Qwen7B-Math-L28-HP / checkpoint-2146 /trainer_state.json
ZMC2019's picture
Training in progress, epoch 2, checkpoint
dc6660d verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 2146,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004659832246039142,
"grad_norm": 63.97114436632525,
"learning_rate": 7.763975155279503e-07,
"loss": 11.009,
"step": 5
},
{
"epoch": 0.009319664492078284,
"grad_norm": 97.6207743013609,
"learning_rate": 1.5527950310559006e-06,
"loss": 10.0558,
"step": 10
},
{
"epoch": 0.013979496738117428,
"grad_norm": 68.74620690530702,
"learning_rate": 2.329192546583851e-06,
"loss": 5.5709,
"step": 15
},
{
"epoch": 0.01863932898415657,
"grad_norm": 7.329953761956908,
"learning_rate": 3.1055900621118013e-06,
"loss": 1.674,
"step": 20
},
{
"epoch": 0.023299161230195712,
"grad_norm": 2.631209184742148,
"learning_rate": 3.881987577639752e-06,
"loss": 1.1266,
"step": 25
},
{
"epoch": 0.027958993476234855,
"grad_norm": 1.2736389968892712,
"learning_rate": 4.658385093167702e-06,
"loss": 0.9104,
"step": 30
},
{
"epoch": 0.032618825722273995,
"grad_norm": 0.7641614678643728,
"learning_rate": 5.4347826086956525e-06,
"loss": 0.7886,
"step": 35
},
{
"epoch": 0.03727865796831314,
"grad_norm": 0.6854613522496884,
"learning_rate": 6.2111801242236025e-06,
"loss": 0.7361,
"step": 40
},
{
"epoch": 0.04193849021435228,
"grad_norm": 0.5502087380767395,
"learning_rate": 6.9875776397515525e-06,
"loss": 0.6884,
"step": 45
},
{
"epoch": 0.046598322460391424,
"grad_norm": 0.4674581907162505,
"learning_rate": 7.763975155279503e-06,
"loss": 0.6644,
"step": 50
},
{
"epoch": 0.05125815470643057,
"grad_norm": 0.40309163851709934,
"learning_rate": 8.540372670807453e-06,
"loss": 0.6316,
"step": 55
},
{
"epoch": 0.05591798695246971,
"grad_norm": 0.3284620469689577,
"learning_rate": 9.316770186335403e-06,
"loss": 0.5978,
"step": 60
},
{
"epoch": 0.06057781919850885,
"grad_norm": 0.30466734688995895,
"learning_rate": 1.0093167701863353e-05,
"loss": 0.5837,
"step": 65
},
{
"epoch": 0.06523765144454799,
"grad_norm": 0.3095353132100952,
"learning_rate": 1.0869565217391305e-05,
"loss": 0.5808,
"step": 70
},
{
"epoch": 0.06989748369058714,
"grad_norm": 0.2742862834044995,
"learning_rate": 1.1645962732919255e-05,
"loss": 0.5639,
"step": 75
},
{
"epoch": 0.07455731593662628,
"grad_norm": 0.28624072357525304,
"learning_rate": 1.2422360248447205e-05,
"loss": 0.5552,
"step": 80
},
{
"epoch": 0.07921714818266543,
"grad_norm": 0.35136842750239394,
"learning_rate": 1.3198757763975155e-05,
"loss": 0.5465,
"step": 85
},
{
"epoch": 0.08387698042870456,
"grad_norm": 0.3554715302423071,
"learning_rate": 1.3975155279503105e-05,
"loss": 0.5572,
"step": 90
},
{
"epoch": 0.08853681267474371,
"grad_norm": 0.30333751588191904,
"learning_rate": 1.4751552795031057e-05,
"loss": 0.5389,
"step": 95
},
{
"epoch": 0.09319664492078285,
"grad_norm": 0.34447253483888585,
"learning_rate": 1.5527950310559007e-05,
"loss": 0.5248,
"step": 100
},
{
"epoch": 0.097856477166822,
"grad_norm": 0.3509763920502021,
"learning_rate": 1.630434782608696e-05,
"loss": 0.5197,
"step": 105
},
{
"epoch": 0.10251630941286113,
"grad_norm": 0.3047065908098556,
"learning_rate": 1.7080745341614907e-05,
"loss": 0.5215,
"step": 110
},
{
"epoch": 0.10717614165890028,
"grad_norm": 0.35772122594235833,
"learning_rate": 1.785714285714286e-05,
"loss": 0.519,
"step": 115
},
{
"epoch": 0.11183597390493942,
"grad_norm": 0.4237939965410238,
"learning_rate": 1.8633540372670807e-05,
"loss": 0.5057,
"step": 120
},
{
"epoch": 0.11649580615097857,
"grad_norm": 0.3052118953673648,
"learning_rate": 1.940993788819876e-05,
"loss": 0.5022,
"step": 125
},
{
"epoch": 0.1211556383970177,
"grad_norm": 0.48506740368145884,
"learning_rate": 2.0186335403726707e-05,
"loss": 0.5083,
"step": 130
},
{
"epoch": 0.12581547064305684,
"grad_norm": 0.4067639562809268,
"learning_rate": 2.096273291925466e-05,
"loss": 0.5074,
"step": 135
},
{
"epoch": 0.13047530288909598,
"grad_norm": 0.44239635913158043,
"learning_rate": 2.173913043478261e-05,
"loss": 0.4919,
"step": 140
},
{
"epoch": 0.13513513513513514,
"grad_norm": 0.3510658763415586,
"learning_rate": 2.2515527950310562e-05,
"loss": 0.5188,
"step": 145
},
{
"epoch": 0.13979496738117428,
"grad_norm": 0.44553534657252714,
"learning_rate": 2.329192546583851e-05,
"loss": 0.4929,
"step": 150
},
{
"epoch": 0.14445479962721341,
"grad_norm": 0.44028135611068503,
"learning_rate": 2.4068322981366462e-05,
"loss": 0.5013,
"step": 155
},
{
"epoch": 0.14911463187325255,
"grad_norm": 0.4559159420874923,
"learning_rate": 2.484472049689441e-05,
"loss": 0.4821,
"step": 160
},
{
"epoch": 0.15377446411929171,
"grad_norm": 0.47309971975261556,
"learning_rate": 2.5621118012422362e-05,
"loss": 0.4933,
"step": 165
},
{
"epoch": 0.15843429636533085,
"grad_norm": 0.4022989429457791,
"learning_rate": 2.639751552795031e-05,
"loss": 0.4694,
"step": 170
},
{
"epoch": 0.16309412861137,
"grad_norm": 0.48867133460334755,
"learning_rate": 2.7173913043478262e-05,
"loss": 0.483,
"step": 175
},
{
"epoch": 0.16775396085740912,
"grad_norm": 0.4405332554716524,
"learning_rate": 2.795031055900621e-05,
"loss": 0.4759,
"step": 180
},
{
"epoch": 0.1724137931034483,
"grad_norm": 0.4170098031366455,
"learning_rate": 2.8726708074534165e-05,
"loss": 0.4792,
"step": 185
},
{
"epoch": 0.17707362534948742,
"grad_norm": 0.5595553852386751,
"learning_rate": 2.9503105590062114e-05,
"loss": 0.4891,
"step": 190
},
{
"epoch": 0.18173345759552656,
"grad_norm": 0.6802692412088472,
"learning_rate": 3.0279503105590062e-05,
"loss": 0.4796,
"step": 195
},
{
"epoch": 0.1863932898415657,
"grad_norm": 0.5052964875363417,
"learning_rate": 3.1055900621118014e-05,
"loss": 0.482,
"step": 200
},
{
"epoch": 0.19105312208760486,
"grad_norm": 1.0027813070497937,
"learning_rate": 3.183229813664597e-05,
"loss": 0.471,
"step": 205
},
{
"epoch": 0.195712954333644,
"grad_norm": 0.6324877868930993,
"learning_rate": 3.260869565217392e-05,
"loss": 0.484,
"step": 210
},
{
"epoch": 0.20037278657968313,
"grad_norm": 0.5481580441751476,
"learning_rate": 3.3385093167701865e-05,
"loss": 0.4658,
"step": 215
},
{
"epoch": 0.20503261882572227,
"grad_norm": 0.5572577049127609,
"learning_rate": 3.4161490683229814e-05,
"loss": 0.4701,
"step": 220
},
{
"epoch": 0.2096924510717614,
"grad_norm": 0.5062995178514071,
"learning_rate": 3.493788819875777e-05,
"loss": 0.4521,
"step": 225
},
{
"epoch": 0.21435228331780057,
"grad_norm": 0.49686277193314815,
"learning_rate": 3.571428571428572e-05,
"loss": 0.4654,
"step": 230
},
{
"epoch": 0.2190121155638397,
"grad_norm": 0.4384456812240532,
"learning_rate": 3.6490683229813665e-05,
"loss": 0.4629,
"step": 235
},
{
"epoch": 0.22367194780987884,
"grad_norm": 0.40181315075678486,
"learning_rate": 3.7267080745341614e-05,
"loss": 0.4604,
"step": 240
},
{
"epoch": 0.22833178005591798,
"grad_norm": 0.4680413353176079,
"learning_rate": 3.804347826086957e-05,
"loss": 0.4658,
"step": 245
},
{
"epoch": 0.23299161230195714,
"grad_norm": 0.4121568463339279,
"learning_rate": 3.881987577639752e-05,
"loss": 0.458,
"step": 250
},
{
"epoch": 0.23765144454799628,
"grad_norm": 0.46257624459214514,
"learning_rate": 3.9596273291925465e-05,
"loss": 0.4664,
"step": 255
},
{
"epoch": 0.2423112767940354,
"grad_norm": 0.4208718113851754,
"learning_rate": 4.0372670807453414e-05,
"loss": 0.4616,
"step": 260
},
{
"epoch": 0.24697110904007455,
"grad_norm": 0.4920194796206926,
"learning_rate": 4.114906832298137e-05,
"loss": 0.4601,
"step": 265
},
{
"epoch": 0.2516309412861137,
"grad_norm": 0.44517914619453425,
"learning_rate": 4.192546583850932e-05,
"loss": 0.4585,
"step": 270
},
{
"epoch": 0.25629077353215285,
"grad_norm": 0.4272832211849931,
"learning_rate": 4.270186335403727e-05,
"loss": 0.4645,
"step": 275
},
{
"epoch": 0.26095060577819196,
"grad_norm": 0.46232694776349703,
"learning_rate": 4.347826086956522e-05,
"loss": 0.4566,
"step": 280
},
{
"epoch": 0.2656104380242311,
"grad_norm": 0.47318397108506427,
"learning_rate": 4.425465838509317e-05,
"loss": 0.4606,
"step": 285
},
{
"epoch": 0.2702702702702703,
"grad_norm": 0.5203346361695503,
"learning_rate": 4.5031055900621124e-05,
"loss": 0.4627,
"step": 290
},
{
"epoch": 0.2749301025163094,
"grad_norm": 0.6727002183656305,
"learning_rate": 4.580745341614907e-05,
"loss": 0.4669,
"step": 295
},
{
"epoch": 0.27958993476234856,
"grad_norm": 0.7920690776017782,
"learning_rate": 4.658385093167702e-05,
"loss": 0.4696,
"step": 300
},
{
"epoch": 0.2842497670083877,
"grad_norm": 0.5549969347516743,
"learning_rate": 4.736024844720497e-05,
"loss": 0.4577,
"step": 305
},
{
"epoch": 0.28890959925442683,
"grad_norm": 0.49025638861057025,
"learning_rate": 4.8136645962732924e-05,
"loss": 0.4493,
"step": 310
},
{
"epoch": 0.293569431500466,
"grad_norm": 0.48227724645400744,
"learning_rate": 4.891304347826087e-05,
"loss": 0.4576,
"step": 315
},
{
"epoch": 0.2982292637465051,
"grad_norm": 0.6932156419493832,
"learning_rate": 4.968944099378882e-05,
"loss": 0.4429,
"step": 320
},
{
"epoch": 0.30288909599254427,
"grad_norm": 0.5932544065676107,
"learning_rate": 4.994822229892993e-05,
"loss": 0.4406,
"step": 325
},
{
"epoch": 0.30754892823858343,
"grad_norm": 0.44595884846414735,
"learning_rate": 4.986192613047981e-05,
"loss": 0.4514,
"step": 330
},
{
"epoch": 0.31220876048462254,
"grad_norm": 0.7093678389168179,
"learning_rate": 4.977562996202969e-05,
"loss": 0.4522,
"step": 335
},
{
"epoch": 0.3168685927306617,
"grad_norm": 0.3992271710082444,
"learning_rate": 4.968933379357957e-05,
"loss": 0.4527,
"step": 340
},
{
"epoch": 0.32152842497670087,
"grad_norm": 0.4322675744616485,
"learning_rate": 4.9603037625129445e-05,
"loss": 0.4474,
"step": 345
},
{
"epoch": 0.32618825722274,
"grad_norm": 0.4894749100842417,
"learning_rate": 4.951674145667933e-05,
"loss": 0.4611,
"step": 350
},
{
"epoch": 0.33084808946877914,
"grad_norm": 0.45545284883251785,
"learning_rate": 4.94304452882292e-05,
"loss": 0.447,
"step": 355
},
{
"epoch": 0.33550792171481825,
"grad_norm": 0.5018284119539658,
"learning_rate": 4.934414911977908e-05,
"loss": 0.4449,
"step": 360
},
{
"epoch": 0.3401677539608574,
"grad_norm": 0.4101489952466933,
"learning_rate": 4.9257852951328965e-05,
"loss": 0.4411,
"step": 365
},
{
"epoch": 0.3448275862068966,
"grad_norm": 0.4492613666448338,
"learning_rate": 4.917155678287884e-05,
"loss": 0.4542,
"step": 370
},
{
"epoch": 0.3494874184529357,
"grad_norm": 0.5065139889539029,
"learning_rate": 4.908526061442872e-05,
"loss": 0.441,
"step": 375
},
{
"epoch": 0.35414725069897485,
"grad_norm": 0.5142632297631841,
"learning_rate": 4.89989644459786e-05,
"loss": 0.4559,
"step": 380
},
{
"epoch": 0.35880708294501396,
"grad_norm": 0.38956464528036483,
"learning_rate": 4.891266827752848e-05,
"loss": 0.4576,
"step": 385
},
{
"epoch": 0.3634669151910531,
"grad_norm": 0.48099181069040514,
"learning_rate": 4.882637210907836e-05,
"loss": 0.4415,
"step": 390
},
{
"epoch": 0.3681267474370923,
"grad_norm": 0.4527262541512366,
"learning_rate": 4.874007594062824e-05,
"loss": 0.4479,
"step": 395
},
{
"epoch": 0.3727865796831314,
"grad_norm": 0.45728910187809607,
"learning_rate": 4.865377977217811e-05,
"loss": 0.4326,
"step": 400
},
{
"epoch": 0.37744641192917056,
"grad_norm": 0.5325567837367734,
"learning_rate": 4.8567483603728e-05,
"loss": 0.4511,
"step": 405
},
{
"epoch": 0.3821062441752097,
"grad_norm": 0.47530962989623887,
"learning_rate": 4.8481187435277875e-05,
"loss": 0.4472,
"step": 410
},
{
"epoch": 0.38676607642124883,
"grad_norm": 0.5492535133414717,
"learning_rate": 4.839489126682776e-05,
"loss": 0.4335,
"step": 415
},
{
"epoch": 0.391425908667288,
"grad_norm": 0.5374577291241773,
"learning_rate": 4.830859509837763e-05,
"loss": 0.4519,
"step": 420
},
{
"epoch": 0.3960857409133271,
"grad_norm": 0.3671416917960664,
"learning_rate": 4.822229892992751e-05,
"loss": 0.4493,
"step": 425
},
{
"epoch": 0.40074557315936626,
"grad_norm": 0.47641546071577745,
"learning_rate": 4.8136002761477395e-05,
"loss": 0.4363,
"step": 430
},
{
"epoch": 0.40540540540540543,
"grad_norm": 0.5060064177084717,
"learning_rate": 4.804970659302727e-05,
"loss": 0.4493,
"step": 435
},
{
"epoch": 0.41006523765144454,
"grad_norm": 0.4421952331981267,
"learning_rate": 4.796341042457715e-05,
"loss": 0.4438,
"step": 440
},
{
"epoch": 0.4147250698974837,
"grad_norm": 0.4912038645027838,
"learning_rate": 4.787711425612703e-05,
"loss": 0.4381,
"step": 445
},
{
"epoch": 0.4193849021435228,
"grad_norm": 0.49994262075500673,
"learning_rate": 4.779081808767691e-05,
"loss": 0.4287,
"step": 450
},
{
"epoch": 0.424044734389562,
"grad_norm": 0.46833774296806413,
"learning_rate": 4.770452191922679e-05,
"loss": 0.4356,
"step": 455
},
{
"epoch": 0.42870456663560114,
"grad_norm": 0.37814791573621137,
"learning_rate": 4.761822575077667e-05,
"loss": 0.4325,
"step": 460
},
{
"epoch": 0.43336439888164024,
"grad_norm": 0.4979334392926385,
"learning_rate": 4.753192958232654e-05,
"loss": 0.4296,
"step": 465
},
{
"epoch": 0.4380242311276794,
"grad_norm": 0.45595286465505486,
"learning_rate": 4.744563341387643e-05,
"loss": 0.4424,
"step": 470
},
{
"epoch": 0.4426840633737186,
"grad_norm": 0.38484174975814034,
"learning_rate": 4.7359337245426306e-05,
"loss": 0.4317,
"step": 475
},
{
"epoch": 0.4473438956197577,
"grad_norm": 0.5318530012205278,
"learning_rate": 4.7273041076976184e-05,
"loss": 0.4437,
"step": 480
},
{
"epoch": 0.45200372786579684,
"grad_norm": 0.5500470128648078,
"learning_rate": 4.718674490852606e-05,
"loss": 0.4364,
"step": 485
},
{
"epoch": 0.45666356011183595,
"grad_norm": 0.46022150094529285,
"learning_rate": 4.710044874007594e-05,
"loss": 0.4307,
"step": 490
},
{
"epoch": 0.4613233923578751,
"grad_norm": 0.4453333531177356,
"learning_rate": 4.7014152571625826e-05,
"loss": 0.4366,
"step": 495
},
{
"epoch": 0.4659832246039143,
"grad_norm": 0.5514063928798473,
"learning_rate": 4.6927856403175704e-05,
"loss": 0.4396,
"step": 500
},
{
"epoch": 0.4706430568499534,
"grad_norm": 0.39385926743432437,
"learning_rate": 4.684156023472558e-05,
"loss": 0.4262,
"step": 505
},
{
"epoch": 0.47530288909599255,
"grad_norm": 0.5044340425827932,
"learning_rate": 4.675526406627546e-05,
"loss": 0.4348,
"step": 510
},
{
"epoch": 0.47996272134203166,
"grad_norm": 0.6595607098484944,
"learning_rate": 4.666896789782534e-05,
"loss": 0.4327,
"step": 515
},
{
"epoch": 0.4846225535880708,
"grad_norm": 0.6115798965441747,
"learning_rate": 4.658267172937522e-05,
"loss": 0.4364,
"step": 520
},
{
"epoch": 0.48928238583411,
"grad_norm": 0.44577176108357613,
"learning_rate": 4.64963755609251e-05,
"loss": 0.4352,
"step": 525
},
{
"epoch": 0.4939422180801491,
"grad_norm": 0.5381399519172929,
"learning_rate": 4.641007939247497e-05,
"loss": 0.4323,
"step": 530
},
{
"epoch": 0.49860205032618826,
"grad_norm": 0.37136037243594944,
"learning_rate": 4.632378322402486e-05,
"loss": 0.4325,
"step": 535
},
{
"epoch": 0.5032618825722274,
"grad_norm": 0.3892034994234061,
"learning_rate": 4.6237487055574736e-05,
"loss": 0.4225,
"step": 540
},
{
"epoch": 0.5079217148182665,
"grad_norm": 0.49386732668387484,
"learning_rate": 4.6151190887124615e-05,
"loss": 0.4287,
"step": 545
},
{
"epoch": 0.5125815470643057,
"grad_norm": 0.4168272091126582,
"learning_rate": 4.606489471867449e-05,
"loss": 0.4289,
"step": 550
},
{
"epoch": 0.5172413793103449,
"grad_norm": 0.3582043405694146,
"learning_rate": 4.597859855022437e-05,
"loss": 0.4231,
"step": 555
},
{
"epoch": 0.5219012115563839,
"grad_norm": 0.41696696369770525,
"learning_rate": 4.589230238177425e-05,
"loss": 0.4443,
"step": 560
},
{
"epoch": 0.5265610438024231,
"grad_norm": 0.4861469903998086,
"learning_rate": 4.5806006213324134e-05,
"loss": 0.4216,
"step": 565
},
{
"epoch": 0.5312208760484622,
"grad_norm": 0.5140354855395411,
"learning_rate": 4.5719710044874006e-05,
"loss": 0.4214,
"step": 570
},
{
"epoch": 0.5358807082945014,
"grad_norm": 0.41866668365773924,
"learning_rate": 4.563341387642389e-05,
"loss": 0.4276,
"step": 575
},
{
"epoch": 0.5405405405405406,
"grad_norm": 0.3584845668041098,
"learning_rate": 4.554711770797377e-05,
"loss": 0.4275,
"step": 580
},
{
"epoch": 0.5452003727865797,
"grad_norm": 0.49174805356075396,
"learning_rate": 4.546082153952365e-05,
"loss": 0.4246,
"step": 585
},
{
"epoch": 0.5498602050326188,
"grad_norm": 0.41185105098746255,
"learning_rate": 4.5374525371073526e-05,
"loss": 0.421,
"step": 590
},
{
"epoch": 0.554520037278658,
"grad_norm": 0.36837391509005896,
"learning_rate": 4.5288229202623404e-05,
"loss": 0.423,
"step": 595
},
{
"epoch": 0.5591798695246971,
"grad_norm": 0.38520965444801625,
"learning_rate": 4.520193303417328e-05,
"loss": 0.4233,
"step": 600
},
{
"epoch": 0.5638397017707363,
"grad_norm": 0.36649198023585905,
"learning_rate": 4.511563686572317e-05,
"loss": 0.4219,
"step": 605
},
{
"epoch": 0.5684995340167754,
"grad_norm": 0.4164679504509725,
"learning_rate": 4.5029340697273045e-05,
"loss": 0.433,
"step": 610
},
{
"epoch": 0.5731593662628145,
"grad_norm": 0.3214873987463138,
"learning_rate": 4.4943044528822923e-05,
"loss": 0.4197,
"step": 615
},
{
"epoch": 0.5778191985088537,
"grad_norm": 0.44600645286913126,
"learning_rate": 4.48567483603728e-05,
"loss": 0.4168,
"step": 620
},
{
"epoch": 0.5824790307548928,
"grad_norm": 0.37318395537297955,
"learning_rate": 4.477045219192268e-05,
"loss": 0.4385,
"step": 625
},
{
"epoch": 0.587138863000932,
"grad_norm": 0.3604844647747845,
"learning_rate": 4.4684156023472565e-05,
"loss": 0.429,
"step": 630
},
{
"epoch": 0.5917986952469712,
"grad_norm": 0.49391135959870375,
"learning_rate": 4.4597859855022436e-05,
"loss": 0.4336,
"step": 635
},
{
"epoch": 0.5964585274930102,
"grad_norm": 0.5003007414297862,
"learning_rate": 4.4511563686572315e-05,
"loss": 0.4244,
"step": 640
},
{
"epoch": 0.6011183597390494,
"grad_norm": 0.4136262369903074,
"learning_rate": 4.44252675181222e-05,
"loss": 0.4239,
"step": 645
},
{
"epoch": 0.6057781919850885,
"grad_norm": 0.4100037093770799,
"learning_rate": 4.433897134967208e-05,
"loss": 0.4329,
"step": 650
},
{
"epoch": 0.6104380242311277,
"grad_norm": 0.40473278171973515,
"learning_rate": 4.4252675181221956e-05,
"loss": 0.4302,
"step": 655
},
{
"epoch": 0.6150978564771669,
"grad_norm": 0.3581991814954745,
"learning_rate": 4.4166379012771834e-05,
"loss": 0.4122,
"step": 660
},
{
"epoch": 0.6197576887232059,
"grad_norm": 0.4380544935583559,
"learning_rate": 4.408008284432171e-05,
"loss": 0.4127,
"step": 665
},
{
"epoch": 0.6244175209692451,
"grad_norm": 0.4100795719599541,
"learning_rate": 4.39937866758716e-05,
"loss": 0.4115,
"step": 670
},
{
"epoch": 0.6290773532152842,
"grad_norm": 0.40367865374562034,
"learning_rate": 4.3907490507421476e-05,
"loss": 0.4201,
"step": 675
},
{
"epoch": 0.6337371854613234,
"grad_norm": 0.4255661108043704,
"learning_rate": 4.382119433897135e-05,
"loss": 0.4232,
"step": 680
},
{
"epoch": 0.6383970177073626,
"grad_norm": 0.43799906091812546,
"learning_rate": 4.373489817052123e-05,
"loss": 0.4326,
"step": 685
},
{
"epoch": 0.6430568499534017,
"grad_norm": 0.4582804945285276,
"learning_rate": 4.364860200207111e-05,
"loss": 0.4147,
"step": 690
},
{
"epoch": 0.6477166821994408,
"grad_norm": 0.3745759468060015,
"learning_rate": 4.356230583362099e-05,
"loss": 0.4199,
"step": 695
},
{
"epoch": 0.65237651444548,
"grad_norm": 0.35894331983296773,
"learning_rate": 4.347600966517087e-05,
"loss": 0.4283,
"step": 700
},
{
"epoch": 0.6570363466915191,
"grad_norm": 0.4107553298840247,
"learning_rate": 4.3389713496720745e-05,
"loss": 0.4219,
"step": 705
},
{
"epoch": 0.6616961789375583,
"grad_norm": 0.4827554934119169,
"learning_rate": 4.330341732827063e-05,
"loss": 0.4225,
"step": 710
},
{
"epoch": 0.6663560111835974,
"grad_norm": 0.3805082888445549,
"learning_rate": 4.321712115982051e-05,
"loss": 0.411,
"step": 715
},
{
"epoch": 0.6710158434296365,
"grad_norm": 0.3764900820609077,
"learning_rate": 4.3130824991370387e-05,
"loss": 0.4214,
"step": 720
},
{
"epoch": 0.6756756756756757,
"grad_norm": 0.3457535674984788,
"learning_rate": 4.3044528822920265e-05,
"loss": 0.42,
"step": 725
},
{
"epoch": 0.6803355079217148,
"grad_norm": 0.44229272421355714,
"learning_rate": 4.295823265447014e-05,
"loss": 0.415,
"step": 730
},
{
"epoch": 0.684995340167754,
"grad_norm": 0.3405071649598411,
"learning_rate": 4.287193648602002e-05,
"loss": 0.4188,
"step": 735
},
{
"epoch": 0.6896551724137931,
"grad_norm": 0.3420276540344356,
"learning_rate": 4.27856403175699e-05,
"loss": 0.4125,
"step": 740
},
{
"epoch": 0.6943150046598322,
"grad_norm": 0.3714271443540526,
"learning_rate": 4.269934414911978e-05,
"loss": 0.4082,
"step": 745
},
{
"epoch": 0.6989748369058714,
"grad_norm": 0.4440957065820136,
"learning_rate": 4.261304798066966e-05,
"loss": 0.4132,
"step": 750
},
{
"epoch": 0.7036346691519105,
"grad_norm": 0.38502725849051433,
"learning_rate": 4.252675181221954e-05,
"loss": 0.4155,
"step": 755
},
{
"epoch": 0.7082945013979497,
"grad_norm": 0.4101980956792011,
"learning_rate": 4.244045564376942e-05,
"loss": 0.4187,
"step": 760
},
{
"epoch": 0.7129543336439889,
"grad_norm": 0.3949116271091045,
"learning_rate": 4.23541594753193e-05,
"loss": 0.4027,
"step": 765
},
{
"epoch": 0.7176141658900279,
"grad_norm": 0.3778946135235895,
"learning_rate": 4.2267863306869176e-05,
"loss": 0.4117,
"step": 770
},
{
"epoch": 0.7222739981360671,
"grad_norm": 0.3840085266541027,
"learning_rate": 4.2181567138419054e-05,
"loss": 0.4244,
"step": 775
},
{
"epoch": 0.7269338303821062,
"grad_norm": 0.35414793238950554,
"learning_rate": 4.209527096996894e-05,
"loss": 0.409,
"step": 780
},
{
"epoch": 0.7315936626281454,
"grad_norm": 0.4017361351425298,
"learning_rate": 4.200897480151881e-05,
"loss": 0.4161,
"step": 785
},
{
"epoch": 0.7362534948741846,
"grad_norm": 0.4577917995677147,
"learning_rate": 4.1922678633068695e-05,
"loss": 0.4194,
"step": 790
},
{
"epoch": 0.7409133271202236,
"grad_norm": 0.4449028327643515,
"learning_rate": 4.1836382464618573e-05,
"loss": 0.4139,
"step": 795
},
{
"epoch": 0.7455731593662628,
"grad_norm": 0.4278118467092459,
"learning_rate": 4.175008629616845e-05,
"loss": 0.4119,
"step": 800
},
{
"epoch": 0.750232991612302,
"grad_norm": 0.41286072521941974,
"learning_rate": 4.166379012771833e-05,
"loss": 0.4118,
"step": 805
},
{
"epoch": 0.7548928238583411,
"grad_norm": 0.36674422319093963,
"learning_rate": 4.157749395926821e-05,
"loss": 0.4301,
"step": 810
},
{
"epoch": 0.7595526561043803,
"grad_norm": 0.3466794250376395,
"learning_rate": 4.1491197790818086e-05,
"loss": 0.4043,
"step": 815
},
{
"epoch": 0.7642124883504194,
"grad_norm": 0.5419218965239389,
"learning_rate": 4.140490162236797e-05,
"loss": 0.4199,
"step": 820
},
{
"epoch": 0.7688723205964585,
"grad_norm": 0.4446206400997879,
"learning_rate": 4.131860545391785e-05,
"loss": 0.4119,
"step": 825
},
{
"epoch": 0.7735321528424977,
"grad_norm": 0.3679372318535966,
"learning_rate": 4.123230928546773e-05,
"loss": 0.4067,
"step": 830
},
{
"epoch": 0.7781919850885368,
"grad_norm": 0.3883744139143354,
"learning_rate": 4.1146013117017606e-05,
"loss": 0.4139,
"step": 835
},
{
"epoch": 0.782851817334576,
"grad_norm": 0.4406909830090548,
"learning_rate": 4.1059716948567484e-05,
"loss": 0.4169,
"step": 840
},
{
"epoch": 0.7875116495806151,
"grad_norm": 0.46114038026303233,
"learning_rate": 4.097342078011737e-05,
"loss": 0.4249,
"step": 845
},
{
"epoch": 0.7921714818266542,
"grad_norm": 0.3915360251493529,
"learning_rate": 4.088712461166724e-05,
"loss": 0.4145,
"step": 850
},
{
"epoch": 0.7968313140726934,
"grad_norm": 0.5167680703260195,
"learning_rate": 4.080082844321712e-05,
"loss": 0.4166,
"step": 855
},
{
"epoch": 0.8014911463187325,
"grad_norm": 0.3754422426152857,
"learning_rate": 4.0714532274767004e-05,
"loss": 0.4234,
"step": 860
},
{
"epoch": 0.8061509785647717,
"grad_norm": 0.3476171939553642,
"learning_rate": 4.062823610631688e-05,
"loss": 0.4044,
"step": 865
},
{
"epoch": 0.8108108108108109,
"grad_norm": 0.362219475724925,
"learning_rate": 4.054193993786676e-05,
"loss": 0.4189,
"step": 870
},
{
"epoch": 0.8154706430568499,
"grad_norm": 0.3820875006288019,
"learning_rate": 4.045564376941664e-05,
"loss": 0.4145,
"step": 875
},
{
"epoch": 0.8201304753028891,
"grad_norm": 0.5867270655165092,
"learning_rate": 4.036934760096652e-05,
"loss": 0.4057,
"step": 880
},
{
"epoch": 0.8247903075489282,
"grad_norm": 0.4109568389025995,
"learning_rate": 4.02830514325164e-05,
"loss": 0.42,
"step": 885
},
{
"epoch": 0.8294501397949674,
"grad_norm": 0.4342697112855388,
"learning_rate": 4.019675526406628e-05,
"loss": 0.4065,
"step": 890
},
{
"epoch": 0.8341099720410066,
"grad_norm": 0.3703770039740134,
"learning_rate": 4.011045909561615e-05,
"loss": 0.426,
"step": 895
},
{
"epoch": 0.8387698042870456,
"grad_norm": 0.44291526775558954,
"learning_rate": 4.0024162927166037e-05,
"loss": 0.4087,
"step": 900
},
{
"epoch": 0.8434296365330848,
"grad_norm": 0.33257023554753873,
"learning_rate": 3.9937866758715915e-05,
"loss": 0.4104,
"step": 905
},
{
"epoch": 0.848089468779124,
"grad_norm": 0.5210603832824299,
"learning_rate": 3.98515705902658e-05,
"loss": 0.4164,
"step": 910
},
{
"epoch": 0.8527493010251631,
"grad_norm": 0.607238634824021,
"learning_rate": 3.976527442181567e-05,
"loss": 0.4094,
"step": 915
},
{
"epoch": 0.8574091332712023,
"grad_norm": 0.4484844822862872,
"learning_rate": 3.967897825336555e-05,
"loss": 0.4157,
"step": 920
},
{
"epoch": 0.8620689655172413,
"grad_norm": 0.3688615067553498,
"learning_rate": 3.9592682084915434e-05,
"loss": 0.4029,
"step": 925
},
{
"epoch": 0.8667287977632805,
"grad_norm": 0.5024805631475334,
"learning_rate": 3.950638591646531e-05,
"loss": 0.4017,
"step": 930
},
{
"epoch": 0.8713886300093197,
"grad_norm": 0.382016518381038,
"learning_rate": 3.942008974801519e-05,
"loss": 0.4135,
"step": 935
},
{
"epoch": 0.8760484622553588,
"grad_norm": 0.30605957215975665,
"learning_rate": 3.933379357956507e-05,
"loss": 0.4091,
"step": 940
},
{
"epoch": 0.880708294501398,
"grad_norm": 0.3877784507885669,
"learning_rate": 3.924749741111495e-05,
"loss": 0.4113,
"step": 945
},
{
"epoch": 0.8853681267474371,
"grad_norm": 0.48733905055236465,
"learning_rate": 3.916120124266483e-05,
"loss": 0.4191,
"step": 950
},
{
"epoch": 0.8900279589934762,
"grad_norm": 0.43034903356385523,
"learning_rate": 3.9074905074214704e-05,
"loss": 0.4182,
"step": 955
},
{
"epoch": 0.8946877912395154,
"grad_norm": 0.3390149226125637,
"learning_rate": 3.898860890576458e-05,
"loss": 0.4134,
"step": 960
},
{
"epoch": 0.8993476234855545,
"grad_norm": 0.3977054990731856,
"learning_rate": 3.890231273731447e-05,
"loss": 0.405,
"step": 965
},
{
"epoch": 0.9040074557315937,
"grad_norm": 0.31681352931839235,
"learning_rate": 3.8816016568864345e-05,
"loss": 0.4028,
"step": 970
},
{
"epoch": 0.9086672879776329,
"grad_norm": 0.32371549087627044,
"learning_rate": 3.8729720400414224e-05,
"loss": 0.4215,
"step": 975
},
{
"epoch": 0.9133271202236719,
"grad_norm": 0.404454156562585,
"learning_rate": 3.86434242319641e-05,
"loss": 0.4088,
"step": 980
},
{
"epoch": 0.9179869524697111,
"grad_norm": 0.5071586018047269,
"learning_rate": 3.855712806351398e-05,
"loss": 0.4138,
"step": 985
},
{
"epoch": 0.9226467847157502,
"grad_norm": 0.4612419638937902,
"learning_rate": 3.8470831895063865e-05,
"loss": 0.4084,
"step": 990
},
{
"epoch": 0.9273066169617894,
"grad_norm": 0.38416516599213146,
"learning_rate": 3.838453572661374e-05,
"loss": 0.4041,
"step": 995
},
{
"epoch": 0.9319664492078286,
"grad_norm": 0.44592015311937055,
"learning_rate": 3.8298239558163615e-05,
"loss": 0.4072,
"step": 1000
},
{
"epoch": 0.9366262814538676,
"grad_norm": 0.39712989940263566,
"learning_rate": 3.82119433897135e-05,
"loss": 0.412,
"step": 1005
},
{
"epoch": 0.9412861136999068,
"grad_norm": 0.38203797576686194,
"learning_rate": 3.812564722126338e-05,
"loss": 0.4043,
"step": 1010
},
{
"epoch": 0.9459459459459459,
"grad_norm": 0.424117954981762,
"learning_rate": 3.8039351052813256e-05,
"loss": 0.4128,
"step": 1015
},
{
"epoch": 0.9506057781919851,
"grad_norm": 0.42226979399269843,
"learning_rate": 3.7953054884363134e-05,
"loss": 0.4064,
"step": 1020
},
{
"epoch": 0.9552656104380243,
"grad_norm": 0.36275341625519136,
"learning_rate": 3.786675871591301e-05,
"loss": 0.41,
"step": 1025
},
{
"epoch": 0.9599254426840633,
"grad_norm": 0.2989486198553155,
"learning_rate": 3.77804625474629e-05,
"loss": 0.415,
"step": 1030
},
{
"epoch": 0.9645852749301025,
"grad_norm": 0.4490357688676246,
"learning_rate": 3.7694166379012776e-05,
"loss": 0.4075,
"step": 1035
},
{
"epoch": 0.9692451071761417,
"grad_norm": 0.4117148830526542,
"learning_rate": 3.7607870210562654e-05,
"loss": 0.405,
"step": 1040
},
{
"epoch": 0.9739049394221808,
"grad_norm": 0.3029562636088919,
"learning_rate": 3.752157404211253e-05,
"loss": 0.3974,
"step": 1045
},
{
"epoch": 0.97856477166822,
"grad_norm": 0.45571770500867936,
"learning_rate": 3.743527787366241e-05,
"loss": 0.4274,
"step": 1050
},
{
"epoch": 0.983224603914259,
"grad_norm": 0.3893154421399455,
"learning_rate": 3.734898170521229e-05,
"loss": 0.4046,
"step": 1055
},
{
"epoch": 0.9878844361602982,
"grad_norm": 0.4126071992656825,
"learning_rate": 3.7262685536762174e-05,
"loss": 0.4039,
"step": 1060
},
{
"epoch": 0.9925442684063374,
"grad_norm": 0.3618208759302852,
"learning_rate": 3.7176389368312045e-05,
"loss": 0.4085,
"step": 1065
},
{
"epoch": 0.9972041006523765,
"grad_norm": 0.37089340832251017,
"learning_rate": 3.709009319986193e-05,
"loss": 0.4083,
"step": 1070
},
{
"epoch": 1.0018639328984156,
"grad_norm": 0.3466767328048872,
"learning_rate": 3.700379703141181e-05,
"loss": 0.3726,
"step": 1075
},
{
"epoch": 1.0065237651444547,
"grad_norm": 0.32629710042784293,
"learning_rate": 3.6917500862961687e-05,
"loss": 0.3473,
"step": 1080
},
{
"epoch": 1.011183597390494,
"grad_norm": 0.4255519477332497,
"learning_rate": 3.6831204694511565e-05,
"loss": 0.3538,
"step": 1085
},
{
"epoch": 1.015843429636533,
"grad_norm": 0.37257752212763834,
"learning_rate": 3.674490852606144e-05,
"loss": 0.3605,
"step": 1090
},
{
"epoch": 1.0205032618825722,
"grad_norm": 0.3188706073756023,
"learning_rate": 3.665861235761132e-05,
"loss": 0.3462,
"step": 1095
},
{
"epoch": 1.0251630941286114,
"grad_norm": 0.3323491863961747,
"learning_rate": 3.6572316189161206e-05,
"loss": 0.3373,
"step": 1100
},
{
"epoch": 1.0298229263746506,
"grad_norm": 0.3147904606802217,
"learning_rate": 3.6486020020711085e-05,
"loss": 0.3435,
"step": 1105
},
{
"epoch": 1.0344827586206897,
"grad_norm": 0.31148716928303666,
"learning_rate": 3.639972385226096e-05,
"loss": 0.3444,
"step": 1110
},
{
"epoch": 1.0391425908667289,
"grad_norm": 0.3115874216444653,
"learning_rate": 3.631342768381084e-05,
"loss": 0.3439,
"step": 1115
},
{
"epoch": 1.0438024231127678,
"grad_norm": 0.3850573772968161,
"learning_rate": 3.622713151536072e-05,
"loss": 0.3445,
"step": 1120
},
{
"epoch": 1.048462255358807,
"grad_norm": 0.331203932647263,
"learning_rate": 3.6140835346910604e-05,
"loss": 0.3479,
"step": 1125
},
{
"epoch": 1.0531220876048462,
"grad_norm": 0.3247246142084058,
"learning_rate": 3.6054539178460476e-05,
"loss": 0.3442,
"step": 1130
},
{
"epoch": 1.0577819198508853,
"grad_norm": 0.39344991446138056,
"learning_rate": 3.5968243010010354e-05,
"loss": 0.3609,
"step": 1135
},
{
"epoch": 1.0624417520969245,
"grad_norm": 0.3313029115877748,
"learning_rate": 3.588194684156024e-05,
"loss": 0.3519,
"step": 1140
},
{
"epoch": 1.0671015843429636,
"grad_norm": 0.3505328862414304,
"learning_rate": 3.579565067311012e-05,
"loss": 0.3432,
"step": 1145
},
{
"epoch": 1.0717614165890028,
"grad_norm": 0.30763353170600233,
"learning_rate": 3.5709354504659995e-05,
"loss": 0.3397,
"step": 1150
},
{
"epoch": 1.076421248835042,
"grad_norm": 0.33720507780205905,
"learning_rate": 3.5623058336209874e-05,
"loss": 0.3508,
"step": 1155
},
{
"epoch": 1.0810810810810811,
"grad_norm": 0.29987986518950155,
"learning_rate": 3.553676216775975e-05,
"loss": 0.342,
"step": 1160
},
{
"epoch": 1.0857409133271203,
"grad_norm": 0.2918824556630535,
"learning_rate": 3.545046599930964e-05,
"loss": 0.354,
"step": 1165
},
{
"epoch": 1.0904007455731595,
"grad_norm": 0.28884108622026256,
"learning_rate": 3.536416983085951e-05,
"loss": 0.3555,
"step": 1170
},
{
"epoch": 1.0950605778191984,
"grad_norm": 0.3492325417202242,
"learning_rate": 3.5277873662409386e-05,
"loss": 0.3419,
"step": 1175
},
{
"epoch": 1.0997204100652376,
"grad_norm": 0.3374356989327302,
"learning_rate": 3.519157749395927e-05,
"loss": 0.3479,
"step": 1180
},
{
"epoch": 1.1043802423112767,
"grad_norm": 0.40694167907836565,
"learning_rate": 3.510528132550915e-05,
"loss": 0.345,
"step": 1185
},
{
"epoch": 1.109040074557316,
"grad_norm": 0.29634913375582494,
"learning_rate": 3.501898515705903e-05,
"loss": 0.347,
"step": 1190
},
{
"epoch": 1.113699906803355,
"grad_norm": 0.33725135215696544,
"learning_rate": 3.4932688988608906e-05,
"loss": 0.3584,
"step": 1195
},
{
"epoch": 1.1183597390493942,
"grad_norm": 0.36441882351153176,
"learning_rate": 3.4846392820158784e-05,
"loss": 0.3475,
"step": 1200
},
{
"epoch": 1.1230195712954334,
"grad_norm": 0.43771868179170575,
"learning_rate": 3.476009665170867e-05,
"loss": 0.3476,
"step": 1205
},
{
"epoch": 1.1276794035414726,
"grad_norm": 0.36715913285296337,
"learning_rate": 3.467380048325855e-05,
"loss": 0.3456,
"step": 1210
},
{
"epoch": 1.1323392357875117,
"grad_norm": 0.3210278084887568,
"learning_rate": 3.458750431480842e-05,
"loss": 0.3528,
"step": 1215
},
{
"epoch": 1.1369990680335509,
"grad_norm": 0.40978152129347956,
"learning_rate": 3.4501208146358304e-05,
"loss": 0.3517,
"step": 1220
},
{
"epoch": 1.14165890027959,
"grad_norm": 0.3501037455697221,
"learning_rate": 3.441491197790818e-05,
"loss": 0.3474,
"step": 1225
},
{
"epoch": 1.146318732525629,
"grad_norm": 0.2949808119143393,
"learning_rate": 3.432861580945806e-05,
"loss": 0.3487,
"step": 1230
},
{
"epoch": 1.1509785647716682,
"grad_norm": 0.33537243519091453,
"learning_rate": 3.424231964100794e-05,
"loss": 0.348,
"step": 1235
},
{
"epoch": 1.1556383970177073,
"grad_norm": 0.34215531389067516,
"learning_rate": 3.415602347255782e-05,
"loss": 0.3465,
"step": 1240
},
{
"epoch": 1.1602982292637465,
"grad_norm": 0.28579102346746615,
"learning_rate": 3.40697273041077e-05,
"loss": 0.3535,
"step": 1245
},
{
"epoch": 1.1649580615097856,
"grad_norm": 0.38985713423130514,
"learning_rate": 3.398343113565758e-05,
"loss": 0.3476,
"step": 1250
},
{
"epoch": 1.1696178937558248,
"grad_norm": 0.2757491186716454,
"learning_rate": 3.389713496720746e-05,
"loss": 0.3265,
"step": 1255
},
{
"epoch": 1.174277726001864,
"grad_norm": 0.26622935533794223,
"learning_rate": 3.381083879875734e-05,
"loss": 0.3463,
"step": 1260
},
{
"epoch": 1.1789375582479031,
"grad_norm": 0.2994815102975485,
"learning_rate": 3.3724542630307215e-05,
"loss": 0.3429,
"step": 1265
},
{
"epoch": 1.1835973904939423,
"grad_norm": 0.30900186681072417,
"learning_rate": 3.363824646185709e-05,
"loss": 0.3522,
"step": 1270
},
{
"epoch": 1.1882572227399812,
"grad_norm": 0.28156243744828824,
"learning_rate": 3.355195029340698e-05,
"loss": 0.3458,
"step": 1275
},
{
"epoch": 1.1929170549860204,
"grad_norm": 0.34048483109754407,
"learning_rate": 3.346565412495685e-05,
"loss": 0.3567,
"step": 1280
},
{
"epoch": 1.1975768872320596,
"grad_norm": 0.3482879519870774,
"learning_rate": 3.3379357956506735e-05,
"loss": 0.3578,
"step": 1285
},
{
"epoch": 1.2022367194780987,
"grad_norm": 0.3433620665119787,
"learning_rate": 3.329306178805661e-05,
"loss": 0.3408,
"step": 1290
},
{
"epoch": 1.206896551724138,
"grad_norm": 0.33173956917096376,
"learning_rate": 3.320676561960649e-05,
"loss": 0.3456,
"step": 1295
},
{
"epoch": 1.211556383970177,
"grad_norm": 0.4271653931916803,
"learning_rate": 3.312046945115637e-05,
"loss": 0.3561,
"step": 1300
},
{
"epoch": 1.2162162162162162,
"grad_norm": 0.38709260099877685,
"learning_rate": 3.303417328270625e-05,
"loss": 0.3527,
"step": 1305
},
{
"epoch": 1.2208760484622554,
"grad_norm": 0.29041529382376463,
"learning_rate": 3.2947877114256126e-05,
"loss": 0.3422,
"step": 1310
},
{
"epoch": 1.2255358807082946,
"grad_norm": 0.31574510099421743,
"learning_rate": 3.286158094580601e-05,
"loss": 0.3619,
"step": 1315
},
{
"epoch": 1.2301957129543337,
"grad_norm": 0.2779026955554108,
"learning_rate": 3.277528477735589e-05,
"loss": 0.348,
"step": 1320
},
{
"epoch": 1.2348555452003729,
"grad_norm": 0.3117138832842607,
"learning_rate": 3.268898860890577e-05,
"loss": 0.3548,
"step": 1325
},
{
"epoch": 1.2395153774464118,
"grad_norm": 0.34556214123661766,
"learning_rate": 3.2602692440455645e-05,
"loss": 0.3427,
"step": 1330
},
{
"epoch": 1.244175209692451,
"grad_norm": 0.3207908460587471,
"learning_rate": 3.2516396272005524e-05,
"loss": 0.3457,
"step": 1335
},
{
"epoch": 1.2488350419384902,
"grad_norm": 0.4579638822801661,
"learning_rate": 3.243010010355541e-05,
"loss": 0.3511,
"step": 1340
},
{
"epoch": 1.2534948741845293,
"grad_norm": 0.3706933498980916,
"learning_rate": 3.234380393510528e-05,
"loss": 0.3385,
"step": 1345
},
{
"epoch": 1.2581547064305685,
"grad_norm": 0.28751428494705217,
"learning_rate": 3.225750776665516e-05,
"loss": 0.3454,
"step": 1350
},
{
"epoch": 1.2628145386766076,
"grad_norm": 0.29088848949940943,
"learning_rate": 3.217121159820504e-05,
"loss": 0.3554,
"step": 1355
},
{
"epoch": 1.2674743709226468,
"grad_norm": 0.32043873791461,
"learning_rate": 3.208491542975492e-05,
"loss": 0.3448,
"step": 1360
},
{
"epoch": 1.272134203168686,
"grad_norm": 0.32135003427013953,
"learning_rate": 3.19986192613048e-05,
"loss": 0.3504,
"step": 1365
},
{
"epoch": 1.2767940354147251,
"grad_norm": 0.34836051639597065,
"learning_rate": 3.191232309285468e-05,
"loss": 0.3491,
"step": 1370
},
{
"epoch": 1.281453867660764,
"grad_norm": 0.3083433720315085,
"learning_rate": 3.1826026924404556e-05,
"loss": 0.339,
"step": 1375
},
{
"epoch": 1.2861136999068035,
"grad_norm": 0.3294061695691229,
"learning_rate": 3.173973075595444e-05,
"loss": 0.3536,
"step": 1380
},
{
"epoch": 1.2907735321528424,
"grad_norm": 0.3243247321823703,
"learning_rate": 3.165343458750431e-05,
"loss": 0.3529,
"step": 1385
},
{
"epoch": 1.2954333643988816,
"grad_norm": 0.34216987102113755,
"learning_rate": 3.156713841905419e-05,
"loss": 0.3455,
"step": 1390
},
{
"epoch": 1.3000931966449207,
"grad_norm": 0.2719200783387942,
"learning_rate": 3.1480842250604076e-05,
"loss": 0.3494,
"step": 1395
},
{
"epoch": 1.30475302889096,
"grad_norm": 0.3505609559139422,
"learning_rate": 3.1394546082153954e-05,
"loss": 0.3437,
"step": 1400
},
{
"epoch": 1.309412861136999,
"grad_norm": 0.33365123410554454,
"learning_rate": 3.130824991370383e-05,
"loss": 0.3544,
"step": 1405
},
{
"epoch": 1.3140726933830382,
"grad_norm": 0.3189436936052233,
"learning_rate": 3.122195374525371e-05,
"loss": 0.3424,
"step": 1410
},
{
"epoch": 1.3187325256290774,
"grad_norm": 0.2863006966606347,
"learning_rate": 3.113565757680359e-05,
"loss": 0.3503,
"step": 1415
},
{
"epoch": 1.3233923578751166,
"grad_norm": 0.28826221521063,
"learning_rate": 3.1049361408353474e-05,
"loss": 0.3376,
"step": 1420
},
{
"epoch": 1.3280521901211557,
"grad_norm": 0.24894240427324923,
"learning_rate": 3.096306523990335e-05,
"loss": 0.3455,
"step": 1425
},
{
"epoch": 1.3327120223671947,
"grad_norm": 0.27568016054738054,
"learning_rate": 3.0876769071453223e-05,
"loss": 0.3541,
"step": 1430
},
{
"epoch": 1.337371854613234,
"grad_norm": 0.35064905716882083,
"learning_rate": 3.079047290300311e-05,
"loss": 0.3519,
"step": 1435
},
{
"epoch": 1.342031686859273,
"grad_norm": 0.3212183182772905,
"learning_rate": 3.070417673455299e-05,
"loss": 0.3505,
"step": 1440
},
{
"epoch": 1.3466915191053122,
"grad_norm": 0.28916536258878456,
"learning_rate": 3.061788056610287e-05,
"loss": 0.3553,
"step": 1445
},
{
"epoch": 1.3513513513513513,
"grad_norm": 0.3235421441470868,
"learning_rate": 3.053158439765274e-05,
"loss": 0.3411,
"step": 1450
},
{
"epoch": 1.3560111835973905,
"grad_norm": 0.2968602289573508,
"learning_rate": 3.0445288229202625e-05,
"loss": 0.3488,
"step": 1455
},
{
"epoch": 1.3606710158434296,
"grad_norm": 0.3184864286908833,
"learning_rate": 3.0358992060752506e-05,
"loss": 0.3548,
"step": 1460
},
{
"epoch": 1.3653308480894688,
"grad_norm": 0.3229863708803619,
"learning_rate": 3.027269589230238e-05,
"loss": 0.3487,
"step": 1465
},
{
"epoch": 1.369990680335508,
"grad_norm": 0.3221853882096656,
"learning_rate": 3.018639972385226e-05,
"loss": 0.3463,
"step": 1470
},
{
"epoch": 1.3746505125815471,
"grad_norm": 0.36492627157017027,
"learning_rate": 3.010010355540214e-05,
"loss": 0.3479,
"step": 1475
},
{
"epoch": 1.3793103448275863,
"grad_norm": 0.26851057138257656,
"learning_rate": 3.001380738695202e-05,
"loss": 0.3317,
"step": 1480
},
{
"epoch": 1.3839701770736252,
"grad_norm": 0.2833599124258928,
"learning_rate": 2.99275112185019e-05,
"loss": 0.349,
"step": 1485
},
{
"epoch": 1.3886300093196646,
"grad_norm": 0.3540057666942934,
"learning_rate": 2.984121505005178e-05,
"loss": 0.3541,
"step": 1490
},
{
"epoch": 1.3932898415657036,
"grad_norm": 0.37206165635102434,
"learning_rate": 2.9754918881601657e-05,
"loss": 0.3625,
"step": 1495
},
{
"epoch": 1.3979496738117427,
"grad_norm": 0.3421043062141437,
"learning_rate": 2.966862271315154e-05,
"loss": 0.3588,
"step": 1500
},
{
"epoch": 1.402609506057782,
"grad_norm": 0.3495424193811505,
"learning_rate": 2.9582326544701417e-05,
"loss": 0.3531,
"step": 1505
},
{
"epoch": 1.407269338303821,
"grad_norm": 0.30714455983322153,
"learning_rate": 2.9496030376251292e-05,
"loss": 0.347,
"step": 1510
},
{
"epoch": 1.4119291705498602,
"grad_norm": 0.271764182801288,
"learning_rate": 2.9409734207801177e-05,
"loss": 0.3504,
"step": 1515
},
{
"epoch": 1.4165890027958994,
"grad_norm": 0.345977791976646,
"learning_rate": 2.9323438039351052e-05,
"loss": 0.35,
"step": 1520
},
{
"epoch": 1.4212488350419386,
"grad_norm": 0.3090209854200775,
"learning_rate": 2.9237141870900937e-05,
"loss": 0.3457,
"step": 1525
},
{
"epoch": 1.4259086672879777,
"grad_norm": 0.2902474757739303,
"learning_rate": 2.9150845702450812e-05,
"loss": 0.3448,
"step": 1530
},
{
"epoch": 1.4305684995340169,
"grad_norm": 0.32236612838715284,
"learning_rate": 2.906454953400069e-05,
"loss": 0.3497,
"step": 1535
},
{
"epoch": 1.4352283317800558,
"grad_norm": 0.32048885565576973,
"learning_rate": 2.897825336555057e-05,
"loss": 0.346,
"step": 1540
},
{
"epoch": 1.439888164026095,
"grad_norm": 0.36763831382668277,
"learning_rate": 2.889195719710045e-05,
"loss": 0.3462,
"step": 1545
},
{
"epoch": 1.4445479962721341,
"grad_norm": 0.3667395673175459,
"learning_rate": 2.8805661028650328e-05,
"loss": 0.3426,
"step": 1550
},
{
"epoch": 1.4492078285181733,
"grad_norm": 0.43244599196751526,
"learning_rate": 2.871936486020021e-05,
"loss": 0.3398,
"step": 1555
},
{
"epoch": 1.4538676607642125,
"grad_norm": 0.32356353846643177,
"learning_rate": 2.8633068691750088e-05,
"loss": 0.3425,
"step": 1560
},
{
"epoch": 1.4585274930102516,
"grad_norm": 0.3412037069282531,
"learning_rate": 2.854677252329997e-05,
"loss": 0.3451,
"step": 1565
},
{
"epoch": 1.4631873252562908,
"grad_norm": 0.3590171697330432,
"learning_rate": 2.8460476354849848e-05,
"loss": 0.3524,
"step": 1570
},
{
"epoch": 1.46784715750233,
"grad_norm": 0.3940544933936819,
"learning_rate": 2.8374180186399723e-05,
"loss": 0.3519,
"step": 1575
},
{
"epoch": 1.4725069897483691,
"grad_norm": 0.3182855473229679,
"learning_rate": 2.8287884017949608e-05,
"loss": 0.3533,
"step": 1580
},
{
"epoch": 1.477166821994408,
"grad_norm": 0.3568443752239294,
"learning_rate": 2.8201587849499482e-05,
"loss": 0.346,
"step": 1585
},
{
"epoch": 1.4818266542404475,
"grad_norm": 0.34974506261722194,
"learning_rate": 2.811529168104936e-05,
"loss": 0.3564,
"step": 1590
},
{
"epoch": 1.4864864864864864,
"grad_norm": 0.31142218614639605,
"learning_rate": 2.8028995512599242e-05,
"loss": 0.3473,
"step": 1595
},
{
"epoch": 1.4911463187325256,
"grad_norm": 0.3634657161601137,
"learning_rate": 2.794269934414912e-05,
"loss": 0.3485,
"step": 1600
},
{
"epoch": 1.4958061509785647,
"grad_norm": 0.29200266930057117,
"learning_rate": 2.7856403175699002e-05,
"loss": 0.3529,
"step": 1605
},
{
"epoch": 1.500465983224604,
"grad_norm": 0.3440371768426119,
"learning_rate": 2.777010700724888e-05,
"loss": 0.3477,
"step": 1610
},
{
"epoch": 1.505125815470643,
"grad_norm": 0.32373391689242287,
"learning_rate": 2.768381083879876e-05,
"loss": 0.3431,
"step": 1615
},
{
"epoch": 1.5097856477166822,
"grad_norm": 0.26950475692133025,
"learning_rate": 2.759751467034864e-05,
"loss": 0.3453,
"step": 1620
},
{
"epoch": 1.5144454799627214,
"grad_norm": 0.2826210501773821,
"learning_rate": 2.751121850189852e-05,
"loss": 0.3397,
"step": 1625
},
{
"epoch": 1.5191053122087603,
"grad_norm": 0.3102161140774324,
"learning_rate": 2.7424922333448393e-05,
"loss": 0.3328,
"step": 1630
},
{
"epoch": 1.5237651444547997,
"grad_norm": 0.33760697921958377,
"learning_rate": 2.7338626164998278e-05,
"loss": 0.3403,
"step": 1635
},
{
"epoch": 1.5284249767008387,
"grad_norm": 0.29166453965201583,
"learning_rate": 2.7252329996548153e-05,
"loss": 0.3434,
"step": 1640
},
{
"epoch": 1.533084808946878,
"grad_norm": 0.2798133072967164,
"learning_rate": 2.7166033828098038e-05,
"loss": 0.3586,
"step": 1645
},
{
"epoch": 1.537744641192917,
"grad_norm": 0.32261771570684744,
"learning_rate": 2.7079737659647913e-05,
"loss": 0.3473,
"step": 1650
},
{
"epoch": 1.5424044734389561,
"grad_norm": 0.32981956943151086,
"learning_rate": 2.699344149119779e-05,
"loss": 0.3543,
"step": 1655
},
{
"epoch": 1.5470643056849953,
"grad_norm": 0.32701364952130885,
"learning_rate": 2.6907145322747673e-05,
"loss": 0.3488,
"step": 1660
},
{
"epoch": 1.5517241379310345,
"grad_norm": 0.3332712178312528,
"learning_rate": 2.682084915429755e-05,
"loss": 0.3398,
"step": 1665
},
{
"epoch": 1.5563839701770736,
"grad_norm": 0.348283845697873,
"learning_rate": 2.673455298584743e-05,
"loss": 0.3447,
"step": 1670
},
{
"epoch": 1.5610438024231128,
"grad_norm": 0.3437847787547748,
"learning_rate": 2.664825681739731e-05,
"loss": 0.352,
"step": 1675
},
{
"epoch": 1.565703634669152,
"grad_norm": 0.35408777756271287,
"learning_rate": 2.6561960648947186e-05,
"loss": 0.3437,
"step": 1680
},
{
"epoch": 1.570363466915191,
"grad_norm": 0.34896651367726,
"learning_rate": 2.647566448049707e-05,
"loss": 0.3398,
"step": 1685
},
{
"epoch": 1.5750232991612303,
"grad_norm": 0.32694492129697167,
"learning_rate": 2.6389368312046945e-05,
"loss": 0.3521,
"step": 1690
},
{
"epoch": 1.5796831314072692,
"grad_norm": 0.2909691544432443,
"learning_rate": 2.6303072143596824e-05,
"loss": 0.3561,
"step": 1695
},
{
"epoch": 1.5843429636533086,
"grad_norm": 0.38721758891012714,
"learning_rate": 2.6216775975146705e-05,
"loss": 0.3486,
"step": 1700
},
{
"epoch": 1.5890027958993476,
"grad_norm": 0.311267472576106,
"learning_rate": 2.6130479806696584e-05,
"loss": 0.3458,
"step": 1705
},
{
"epoch": 1.5936626281453867,
"grad_norm": 0.3757471184213451,
"learning_rate": 2.6044183638246462e-05,
"loss": 0.3518,
"step": 1710
},
{
"epoch": 1.598322460391426,
"grad_norm": 0.2688206025720871,
"learning_rate": 2.5957887469796343e-05,
"loss": 0.341,
"step": 1715
},
{
"epoch": 1.602982292637465,
"grad_norm": 0.3092285021432447,
"learning_rate": 2.587159130134622e-05,
"loss": 0.3451,
"step": 1720
},
{
"epoch": 1.6076421248835042,
"grad_norm": 0.28400015059271655,
"learning_rate": 2.5785295132896096e-05,
"loss": 0.3451,
"step": 1725
},
{
"epoch": 1.6123019571295434,
"grad_norm": 0.3445231051932557,
"learning_rate": 2.569899896444598e-05,
"loss": 0.339,
"step": 1730
},
{
"epoch": 1.6169617893755825,
"grad_norm": 0.38314927018404166,
"learning_rate": 2.5612702795995856e-05,
"loss": 0.3501,
"step": 1735
},
{
"epoch": 1.6216216216216215,
"grad_norm": 0.29012666606461185,
"learning_rate": 2.552640662754574e-05,
"loss": 0.3442,
"step": 1740
},
{
"epoch": 1.6262814538676609,
"grad_norm": 0.3288769112109895,
"learning_rate": 2.5440110459095616e-05,
"loss": 0.3512,
"step": 1745
},
{
"epoch": 1.6309412861136998,
"grad_norm": 0.3868687049946097,
"learning_rate": 2.5353814290645494e-05,
"loss": 0.3498,
"step": 1750
},
{
"epoch": 1.6356011183597392,
"grad_norm": 0.34707374673996927,
"learning_rate": 2.5267518122195376e-05,
"loss": 0.3419,
"step": 1755
},
{
"epoch": 1.6402609506057781,
"grad_norm": 0.38650219982852274,
"learning_rate": 2.5181221953745254e-05,
"loss": 0.3502,
"step": 1760
},
{
"epoch": 1.6449207828518173,
"grad_norm": 0.30442621914305384,
"learning_rate": 2.5094925785295132e-05,
"loss": 0.352,
"step": 1765
},
{
"epoch": 1.6495806150978565,
"grad_norm": 0.295831906556078,
"learning_rate": 2.5008629616845014e-05,
"loss": 0.3451,
"step": 1770
},
{
"epoch": 1.6542404473438956,
"grad_norm": 0.41956198605419515,
"learning_rate": 2.4922333448394892e-05,
"loss": 0.338,
"step": 1775
},
{
"epoch": 1.6589002795899348,
"grad_norm": 0.34229517285638916,
"learning_rate": 2.483603727994477e-05,
"loss": 0.3475,
"step": 1780
},
{
"epoch": 1.6635601118359737,
"grad_norm": 0.31924807835467234,
"learning_rate": 2.4749741111494652e-05,
"loss": 0.3446,
"step": 1785
},
{
"epoch": 1.6682199440820131,
"grad_norm": 0.3284589327531291,
"learning_rate": 2.466344494304453e-05,
"loss": 0.3438,
"step": 1790
},
{
"epoch": 1.672879776328052,
"grad_norm": 0.28703634586528093,
"learning_rate": 2.457714877459441e-05,
"loss": 0.3366,
"step": 1795
},
{
"epoch": 1.6775396085740915,
"grad_norm": 0.2932607124323887,
"learning_rate": 2.4490852606144287e-05,
"loss": 0.3366,
"step": 1800
},
{
"epoch": 1.6821994408201304,
"grad_norm": 0.296583913776671,
"learning_rate": 2.440455643769417e-05,
"loss": 0.34,
"step": 1805
},
{
"epoch": 1.6868592730661698,
"grad_norm": 0.2687683226730502,
"learning_rate": 2.4318260269244047e-05,
"loss": 0.3434,
"step": 1810
},
{
"epoch": 1.6915191053122087,
"grad_norm": 0.25347808663410265,
"learning_rate": 2.4231964100793925e-05,
"loss": 0.331,
"step": 1815
},
{
"epoch": 1.696178937558248,
"grad_norm": 0.2713693960239594,
"learning_rate": 2.4145667932343803e-05,
"loss": 0.3426,
"step": 1820
},
{
"epoch": 1.700838769804287,
"grad_norm": 0.26766349202361095,
"learning_rate": 2.4059371763893685e-05,
"loss": 0.3468,
"step": 1825
},
{
"epoch": 1.7054986020503262,
"grad_norm": 0.2761857212120324,
"learning_rate": 2.3973075595443563e-05,
"loss": 0.3524,
"step": 1830
},
{
"epoch": 1.7101584342963654,
"grad_norm": 0.3533387971020311,
"learning_rate": 2.388677942699344e-05,
"loss": 0.347,
"step": 1835
},
{
"epoch": 1.7148182665424043,
"grad_norm": 0.3304465470183926,
"learning_rate": 2.3800483258543323e-05,
"loss": 0.3433,
"step": 1840
},
{
"epoch": 1.7194780987884437,
"grad_norm": 0.2838212354941664,
"learning_rate": 2.37141870900932e-05,
"loss": 0.3487,
"step": 1845
},
{
"epoch": 1.7241379310344827,
"grad_norm": 0.27182912167501216,
"learning_rate": 2.3627890921643083e-05,
"loss": 0.3307,
"step": 1850
},
{
"epoch": 1.728797763280522,
"grad_norm": 0.3463275642546345,
"learning_rate": 2.3541594753192957e-05,
"loss": 0.3407,
"step": 1855
},
{
"epoch": 1.733457595526561,
"grad_norm": 0.33680897836510765,
"learning_rate": 2.345529858474284e-05,
"loss": 0.341,
"step": 1860
},
{
"epoch": 1.7381174277726001,
"grad_norm": 0.33727330449578286,
"learning_rate": 2.3369002416292717e-05,
"loss": 0.3405,
"step": 1865
},
{
"epoch": 1.7427772600186393,
"grad_norm": 0.3100575835627644,
"learning_rate": 2.32827062478426e-05,
"loss": 0.3394,
"step": 1870
},
{
"epoch": 1.7474370922646785,
"grad_norm": 0.2826440437227791,
"learning_rate": 2.3196410079392474e-05,
"loss": 0.348,
"step": 1875
},
{
"epoch": 1.7520969245107176,
"grad_norm": 0.32083683767973736,
"learning_rate": 2.3110113910942355e-05,
"loss": 0.3425,
"step": 1880
},
{
"epoch": 1.7567567567567568,
"grad_norm": 0.29227267913176436,
"learning_rate": 2.3023817742492234e-05,
"loss": 0.3482,
"step": 1885
},
{
"epoch": 1.761416589002796,
"grad_norm": 0.2698620187901574,
"learning_rate": 2.2937521574042115e-05,
"loss": 0.3457,
"step": 1890
},
{
"epoch": 1.766076421248835,
"grad_norm": 0.3151861063948914,
"learning_rate": 2.285122540559199e-05,
"loss": 0.3431,
"step": 1895
},
{
"epoch": 1.7707362534948743,
"grad_norm": 0.3538300373789918,
"learning_rate": 2.276492923714187e-05,
"loss": 0.3353,
"step": 1900
},
{
"epoch": 1.7753960857409132,
"grad_norm": 0.27314570126621124,
"learning_rate": 2.267863306869175e-05,
"loss": 0.3347,
"step": 1905
},
{
"epoch": 1.7800559179869526,
"grad_norm": 0.260828982118303,
"learning_rate": 2.259233690024163e-05,
"loss": 0.3476,
"step": 1910
},
{
"epoch": 1.7847157502329916,
"grad_norm": 0.293875378087096,
"learning_rate": 2.250604073179151e-05,
"loss": 0.3308,
"step": 1915
},
{
"epoch": 1.7893755824790307,
"grad_norm": 0.3118963241797147,
"learning_rate": 2.2419744563341388e-05,
"loss": 0.3469,
"step": 1920
},
{
"epoch": 1.7940354147250699,
"grad_norm": 0.25482474228286517,
"learning_rate": 2.233344839489127e-05,
"loss": 0.3435,
"step": 1925
},
{
"epoch": 1.798695246971109,
"grad_norm": 0.29389984211878534,
"learning_rate": 2.2247152226441148e-05,
"loss": 0.34,
"step": 1930
},
{
"epoch": 1.8033550792171482,
"grad_norm": 0.29489566146760227,
"learning_rate": 2.2160856057991026e-05,
"loss": 0.345,
"step": 1935
},
{
"epoch": 1.8080149114631874,
"grad_norm": 0.2852643696421498,
"learning_rate": 2.2074559889540904e-05,
"loss": 0.3327,
"step": 1940
},
{
"epoch": 1.8126747437092265,
"grad_norm": 0.26648292779453114,
"learning_rate": 2.1988263721090786e-05,
"loss": 0.3495,
"step": 1945
},
{
"epoch": 1.8173345759552655,
"grad_norm": 0.27853154206944286,
"learning_rate": 2.1901967552640664e-05,
"loss": 0.333,
"step": 1950
},
{
"epoch": 1.8219944082013049,
"grad_norm": 0.3338574731573854,
"learning_rate": 2.1815671384190542e-05,
"loss": 0.341,
"step": 1955
},
{
"epoch": 1.8266542404473438,
"grad_norm": 0.3269171924058155,
"learning_rate": 2.172937521574042e-05,
"loss": 0.3447,
"step": 1960
},
{
"epoch": 1.8313140726933832,
"grad_norm": 0.30017033096794987,
"learning_rate": 2.1643079047290302e-05,
"loss": 0.3458,
"step": 1965
},
{
"epoch": 1.8359739049394221,
"grad_norm": 0.2850784608066985,
"learning_rate": 2.155678287884018e-05,
"loss": 0.3434,
"step": 1970
},
{
"epoch": 1.8406337371854613,
"grad_norm": 0.2656541385829881,
"learning_rate": 2.147048671039006e-05,
"loss": 0.3516,
"step": 1975
},
{
"epoch": 1.8452935694315005,
"grad_norm": 0.2471203909463743,
"learning_rate": 2.1384190541939937e-05,
"loss": 0.3418,
"step": 1980
},
{
"epoch": 1.8499534016775396,
"grad_norm": 0.2650899753963081,
"learning_rate": 2.129789437348982e-05,
"loss": 0.3377,
"step": 1985
},
{
"epoch": 1.8546132339235788,
"grad_norm": 0.2647940247061497,
"learning_rate": 2.1211598205039697e-05,
"loss": 0.3564,
"step": 1990
},
{
"epoch": 1.8592730661696177,
"grad_norm": 0.3227812765644576,
"learning_rate": 2.1125302036589575e-05,
"loss": 0.3421,
"step": 1995
},
{
"epoch": 1.8639328984156571,
"grad_norm": 0.297267058414044,
"learning_rate": 2.1039005868139457e-05,
"loss": 0.3409,
"step": 2000
},
{
"epoch": 1.868592730661696,
"grad_norm": 0.2805203477278729,
"learning_rate": 2.0952709699689335e-05,
"loss": 0.3344,
"step": 2005
},
{
"epoch": 1.8732525629077355,
"grad_norm": 0.2885697842925641,
"learning_rate": 2.0866413531239216e-05,
"loss": 0.3499,
"step": 2010
},
{
"epoch": 1.8779123951537744,
"grad_norm": 0.2864142255057866,
"learning_rate": 2.078011736278909e-05,
"loss": 0.336,
"step": 2015
},
{
"epoch": 1.8825722273998136,
"grad_norm": 0.27559826462867176,
"learning_rate": 2.0693821194338973e-05,
"loss": 0.3308,
"step": 2020
},
{
"epoch": 1.8872320596458527,
"grad_norm": 0.3245194476026471,
"learning_rate": 2.060752502588885e-05,
"loss": 0.337,
"step": 2025
},
{
"epoch": 1.8918918918918919,
"grad_norm": 0.24340386122312743,
"learning_rate": 2.0521228857438733e-05,
"loss": 0.3441,
"step": 2030
},
{
"epoch": 1.896551724137931,
"grad_norm": 0.24501266752590212,
"learning_rate": 2.0434932688988608e-05,
"loss": 0.3371,
"step": 2035
},
{
"epoch": 1.9012115563839702,
"grad_norm": 0.24503608734019988,
"learning_rate": 2.034863652053849e-05,
"loss": 0.3398,
"step": 2040
},
{
"epoch": 1.9058713886300094,
"grad_norm": 0.24066133214971583,
"learning_rate": 2.0262340352088367e-05,
"loss": 0.3556,
"step": 2045
},
{
"epoch": 1.9105312208760483,
"grad_norm": 0.2932258663979882,
"learning_rate": 2.017604418363825e-05,
"loss": 0.3461,
"step": 2050
},
{
"epoch": 1.9151910531220877,
"grad_norm": 0.2923717720171897,
"learning_rate": 2.0089748015188127e-05,
"loss": 0.3407,
"step": 2055
},
{
"epoch": 1.9198508853681266,
"grad_norm": 0.2440018819032557,
"learning_rate": 2.0003451846738005e-05,
"loss": 0.3481,
"step": 2060
},
{
"epoch": 1.924510717614166,
"grad_norm": 0.3006737152563499,
"learning_rate": 1.9917155678287887e-05,
"loss": 0.3375,
"step": 2065
},
{
"epoch": 1.929170549860205,
"grad_norm": 0.26381943626428767,
"learning_rate": 1.9830859509837765e-05,
"loss": 0.345,
"step": 2070
},
{
"epoch": 1.9338303821062441,
"grad_norm": 0.2559642563887702,
"learning_rate": 1.9744563341387643e-05,
"loss": 0.3463,
"step": 2075
},
{
"epoch": 1.9384902143522833,
"grad_norm": 0.2854146175559252,
"learning_rate": 1.9658267172937522e-05,
"loss": 0.3415,
"step": 2080
},
{
"epoch": 1.9431500465983225,
"grad_norm": 0.3551942925927518,
"learning_rate": 1.9571971004487403e-05,
"loss": 0.3411,
"step": 2085
},
{
"epoch": 1.9478098788443616,
"grad_norm": 0.30092223062194995,
"learning_rate": 1.948567483603728e-05,
"loss": 0.3472,
"step": 2090
},
{
"epoch": 1.9524697110904008,
"grad_norm": 0.30697657331539163,
"learning_rate": 1.939937866758716e-05,
"loss": 0.3295,
"step": 2095
},
{
"epoch": 1.95712954333644,
"grad_norm": 0.2980503359064689,
"learning_rate": 1.9313082499137038e-05,
"loss": 0.3468,
"step": 2100
},
{
"epoch": 1.961789375582479,
"grad_norm": 0.3059325972042569,
"learning_rate": 1.922678633068692e-05,
"loss": 0.3412,
"step": 2105
},
{
"epoch": 1.9664492078285183,
"grad_norm": 0.2960890819228813,
"learning_rate": 1.9140490162236798e-05,
"loss": 0.3466,
"step": 2110
},
{
"epoch": 1.9711090400745572,
"grad_norm": 0.2851735506449098,
"learning_rate": 1.9054193993786676e-05,
"loss": 0.354,
"step": 2115
},
{
"epoch": 1.9757688723205966,
"grad_norm": 0.2643197569523461,
"learning_rate": 1.8967897825336554e-05,
"loss": 0.342,
"step": 2120
},
{
"epoch": 1.9804287045666356,
"grad_norm": 0.2544038073538777,
"learning_rate": 1.8881601656886436e-05,
"loss": 0.3393,
"step": 2125
},
{
"epoch": 1.9850885368126747,
"grad_norm": 0.2713719078498063,
"learning_rate": 1.8795305488436314e-05,
"loss": 0.3428,
"step": 2130
},
{
"epoch": 1.9897483690587139,
"grad_norm": 0.3067483498066776,
"learning_rate": 1.8709009319986192e-05,
"loss": 0.3315,
"step": 2135
},
{
"epoch": 1.994408201304753,
"grad_norm": 0.2733634463212891,
"learning_rate": 1.8622713151536074e-05,
"loss": 0.3385,
"step": 2140
},
{
"epoch": 1.9990680335507922,
"grad_norm": 0.2646728305606801,
"learning_rate": 1.8536416983085952e-05,
"loss": 0.3432,
"step": 2145
}
],
"logging_steps": 5,
"max_steps": 3219,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.8376286011392e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}