": 3810
},
{
"epoch": 0.03,
"grad_norm": 0.770865797996521,
"learning_rate": 9.978764729362766e-05,
"loss": 2.2654,
"step": 3815
},
{
"epoch": 0.03,
"grad_norm": 0.6992185711860657,
"learning_rate": 9.978707570638162e-05,
"loss": 2.1689,
"step": 3820
},
{
"epoch": 0.03,
"grad_norm": 0.7318618297576904,
"learning_rate": 9.978650335254332e-05,
"loss": 2.269,
"step": 3825
},
{
"epoch": 0.03,
"grad_norm": 0.7445050477981567,
"learning_rate": 9.978593023212155e-05,
"loss": 2.2782,
"step": 3830
},
{
"epoch": 0.03,
"grad_norm": 0.7475794553756714,
"learning_rate": 9.978535634512515e-05,
"loss": 2.2468,
"step": 3835
},
{
"epoch": 0.03,
"grad_norm": 0.7316878437995911,
"learning_rate": 9.978478169156298e-05,
"loss": 2.1979,
"step": 3840
},
{
"epoch": 0.03,
"grad_norm": 0.7656149864196777,
"learning_rate": 9.978420627144385e-05,
"loss": 2.2399,
"step": 3845
},
{
"epoch": 0.03,
"grad_norm": 0.7177276015281677,
"learning_rate": 9.978363008477665e-05,
"loss": 2.3607,
"step": 3850
},
{
"epoch": 0.03,
"grad_norm": 0.721457839012146,
"learning_rate": 9.978305313157022e-05,
"loss": 2.1747,
"step": 3855
},
{
"epoch": 0.03,
"grad_norm": 0.741750180721283,
"learning_rate": 9.978247541183348e-05,
"loss": 2.2823,
"step": 3860
},
{
"epoch": 0.03,
"grad_norm": 0.7345651984214783,
"learning_rate": 9.978189692557529e-05,
"loss": 2.2305,
"step": 3865
},
{
"epoch": 0.03,
"grad_norm": 0.7265893220901489,
"learning_rate": 9.978131767280456e-05,
"loss": 2.2205,
"step": 3870
},
{
"epoch": 0.03,
"grad_norm": 0.7322306632995605,
"learning_rate": 9.978073765353026e-05,
"loss": 2.119,
"step": 3875
},
{
"epoch": 0.03,
"grad_norm": 0.7472878694534302,
"learning_rate": 9.978015686776125e-05,
"loss": 2.3233,
"step": 3880
},
{
"epoch": 0.03,
"grad_norm": 0.7837201356887817,
"learning_rate": 9.977957531550652e-05,
"loss": 2.147,
"step": 3885
},
{
"epoch": 0.03,
"grad_norm": 0.7640808820724487,
"learning_rate": 9.977899299677502e-05,
"loss": 2.2707,
"step": 3890
},
{
"epoch": 0.03,
"grad_norm": 0.7678388953208923,
"learning_rate": 9.97784099115757e-05,
"loss": 2.2525,
"step": 3895
},
{
"epoch": 0.03,
"grad_norm": 0.7076819539070129,
"learning_rate": 9.977782605991754e-05,
"loss": 2.2119,
"step": 3900
},
{
"epoch": 0.03,
"grad_norm": 0.7177622318267822,
"learning_rate": 9.977724144180953e-05,
"loss": 2.2332,
"step": 3905
},
{
"epoch": 0.03,
"grad_norm": 0.7158358693122864,
"learning_rate": 9.977665605726067e-05,
"loss": 2.2276,
"step": 3910
},
{
"epoch": 0.03,
"grad_norm": 0.728640079498291,
"learning_rate": 9.977606990628001e-05,
"loss": 2.3393,
"step": 3915
},
{
"epoch": 0.03,
"grad_norm": 0.7213046550750732,
"learning_rate": 9.977548298887651e-05,
"loss": 2.2253,
"step": 3920
},
{
"epoch": 0.03,
"grad_norm": 0.7271808981895447,
"learning_rate": 9.977489530505927e-05,
"loss": 2.1872,
"step": 3925
},
{
"epoch": 0.03,
"grad_norm": 0.7426838278770447,
"learning_rate": 9.97743068548373e-05,
"loss": 2.2395,
"step": 3930
},
{
"epoch": 0.03,
"grad_norm": 0.7302041053771973,
"learning_rate": 9.977371763821966e-05,
"loss": 2.2654,
"step": 3935
},
{
"epoch": 0.03,
"grad_norm": 0.805589497089386,
"learning_rate": 9.977312765521545e-05,
"loss": 2.2427,
"step": 3940
},
{
"epoch": 0.03,
"grad_norm": 0.7397799491882324,
"learning_rate": 9.977253690583372e-05,
"loss": 2.2687,
"step": 3945
},
{
"epoch": 0.03,
"grad_norm": 0.9142820239067078,
"learning_rate": 9.97719453900836e-05,
"loss": 2.1588,
"step": 3950
},
{
"epoch": 0.03,
"grad_norm": 0.7197009325027466,
"learning_rate": 9.977135310797418e-05,
"loss": 2.1895,
"step": 3955
},
{
"epoch": 0.03,
"grad_norm": 0.7083832025527954,
"learning_rate": 9.977076005951458e-05,
"loss": 2.2337,
"step": 3960
},
{
"epoch": 0.03,
"grad_norm": 0.7392210960388184,
"learning_rate": 9.977016624471392e-05,
"loss": 2.2,
"step": 3965
},
{
"epoch": 0.03,
"grad_norm": 0.7502233386039734,
"learning_rate": 9.976957166358137e-05,
"loss": 2.322,
"step": 3970
},
{
"epoch": 0.03,
"grad_norm": 0.714421272277832,
"learning_rate": 9.976897631612607e-05,
"loss": 2.2356,
"step": 3975
},
{
"epoch": 0.03,
"grad_norm": 0.7568963766098022,
"learning_rate": 9.97683802023572e-05,
"loss": 2.2557,
"step": 3980
},
{
"epoch": 0.03,
"grad_norm": 0.7236943244934082,
"learning_rate": 9.976778332228389e-05,
"loss": 2.2333,
"step": 3985
},
{
"epoch": 0.03,
"grad_norm": 0.7373430728912354,
"learning_rate": 9.97671856759154e-05,
"loss": 2.2458,
"step": 3990
},
{
"epoch": 0.03,
"grad_norm": 0.7039047479629517,
"learning_rate": 9.976658726326088e-05,
"loss": 2.3006,
"step": 3995
},
{
"epoch": 0.03,
"grad_norm": 0.7171424031257629,
"learning_rate": 9.976598808432957e-05,
"loss": 2.2582,
"step": 4000
},
{
"epoch": 0.03,
"grad_norm": 0.7468255162239075,
"learning_rate": 9.976538813913068e-05,
"loss": 2.3099,
"step": 4005
},
{
"epoch": 0.03,
"grad_norm": 0.721000611782074,
"learning_rate": 9.976478742767347e-05,
"loss": 2.2082,
"step": 4010
},
{
"epoch": 0.03,
"grad_norm": 0.764666736125946,
"learning_rate": 9.976418594996716e-05,
"loss": 2.2257,
"step": 4015
},
{
"epoch": 0.03,
"grad_norm": 0.737481415271759,
"learning_rate": 9.976358370602103e-05,
"loss": 2.2112,
"step": 4020
},
{
"epoch": 0.03,
"grad_norm": 0.7521570920944214,
"learning_rate": 9.976298069584438e-05,
"loss": 2.2573,
"step": 4025
},
{
"epoch": 0.03,
"grad_norm": 0.7448917031288147,
"learning_rate": 9.976237691944643e-05,
"loss": 2.2873,
"step": 4030
},
{
"epoch": 0.03,
"grad_norm": 0.7232655882835388,
"learning_rate": 9.976177237683653e-05,
"loss": 2.1168,
"step": 4035
},
{
"epoch": 0.03,
"grad_norm": 0.8181683421134949,
"learning_rate": 9.976116706802395e-05,
"loss": 2.2106,
"step": 4040
},
{
"epoch": 0.03,
"grad_norm": 0.7300477623939514,
"learning_rate": 9.976056099301805e-05,
"loss": 2.2953,
"step": 4045
},
{
"epoch": 0.03,
"grad_norm": 0.7364791631698608,
"learning_rate": 9.975995415182812e-05,
"loss": 2.2535,
"step": 4050
},
{
"epoch": 0.03,
"grad_norm": 0.7359514832496643,
"learning_rate": 9.975934654446355e-05,
"loss": 2.2029,
"step": 4055
},
{
"epoch": 0.03,
"grad_norm": 0.782615065574646,
"learning_rate": 9.975873817093367e-05,
"loss": 2.2934,
"step": 4060
},
{
"epoch": 0.03,
"grad_norm": 0.7512704133987427,
"learning_rate": 9.975812903124782e-05,
"loss": 2.2582,
"step": 4065
},
{
"epoch": 0.03,
"grad_norm": 0.787928581237793,
"learning_rate": 9.975751912541544e-05,
"loss": 2.2616,
"step": 4070
},
{
"epoch": 0.03,
"grad_norm": 0.7857593297958374,
"learning_rate": 9.975690845344588e-05,
"loss": 2.2038,
"step": 4075
},
{
"epoch": 0.03,
"grad_norm": 0.7516389489173889,
"learning_rate": 9.975629701534854e-05,
"loss": 2.1841,
"step": 4080
},
{
"epoch": 0.03,
"grad_norm": 0.7074551582336426,
"learning_rate": 9.975568481113285e-05,
"loss": 2.1323,
"step": 4085
},
{
"epoch": 0.03,
"grad_norm": 0.6899214386940002,
"learning_rate": 9.975507184080823e-05,
"loss": 2.2485,
"step": 4090
},
{
"epoch": 0.03,
"grad_norm": 0.7326139211654663,
"learning_rate": 9.975445810438413e-05,
"loss": 2.205,
"step": 4095
},
{
"epoch": 0.03,
"grad_norm": 0.7302681803703308,
"learning_rate": 9.975384360186999e-05,
"loss": 2.2466,
"step": 4100
},
{
"epoch": 0.03,
"grad_norm": 0.7665270566940308,
"learning_rate": 9.975322833327527e-05,
"loss": 2.2778,
"step": 4105
},
{
"epoch": 0.03,
"grad_norm": 0.7606353163719177,
"learning_rate": 9.975261229860944e-05,
"loss": 2.2396,
"step": 4110
},
{
"epoch": 0.03,
"grad_norm": 0.7223678231239319,
"learning_rate": 9.975199549788199e-05,
"loss": 2.1072,
"step": 4115
},
{
"epoch": 0.03,
"grad_norm": 0.7699795365333557,
"learning_rate": 9.975137793110243e-05,
"loss": 2.2839,
"step": 4120
},
{
"epoch": 0.03,
"grad_norm": 0.7184005975723267,
"learning_rate": 9.975075959828026e-05,
"loss": 2.2409,
"step": 4125
},
{
"epoch": 0.03,
"grad_norm": 0.7539365887641907,
"learning_rate": 9.975014049942499e-05,
"loss": 2.2234,
"step": 4130
},
{
"epoch": 0.03,
"grad_norm": 0.7457564473152161,
"learning_rate": 9.974952063454615e-05,
"loss": 2.2561,
"step": 4135
},
{
"epoch": 0.03,
"grad_norm": 0.7375931143760681,
"learning_rate": 9.974890000365332e-05,
"loss": 2.199,
"step": 4140
},
{
"epoch": 0.03,
"grad_norm": 0.7635476589202881,
"learning_rate": 9.9748278606756e-05,
"loss": 2.2927,
"step": 4145
},
{
"epoch": 0.03,
"grad_norm": 0.7252872586250305,
"learning_rate": 9.974765644386382e-05,
"loss": 2.3524,
"step": 4150
},
{
"epoch": 0.03,
"grad_norm": 0.7200140953063965,
"learning_rate": 9.974703351498631e-05,
"loss": 2.1936,
"step": 4155
},
{
"epoch": 0.03,
"grad_norm": 0.7847145795822144,
"learning_rate": 9.974640982013307e-05,
"loss": 2.28,
"step": 4160
},
{
"epoch": 0.03,
"grad_norm": 0.7265502214431763,
"learning_rate": 9.974578535931374e-05,
"loss": 2.2221,
"step": 4165
},
{
"epoch": 0.03,
"grad_norm": 0.7368555068969727,
"learning_rate": 9.974516013253789e-05,
"loss": 2.2592,
"step": 4170
},
{
"epoch": 0.03,
"grad_norm": 0.7392286658287048,
"learning_rate": 9.974453413981517e-05,
"loss": 2.2359,
"step": 4175
},
{
"epoch": 0.03,
"grad_norm": 0.740394115447998,
"learning_rate": 9.974390738115521e-05,
"loss": 2.3395,
"step": 4180
},
{
"epoch": 0.03,
"grad_norm": 0.7096404433250427,
"learning_rate": 9.974327985656767e-05,
"loss": 2.2096,
"step": 4185
},
{
"epoch": 0.03,
"grad_norm": 0.7395580410957336,
"learning_rate": 9.974265156606221e-05,
"loss": 2.2403,
"step": 4190
},
{
"epoch": 0.03,
"grad_norm": 0.7067394256591797,
"learning_rate": 9.974202250964849e-05,
"loss": 2.2042,
"step": 4195
},
{
"epoch": 0.03,
"grad_norm": 0.7257940769195557,
"learning_rate": 9.974139268733623e-05,
"loss": 2.1729,
"step": 4200
},
{
"epoch": 0.03,
"grad_norm": 0.7475812435150146,
"learning_rate": 9.974076209913508e-05,
"loss": 2.3245,
"step": 4205
},
{
"epoch": 0.03,
"grad_norm": 0.7111890316009521,
"learning_rate": 9.974013074505477e-05,
"loss": 2.1928,
"step": 4210
},
{
"epoch": 0.03,
"grad_norm": 0.703853964805603,
"learning_rate": 9.973949862510503e-05,
"loss": 2.2108,
"step": 4215
},
{
"epoch": 0.03,
"grad_norm": 0.7286904454231262,
"learning_rate": 9.97388657392956e-05,
"loss": 2.2284,
"step": 4220
},
{
"epoch": 0.03,
"grad_norm": 0.7088801860809326,
"learning_rate": 9.97382320876362e-05,
"loss": 2.2377,
"step": 4225
},
{
"epoch": 0.03,
"grad_norm": 0.683827817440033,
"learning_rate": 9.973759767013662e-05,
"loss": 2.151,
"step": 4230
},
{
"epoch": 0.03,
"grad_norm": 0.7164034247398376,
"learning_rate": 9.973696248680659e-05,
"loss": 2.2687,
"step": 4235
},
{
"epoch": 0.03,
"grad_norm": 0.7517808079719543,
"learning_rate": 9.97363265376559e-05,
"loss": 2.277,
"step": 4240
},
{
"epoch": 0.03,
"grad_norm": 0.7428967952728271,
"learning_rate": 9.973568982269436e-05,
"loss": 2.1804,
"step": 4245
},
{
"epoch": 0.03,
"grad_norm": 0.7158821821212769,
"learning_rate": 9.973505234193177e-05,
"loss": 2.2149,
"step": 4250
},
{
"epoch": 0.03,
"grad_norm": 0.7148890495300293,
"learning_rate": 9.973441409537793e-05,
"loss": 2.2363,
"step": 4255
},
{
"epoch": 0.03,
"grad_norm": 0.7264882326126099,
"learning_rate": 9.973377508304269e-05,
"loss": 2.2475,
"step": 4260
},
{
"epoch": 0.03,
"grad_norm": 0.7632892727851868,
"learning_rate": 9.973313530493586e-05,
"loss": 2.2491,
"step": 4265
},
{
"epoch": 0.03,
"grad_norm": 0.7395174503326416,
"learning_rate": 9.973249476106732e-05,
"loss": 2.2347,
"step": 4270
},
{
"epoch": 0.03,
"grad_norm": 0.7253788113594055,
"learning_rate": 9.97318534514469e-05,
"loss": 2.1415,
"step": 4275
},
{
"epoch": 0.03,
"grad_norm": 0.7308273911476135,
"learning_rate": 9.97312113760845e-05,
"loss": 2.165,
"step": 4280
},
{
"epoch": 0.03,
"grad_norm": 0.7150307297706604,
"learning_rate": 9.973056853499001e-05,
"loss": 2.2543,
"step": 4285
},
{
"epoch": 0.03,
"grad_norm": 0.821225106716156,
"learning_rate": 9.972992492817331e-05,
"loss": 2.2584,
"step": 4290
},
{
"epoch": 0.03,
"grad_norm": 0.754322350025177,
"learning_rate": 9.972928055564432e-05,
"loss": 2.1977,
"step": 4295
},
{
"epoch": 0.03,
"grad_norm": 0.725344717502594,
"learning_rate": 9.972863541741297e-05,
"loss": 2.1861,
"step": 4300
},
{
"epoch": 0.03,
"grad_norm": 0.7322962284088135,
"learning_rate": 9.972798951348918e-05,
"loss": 2.1895,
"step": 4305
},
{
"epoch": 0.03,
"grad_norm": 0.7273723483085632,
"learning_rate": 9.972734284388288e-05,
"loss": 2.2539,
"step": 4310
},
{
"epoch": 0.03,
"grad_norm": 0.6831798553466797,
"learning_rate": 9.972669540860406e-05,
"loss": 2.2428,
"step": 4315
},
{
"epoch": 0.03,
"grad_norm": 0.709976077079773,
"learning_rate": 9.972604720766268e-05,
"loss": 2.2179,
"step": 4320
},
{
"epoch": 0.03,
"grad_norm": 0.7683183550834656,
"learning_rate": 9.97253982410687e-05,
"loss": 2.2163,
"step": 4325
},
{
"epoch": 0.03,
"grad_norm": 0.750195324420929,
"learning_rate": 9.972474850883213e-05,
"loss": 2.2668,
"step": 4330
},
{
"epoch": 0.03,
"grad_norm": 0.7190375924110413,
"learning_rate": 9.972409801096299e-05,
"loss": 2.2257,
"step": 4335
},
{
"epoch": 0.03,
"grad_norm": 0.7484508752822876,
"learning_rate": 9.972344674747125e-05,
"loss": 2.2467,
"step": 4340
},
{
"epoch": 0.03,
"grad_norm": 0.748528003692627,
"learning_rate": 9.972279471836698e-05,
"loss": 2.2461,
"step": 4345
},
{
"epoch": 0.03,
"grad_norm": 0.758016049861908,
"learning_rate": 9.972214192366021e-05,
"loss": 2.2537,
"step": 4350
},
{
"epoch": 0.03,
"grad_norm": 0.7227737903594971,
"learning_rate": 9.972148836336098e-05,
"loss": 2.2059,
"step": 4355
},
{
"epoch": 0.03,
"grad_norm": 0.7263017296791077,
"learning_rate": 9.972083403747937e-05,
"loss": 2.2836,
"step": 4360
},
{
"epoch": 0.03,
"grad_norm": 0.71320641040802,
"learning_rate": 9.972017894602542e-05,
"loss": 2.2828,
"step": 4365
},
{
"epoch": 0.03,
"grad_norm": 0.8004863858222961,
"learning_rate": 9.971952308900925e-05,
"loss": 2.2496,
"step": 4370
},
{
"epoch": 0.03,
"grad_norm": 0.7189190983772278,
"learning_rate": 9.971886646644096e-05,
"loss": 2.2947,
"step": 4375
},
{
"epoch": 0.03,
"grad_norm": 0.714165985584259,
"learning_rate": 9.971820907833065e-05,
"loss": 2.2711,
"step": 4380
},
{
"epoch": 0.03,
"grad_norm": 0.7419973611831665,
"learning_rate": 9.971755092468843e-05,
"loss": 2.2671,
"step": 4385
},
{
"epoch": 0.03,
"grad_norm": 0.7520475387573242,
"learning_rate": 9.971689200552445e-05,
"loss": 2.1696,
"step": 4390
},
{
"epoch": 0.03,
"grad_norm": 0.6914311051368713,
"learning_rate": 9.971623232084886e-05,
"loss": 2.2245,
"step": 4395
},
{
"epoch": 0.03,
"grad_norm": 0.7650794386863708,
"learning_rate": 9.971557187067179e-05,
"loss": 2.2847,
"step": 4400
},
{
"epoch": 0.03,
"grad_norm": 0.7144673466682434,
"learning_rate": 9.971491065500345e-05,
"loss": 2.1346,
"step": 4405
},
{
"epoch": 0.03,
"grad_norm": 0.7397557497024536,
"learning_rate": 9.9714248673854e-05,
"loss": 2.2399,
"step": 4410
},
{
"epoch": 0.03,
"grad_norm": 0.6969753503799438,
"learning_rate": 9.971358592723361e-05,
"loss": 2.2931,
"step": 4415
},
{
"epoch": 0.03,
"grad_norm": 0.7484560012817383,
"learning_rate": 9.97129224151525e-05,
"loss": 2.1705,
"step": 4420
},
{
"epoch": 0.03,
"grad_norm": 0.7202750444412231,
"learning_rate": 9.971225813762094e-05,
"loss": 2.222,
"step": 4425
},
{
"epoch": 0.03,
"grad_norm": 0.7263026237487793,
"learning_rate": 9.971159309464906e-05,
"loss": 2.2057,
"step": 4430
},
{
"epoch": 0.04,
"grad_norm": 0.7260299921035767,
"learning_rate": 9.971092728624718e-05,
"loss": 2.276,
"step": 4435
},
{
"epoch": 0.04,
"grad_norm": 0.6578490138053894,
"learning_rate": 9.97102607124255e-05,
"loss": 2.2337,
"step": 4440
},
{
"epoch": 0.04,
"grad_norm": 0.6938652396202087,
"learning_rate": 9.970959337319433e-05,
"loss": 2.3031,
"step": 4445
},
{
"epoch": 0.04,
"grad_norm": 0.7474022507667542,
"learning_rate": 9.970892526856392e-05,
"loss": 2.2649,
"step": 4450
},
{
"epoch": 0.04,
"grad_norm": 0.7217170000076294,
"learning_rate": 9.970825639854454e-05,
"loss": 2.2593,
"step": 4455
},
{
"epoch": 0.04,
"grad_norm": 0.7258909940719604,
"learning_rate": 9.970758676314652e-05,
"loss": 2.3052,
"step": 4460
},
{
"epoch": 0.04,
"grad_norm": 0.7264653444290161,
"learning_rate": 9.970691636238017e-05,
"loss": 2.2172,
"step": 4465
},
{
"epoch": 0.04,
"grad_norm": 0.7618257403373718,
"learning_rate": 9.970624519625578e-05,
"loss": 2.1979,
"step": 4470
},
{
"epoch": 0.04,
"grad_norm": 0.7145951390266418,
"learning_rate": 9.970557326478371e-05,
"loss": 2.2414,
"step": 4475
},
{
"epoch": 0.04,
"grad_norm": 0.7192901372909546,
"learning_rate": 9.97049005679743e-05,
"loss": 2.2456,
"step": 4480
},
{
"epoch": 0.04,
"grad_norm": 0.7403134107589722,
"learning_rate": 9.970422710583791e-05,
"loss": 2.2138,
"step": 4485
},
{
"epoch": 0.04,
"grad_norm": 0.6975427269935608,
"learning_rate": 9.970355287838491e-05,
"loss": 2.2344,
"step": 4490
},
{
"epoch": 0.04,
"grad_norm": 0.7178293466567993,
"learning_rate": 9.970287788562567e-05,
"loss": 2.1711,
"step": 4495
},
{
"epoch": 0.04,
"grad_norm": 0.6998232007026672,
"learning_rate": 9.97022021275706e-05,
"loss": 2.3119,
"step": 4500
},
{
"epoch": 0.04,
"grad_norm": 0.7178532481193542,
"learning_rate": 9.97015256042301e-05,
"loss": 2.2646,
"step": 4505
},
{
"epoch": 0.04,
"grad_norm": 0.732990562915802,
"learning_rate": 9.97008483156146e-05,
"loss": 2.177,
"step": 4510
},
{
"epoch": 0.04,
"grad_norm": 0.7072612643241882,
"learning_rate": 9.970017026173448e-05,
"loss": 2.1777,
"step": 4515
},
{
"epoch": 0.04,
"grad_norm": 0.7688290476799011,
"learning_rate": 9.969949144260022e-05,
"loss": 2.298,
"step": 4520
},
{
"epoch": 0.04,
"grad_norm": 0.7091464400291443,
"learning_rate": 9.969881185822227e-05,
"loss": 2.2852,
"step": 4525
},
{
"epoch": 0.04,
"grad_norm": 0.7497336268424988,
"learning_rate": 9.969813150861111e-05,
"loss": 2.229,
"step": 4530
},
{
"epoch": 0.04,
"grad_norm": 0.7008174061775208,
"learning_rate": 9.969745039377717e-05,
"loss": 2.2518,
"step": 4535
},
{
"epoch": 0.04,
"grad_norm": 0.7159837484359741,
"learning_rate": 9.969676851373098e-05,
"loss": 2.2963,
"step": 4540
},
{
"epoch": 0.04,
"grad_norm": 0.6983662247657776,
"learning_rate": 9.969608586848301e-05,
"loss": 2.2575,
"step": 4545
},
{
"epoch": 0.04,
"grad_norm": 0.7038892507553101,
"learning_rate": 9.969540245804377e-05,
"loss": 2.2714,
"step": 4550
},
{
"epoch": 0.04,
"grad_norm": 0.7668149471282959,
"learning_rate": 9.96947182824238e-05,
"loss": 2.2092,
"step": 4555
},
{
"epoch": 0.04,
"grad_norm": 0.7428255081176758,
"learning_rate": 9.969403334163364e-05,
"loss": 2.1279,
"step": 4560
},
{
"epoch": 0.04,
"grad_norm": 0.7532406449317932,
"learning_rate": 9.969334763568383e-05,
"loss": 2.264,
"step": 4565
},
{
"epoch": 0.04,
"grad_norm": 0.7155984044075012,
"learning_rate": 9.969266116458491e-05,
"loss": 2.1568,
"step": 4570
},
{
"epoch": 0.04,
"grad_norm": 0.7309367060661316,
"learning_rate": 9.969197392834747e-05,
"loss": 2.2181,
"step": 4575
},
{
"epoch": 0.04,
"grad_norm": 0.7706004977226257,
"learning_rate": 9.969128592698207e-05,
"loss": 2.1938,
"step": 4580
},
{
"epoch": 0.04,
"grad_norm": 0.7427154183387756,
"learning_rate": 9.969059716049934e-05,
"loss": 2.2491,
"step": 4585
},
{
"epoch": 0.04,
"grad_norm": 0.7017691135406494,
"learning_rate": 9.968990762890986e-05,
"loss": 2.1962,
"step": 4590
},
{
"epoch": 0.04,
"grad_norm": 0.742956280708313,
"learning_rate": 9.968921733222425e-05,
"loss": 2.2426,
"step": 4595
},
{
"epoch": 0.04,
"grad_norm": 0.7034217715263367,
"learning_rate": 9.968852627045313e-05,
"loss": 2.2006,
"step": 4600
},
{
"epoch": 0.04,
"grad_norm": 0.7533945441246033,
"learning_rate": 9.968783444360716e-05,
"loss": 2.197,
"step": 4605
},
{
"epoch": 0.04,
"grad_norm": 0.737179696559906,
"learning_rate": 9.968714185169698e-05,
"loss": 2.1797,
"step": 4610
},
{
"epoch": 0.04,
"grad_norm": 0.7167485356330872,
"learning_rate": 9.968644849473326e-05,
"loss": 2.1809,
"step": 4615
},
{
"epoch": 0.04,
"grad_norm": 0.7324991226196289,
"learning_rate": 9.968575437272667e-05,
"loss": 2.2634,
"step": 4620
},
{
"epoch": 0.04,
"grad_norm": 0.7531011700630188,
"learning_rate": 9.968505948568792e-05,
"loss": 2.218,
"step": 4625
},
{
"epoch": 0.04,
"grad_norm": 0.7397653460502625,
"learning_rate": 9.968436383362766e-05,
"loss": 2.2372,
"step": 4630
},
{
"epoch": 0.04,
"grad_norm": 0.7288955450057983,
"learning_rate": 9.968366741655663e-05,
"loss": 2.2111,
"step": 4635
},
{
"epoch": 0.04,
"grad_norm": 0.7249389290809631,
"learning_rate": 9.968297023448558e-05,
"loss": 2.2678,
"step": 4640
},
{
"epoch": 0.04,
"grad_norm": 0.7195791602134705,
"learning_rate": 9.96822722874252e-05,
"loss": 2.2444,
"step": 4645
},
{
"epoch": 0.04,
"grad_norm": 0.727279782295227,
"learning_rate": 9.968157357538625e-05,
"loss": 2.2285,
"step": 4650
},
{
"epoch": 0.04,
"grad_norm": 0.7133840918540955,
"learning_rate": 9.968087409837951e-05,
"loss": 2.2741,
"step": 4655
},
{
"epoch": 0.04,
"grad_norm": 0.681530773639679,
"learning_rate": 9.968017385641574e-05,
"loss": 2.2366,
"step": 4660
},
{
"epoch": 0.04,
"grad_norm": 0.7042701244354248,
"learning_rate": 9.967947284950569e-05,
"loss": 2.1761,
"step": 4665
},
{
"epoch": 0.04,
"grad_norm": 0.6984108090400696,
"learning_rate": 9.967877107766021e-05,
"loss": 2.1257,
"step": 4670
},
{
"epoch": 0.04,
"grad_norm": 0.7614035606384277,
"learning_rate": 9.967806854089006e-05,
"loss": 2.245,
"step": 4675
},
{
"epoch": 0.04,
"grad_norm": 0.7259561419487,
"learning_rate": 9.967736523920607e-05,
"loss": 2.24,
"step": 4680
},
{
"epoch": 0.04,
"grad_norm": 0.724139928817749,
"learning_rate": 9.967666117261907e-05,
"loss": 2.2996,
"step": 4685
},
{
"epoch": 0.04,
"grad_norm": 0.7268062233924866,
"learning_rate": 9.96759563411399e-05,
"loss": 2.172,
"step": 4690
},
{
"epoch": 0.04,
"grad_norm": 0.70979905128479,
"learning_rate": 9.967525074477943e-05,
"loss": 2.2257,
"step": 4695
},
{
"epoch": 0.04,
"grad_norm": 0.7039162516593933,
"learning_rate": 9.96745443835485e-05,
"loss": 2.246,
"step": 4700
},
{
"epoch": 0.04,
"grad_norm": 0.7254780530929565,
"learning_rate": 9.967383725745799e-05,
"loss": 2.1863,
"step": 4705
},
{
"epoch": 0.04,
"grad_norm": 0.767579972743988,
"learning_rate": 9.96731293665188e-05,
"loss": 2.178,
"step": 4710
},
{
"epoch": 0.04,
"grad_norm": 0.7310457825660706,
"learning_rate": 9.967242071074181e-05,
"loss": 2.2462,
"step": 4715
},
{
"epoch": 0.04,
"grad_norm": 0.7475757598876953,
"learning_rate": 9.967171129013797e-05,
"loss": 2.1078,
"step": 4720
},
{
"epoch": 0.04,
"grad_norm": 0.742080807685852,
"learning_rate": 9.967100110471817e-05,
"loss": 2.2506,
"step": 4725
},
{
"epoch": 0.04,
"grad_norm": 0.7402095198631287,
"learning_rate": 9.967029015449334e-05,
"loss": 2.2946,
"step": 4730
},
{
"epoch": 0.04,
"grad_norm": 0.8603042960166931,
"learning_rate": 9.966957843947444e-05,
"loss": 2.2475,
"step": 4735
},
{
"epoch": 0.04,
"grad_norm": 0.7042680978775024,
"learning_rate": 9.966886595967243e-05,
"loss": 2.2286,
"step": 4740
},
{
"epoch": 0.04,
"grad_norm": 0.7266623973846436,
"learning_rate": 9.966815271509828e-05,
"loss": 2.2,
"step": 4745
},
{
"epoch": 0.04,
"grad_norm": 0.7678672671318054,
"learning_rate": 9.966743870576297e-05,
"loss": 2.2005,
"step": 4750
},
{
"epoch": 0.04,
"grad_norm": 0.7180349230766296,
"learning_rate": 9.96667239316775e-05,
"loss": 2.1854,
"step": 4755
},
{
"epoch": 0.04,
"grad_norm": 0.6994274854660034,
"learning_rate": 9.966600839285286e-05,
"loss": 2.168,
"step": 4760
},
{
"epoch": 0.04,
"grad_norm": 0.7370887398719788,
"learning_rate": 9.966529208930008e-05,
"loss": 2.1542,
"step": 4765
},
{
"epoch": 0.04,
"grad_norm": 0.7234534025192261,
"learning_rate": 9.96645750210302e-05,
"loss": 2.2899,
"step": 4770
},
{
"epoch": 0.04,
"grad_norm": 0.7281872034072876,
"learning_rate": 9.966385718805423e-05,
"loss": 2.1794,
"step": 4775
},
{
"epoch": 0.04,
"grad_norm": 0.7975899577140808,
"learning_rate": 9.966313859038325e-05,
"loss": 2.2268,
"step": 4780
},
{
"epoch": 0.04,
"grad_norm": 0.6825891733169556,
"learning_rate": 9.966241922802832e-05,
"loss": 2.1266,
"step": 4785
},
{
"epoch": 0.04,
"grad_norm": 0.7317417860031128,
"learning_rate": 9.96616991010005e-05,
"loss": 2.2099,
"step": 4790
},
{
"epoch": 0.04,
"grad_norm": 0.6832736134529114,
"learning_rate": 9.96609782093109e-05,
"loss": 2.1494,
"step": 4795
},
{
"epoch": 0.04,
"grad_norm": 0.738480269908905,
"learning_rate": 9.96602565529706e-05,
"loss": 2.2318,
"step": 4800
},
{
"epoch": 0.04,
"grad_norm": 0.7467361688613892,
"learning_rate": 9.965953413199071e-05,
"loss": 2.2183,
"step": 4805
},
{
"epoch": 0.04,
"grad_norm": 0.8224136233329773,
"learning_rate": 9.965881094638238e-05,
"loss": 2.2255,
"step": 4810
},
{
"epoch": 0.04,
"grad_norm": 0.681725263595581,
"learning_rate": 9.965808699615674e-05,
"loss": 2.2826,
"step": 4815
},
{
"epoch": 0.04,
"grad_norm": 0.7035905122756958,
"learning_rate": 9.965736228132492e-05,
"loss": 2.2454,
"step": 4820
},
{
"epoch": 0.04,
"grad_norm": 0.8092908263206482,
"learning_rate": 9.965663680189807e-05,
"loss": 2.2435,
"step": 4825
},
{
"epoch": 0.04,
"grad_norm": 0.6941966414451599,
"learning_rate": 9.965591055788738e-05,
"loss": 2.2705,
"step": 4830
},
{
"epoch": 0.04,
"grad_norm": 0.692603349685669,
"learning_rate": 9.965518354930405e-05,
"loss": 2.2345,
"step": 4835
},
{
"epoch": 0.04,
"grad_norm": 0.6880177855491638,
"learning_rate": 9.965445577615922e-05,
"loss": 2.2439,
"step": 4840
},
{
"epoch": 0.04,
"grad_norm": 0.8022019863128662,
"learning_rate": 9.965372723846415e-05,
"loss": 2.187,
"step": 4845
},
{
"epoch": 0.04,
"grad_norm": 0.7054466605186462,
"learning_rate": 9.965299793623003e-05,
"loss": 2.2986,
"step": 4850
},
{
"epoch": 0.04,
"grad_norm": 0.7252600193023682,
"learning_rate": 9.96522678694681e-05,
"loss": 2.2328,
"step": 4855
},
{
"epoch": 0.04,
"grad_norm": 0.7630112767219543,
"learning_rate": 9.965153703818959e-05,
"loss": 2.2719,
"step": 4860
},
{
"epoch": 0.04,
"grad_norm": 0.7513236999511719,
"learning_rate": 9.965080544240575e-05,
"loss": 2.2807,
"step": 4865
},
{
"epoch": 0.04,
"grad_norm": 0.7120503783226013,
"learning_rate": 9.965007308212786e-05,
"loss": 2.2084,
"step": 4870
},
{
"epoch": 0.04,
"grad_norm": 0.7394705414772034,
"learning_rate": 9.964933995736721e-05,
"loss": 2.2512,
"step": 4875
},
{
"epoch": 0.04,
"grad_norm": 0.6862260103225708,
"learning_rate": 9.964860606813504e-05,
"loss": 2.1711,
"step": 4880
},
{
"epoch": 0.04,
"grad_norm": 0.7489215731620789,
"learning_rate": 9.96478714144427e-05,
"loss": 2.3017,
"step": 4885
},
{
"epoch": 0.04,
"grad_norm": 0.6942036151885986,
"learning_rate": 9.964713599630146e-05,
"loss": 2.1424,
"step": 4890
},
{
"epoch": 0.04,
"grad_norm": 0.7072425484657288,
"learning_rate": 9.964639981372267e-05,
"loss": 2.2107,
"step": 4895
},
{
"epoch": 0.04,
"grad_norm": 0.6967791318893433,
"learning_rate": 9.964566286671766e-05,
"loss": 2.2539,
"step": 4900
},
{
"epoch": 0.04,
"grad_norm": 0.7189387083053589,
"learning_rate": 9.964492515529777e-05,
"loss": 2.2659,
"step": 4905
},
{
"epoch": 0.04,
"grad_norm": 0.7067403197288513,
"learning_rate": 9.964418667947438e-05,
"loss": 2.2221,
"step": 4910
},
{
"epoch": 0.04,
"grad_norm": 0.7799524068832397,
"learning_rate": 9.964344743925884e-05,
"loss": 2.2289,
"step": 4915
},
{
"epoch": 0.04,
"grad_norm": 0.7533849477767944,
"learning_rate": 9.964270743466253e-05,
"loss": 2.2054,
"step": 4920
},
{
"epoch": 0.04,
"grad_norm": 0.71812903881073,
"learning_rate": 9.964196666569686e-05,
"loss": 2.2163,
"step": 4925
},
{
"epoch": 0.04,
"grad_norm": 0.702627420425415,
"learning_rate": 9.964122513237322e-05,
"loss": 2.2544,
"step": 4930
},
{
"epoch": 0.04,
"grad_norm": 0.7249962687492371,
"learning_rate": 9.964048283470305e-05,
"loss": 2.2429,
"step": 4935
},
{
"epoch": 0.04,
"grad_norm": 0.7425085306167603,
"learning_rate": 9.963973977269776e-05,
"loss": 2.3028,
"step": 4940
},
{
"epoch": 0.04,
"grad_norm": 0.7217290997505188,
"learning_rate": 9.963899594636879e-05,
"loss": 2.2321,
"step": 4945
},
{
"epoch": 0.04,
"grad_norm": 0.7092394232749939,
"learning_rate": 9.96382513557276e-05,
"loss": 2.2403,
"step": 4950
},
{
"epoch": 0.04,
"grad_norm": 0.695550799369812,
"learning_rate": 9.963750600078565e-05,
"loss": 2.276,
"step": 4955
},
{
"epoch": 0.04,
"grad_norm": 0.7182446122169495,
"learning_rate": 9.963675988155443e-05,
"loss": 2.2516,
"step": 4960
},
{
"epoch": 0.04,
"grad_norm": 0.7196715474128723,
"learning_rate": 9.963601299804541e-05,
"loss": 2.24,
"step": 4965
},
{
"epoch": 0.04,
"grad_norm": 0.7121938467025757,
"learning_rate": 9.963526535027012e-05,
"loss": 2.2156,
"step": 4970
},
{
"epoch": 0.04,
"grad_norm": 0.6960738301277161,
"learning_rate": 9.963451693824003e-05,
"loss": 2.2592,
"step": 4975
},
{
"epoch": 0.04,
"grad_norm": 0.7596395015716553,
"learning_rate": 9.963376776196669e-05,
"loss": 2.202,
"step": 4980
},
{
"epoch": 0.04,
"grad_norm": 0.70936518907547,
"learning_rate": 9.963301782146163e-05,
"loss": 2.2367,
"step": 4985
},
{
"epoch": 0.04,
"grad_norm": 0.7160218954086304,
"learning_rate": 9.96322671167364e-05,
"loss": 2.1929,
"step": 4990
},
{
"epoch": 0.04,
"grad_norm": 0.9338270425796509,
"learning_rate": 9.963151564780256e-05,
"loss": 2.168,
"step": 4995
},
{
"epoch": 0.04,
"grad_norm": 0.7260292172431946,
"learning_rate": 9.963076341467165e-05,
"loss": 2.2334,
"step": 5000
},
{
"epoch": 0.04,
"grad_norm": 0.7245514392852783,
"learning_rate": 9.963001041735531e-05,
"loss": 2.2504,
"step": 5005
},
{
"epoch": 0.04,
"grad_norm": 0.6986328959465027,
"learning_rate": 9.962925665586509e-05,
"loss": 2.2057,
"step": 5010
},
{
"epoch": 0.04,
"grad_norm": 0.713405966758728,
"learning_rate": 9.962850213021262e-05,
"loss": 2.1807,
"step": 5015
},
{
"epoch": 0.04,
"grad_norm": 0.7159180641174316,
"learning_rate": 9.96277468404095e-05,
"loss": 2.1457,
"step": 5020
},
{
"epoch": 0.04,
"grad_norm": 0.6760283708572388,
"learning_rate": 9.962699078646735e-05,
"loss": 2.1942,
"step": 5025
},
{
"epoch": 0.04,
"grad_norm": 0.7030491828918457,
"learning_rate": 9.962623396839785e-05,
"loss": 2.2306,
"step": 5030
},
{
"epoch": 0.04,
"grad_norm": 0.688221275806427,
"learning_rate": 9.962547638621262e-05,
"loss": 2.192,
"step": 5035
},
{
"epoch": 0.04,
"grad_norm": 0.708469033241272,
"learning_rate": 9.962471803992334e-05,
"loss": 2.203,
"step": 5040
},
{
"epoch": 0.04,
"grad_norm": 0.7042374610900879,
"learning_rate": 9.962395892954168e-05,
"loss": 2.1659,
"step": 5045
},
{
"epoch": 0.04,
"grad_norm": 0.6953795552253723,
"learning_rate": 9.962319905507934e-05,
"loss": 2.2556,
"step": 5050
},
{
"epoch": 0.04,
"grad_norm": 0.6869887113571167,
"learning_rate": 9.9622438416548e-05,
"loss": 2.1914,
"step": 5055
},
{
"epoch": 0.04,
"grad_norm": 0.7187615036964417,
"learning_rate": 9.962167701395937e-05,
"loss": 2.2061,
"step": 5060
},
{
"epoch": 0.04,
"grad_norm": 0.6694822311401367,
"learning_rate": 9.962091484732521e-05,
"loss": 2.2175,
"step": 5065
},
{
"epoch": 0.04,
"grad_norm": 0.8211686611175537,
"learning_rate": 9.962015191665722e-05,
"loss": 2.2099,
"step": 5070
},
{
"epoch": 0.04,
"grad_norm": 0.7291555404663086,
"learning_rate": 9.961938822196716e-05,
"loss": 2.2419,
"step": 5075
},
{
"epoch": 0.04,
"grad_norm": 0.6889863610267639,
"learning_rate": 9.961862376326678e-05,
"loss": 2.2326,
"step": 5080
},
{
"epoch": 0.04,
"grad_norm": 0.7056418657302856,
"learning_rate": 9.961785854056788e-05,
"loss": 2.164,
"step": 5085
},
{
"epoch": 0.04,
"grad_norm": 0.7104595303535461,
"learning_rate": 9.961709255388221e-05,
"loss": 2.172,
"step": 5090
},
{
"epoch": 0.04,
"grad_norm": 0.7055512070655823,
"learning_rate": 9.961632580322156e-05,
"loss": 2.1341,
"step": 5095
},
{
"epoch": 0.04,
"grad_norm": 0.7353788614273071,
"learning_rate": 9.961555828859777e-05,
"loss": 2.185,
"step": 5100
},
{
"epoch": 0.04,
"grad_norm": 0.6913115978240967,
"learning_rate": 9.961479001002264e-05,
"loss": 2.2055,
"step": 5105
},
{
"epoch": 0.04,
"grad_norm": 0.7767313718795776,
"learning_rate": 9.961402096750799e-05,
"loss": 2.207,
"step": 5110
},
{
"epoch": 0.04,
"grad_norm": 0.708200991153717,
"learning_rate": 9.961325116106566e-05,
"loss": 2.1826,
"step": 5115
},
{
"epoch": 0.04,
"grad_norm": 0.7543495893478394,
"learning_rate": 9.961248059070751e-05,
"loss": 2.3205,
"step": 5120
},
{
"epoch": 0.04,
"grad_norm": 0.7465729713439941,
"learning_rate": 9.961170925644542e-05,
"loss": 2.2511,
"step": 5125
},
{
"epoch": 0.04,
"grad_norm": 0.7517914175987244,
"learning_rate": 9.961093715829127e-05,
"loss": 2.2091,
"step": 5130
},
{
"epoch": 0.04,
"grad_norm": 0.730800211429596,
"learning_rate": 9.961016429625691e-05,
"loss": 2.134,
"step": 5135
},
{
"epoch": 0.04,
"grad_norm": 0.7363988757133484,
"learning_rate": 9.960939067035426e-05,
"loss": 2.2212,
"step": 5140
},
{
"epoch": 0.04,
"grad_norm": 0.7111669182777405,
"learning_rate": 9.960861628059524e-05,
"loss": 2.2031,
"step": 5145
},
{
"epoch": 0.04,
"grad_norm": 0.6847307682037354,
"learning_rate": 9.960784112699177e-05,
"loss": 2.1494,
"step": 5150
},
{
"epoch": 0.04,
"grad_norm": 0.7179961204528809,
"learning_rate": 9.960706520955578e-05,
"loss": 2.1818,
"step": 5155
},
{
"epoch": 0.04,
"grad_norm": 0.7111839652061462,
"learning_rate": 9.960628852829923e-05,
"loss": 2.212,
"step": 5160
},
{
"epoch": 0.04,
"grad_norm": 0.7657085657119751,
"learning_rate": 9.960551108323406e-05,
"loss": 2.2171,
"step": 5165
},
{
"epoch": 0.04,
"grad_norm": 0.6852431297302246,
"learning_rate": 9.960473287437224e-05,
"loss": 2.2191,
"step": 5170
},
{
"epoch": 0.04,
"grad_norm": 0.7360872030258179,
"learning_rate": 9.960395390172579e-05,
"loss": 2.16,
"step": 5175
},
{
"epoch": 0.04,
"grad_norm": 0.7447972297668457,
"learning_rate": 9.960317416530665e-05,
"loss": 2.1885,
"step": 5180
},
{
"epoch": 0.04,
"grad_norm": 0.6946461200714111,
"learning_rate": 9.960239366512686e-05,
"loss": 2.213,
"step": 5185
},
{
"epoch": 0.04,
"grad_norm": 0.7004071474075317,
"learning_rate": 9.960161240119843e-05,
"loss": 2.271,
"step": 5190
},
{
"epoch": 0.04,
"grad_norm": 0.7137370705604553,
"learning_rate": 9.96008303735334e-05,
"loss": 2.226,
"step": 5195
},
{
"epoch": 0.04,
"grad_norm": 0.6999104619026184,
"learning_rate": 9.960004758214379e-05,
"loss": 2.1694,
"step": 5200
},
{
"epoch": 0.04,
"grad_norm": 0.6922555565834045,
"learning_rate": 9.959926402704165e-05,
"loss": 2.2246,
"step": 5205
},
{
"epoch": 0.04,
"grad_norm": 0.6644160747528076,
"learning_rate": 9.959847970823908e-05,
"loss": 2.2148,
"step": 5210
},
{
"epoch": 0.04,
"grad_norm": 0.726941704750061,
"learning_rate": 9.959769462574813e-05,
"loss": 2.0995,
"step": 5215
},
{
"epoch": 0.04,
"grad_norm": 0.7327321767807007,
"learning_rate": 9.959690877958088e-05,
"loss": 2.1338,
"step": 5220
},
{
"epoch": 0.04,
"grad_norm": 0.7129268646240234,
"learning_rate": 9.959612216974946e-05,
"loss": 2.1489,
"step": 5225
},
{
"epoch": 0.04,
"grad_norm": 0.6836938261985779,
"learning_rate": 9.959533479626596e-05,
"loss": 2.2023,
"step": 5230
},
{
"epoch": 0.04,
"grad_norm": 0.708266019821167,
"learning_rate": 9.95945466591425e-05,
"loss": 2.1119,
"step": 5235
},
{
"epoch": 0.04,
"grad_norm": 0.7388333678245544,
"learning_rate": 9.959375775839124e-05,
"loss": 2.1814,
"step": 5240
},
{
"epoch": 0.04,
"grad_norm": 0.6854376792907715,
"learning_rate": 9.95929680940243e-05,
"loss": 2.0302,
"step": 5245
},
{
"epoch": 0.04,
"grad_norm": 0.7159355282783508,
"learning_rate": 9.959217766605384e-05,
"loss": 2.175,
"step": 5250
},
{
"epoch": 0.04,
"grad_norm": 0.7089496850967407,
"learning_rate": 9.959138647449205e-05,
"loss": 2.2657,
"step": 5255
},
{
"epoch": 0.04,
"grad_norm": 0.7049574851989746,
"learning_rate": 9.95905945193511e-05,
"loss": 2.2674,
"step": 5260
},
{
"epoch": 0.04,
"grad_norm": 0.7326235771179199,
"learning_rate": 9.958980180064318e-05,
"loss": 2.2012,
"step": 5265
},
{
"epoch": 0.04,
"grad_norm": 0.7122628688812256,
"learning_rate": 9.958900831838051e-05,
"loss": 2.2198,
"step": 5270
},
{
"epoch": 0.04,
"grad_norm": 0.7207251787185669,
"learning_rate": 9.958821407257531e-05,
"loss": 2.1596,
"step": 5275
},
{
"epoch": 0.04,
"grad_norm": 0.6934665441513062,
"learning_rate": 9.958741906323978e-05,
"loss": 2.1647,
"step": 5280
},
{
"epoch": 0.04,
"grad_norm": 0.690761923789978,
"learning_rate": 9.95866232903862e-05,
"loss": 2.2203,
"step": 5285
},
{
"epoch": 0.04,
"grad_norm": 0.6896169781684875,
"learning_rate": 9.958582675402677e-05,
"loss": 2.1285,
"step": 5290
},
{
"epoch": 0.04,
"grad_norm": 0.6797572374343872,
"learning_rate": 9.958502945417382e-05,
"loss": 2.2624,
"step": 5295
},
{
"epoch": 0.04,
"grad_norm": 0.6790597438812256,
"learning_rate": 9.958423139083958e-05,
"loss": 2.1676,
"step": 5300
},
{
"epoch": 0.04,
"grad_norm": 0.7130025029182434,
"learning_rate": 9.958343256403635e-05,
"loss": 2.1797,
"step": 5305
},
{
"epoch": 0.04,
"grad_norm": 0.7241025567054749,
"learning_rate": 9.958263297377643e-05,
"loss": 2.2552,
"step": 5310
},
{
"epoch": 0.04,
"grad_norm": 0.715133011341095,
"learning_rate": 9.958183262007215e-05,
"loss": 2.1525,
"step": 5315
},
{
"epoch": 0.04,
"grad_norm": 0.7171980142593384,
"learning_rate": 9.95810315029358e-05,
"loss": 2.2468,
"step": 5320
},
{
"epoch": 0.04,
"grad_norm": 0.7376137375831604,
"learning_rate": 9.958022962237974e-05,
"loss": 2.2119,
"step": 5325
},
{
"epoch": 0.04,
"grad_norm": 0.7065651416778564,
"learning_rate": 9.957942697841632e-05,
"loss": 2.2398,
"step": 5330
},
{
"epoch": 0.04,
"grad_norm": 0.720744788646698,
"learning_rate": 9.957862357105785e-05,
"loss": 2.141,
"step": 5335
},
{
"epoch": 0.04,
"grad_norm": 0.7155617475509644,
"learning_rate": 9.957781940031676e-05,
"loss": 2.1679,
"step": 5340
},
{
"epoch": 0.04,
"grad_norm": 0.6631587743759155,
"learning_rate": 9.957701446620543e-05,
"loss": 2.2492,
"step": 5345
},
{
"epoch": 0.04,
"grad_norm": 0.7102684378623962,
"learning_rate": 9.957620876873621e-05,
"loss": 2.2144,
"step": 5350
},
{
"epoch": 0.04,
"grad_norm": 0.7251484394073486,
"learning_rate": 9.957540230792152e-05,
"loss": 2.2252,
"step": 5355
},
{
"epoch": 0.04,
"grad_norm": 0.705807089805603,
"learning_rate": 9.957459508377381e-05,
"loss": 2.2182,
"step": 5360
},
{
"epoch": 0.04,
"grad_norm": 0.7030643820762634,
"learning_rate": 9.957378709630548e-05,
"loss": 2.1884,
"step": 5365
},
{
"epoch": 0.04,
"grad_norm": 0.7358178496360779,
"learning_rate": 9.957297834552898e-05,
"loss": 2.2156,
"step": 5370
},
{
"epoch": 0.04,
"grad_norm": 0.7052801251411438,
"learning_rate": 9.957216883145675e-05,
"loss": 2.2024,
"step": 5375
},
{
"epoch": 0.04,
"grad_norm": 0.7384359836578369,
"learning_rate": 9.957135855410127e-05,
"loss": 2.2846,
"step": 5380
},
{
"epoch": 0.04,
"grad_norm": 0.7037367224693298,
"learning_rate": 9.957054751347501e-05,
"loss": 2.241,
"step": 5385
},
{
"epoch": 0.04,
"grad_norm": 0.7225118279457092,
"learning_rate": 9.956973570959045e-05,
"loss": 2.1166,
"step": 5390
},
{
"epoch": 0.04,
"grad_norm": 0.7181528210639954,
"learning_rate": 9.956892314246012e-05,
"loss": 2.2003,
"step": 5395
},
{
"epoch": 0.04,
"grad_norm": 0.682701051235199,
"learning_rate": 9.956810981209647e-05,
"loss": 2.2509,
"step": 5400
},
{
"epoch": 0.04,
"grad_norm": 0.7236567139625549,
"learning_rate": 9.956729571851211e-05,
"loss": 2.1095,
"step": 5405
},
{
"epoch": 0.04,
"grad_norm": 0.7214797735214233,
"learning_rate": 9.956648086171949e-05,
"loss": 2.3123,
"step": 5410
},
{
"epoch": 0.04,
"grad_norm": 0.7397012710571289,
"learning_rate": 9.956566524173121e-05,
"loss": 2.2253,
"step": 5415
},
{
"epoch": 0.04,
"grad_norm": 0.7206079363822937,
"learning_rate": 9.956484885855982e-05,
"loss": 2.1462,
"step": 5420
},
{
"epoch": 0.04,
"grad_norm": 0.7026817202568054,
"learning_rate": 9.956403171221786e-05,
"loss": 2.1525,
"step": 5425
},
{
"epoch": 0.04,
"grad_norm": 0.670805037021637,
"learning_rate": 9.956321380271796e-05,
"loss": 2.1785,
"step": 5430
},
{
"epoch": 0.04,
"grad_norm": 0.6676881313323975,
"learning_rate": 9.956239513007268e-05,
"loss": 2.1308,
"step": 5435
},
{
"epoch": 0.04,
"grad_norm": 0.712836503982544,
"learning_rate": 9.956157569429464e-05,
"loss": 2.1776,
"step": 5440
},
{
"epoch": 0.04,
"grad_norm": 0.6621427536010742,
"learning_rate": 9.956075549539643e-05,
"loss": 2.152,
"step": 5445
},
{
"epoch": 0.04,
"grad_norm": 0.7493962645530701,
"learning_rate": 9.955993453339072e-05,
"loss": 2.1906,
"step": 5450
},
{
"epoch": 0.04,
"grad_norm": 0.7056708335876465,
"learning_rate": 9.955911280829011e-05,
"loss": 2.2746,
"step": 5455
},
{
"epoch": 0.04,
"grad_norm": 0.6907510161399841,
"learning_rate": 9.95582903201073e-05,
"loss": 2.1648,
"step": 5460
},
{
"epoch": 0.04,
"grad_norm": 0.691606879234314,
"learning_rate": 9.955746706885492e-05,
"loss": 2.209,
"step": 5465
},
{
"epoch": 0.04,
"grad_norm": 0.7153112292289734,
"learning_rate": 9.955664305454564e-05,
"loss": 2.1979,
"step": 5470
},
{
"epoch": 0.04,
"grad_norm": 0.6961346864700317,
"learning_rate": 9.955581827719216e-05,
"loss": 2.2173,
"step": 5475
},
{
"epoch": 0.04,
"grad_norm": 0.7145130038261414,
"learning_rate": 9.955499273680719e-05,
"loss": 2.0979,
"step": 5480
},
{
"epoch": 0.04,
"grad_norm": 0.7034791111946106,
"learning_rate": 9.955416643340343e-05,
"loss": 2.1475,
"step": 5485
},
{
"epoch": 0.04,
"grad_norm": 0.7046310901641846,
"learning_rate": 9.95533393669936e-05,
"loss": 2.1544,
"step": 5490
},
{
"epoch": 0.04,
"grad_norm": 0.7015929222106934,
"learning_rate": 9.955251153759044e-05,
"loss": 2.2323,
"step": 5495
},
{
"epoch": 0.04,
"grad_norm": 0.8292680978775024,
"learning_rate": 9.955168294520672e-05,
"loss": 2.1497,
"step": 5500
},
{
"epoch": 0.04,
"grad_norm": 0.6957909464836121,
"learning_rate": 9.955085358985514e-05,
"loss": 2.0698,
"step": 5505
},
{
"epoch": 0.04,
"grad_norm": 0.6980013847351074,
"learning_rate": 9.955002347154851e-05,
"loss": 2.228,
"step": 5510
},
{
"epoch": 0.04,
"grad_norm": 0.7237197756767273,
"learning_rate": 9.954919259029962e-05,
"loss": 2.1316,
"step": 5515
},
{
"epoch": 0.04,
"grad_norm": 0.7054796814918518,
"learning_rate": 9.954836094612123e-05,
"loss": 2.2475,
"step": 5520
},
{
"epoch": 0.04,
"grad_norm": 0.6966119408607483,
"learning_rate": 9.954752853902619e-05,
"loss": 2.1714,
"step": 5525
},
{
"epoch": 0.04,
"grad_norm": 0.7374390959739685,
"learning_rate": 9.954669536902728e-05,
"loss": 2.2188,
"step": 5530
},
{
"epoch": 0.04,
"grad_norm": 0.6977290511131287,
"learning_rate": 9.954586143613733e-05,
"loss": 2.1777,
"step": 5535
},
{
"epoch": 0.04,
"grad_norm": 0.6929000616073608,
"learning_rate": 9.954502674036919e-05,
"loss": 2.2685,
"step": 5540
},
{
"epoch": 0.04,
"grad_norm": 0.6713245511054993,
"learning_rate": 9.954419128173572e-05,
"loss": 2.1893,
"step": 5545
},
{
"epoch": 0.04,
"grad_norm": 0.7022677659988403,
"learning_rate": 9.954335506024978e-05,
"loss": 2.1864,
"step": 5550
},
{
"epoch": 0.04,
"grad_norm": 0.7587020397186279,
"learning_rate": 9.954251807592423e-05,
"loss": 2.1889,
"step": 5555
},
{
"epoch": 0.04,
"grad_norm": 0.7291474938392639,
"learning_rate": 9.954168032877198e-05,
"loss": 2.1775,
"step": 5560
},
{
"epoch": 0.04,
"grad_norm": 0.6754527688026428,
"learning_rate": 9.95408418188059e-05,
"loss": 2.1879,
"step": 5565
},
{
"epoch": 0.04,
"grad_norm": 0.7864285111427307,
"learning_rate": 9.954000254603894e-05,
"loss": 2.2016,
"step": 5570
},
{
"epoch": 0.04,
"grad_norm": 0.6909163594245911,
"learning_rate": 9.9539162510484e-05,
"loss": 2.1667,
"step": 5575
},
{
"epoch": 0.04,
"grad_norm": 0.7204089760780334,
"learning_rate": 9.953832171215401e-05,
"loss": 2.2056,
"step": 5580
},
{
"epoch": 0.04,
"grad_norm": 0.6933563947677612,
"learning_rate": 9.953748015106193e-05,
"loss": 2.2017,
"step": 5585
},
{
"epoch": 0.04,
"grad_norm": 0.7131657600402832,
"learning_rate": 9.953663782722069e-05,
"loss": 2.1927,
"step": 5590
},
{
"epoch": 0.04,
"grad_norm": 0.7181323170661926,
"learning_rate": 9.953579474064329e-05,
"loss": 2.1474,
"step": 5595
},
{
"epoch": 0.04,
"grad_norm": 0.7145206332206726,
"learning_rate": 9.95349508913427e-05,
"loss": 2.2216,
"step": 5600
},
{
"epoch": 0.04,
"grad_norm": 0.7319287061691284,
"learning_rate": 9.953410627933193e-05,
"loss": 2.2024,
"step": 5605
},
{
"epoch": 0.04,
"grad_norm": 0.7743222117424011,
"learning_rate": 9.953326090462397e-05,
"loss": 2.2221,
"step": 5610
},
{
"epoch": 0.04,
"grad_norm": 0.7674858570098877,
"learning_rate": 9.953241476723181e-05,
"loss": 2.1211,
"step": 5615
},
{
"epoch": 0.04,
"grad_norm": 0.6990914344787598,
"learning_rate": 9.953156786716851e-05,
"loss": 2.1971,
"step": 5620
},
{
"epoch": 0.04,
"grad_norm": 0.7289800047874451,
"learning_rate": 9.953072020444711e-05,
"loss": 2.1779,
"step": 5625
},
{
"epoch": 0.04,
"grad_norm": 0.6719769239425659,
"learning_rate": 9.952987177908066e-05,
"loss": 2.0944,
"step": 5630
},
{
"epoch": 0.04,
"grad_norm": 0.6976754665374756,
"learning_rate": 9.95290225910822e-05,
"loss": 2.2491,
"step": 5635
},
{
"epoch": 0.04,
"grad_norm": 0.678924024105072,
"learning_rate": 9.952817264046486e-05,
"loss": 2.1459,
"step": 5640
},
{
"epoch": 0.04,
"grad_norm": 0.7239236235618591,
"learning_rate": 9.952732192724167e-05,
"loss": 2.2034,
"step": 5645
},
{
"epoch": 0.04,
"grad_norm": 0.7184574604034424,
"learning_rate": 9.952647045142576e-05,
"loss": 2.2519,
"step": 5650
},
{
"epoch": 0.04,
"grad_norm": 0.6768973469734192,
"learning_rate": 9.952561821303021e-05,
"loss": 2.256,
"step": 5655
},
{
"epoch": 0.04,
"grad_norm": 0.6629990935325623,
"learning_rate": 9.952476521206818e-05,
"loss": 2.1622,
"step": 5660
},
{
"epoch": 0.04,
"grad_norm": 0.727118730545044,
"learning_rate": 9.952391144855278e-05,
"loss": 2.2534,
"step": 5665
},
{
"epoch": 0.04,
"grad_norm": 0.7415956854820251,
"learning_rate": 9.952305692249717e-05,
"loss": 2.2391,
"step": 5670
},
{
"epoch": 0.04,
"grad_norm": 0.7168838977813721,
"learning_rate": 9.952220163391451e-05,
"loss": 2.1801,
"step": 5675
},
{
"epoch": 0.04,
"grad_norm": 0.7003626227378845,
"learning_rate": 9.952134558281793e-05,
"loss": 2.2208,
"step": 5680
},
{
"epoch": 0.04,
"grad_norm": 0.6913336515426636,
"learning_rate": 9.952048876922066e-05,
"loss": 2.1485,
"step": 5685
},
{
"epoch": 0.04,
"grad_norm": 0.7206845283508301,
"learning_rate": 9.951963119313588e-05,
"loss": 2.1369,
"step": 5690
},
{
"epoch": 0.04,
"grad_norm": 0.7049555778503418,
"learning_rate": 9.951877285457678e-05,
"loss": 2.2845,
"step": 5695
},
{
"epoch": 0.04,
"grad_norm": 0.6810377240180969,
"learning_rate": 9.951791375355659e-05,
"loss": 2.1662,
"step": 5700
},
{
"epoch": 0.05,
"grad_norm": 0.7140889167785645,
"learning_rate": 9.951705389008852e-05,
"loss": 2.2597,
"step": 5705
},
{
"epoch": 0.05,
"grad_norm": 0.6851275563240051,
"learning_rate": 9.951619326418582e-05,
"loss": 2.2191,
"step": 5710
},
{
"epoch": 0.05,
"grad_norm": 0.6685874462127686,
"learning_rate": 9.951533187586178e-05,
"loss": 2.1987,
"step": 5715
},
{
"epoch": 0.05,
"grad_norm": 0.6668651700019836,
"learning_rate": 9.951446972512958e-05,
"loss": 2.1753,
"step": 5720
},
{
"epoch": 0.05,
"grad_norm": 0.6967094540596008,
"learning_rate": 9.951360681200256e-05,
"loss": 2.1803,
"step": 5725
},
{
"epoch": 0.05,
"grad_norm": 0.7078267335891724,
"learning_rate": 9.9512743136494e-05,
"loss": 2.1539,
"step": 5730
},
{
"epoch": 0.05,
"grad_norm": 0.7270606756210327,
"learning_rate": 9.951187869861715e-05,
"loss": 2.1528,
"step": 5735
},
{
"epoch": 0.05,
"grad_norm": 0.6807758808135986,
"learning_rate": 9.951101349838538e-05,
"loss": 2.2355,
"step": 5740
},
{
"epoch": 0.05,
"grad_norm": 0.7208414077758789,
"learning_rate": 9.9510147535812e-05,
"loss": 2.1905,
"step": 5745
},
{
"epoch": 0.05,
"grad_norm": 0.818071186542511,
"learning_rate": 9.950928081091032e-05,
"loss": 2.2079,
"step": 5750
},
{
"epoch": 0.05,
"grad_norm": 0.7369307279586792,
"learning_rate": 9.95084133236937e-05,
"loss": 2.1918,
"step": 5755
},
{
"epoch": 0.05,
"grad_norm": 0.7024269700050354,
"learning_rate": 9.950754507417547e-05,
"loss": 2.1574,
"step": 5760
},
{
"epoch": 0.05,
"grad_norm": 0.7381339073181152,
"learning_rate": 9.950667606236906e-05,
"loss": 2.2553,
"step": 5765
},
{
"epoch": 0.05,
"grad_norm": 0.7220199704170227,
"learning_rate": 9.95058062882878e-05,
"loss": 2.1742,
"step": 5770
},
{
"epoch": 0.05,
"grad_norm": 0.7041569352149963,
"learning_rate": 9.950493575194509e-05,
"loss": 2.2787,
"step": 5775
},
{
"epoch": 0.05,
"grad_norm": 0.7281708121299744,
"learning_rate": 9.950406445335434e-05,
"loss": 2.2052,
"step": 5780
},
{
"epoch": 0.05,
"grad_norm": 0.7005820870399475,
"learning_rate": 9.950319239252896e-05,
"loss": 2.188,
"step": 5785
},
{
"epoch": 0.05,
"grad_norm": 0.7739869356155396,
"learning_rate": 9.95023195694824e-05,
"loss": 2.2193,
"step": 5790
},
{
"epoch": 0.05,
"grad_norm": 0.7493504285812378,
"learning_rate": 9.950144598422809e-05,
"loss": 2.245,
"step": 5795
},
{
"epoch": 0.05,
"grad_norm": 0.6972953081130981,
"learning_rate": 9.950057163677944e-05,
"loss": 2.1962,
"step": 5800
},
{
"epoch": 0.05,
"grad_norm": 0.7097756862640381,
"learning_rate": 9.949969652714997e-05,
"loss": 2.2669,
"step": 5805
},
{
"epoch": 0.05,
"grad_norm": 0.7299678921699524,
"learning_rate": 9.949882065535312e-05,
"loss": 2.199,
"step": 5810
},
{
"epoch": 0.05,
"grad_norm": 0.7291364669799805,
"learning_rate": 9.949794402140239e-05,
"loss": 2.1877,
"step": 5815
},
{
"epoch": 0.05,
"grad_norm": 0.7208821177482605,
"learning_rate": 9.949706662531127e-05,
"loss": 2.1767,
"step": 5820
},
{
"epoch": 0.05,
"grad_norm": 0.712274968624115,
"learning_rate": 9.949618846709327e-05,
"loss": 2.2043,
"step": 5825
},
{
"epoch": 0.05,
"grad_norm": 0.6932068467140198,
"learning_rate": 9.949530954676194e-05,
"loss": 2.2163,
"step": 5830
},
{
"epoch": 0.05,
"grad_norm": 0.6805621385574341,
"learning_rate": 9.949442986433076e-05,
"loss": 2.1742,
"step": 5835
},
{
"epoch": 0.05,
"grad_norm": 0.6987532377243042,
"learning_rate": 9.949354941981332e-05,
"loss": 2.1832,
"step": 5840
},
{
"epoch": 0.05,
"grad_norm": 0.7250333428382874,
"learning_rate": 9.949266821322315e-05,
"loss": 2.1407,
"step": 5845
},
{
"epoch": 0.05,
"grad_norm": 0.7014710307121277,
"learning_rate": 9.949178624457384e-05,
"loss": 2.2408,
"step": 5850
},
{
"epoch": 0.05,
"grad_norm": 0.690898060798645,
"learning_rate": 9.949090351387895e-05,
"loss": 2.2061,
"step": 5855
},
{
"epoch": 0.05,
"grad_norm": 0.6809883117675781,
"learning_rate": 9.949002002115208e-05,
"loss": 2.152,
"step": 5860
},
{
"epoch": 0.05,
"grad_norm": 0.7302233576774597,
"learning_rate": 9.948913576640683e-05,
"loss": 2.2029,
"step": 5865
},
{
"epoch": 0.05,
"grad_norm": 0.7010434865951538,
"learning_rate": 9.948825074965682e-05,
"loss": 2.24,
"step": 5870
},
{
"epoch": 0.05,
"grad_norm": 0.6678712368011475,
"learning_rate": 9.948736497091569e-05,
"loss": 2.1418,
"step": 5875
},
{
"epoch": 0.05,
"grad_norm": 0.714667558670044,
"learning_rate": 9.948647843019705e-05,
"loss": 2.1778,
"step": 5880
},
{
"epoch": 0.05,
"grad_norm": 0.6870317459106445,
"learning_rate": 9.948559112751456e-05,
"loss": 2.115,
"step": 5885
},
{
"epoch": 0.05,
"grad_norm": 0.70961993932724,
"learning_rate": 9.948470306288189e-05,
"loss": 2.2573,
"step": 5890
},
{
"epoch": 0.05,
"grad_norm": 0.7190218567848206,
"learning_rate": 9.948381423631273e-05,
"loss": 2.1953,
"step": 5895
},
{
"epoch": 0.05,
"grad_norm": 0.690999448299408,
"learning_rate": 9.948292464782072e-05,
"loss": 2.2217,
"step": 5900
},
{
"epoch": 0.05,
"grad_norm": 0.7046534419059753,
"learning_rate": 9.948203429741961e-05,
"loss": 2.1961,
"step": 5905
},
{
"epoch": 0.05,
"grad_norm": 0.7513294219970703,
"learning_rate": 9.948114318512307e-05,
"loss": 2.1833,
"step": 5910
},
{
"epoch": 0.05,
"grad_norm": 0.6842653155326843,
"learning_rate": 9.948025131094482e-05,
"loss": 2.2726,
"step": 5915
},
{
"epoch": 0.05,
"grad_norm": 0.6850530505180359,
"learning_rate": 9.947935867489862e-05,
"loss": 2.1428,
"step": 5920
},
{
"epoch": 0.05,
"grad_norm": 0.7486492395401001,
"learning_rate": 9.94784652769982e-05,
"loss": 2.2039,
"step": 5925
},
{
"epoch": 0.05,
"grad_norm": 0.7261675000190735,
"learning_rate": 9.947757111725732e-05,
"loss": 2.1859,
"step": 5930
},
{
"epoch": 0.05,
"grad_norm": 0.6757484078407288,
"learning_rate": 9.947667619568974e-05,
"loss": 2.2347,
"step": 5935
},
{
"epoch": 0.05,
"grad_norm": 0.70115727186203,
"learning_rate": 9.947578051230925e-05,
"loss": 2.3561,
"step": 5940
},
{
"epoch": 0.05,
"grad_norm": 0.6758916974067688,
"learning_rate": 9.947488406712962e-05,
"loss": 2.2483,
"step": 5945
},
{
"epoch": 0.05,
"grad_norm": 0.7049750089645386,
"learning_rate": 9.947398686016469e-05,
"loss": 2.2705,
"step": 5950
},
{
"epoch": 0.05,
"grad_norm": 0.6915125250816345,
"learning_rate": 9.947308889142824e-05,
"loss": 2.2085,
"step": 5955
},
{
"epoch": 0.05,
"grad_norm": 0.6960284113883972,
"learning_rate": 9.94721901609341e-05,
"loss": 2.1538,
"step": 5960
},
{
"epoch": 0.05,
"grad_norm": 0.7163115739822388,
"learning_rate": 9.947129066869614e-05,
"loss": 2.1027,
"step": 5965
},
{
"epoch": 0.05,
"grad_norm": 0.9278883337974548,
"learning_rate": 9.947039041472817e-05,
"loss": 2.1775,
"step": 5970
},
{
"epoch": 0.05,
"grad_norm": 0.7370837926864624,
"learning_rate": 9.946948939904407e-05,
"loss": 2.1818,
"step": 5975
},
{
"epoch": 0.05,
"grad_norm": 0.731526255607605,
"learning_rate": 9.946858762165772e-05,
"loss": 2.2164,
"step": 5980
},
{
"epoch": 0.05,
"grad_norm": 0.7468245625495911,
"learning_rate": 9.946768508258299e-05,
"loss": 2.2517,
"step": 5985
},
{
"epoch": 0.05,
"grad_norm": 0.710542619228363,
"learning_rate": 9.946678178183377e-05,
"loss": 2.2253,
"step": 5990
},
{
"epoch": 0.05,
"grad_norm": 0.7063456773757935,
"learning_rate": 9.946587771942401e-05,
"loss": 2.1281,
"step": 5995
},
{
"epoch": 0.05,
"grad_norm": 0.7552475333213806,
"learning_rate": 9.946497289536758e-05,
"loss": 2.2396,
"step": 6000
}
],
"logging_steps": 5,
"max_steps": 126689,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 1000,
"total_flos": 1.7487862362537984e+19,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}