Qwen7B-Math-L32 / checkpoint-2146 /trainer_state.json
ZMC2019's picture
Training in progress, epoch 2, checkpoint
b6a4377 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": 59.914947522200755,
"learning_rate": 7.763975155279503e-07,
"loss": 10.6862,
"step": 5
},
{
"epoch": 0.009319664492078284,
"grad_norm": 61.506550472016244,
"learning_rate": 1.5527950310559006e-06,
"loss": 10.4784,
"step": 10
},
{
"epoch": 0.013979496738117428,
"grad_norm": 100.03132681639305,
"learning_rate": 2.329192546583851e-06,
"loss": 8.8593,
"step": 15
},
{
"epoch": 0.01863932898415657,
"grad_norm": 28.64812769413447,
"learning_rate": 3.1055900621118013e-06,
"loss": 2.9689,
"step": 20
},
{
"epoch": 0.023299161230195712,
"grad_norm": 3.1458673773327392,
"learning_rate": 3.881987577639752e-06,
"loss": 1.3154,
"step": 25
},
{
"epoch": 0.027958993476234855,
"grad_norm": 1.2888476895438312,
"learning_rate": 4.658385093167702e-06,
"loss": 0.9997,
"step": 30
},
{
"epoch": 0.032618825722273995,
"grad_norm": 0.8962553376010008,
"learning_rate": 5.4347826086956525e-06,
"loss": 0.8452,
"step": 35
},
{
"epoch": 0.03727865796831314,
"grad_norm": 0.6393961134946683,
"learning_rate": 6.2111801242236025e-06,
"loss": 0.7803,
"step": 40
},
{
"epoch": 0.04193849021435228,
"grad_norm": 0.513465541203958,
"learning_rate": 6.9875776397515525e-06,
"loss": 0.7182,
"step": 45
},
{
"epoch": 0.046598322460391424,
"grad_norm": 0.401362860410936,
"learning_rate": 7.763975155279503e-06,
"loss": 0.686,
"step": 50
},
{
"epoch": 0.05125815470643057,
"grad_norm": 0.4467110756469191,
"learning_rate": 8.540372670807453e-06,
"loss": 0.6477,
"step": 55
},
{
"epoch": 0.05591798695246971,
"grad_norm": 0.3260347816467283,
"learning_rate": 9.316770186335403e-06,
"loss": 0.6096,
"step": 60
},
{
"epoch": 0.06057781919850885,
"grad_norm": 0.3077539007822434,
"learning_rate": 1.0093167701863353e-05,
"loss": 0.5929,
"step": 65
},
{
"epoch": 0.06523765144454799,
"grad_norm": 0.32304663164813313,
"learning_rate": 1.0869565217391305e-05,
"loss": 0.5885,
"step": 70
},
{
"epoch": 0.06989748369058714,
"grad_norm": 0.29376236300157593,
"learning_rate": 1.1645962732919255e-05,
"loss": 0.5703,
"step": 75
},
{
"epoch": 0.07455731593662628,
"grad_norm": 0.3348078948651445,
"learning_rate": 1.2422360248447205e-05,
"loss": 0.5604,
"step": 80
},
{
"epoch": 0.07921714818266543,
"grad_norm": 0.3426802151879899,
"learning_rate": 1.3198757763975155e-05,
"loss": 0.5507,
"step": 85
},
{
"epoch": 0.08387698042870456,
"grad_norm": 0.28857538776714187,
"learning_rate": 1.3975155279503105e-05,
"loss": 0.5607,
"step": 90
},
{
"epoch": 0.08853681267474371,
"grad_norm": 0.26144673946425695,
"learning_rate": 1.4751552795031057e-05,
"loss": 0.5421,
"step": 95
},
{
"epoch": 0.09319664492078285,
"grad_norm": 0.31140980297277515,
"learning_rate": 1.5527950310559007e-05,
"loss": 0.5272,
"step": 100
},
{
"epoch": 0.097856477166822,
"grad_norm": 0.3266343156750877,
"learning_rate": 1.630434782608696e-05,
"loss": 0.5215,
"step": 105
},
{
"epoch": 0.10251630941286113,
"grad_norm": 0.29918673347327257,
"learning_rate": 1.7080745341614907e-05,
"loss": 0.523,
"step": 110
},
{
"epoch": 0.10717614165890028,
"grad_norm": 0.3310333978811669,
"learning_rate": 1.785714285714286e-05,
"loss": 0.5202,
"step": 115
},
{
"epoch": 0.11183597390493942,
"grad_norm": 0.36321738118381847,
"learning_rate": 1.8633540372670807e-05,
"loss": 0.5067,
"step": 120
},
{
"epoch": 0.11649580615097857,
"grad_norm": 0.2797666994370987,
"learning_rate": 1.940993788819876e-05,
"loss": 0.5026,
"step": 125
},
{
"epoch": 0.1211556383970177,
"grad_norm": 0.32298025318616186,
"learning_rate": 2.0186335403726707e-05,
"loss": 0.5084,
"step": 130
},
{
"epoch": 0.12581547064305684,
"grad_norm": 0.4000752827733501,
"learning_rate": 2.096273291925466e-05,
"loss": 0.5069,
"step": 135
},
{
"epoch": 0.13047530288909598,
"grad_norm": 0.4922924496261864,
"learning_rate": 2.173913043478261e-05,
"loss": 0.4919,
"step": 140
},
{
"epoch": 0.13513513513513514,
"grad_norm": 0.3528459265317262,
"learning_rate": 2.2515527950310562e-05,
"loss": 0.5182,
"step": 145
},
{
"epoch": 0.13979496738117428,
"grad_norm": 0.41310907318766044,
"learning_rate": 2.329192546583851e-05,
"loss": 0.4924,
"step": 150
},
{
"epoch": 0.14445479962721341,
"grad_norm": 0.4269460447807956,
"learning_rate": 2.4068322981366462e-05,
"loss": 0.5005,
"step": 155
},
{
"epoch": 0.14911463187325255,
"grad_norm": 0.4188942709163961,
"learning_rate": 2.484472049689441e-05,
"loss": 0.4809,
"step": 160
},
{
"epoch": 0.15377446411929171,
"grad_norm": 0.4224538737371056,
"learning_rate": 2.5621118012422362e-05,
"loss": 0.4922,
"step": 165
},
{
"epoch": 0.15843429636533085,
"grad_norm": 0.36451695943150036,
"learning_rate": 2.639751552795031e-05,
"loss": 0.4682,
"step": 170
},
{
"epoch": 0.16309412861137,
"grad_norm": 0.44880938962766675,
"learning_rate": 2.7173913043478262e-05,
"loss": 0.4812,
"step": 175
},
{
"epoch": 0.16775396085740912,
"grad_norm": 0.399985899809391,
"learning_rate": 2.795031055900621e-05,
"loss": 0.4744,
"step": 180
},
{
"epoch": 0.1724137931034483,
"grad_norm": 0.3710762382054236,
"learning_rate": 2.8726708074534165e-05,
"loss": 0.4773,
"step": 185
},
{
"epoch": 0.17707362534948742,
"grad_norm": 0.5067423386101156,
"learning_rate": 2.9503105590062114e-05,
"loss": 0.4874,
"step": 190
},
{
"epoch": 0.18173345759552656,
"grad_norm": 0.6497058112525416,
"learning_rate": 3.0279503105590062e-05,
"loss": 0.4774,
"step": 195
},
{
"epoch": 0.1863932898415657,
"grad_norm": 0.5885066139321359,
"learning_rate": 3.1055900621118014e-05,
"loss": 0.4796,
"step": 200
},
{
"epoch": 0.19105312208760486,
"grad_norm": 1.0003664794306601,
"learning_rate": 3.183229813664597e-05,
"loss": 0.4687,
"step": 205
},
{
"epoch": 0.195712954333644,
"grad_norm": 0.5038665173622903,
"learning_rate": 3.260869565217392e-05,
"loss": 0.4813,
"step": 210
},
{
"epoch": 0.20037278657968313,
"grad_norm": 0.39066674417290936,
"learning_rate": 3.3385093167701865e-05,
"loss": 0.4632,
"step": 215
},
{
"epoch": 0.20503261882572227,
"grad_norm": 0.5461800355437195,
"learning_rate": 3.4161490683229814e-05,
"loss": 0.4673,
"step": 220
},
{
"epoch": 0.2096924510717614,
"grad_norm": 0.43194458883726633,
"learning_rate": 3.493788819875777e-05,
"loss": 0.4496,
"step": 225
},
{
"epoch": 0.21435228331780057,
"grad_norm": 0.4440833071381681,
"learning_rate": 3.571428571428572e-05,
"loss": 0.4628,
"step": 230
},
{
"epoch": 0.2190121155638397,
"grad_norm": 0.41566446947186825,
"learning_rate": 3.6490683229813665e-05,
"loss": 0.4604,
"step": 235
},
{
"epoch": 0.22367194780987884,
"grad_norm": 0.358310843676591,
"learning_rate": 3.7267080745341614e-05,
"loss": 0.4578,
"step": 240
},
{
"epoch": 0.22833178005591798,
"grad_norm": 0.44225923549559465,
"learning_rate": 3.804347826086957e-05,
"loss": 0.4627,
"step": 245
},
{
"epoch": 0.23299161230195714,
"grad_norm": 0.3992256291237462,
"learning_rate": 3.881987577639752e-05,
"loss": 0.4552,
"step": 250
},
{
"epoch": 0.23765144454799628,
"grad_norm": 0.4694080889691004,
"learning_rate": 3.9596273291925465e-05,
"loss": 0.4635,
"step": 255
},
{
"epoch": 0.2423112767940354,
"grad_norm": 0.42023599721646526,
"learning_rate": 4.0372670807453414e-05,
"loss": 0.4584,
"step": 260
},
{
"epoch": 0.24697110904007455,
"grad_norm": 0.46533332747349704,
"learning_rate": 4.114906832298137e-05,
"loss": 0.4568,
"step": 265
},
{
"epoch": 0.2516309412861137,
"grad_norm": 0.4866315434700336,
"learning_rate": 4.192546583850932e-05,
"loss": 0.4552,
"step": 270
},
{
"epoch": 0.25629077353215285,
"grad_norm": 0.4846502580427653,
"learning_rate": 4.270186335403727e-05,
"loss": 0.4606,
"step": 275
},
{
"epoch": 0.26095060577819196,
"grad_norm": 0.4022476259260207,
"learning_rate": 4.347826086956522e-05,
"loss": 0.4528,
"step": 280
},
{
"epoch": 0.2656104380242311,
"grad_norm": 0.43388622580915626,
"learning_rate": 4.425465838509317e-05,
"loss": 0.4567,
"step": 285
},
{
"epoch": 0.2702702702702703,
"grad_norm": 0.4856594556844892,
"learning_rate": 4.5031055900621124e-05,
"loss": 0.4585,
"step": 290
},
{
"epoch": 0.2749301025163094,
"grad_norm": 0.6213697727877028,
"learning_rate": 4.580745341614907e-05,
"loss": 0.4624,
"step": 295
},
{
"epoch": 0.27958993476234856,
"grad_norm": 0.660785708911612,
"learning_rate": 4.658385093167702e-05,
"loss": 0.4643,
"step": 300
},
{
"epoch": 0.2842497670083877,
"grad_norm": 0.5094248475675451,
"learning_rate": 4.736024844720497e-05,
"loss": 0.453,
"step": 305
},
{
"epoch": 0.28890959925442683,
"grad_norm": 0.5162843441070357,
"learning_rate": 4.8136645962732924e-05,
"loss": 0.4454,
"step": 310
},
{
"epoch": 0.293569431500466,
"grad_norm": 0.5558477904513442,
"learning_rate": 4.891304347826087e-05,
"loss": 0.4533,
"step": 315
},
{
"epoch": 0.2982292637465051,
"grad_norm": 0.6726916910854334,
"learning_rate": 4.968944099378882e-05,
"loss": 0.4388,
"step": 320
},
{
"epoch": 0.30288909599254427,
"grad_norm": 0.5608394756921781,
"learning_rate": 4.994822229892993e-05,
"loss": 0.4368,
"step": 325
},
{
"epoch": 0.30754892823858343,
"grad_norm": 0.5247959113527437,
"learning_rate": 4.986192613047981e-05,
"loss": 0.4468,
"step": 330
},
{
"epoch": 0.31220876048462254,
"grad_norm": 0.7466581909085609,
"learning_rate": 4.977562996202969e-05,
"loss": 0.4473,
"step": 335
},
{
"epoch": 0.3168685927306617,
"grad_norm": 0.40069693827476377,
"learning_rate": 4.968933379357957e-05,
"loss": 0.4488,
"step": 340
},
{
"epoch": 0.32152842497670087,
"grad_norm": 0.4433073418733753,
"learning_rate": 4.9603037625129445e-05,
"loss": 0.4433,
"step": 345
},
{
"epoch": 0.32618825722274,
"grad_norm": 0.4402804602848506,
"learning_rate": 4.951674145667933e-05,
"loss": 0.4566,
"step": 350
},
{
"epoch": 0.33084808946877914,
"grad_norm": 0.5561916690794446,
"learning_rate": 4.94304452882292e-05,
"loss": 0.4431,
"step": 355
},
{
"epoch": 0.33550792171481825,
"grad_norm": 0.5845890551087048,
"learning_rate": 4.934414911977908e-05,
"loss": 0.4405,
"step": 360
},
{
"epoch": 0.3401677539608574,
"grad_norm": 0.3866415610222162,
"learning_rate": 4.9257852951328965e-05,
"loss": 0.437,
"step": 365
},
{
"epoch": 0.3448275862068966,
"grad_norm": 0.4523706035604396,
"learning_rate": 4.917155678287884e-05,
"loss": 0.45,
"step": 370
},
{
"epoch": 0.3494874184529357,
"grad_norm": 0.45303755908012494,
"learning_rate": 4.908526061442872e-05,
"loss": 0.4364,
"step": 375
},
{
"epoch": 0.35414725069897485,
"grad_norm": 0.4759810127104163,
"learning_rate": 4.89989644459786e-05,
"loss": 0.4508,
"step": 380
},
{
"epoch": 0.35880708294501396,
"grad_norm": 0.37302493435804995,
"learning_rate": 4.891266827752848e-05,
"loss": 0.4528,
"step": 385
},
{
"epoch": 0.3634669151910531,
"grad_norm": 0.39553345753859565,
"learning_rate": 4.882637210907836e-05,
"loss": 0.4374,
"step": 390
},
{
"epoch": 0.3681267474370923,
"grad_norm": 0.517029746917124,
"learning_rate": 4.874007594062824e-05,
"loss": 0.4434,
"step": 395
},
{
"epoch": 0.3727865796831314,
"grad_norm": 0.43650791449383186,
"learning_rate": 4.865377977217811e-05,
"loss": 0.4286,
"step": 400
},
{
"epoch": 0.37744641192917056,
"grad_norm": 0.4287102270886215,
"learning_rate": 4.8567483603728e-05,
"loss": 0.4468,
"step": 405
},
{
"epoch": 0.3821062441752097,
"grad_norm": 0.43073044657348636,
"learning_rate": 4.8481187435277875e-05,
"loss": 0.4426,
"step": 410
},
{
"epoch": 0.38676607642124883,
"grad_norm": 0.5387025600608393,
"learning_rate": 4.839489126682776e-05,
"loss": 0.4296,
"step": 415
},
{
"epoch": 0.391425908667288,
"grad_norm": 0.5153252513836688,
"learning_rate": 4.830859509837763e-05,
"loss": 0.4475,
"step": 420
},
{
"epoch": 0.3960857409133271,
"grad_norm": 0.3664353088914084,
"learning_rate": 4.822229892992751e-05,
"loss": 0.445,
"step": 425
},
{
"epoch": 0.40074557315936626,
"grad_norm": 0.4646069378801538,
"learning_rate": 4.8136002761477395e-05,
"loss": 0.4321,
"step": 430
},
{
"epoch": 0.40540540540540543,
"grad_norm": 0.5337046566267092,
"learning_rate": 4.804970659302727e-05,
"loss": 0.4452,
"step": 435
},
{
"epoch": 0.41006523765144454,
"grad_norm": 0.5277173274418405,
"learning_rate": 4.796341042457715e-05,
"loss": 0.4393,
"step": 440
},
{
"epoch": 0.4147250698974837,
"grad_norm": 0.5328906770207152,
"learning_rate": 4.787711425612703e-05,
"loss": 0.4339,
"step": 445
},
{
"epoch": 0.4193849021435228,
"grad_norm": 0.4906064823628884,
"learning_rate": 4.779081808767691e-05,
"loss": 0.4246,
"step": 450
},
{
"epoch": 0.424044734389562,
"grad_norm": 0.4871704896808195,
"learning_rate": 4.770452191922679e-05,
"loss": 0.4316,
"step": 455
},
{
"epoch": 0.42870456663560114,
"grad_norm": 0.3872613851854932,
"learning_rate": 4.761822575077667e-05,
"loss": 0.4283,
"step": 460
},
{
"epoch": 0.43336439888164024,
"grad_norm": 0.48560447593847655,
"learning_rate": 4.753192958232654e-05,
"loss": 0.4255,
"step": 465
},
{
"epoch": 0.4380242311276794,
"grad_norm": 0.43287841878863453,
"learning_rate": 4.744563341387643e-05,
"loss": 0.4377,
"step": 470
},
{
"epoch": 0.4426840633737186,
"grad_norm": 0.46424167359241597,
"learning_rate": 4.7359337245426306e-05,
"loss": 0.4277,
"step": 475
},
{
"epoch": 0.4473438956197577,
"grad_norm": 0.39322918496511977,
"learning_rate": 4.7273041076976184e-05,
"loss": 0.4393,
"step": 480
},
{
"epoch": 0.45200372786579684,
"grad_norm": 0.38542293023219015,
"learning_rate": 4.718674490852606e-05,
"loss": 0.432,
"step": 485
},
{
"epoch": 0.45666356011183595,
"grad_norm": 0.41246395275559605,
"learning_rate": 4.710044874007594e-05,
"loss": 0.4263,
"step": 490
},
{
"epoch": 0.4613233923578751,
"grad_norm": 0.3975264680297866,
"learning_rate": 4.7014152571625826e-05,
"loss": 0.4323,
"step": 495
},
{
"epoch": 0.4659832246039143,
"grad_norm": 0.45318918462749547,
"learning_rate": 4.6927856403175704e-05,
"loss": 0.4348,
"step": 500
},
{
"epoch": 0.4706430568499534,
"grad_norm": 0.36573583820356864,
"learning_rate": 4.684156023472558e-05,
"loss": 0.4219,
"step": 505
},
{
"epoch": 0.47530288909599255,
"grad_norm": 0.44294897087649004,
"learning_rate": 4.675526406627546e-05,
"loss": 0.4301,
"step": 510
},
{
"epoch": 0.47996272134203166,
"grad_norm": 0.6261618649064988,
"learning_rate": 4.666896789782534e-05,
"loss": 0.4281,
"step": 515
},
{
"epoch": 0.4846225535880708,
"grad_norm": 0.5618813957954623,
"learning_rate": 4.658267172937522e-05,
"loss": 0.432,
"step": 520
},
{
"epoch": 0.48928238583411,
"grad_norm": 0.4221390092567116,
"learning_rate": 4.64963755609251e-05,
"loss": 0.4309,
"step": 525
},
{
"epoch": 0.4939422180801491,
"grad_norm": 0.5054553118583148,
"learning_rate": 4.641007939247497e-05,
"loss": 0.428,
"step": 530
},
{
"epoch": 0.49860205032618826,
"grad_norm": 0.3674283025881192,
"learning_rate": 4.632378322402486e-05,
"loss": 0.4282,
"step": 535
},
{
"epoch": 0.5032618825722274,
"grad_norm": 0.37646280701753526,
"learning_rate": 4.6237487055574736e-05,
"loss": 0.4176,
"step": 540
},
{
"epoch": 0.5079217148182665,
"grad_norm": 0.4667754579895071,
"learning_rate": 4.6151190887124615e-05,
"loss": 0.4243,
"step": 545
},
{
"epoch": 0.5125815470643057,
"grad_norm": 0.43689496394898064,
"learning_rate": 4.606489471867449e-05,
"loss": 0.4245,
"step": 550
},
{
"epoch": 0.5172413793103449,
"grad_norm": 0.35100380853775026,
"learning_rate": 4.597859855022437e-05,
"loss": 0.4192,
"step": 555
},
{
"epoch": 0.5219012115563839,
"grad_norm": 0.4310432615306107,
"learning_rate": 4.589230238177425e-05,
"loss": 0.4395,
"step": 560
},
{
"epoch": 0.5265610438024231,
"grad_norm": 0.4828668909643852,
"learning_rate": 4.5806006213324134e-05,
"loss": 0.4174,
"step": 565
},
{
"epoch": 0.5312208760484622,
"grad_norm": 0.5359345219410544,
"learning_rate": 4.5719710044874006e-05,
"loss": 0.4172,
"step": 570
},
{
"epoch": 0.5358807082945014,
"grad_norm": 0.44233747611422647,
"learning_rate": 4.563341387642389e-05,
"loss": 0.423,
"step": 575
},
{
"epoch": 0.5405405405405406,
"grad_norm": 0.35787489087133517,
"learning_rate": 4.554711770797377e-05,
"loss": 0.4235,
"step": 580
},
{
"epoch": 0.5452003727865797,
"grad_norm": 0.4480426020707225,
"learning_rate": 4.546082153952365e-05,
"loss": 0.4206,
"step": 585
},
{
"epoch": 0.5498602050326188,
"grad_norm": 0.40460727804169483,
"learning_rate": 4.5374525371073526e-05,
"loss": 0.4168,
"step": 590
},
{
"epoch": 0.554520037278658,
"grad_norm": 0.37469384681867557,
"learning_rate": 4.5288229202623404e-05,
"loss": 0.4188,
"step": 595
},
{
"epoch": 0.5591798695246971,
"grad_norm": 0.38124595301547787,
"learning_rate": 4.520193303417328e-05,
"loss": 0.4192,
"step": 600
},
{
"epoch": 0.5638397017707363,
"grad_norm": 0.3602652491312147,
"learning_rate": 4.511563686572317e-05,
"loss": 0.4179,
"step": 605
},
{
"epoch": 0.5684995340167754,
"grad_norm": 0.39572056728413657,
"learning_rate": 4.5029340697273045e-05,
"loss": 0.4287,
"step": 610
},
{
"epoch": 0.5731593662628145,
"grad_norm": 0.33403742184402074,
"learning_rate": 4.4943044528822923e-05,
"loss": 0.4155,
"step": 615
},
{
"epoch": 0.5778191985088537,
"grad_norm": 0.40306442434828443,
"learning_rate": 4.48567483603728e-05,
"loss": 0.4125,
"step": 620
},
{
"epoch": 0.5824790307548928,
"grad_norm": 0.3774250530609767,
"learning_rate": 4.477045219192268e-05,
"loss": 0.4341,
"step": 625
},
{
"epoch": 0.587138863000932,
"grad_norm": 0.39959005575592826,
"learning_rate": 4.4684156023472565e-05,
"loss": 0.4245,
"step": 630
},
{
"epoch": 0.5917986952469712,
"grad_norm": 0.47668132423182985,
"learning_rate": 4.4597859855022436e-05,
"loss": 0.4293,
"step": 635
},
{
"epoch": 0.5964585274930102,
"grad_norm": 0.5031461055880807,
"learning_rate": 4.4511563686572315e-05,
"loss": 0.42,
"step": 640
},
{
"epoch": 0.6011183597390494,
"grad_norm": 0.41239093059640153,
"learning_rate": 4.44252675181222e-05,
"loss": 0.4196,
"step": 645
},
{
"epoch": 0.6057781919850885,
"grad_norm": 0.40854847004740863,
"learning_rate": 4.433897134967208e-05,
"loss": 0.428,
"step": 650
},
{
"epoch": 0.6104380242311277,
"grad_norm": 0.41242124461222524,
"learning_rate": 4.4252675181221956e-05,
"loss": 0.426,
"step": 655
},
{
"epoch": 0.6150978564771669,
"grad_norm": 0.3573616803868084,
"learning_rate": 4.4166379012771834e-05,
"loss": 0.4078,
"step": 660
},
{
"epoch": 0.6197576887232059,
"grad_norm": 0.4985584033933677,
"learning_rate": 4.408008284432171e-05,
"loss": 0.4089,
"step": 665
},
{
"epoch": 0.6244175209692451,
"grad_norm": 0.41754039223029704,
"learning_rate": 4.39937866758716e-05,
"loss": 0.4075,
"step": 670
},
{
"epoch": 0.6290773532152842,
"grad_norm": 0.42762641741838325,
"learning_rate": 4.3907490507421476e-05,
"loss": 0.4157,
"step": 675
},
{
"epoch": 0.6337371854613234,
"grad_norm": 0.4381716045704423,
"learning_rate": 4.382119433897135e-05,
"loss": 0.4188,
"step": 680
},
{
"epoch": 0.6383970177073626,
"grad_norm": 0.42615839208636525,
"learning_rate": 4.373489817052123e-05,
"loss": 0.4283,
"step": 685
},
{
"epoch": 0.6430568499534017,
"grad_norm": 0.41547768813271807,
"learning_rate": 4.364860200207111e-05,
"loss": 0.4107,
"step": 690
},
{
"epoch": 0.6477166821994408,
"grad_norm": 0.4257514042629888,
"learning_rate": 4.356230583362099e-05,
"loss": 0.4152,
"step": 695
},
{
"epoch": 0.65237651444548,
"grad_norm": 0.3718564529022634,
"learning_rate": 4.347600966517087e-05,
"loss": 0.4245,
"step": 700
},
{
"epoch": 0.6570363466915191,
"grad_norm": 0.4097816553154767,
"learning_rate": 4.3389713496720745e-05,
"loss": 0.4178,
"step": 705
},
{
"epoch": 0.6616961789375583,
"grad_norm": 0.5185129505497016,
"learning_rate": 4.330341732827063e-05,
"loss": 0.4184,
"step": 710
},
{
"epoch": 0.6663560111835974,
"grad_norm": 0.38356789237987543,
"learning_rate": 4.321712115982051e-05,
"loss": 0.4066,
"step": 715
},
{
"epoch": 0.6710158434296365,
"grad_norm": 0.39820661867372364,
"learning_rate": 4.3130824991370387e-05,
"loss": 0.4169,
"step": 720
},
{
"epoch": 0.6756756756756757,
"grad_norm": 0.3468550490700257,
"learning_rate": 4.3044528822920265e-05,
"loss": 0.4159,
"step": 725
},
{
"epoch": 0.6803355079217148,
"grad_norm": 0.42714803174042604,
"learning_rate": 4.295823265447014e-05,
"loss": 0.4107,
"step": 730
},
{
"epoch": 0.684995340167754,
"grad_norm": 0.35975622295120474,
"learning_rate": 4.287193648602002e-05,
"loss": 0.4148,
"step": 735
},
{
"epoch": 0.6896551724137931,
"grad_norm": 0.3398454105039195,
"learning_rate": 4.27856403175699e-05,
"loss": 0.4085,
"step": 740
},
{
"epoch": 0.6943150046598322,
"grad_norm": 0.402262044573312,
"learning_rate": 4.269934414911978e-05,
"loss": 0.4044,
"step": 745
},
{
"epoch": 0.6989748369058714,
"grad_norm": 0.42023295565874197,
"learning_rate": 4.261304798066966e-05,
"loss": 0.409,
"step": 750
},
{
"epoch": 0.7036346691519105,
"grad_norm": 0.3886074714281147,
"learning_rate": 4.252675181221954e-05,
"loss": 0.4113,
"step": 755
},
{
"epoch": 0.7082945013979497,
"grad_norm": 0.42499664467971804,
"learning_rate": 4.244045564376942e-05,
"loss": 0.4146,
"step": 760
},
{
"epoch": 0.7129543336439889,
"grad_norm": 0.3605554838945289,
"learning_rate": 4.23541594753193e-05,
"loss": 0.399,
"step": 765
},
{
"epoch": 0.7176141658900279,
"grad_norm": 0.35331727829864007,
"learning_rate": 4.2267863306869176e-05,
"loss": 0.4076,
"step": 770
},
{
"epoch": 0.7222739981360671,
"grad_norm": 0.4047141050647308,
"learning_rate": 4.2181567138419054e-05,
"loss": 0.4203,
"step": 775
},
{
"epoch": 0.7269338303821062,
"grad_norm": 0.370214336816459,
"learning_rate": 4.209527096996894e-05,
"loss": 0.4047,
"step": 780
},
{
"epoch": 0.7315936626281454,
"grad_norm": 0.39480883644980663,
"learning_rate": 4.200897480151881e-05,
"loss": 0.4121,
"step": 785
},
{
"epoch": 0.7362534948741846,
"grad_norm": 0.5146939285368787,
"learning_rate": 4.1922678633068695e-05,
"loss": 0.4151,
"step": 790
},
{
"epoch": 0.7409133271202236,
"grad_norm": 0.43348578379113595,
"learning_rate": 4.1836382464618573e-05,
"loss": 0.4096,
"step": 795
},
{
"epoch": 0.7455731593662628,
"grad_norm": 0.3922926303913187,
"learning_rate": 4.175008629616845e-05,
"loss": 0.4078,
"step": 800
},
{
"epoch": 0.750232991612302,
"grad_norm": 0.3640607241449352,
"learning_rate": 4.166379012771833e-05,
"loss": 0.4078,
"step": 805
},
{
"epoch": 0.7548928238583411,
"grad_norm": 0.349690850963141,
"learning_rate": 4.157749395926821e-05,
"loss": 0.4256,
"step": 810
},
{
"epoch": 0.7595526561043803,
"grad_norm": 0.36285524732493346,
"learning_rate": 4.1491197790818086e-05,
"loss": 0.4004,
"step": 815
},
{
"epoch": 0.7642124883504194,
"grad_norm": 0.5210490846235861,
"learning_rate": 4.140490162236797e-05,
"loss": 0.4151,
"step": 820
},
{
"epoch": 0.7688723205964585,
"grad_norm": 0.41942137907605526,
"learning_rate": 4.131860545391785e-05,
"loss": 0.4077,
"step": 825
},
{
"epoch": 0.7735321528424977,
"grad_norm": 0.3901545056499728,
"learning_rate": 4.123230928546773e-05,
"loss": 0.4024,
"step": 830
},
{
"epoch": 0.7781919850885368,
"grad_norm": 0.4038340272785345,
"learning_rate": 4.1146013117017606e-05,
"loss": 0.4095,
"step": 835
},
{
"epoch": 0.782851817334576,
"grad_norm": 0.4532227365804893,
"learning_rate": 4.1059716948567484e-05,
"loss": 0.4127,
"step": 840
},
{
"epoch": 0.7875116495806151,
"grad_norm": 0.4770900996270386,
"learning_rate": 4.097342078011737e-05,
"loss": 0.4204,
"step": 845
},
{
"epoch": 0.7921714818266542,
"grad_norm": 0.37619746602800846,
"learning_rate": 4.088712461166724e-05,
"loss": 0.4105,
"step": 850
},
{
"epoch": 0.7968313140726934,
"grad_norm": 0.5274698191091436,
"learning_rate": 4.080082844321712e-05,
"loss": 0.4121,
"step": 855
},
{
"epoch": 0.8014911463187325,
"grad_norm": 0.41457417749149855,
"learning_rate": 4.0714532274767004e-05,
"loss": 0.419,
"step": 860
},
{
"epoch": 0.8061509785647717,
"grad_norm": 0.3476134180441984,
"learning_rate": 4.062823610631688e-05,
"loss": 0.4001,
"step": 865
},
{
"epoch": 0.8108108108108109,
"grad_norm": 0.3018646760101771,
"learning_rate": 4.054193993786676e-05,
"loss": 0.4144,
"step": 870
},
{
"epoch": 0.8154706430568499,
"grad_norm": 0.3505985836303181,
"learning_rate": 4.045564376941664e-05,
"loss": 0.4099,
"step": 875
},
{
"epoch": 0.8201304753028891,
"grad_norm": 0.37454036950977276,
"learning_rate": 4.036934760096652e-05,
"loss": 0.4015,
"step": 880
},
{
"epoch": 0.8247903075489282,
"grad_norm": 0.412814499179022,
"learning_rate": 4.02830514325164e-05,
"loss": 0.4152,
"step": 885
},
{
"epoch": 0.8294501397949674,
"grad_norm": 0.43525980740205655,
"learning_rate": 4.019675526406628e-05,
"loss": 0.402,
"step": 890
},
{
"epoch": 0.8341099720410066,
"grad_norm": 0.3504008123184863,
"learning_rate": 4.011045909561615e-05,
"loss": 0.4211,
"step": 895
},
{
"epoch": 0.8387698042870456,
"grad_norm": 0.3451666606741711,
"learning_rate": 4.0024162927166037e-05,
"loss": 0.4042,
"step": 900
},
{
"epoch": 0.8434296365330848,
"grad_norm": 0.3412474402734993,
"learning_rate": 3.9937866758715915e-05,
"loss": 0.4057,
"step": 905
},
{
"epoch": 0.848089468779124,
"grad_norm": 0.47725878975243735,
"learning_rate": 3.98515705902658e-05,
"loss": 0.4116,
"step": 910
},
{
"epoch": 0.8527493010251631,
"grad_norm": 0.6035020610905382,
"learning_rate": 3.976527442181567e-05,
"loss": 0.4046,
"step": 915
},
{
"epoch": 0.8574091332712023,
"grad_norm": 0.45947885014374223,
"learning_rate": 3.967897825336555e-05,
"loss": 0.4109,
"step": 920
},
{
"epoch": 0.8620689655172413,
"grad_norm": 0.3346849536418383,
"learning_rate": 3.9592682084915434e-05,
"loss": 0.3976,
"step": 925
},
{
"epoch": 0.8667287977632805,
"grad_norm": 0.47015004763550144,
"learning_rate": 3.950638591646531e-05,
"loss": 0.3977,
"step": 930
},
{
"epoch": 0.8713886300093197,
"grad_norm": 0.3751739393328928,
"learning_rate": 3.942008974801519e-05,
"loss": 0.4088,
"step": 935
},
{
"epoch": 0.8760484622553588,
"grad_norm": 0.31035212858665634,
"learning_rate": 3.933379357956507e-05,
"loss": 0.4049,
"step": 940
},
{
"epoch": 0.880708294501398,
"grad_norm": 0.37862709572624254,
"learning_rate": 3.924749741111495e-05,
"loss": 0.407,
"step": 945
},
{
"epoch": 0.8853681267474371,
"grad_norm": 0.4375453312898591,
"learning_rate": 3.916120124266483e-05,
"loss": 0.4146,
"step": 950
},
{
"epoch": 0.8900279589934762,
"grad_norm": 0.41556861405255147,
"learning_rate": 3.9074905074214704e-05,
"loss": 0.4135,
"step": 955
},
{
"epoch": 0.8946877912395154,
"grad_norm": 0.34721066251213883,
"learning_rate": 3.898860890576458e-05,
"loss": 0.4091,
"step": 960
},
{
"epoch": 0.8993476234855545,
"grad_norm": 0.4189318356066771,
"learning_rate": 3.890231273731447e-05,
"loss": 0.4007,
"step": 965
},
{
"epoch": 0.9040074557315937,
"grad_norm": 0.3129065312499714,
"learning_rate": 3.8816016568864345e-05,
"loss": 0.3985,
"step": 970
},
{
"epoch": 0.9086672879776329,
"grad_norm": 0.339604217707666,
"learning_rate": 3.8729720400414224e-05,
"loss": 0.4169,
"step": 975
},
{
"epoch": 0.9133271202236719,
"grad_norm": 0.425269526417587,
"learning_rate": 3.86434242319641e-05,
"loss": 0.404,
"step": 980
},
{
"epoch": 0.9179869524697111,
"grad_norm": 0.45528213892828173,
"learning_rate": 3.855712806351398e-05,
"loss": 0.4058,
"step": 985
},
{
"epoch": 0.9226467847157502,
"grad_norm": 0.43623028648882733,
"learning_rate": 3.8470831895063865e-05,
"loss": 0.4032,
"step": 990
},
{
"epoch": 0.9273066169617894,
"grad_norm": 0.3755484829385926,
"learning_rate": 3.838453572661374e-05,
"loss": 0.399,
"step": 995
},
{
"epoch": 0.9319664492078286,
"grad_norm": 0.4208532884531691,
"learning_rate": 3.8298239558163615e-05,
"loss": 0.402,
"step": 1000
},
{
"epoch": 0.9366262814538676,
"grad_norm": 0.4006384178165892,
"learning_rate": 3.82119433897135e-05,
"loss": 0.4072,
"step": 1005
},
{
"epoch": 0.9412861136999068,
"grad_norm": 0.3959495097559784,
"learning_rate": 3.812564722126338e-05,
"loss": 0.3996,
"step": 1010
},
{
"epoch": 0.9459459459459459,
"grad_norm": 0.412986020520787,
"learning_rate": 3.8039351052813256e-05,
"loss": 0.408,
"step": 1015
},
{
"epoch": 0.9506057781919851,
"grad_norm": 0.39141710105069416,
"learning_rate": 3.7953054884363134e-05,
"loss": 0.4019,
"step": 1020
},
{
"epoch": 0.9552656104380243,
"grad_norm": 0.3681512788462775,
"learning_rate": 3.786675871591301e-05,
"loss": 0.4053,
"step": 1025
},
{
"epoch": 0.9599254426840633,
"grad_norm": 0.2849981991126552,
"learning_rate": 3.77804625474629e-05,
"loss": 0.4103,
"step": 1030
},
{
"epoch": 0.9645852749301025,
"grad_norm": 0.4064379280880787,
"learning_rate": 3.7694166379012776e-05,
"loss": 0.4027,
"step": 1035
},
{
"epoch": 0.9692451071761417,
"grad_norm": 0.38383331397218684,
"learning_rate": 3.7607870210562654e-05,
"loss": 0.4002,
"step": 1040
},
{
"epoch": 0.9739049394221808,
"grad_norm": 0.27852308304758455,
"learning_rate": 3.752157404211253e-05,
"loss": 0.3932,
"step": 1045
},
{
"epoch": 0.97856477166822,
"grad_norm": 0.38819077991615814,
"learning_rate": 3.743527787366241e-05,
"loss": 0.4221,
"step": 1050
},
{
"epoch": 0.983224603914259,
"grad_norm": 0.3660717452223031,
"learning_rate": 3.734898170521229e-05,
"loss": 0.4002,
"step": 1055
},
{
"epoch": 0.9878844361602982,
"grad_norm": 0.3950869687400394,
"learning_rate": 3.7262685536762174e-05,
"loss": 0.3992,
"step": 1060
},
{
"epoch": 0.9925442684063374,
"grad_norm": 0.37399341432624805,
"learning_rate": 3.7176389368312045e-05,
"loss": 0.4035,
"step": 1065
},
{
"epoch": 0.9972041006523765,
"grad_norm": 0.4579382708170151,
"learning_rate": 3.709009319986193e-05,
"loss": 0.4037,
"step": 1070
},
{
"epoch": 1.0018639328984156,
"grad_norm": 0.3901679201113807,
"learning_rate": 3.700379703141181e-05,
"loss": 0.3687,
"step": 1075
},
{
"epoch": 1.0065237651444547,
"grad_norm": 0.29628791264121573,
"learning_rate": 3.6917500862961687e-05,
"loss": 0.3443,
"step": 1080
},
{
"epoch": 1.011183597390494,
"grad_norm": 0.41463234069815796,
"learning_rate": 3.6831204694511565e-05,
"loss": 0.3514,
"step": 1085
},
{
"epoch": 1.015843429636533,
"grad_norm": 0.3643235278539243,
"learning_rate": 3.674490852606144e-05,
"loss": 0.3581,
"step": 1090
},
{
"epoch": 1.0205032618825722,
"grad_norm": 0.329351162941022,
"learning_rate": 3.665861235761132e-05,
"loss": 0.3441,
"step": 1095
},
{
"epoch": 1.0251630941286114,
"grad_norm": 0.360571953515733,
"learning_rate": 3.6572316189161206e-05,
"loss": 0.3358,
"step": 1100
},
{
"epoch": 1.0298229263746506,
"grad_norm": 0.2863747162313705,
"learning_rate": 3.6486020020711085e-05,
"loss": 0.3418,
"step": 1105
},
{
"epoch": 1.0344827586206897,
"grad_norm": 0.31625372426255294,
"learning_rate": 3.639972385226096e-05,
"loss": 0.3422,
"step": 1110
},
{
"epoch": 1.0391425908667289,
"grad_norm": 0.3120426481511544,
"learning_rate": 3.631342768381084e-05,
"loss": 0.3423,
"step": 1115
},
{
"epoch": 1.0438024231127678,
"grad_norm": 0.36483136953347134,
"learning_rate": 3.622713151536072e-05,
"loss": 0.3425,
"step": 1120
},
{
"epoch": 1.048462255358807,
"grad_norm": 0.28619085512198084,
"learning_rate": 3.6140835346910604e-05,
"loss": 0.346,
"step": 1125
},
{
"epoch": 1.0531220876048462,
"grad_norm": 0.29712001182062725,
"learning_rate": 3.6054539178460476e-05,
"loss": 0.3421,
"step": 1130
},
{
"epoch": 1.0577819198508853,
"grad_norm": 0.3552761659836407,
"learning_rate": 3.5968243010010354e-05,
"loss": 0.3588,
"step": 1135
},
{
"epoch": 1.0624417520969245,
"grad_norm": 0.3461417780753448,
"learning_rate": 3.588194684156024e-05,
"loss": 0.3495,
"step": 1140
},
{
"epoch": 1.0671015843429636,
"grad_norm": 0.39060265095293556,
"learning_rate": 3.579565067311012e-05,
"loss": 0.341,
"step": 1145
},
{
"epoch": 1.0717614165890028,
"grad_norm": 0.3674125535327584,
"learning_rate": 3.5709354504659995e-05,
"loss": 0.3377,
"step": 1150
},
{
"epoch": 1.076421248835042,
"grad_norm": 0.3721009981706529,
"learning_rate": 3.5623058336209874e-05,
"loss": 0.3486,
"step": 1155
},
{
"epoch": 1.0810810810810811,
"grad_norm": 0.3026290235168691,
"learning_rate": 3.553676216775975e-05,
"loss": 0.3399,
"step": 1160
},
{
"epoch": 1.0857409133271203,
"grad_norm": 0.274433360166586,
"learning_rate": 3.545046599930964e-05,
"loss": 0.3518,
"step": 1165
},
{
"epoch": 1.0904007455731595,
"grad_norm": 0.2799065216591901,
"learning_rate": 3.536416983085951e-05,
"loss": 0.3532,
"step": 1170
},
{
"epoch": 1.0950605778191984,
"grad_norm": 0.3733253601455716,
"learning_rate": 3.5277873662409386e-05,
"loss": 0.3399,
"step": 1175
},
{
"epoch": 1.0997204100652376,
"grad_norm": 0.342876802963155,
"learning_rate": 3.519157749395927e-05,
"loss": 0.3459,
"step": 1180
},
{
"epoch": 1.1043802423112767,
"grad_norm": 0.4232909286754048,
"learning_rate": 3.510528132550915e-05,
"loss": 0.343,
"step": 1185
},
{
"epoch": 1.109040074557316,
"grad_norm": 0.30970899413001823,
"learning_rate": 3.501898515705903e-05,
"loss": 0.3448,
"step": 1190
},
{
"epoch": 1.113699906803355,
"grad_norm": 0.32237922787903955,
"learning_rate": 3.4932688988608906e-05,
"loss": 0.3564,
"step": 1195
},
{
"epoch": 1.1183597390493942,
"grad_norm": 0.3591063615488881,
"learning_rate": 3.4846392820158784e-05,
"loss": 0.3448,
"step": 1200
},
{
"epoch": 1.1230195712954334,
"grad_norm": 0.3678506413492932,
"learning_rate": 3.476009665170867e-05,
"loss": 0.3451,
"step": 1205
},
{
"epoch": 1.1276794035414726,
"grad_norm": 0.33625373443101364,
"learning_rate": 3.467380048325855e-05,
"loss": 0.3438,
"step": 1210
},
{
"epoch": 1.1323392357875117,
"grad_norm": 0.26997982769604256,
"learning_rate": 3.458750431480842e-05,
"loss": 0.3501,
"step": 1215
},
{
"epoch": 1.1369990680335509,
"grad_norm": 0.3478270414118411,
"learning_rate": 3.4501208146358304e-05,
"loss": 0.3492,
"step": 1220
},
{
"epoch": 1.14165890027959,
"grad_norm": 0.3126170234508818,
"learning_rate": 3.441491197790818e-05,
"loss": 0.3455,
"step": 1225
},
{
"epoch": 1.146318732525629,
"grad_norm": 0.28448609515784845,
"learning_rate": 3.432861580945806e-05,
"loss": 0.3466,
"step": 1230
},
{
"epoch": 1.1509785647716682,
"grad_norm": 0.32725952329088837,
"learning_rate": 3.424231964100794e-05,
"loss": 0.3455,
"step": 1235
},
{
"epoch": 1.1556383970177073,
"grad_norm": 0.37143166952563245,
"learning_rate": 3.415602347255782e-05,
"loss": 0.3441,
"step": 1240
},
{
"epoch": 1.1602982292637465,
"grad_norm": 0.2766260428806357,
"learning_rate": 3.40697273041077e-05,
"loss": 0.3513,
"step": 1245
},
{
"epoch": 1.1649580615097856,
"grad_norm": 0.36190414573156077,
"learning_rate": 3.398343113565758e-05,
"loss": 0.3452,
"step": 1250
},
{
"epoch": 1.1696178937558248,
"grad_norm": 0.26113533170036535,
"learning_rate": 3.389713496720746e-05,
"loss": 0.3247,
"step": 1255
},
{
"epoch": 1.174277726001864,
"grad_norm": 0.28989566457895166,
"learning_rate": 3.381083879875734e-05,
"loss": 0.3444,
"step": 1260
},
{
"epoch": 1.1789375582479031,
"grad_norm": 0.30414274182010975,
"learning_rate": 3.3724542630307215e-05,
"loss": 0.3406,
"step": 1265
},
{
"epoch": 1.1835973904939423,
"grad_norm": 0.2974836028638616,
"learning_rate": 3.363824646185709e-05,
"loss": 0.3495,
"step": 1270
},
{
"epoch": 1.1882572227399812,
"grad_norm": 0.2712492681195211,
"learning_rate": 3.355195029340698e-05,
"loss": 0.3438,
"step": 1275
},
{
"epoch": 1.1929170549860204,
"grad_norm": 0.3282049984515077,
"learning_rate": 3.346565412495685e-05,
"loss": 0.354,
"step": 1280
},
{
"epoch": 1.1975768872320596,
"grad_norm": 0.32284102681932625,
"learning_rate": 3.3379357956506735e-05,
"loss": 0.355,
"step": 1285
},
{
"epoch": 1.2022367194780987,
"grad_norm": 0.3259254864078388,
"learning_rate": 3.329306178805661e-05,
"loss": 0.3386,
"step": 1290
},
{
"epoch": 1.206896551724138,
"grad_norm": 0.29283997653934446,
"learning_rate": 3.320676561960649e-05,
"loss": 0.3434,
"step": 1295
},
{
"epoch": 1.211556383970177,
"grad_norm": 0.42858355557664335,
"learning_rate": 3.312046945115637e-05,
"loss": 0.3516,
"step": 1300
},
{
"epoch": 1.2162162162162162,
"grad_norm": 0.4080026216044187,
"learning_rate": 3.303417328270625e-05,
"loss": 0.3504,
"step": 1305
},
{
"epoch": 1.2208760484622554,
"grad_norm": 0.3271459642232697,
"learning_rate": 3.2947877114256126e-05,
"loss": 0.3398,
"step": 1310
},
{
"epoch": 1.2255358807082946,
"grad_norm": 0.3425190591556552,
"learning_rate": 3.286158094580601e-05,
"loss": 0.3597,
"step": 1315
},
{
"epoch": 1.2301957129543337,
"grad_norm": 0.3413622204212518,
"learning_rate": 3.277528477735589e-05,
"loss": 0.3455,
"step": 1320
},
{
"epoch": 1.2348555452003729,
"grad_norm": 0.3084612482961901,
"learning_rate": 3.268898860890577e-05,
"loss": 0.3521,
"step": 1325
},
{
"epoch": 1.2395153774464118,
"grad_norm": 0.29664430467349057,
"learning_rate": 3.2602692440455645e-05,
"loss": 0.3405,
"step": 1330
},
{
"epoch": 1.244175209692451,
"grad_norm": 0.27270621992905014,
"learning_rate": 3.2516396272005524e-05,
"loss": 0.3431,
"step": 1335
},
{
"epoch": 1.2488350419384902,
"grad_norm": 0.4410507590623693,
"learning_rate": 3.243010010355541e-05,
"loss": 0.3488,
"step": 1340
},
{
"epoch": 1.2534948741845293,
"grad_norm": 0.3339321094554055,
"learning_rate": 3.234380393510528e-05,
"loss": 0.3363,
"step": 1345
},
{
"epoch": 1.2581547064305685,
"grad_norm": 0.2556993234710388,
"learning_rate": 3.225750776665516e-05,
"loss": 0.3434,
"step": 1350
},
{
"epoch": 1.2628145386766076,
"grad_norm": 0.31690401443797217,
"learning_rate": 3.217121159820504e-05,
"loss": 0.3529,
"step": 1355
},
{
"epoch": 1.2674743709226468,
"grad_norm": 0.32059679347647235,
"learning_rate": 3.208491542975492e-05,
"loss": 0.343,
"step": 1360
},
{
"epoch": 1.272134203168686,
"grad_norm": 0.32959262578712795,
"learning_rate": 3.19986192613048e-05,
"loss": 0.3482,
"step": 1365
},
{
"epoch": 1.2767940354147251,
"grad_norm": 0.3350700905653285,
"learning_rate": 3.191232309285468e-05,
"loss": 0.3467,
"step": 1370
},
{
"epoch": 1.281453867660764,
"grad_norm": 0.32565422741384376,
"learning_rate": 3.1826026924404556e-05,
"loss": 0.337,
"step": 1375
},
{
"epoch": 1.2861136999068035,
"grad_norm": 0.31874076945811086,
"learning_rate": 3.173973075595444e-05,
"loss": 0.3515,
"step": 1380
},
{
"epoch": 1.2907735321528424,
"grad_norm": 0.3412076460955754,
"learning_rate": 3.165343458750431e-05,
"loss": 0.3508,
"step": 1385
},
{
"epoch": 1.2954333643988816,
"grad_norm": 0.38414335906486796,
"learning_rate": 3.156713841905419e-05,
"loss": 0.3435,
"step": 1390
},
{
"epoch": 1.3000931966449207,
"grad_norm": 0.2810335445001914,
"learning_rate": 3.1480842250604076e-05,
"loss": 0.3477,
"step": 1395
},
{
"epoch": 1.30475302889096,
"grad_norm": 0.3542281966835823,
"learning_rate": 3.1394546082153954e-05,
"loss": 0.3415,
"step": 1400
},
{
"epoch": 1.309412861136999,
"grad_norm": 0.3121166663544754,
"learning_rate": 3.130824991370383e-05,
"loss": 0.3524,
"step": 1405
},
{
"epoch": 1.3140726933830382,
"grad_norm": 0.3090627381030505,
"learning_rate": 3.122195374525371e-05,
"loss": 0.3407,
"step": 1410
},
{
"epoch": 1.3187325256290774,
"grad_norm": 0.29972052143775835,
"learning_rate": 3.113565757680359e-05,
"loss": 0.3481,
"step": 1415
},
{
"epoch": 1.3233923578751166,
"grad_norm": 0.33138264909466075,
"learning_rate": 3.1049361408353474e-05,
"loss": 0.3355,
"step": 1420
},
{
"epoch": 1.3280521901211557,
"grad_norm": 0.26809112074287433,
"learning_rate": 3.096306523990335e-05,
"loss": 0.3436,
"step": 1425
},
{
"epoch": 1.3327120223671947,
"grad_norm": 0.2791181212346365,
"learning_rate": 3.0876769071453223e-05,
"loss": 0.3518,
"step": 1430
},
{
"epoch": 1.337371854613234,
"grad_norm": 0.31060675564494544,
"learning_rate": 3.079047290300311e-05,
"loss": 0.3499,
"step": 1435
},
{
"epoch": 1.342031686859273,
"grad_norm": 0.3122211444665765,
"learning_rate": 3.070417673455299e-05,
"loss": 0.3483,
"step": 1440
},
{
"epoch": 1.3466915191053122,
"grad_norm": 0.27439186875203364,
"learning_rate": 3.061788056610287e-05,
"loss": 0.3529,
"step": 1445
},
{
"epoch": 1.3513513513513513,
"grad_norm": 0.3253426338270861,
"learning_rate": 3.053158439765274e-05,
"loss": 0.3387,
"step": 1450
},
{
"epoch": 1.3560111835973905,
"grad_norm": 0.3003534834950553,
"learning_rate": 3.0445288229202625e-05,
"loss": 0.3462,
"step": 1455
},
{
"epoch": 1.3606710158434296,
"grad_norm": 0.29501129699870093,
"learning_rate": 3.0358992060752506e-05,
"loss": 0.3519,
"step": 1460
},
{
"epoch": 1.3653308480894688,
"grad_norm": 0.3202277949447824,
"learning_rate": 3.027269589230238e-05,
"loss": 0.3462,
"step": 1465
},
{
"epoch": 1.369990680335508,
"grad_norm": 0.2981395273102546,
"learning_rate": 3.018639972385226e-05,
"loss": 0.3432,
"step": 1470
},
{
"epoch": 1.3746505125815471,
"grad_norm": 0.3407288422254295,
"learning_rate": 3.010010355540214e-05,
"loss": 0.3454,
"step": 1475
},
{
"epoch": 1.3793103448275863,
"grad_norm": 0.24456230322288272,
"learning_rate": 3.001380738695202e-05,
"loss": 0.3297,
"step": 1480
},
{
"epoch": 1.3839701770736252,
"grad_norm": 0.2790325704839491,
"learning_rate": 2.99275112185019e-05,
"loss": 0.3463,
"step": 1485
},
{
"epoch": 1.3886300093196646,
"grad_norm": 0.33686334122298167,
"learning_rate": 2.984121505005178e-05,
"loss": 0.3514,
"step": 1490
},
{
"epoch": 1.3932898415657036,
"grad_norm": 0.3381134921260071,
"learning_rate": 2.9754918881601657e-05,
"loss": 0.3598,
"step": 1495
},
{
"epoch": 1.3979496738117427,
"grad_norm": 0.32838829943492076,
"learning_rate": 2.966862271315154e-05,
"loss": 0.3556,
"step": 1500
},
{
"epoch": 1.402609506057782,
"grad_norm": 0.34529069200871637,
"learning_rate": 2.9582326544701417e-05,
"loss": 0.3489,
"step": 1505
},
{
"epoch": 1.407269338303821,
"grad_norm": 0.30657286831685626,
"learning_rate": 2.9496030376251292e-05,
"loss": 0.3445,
"step": 1510
},
{
"epoch": 1.4119291705498602,
"grad_norm": 0.25392578712740965,
"learning_rate": 2.9409734207801177e-05,
"loss": 0.3477,
"step": 1515
},
{
"epoch": 1.4165890027958994,
"grad_norm": 0.3091658753936205,
"learning_rate": 2.9323438039351052e-05,
"loss": 0.3473,
"step": 1520
},
{
"epoch": 1.4212488350419386,
"grad_norm": 0.2977398136228978,
"learning_rate": 2.9237141870900937e-05,
"loss": 0.3435,
"step": 1525
},
{
"epoch": 1.4259086672879777,
"grad_norm": 0.30031683880506843,
"learning_rate": 2.9150845702450812e-05,
"loss": 0.3426,
"step": 1530
},
{
"epoch": 1.4305684995340169,
"grad_norm": 0.2915496277557661,
"learning_rate": 2.906454953400069e-05,
"loss": 0.3472,
"step": 1535
},
{
"epoch": 1.4352283317800558,
"grad_norm": 0.29419047279424887,
"learning_rate": 2.897825336555057e-05,
"loss": 0.3434,
"step": 1540
},
{
"epoch": 1.439888164026095,
"grad_norm": 0.3763094540207104,
"learning_rate": 2.889195719710045e-05,
"loss": 0.3439,
"step": 1545
},
{
"epoch": 1.4445479962721341,
"grad_norm": 0.3354009542668028,
"learning_rate": 2.8805661028650328e-05,
"loss": 0.3403,
"step": 1550
},
{
"epoch": 1.4492078285181733,
"grad_norm": 0.39848398270218305,
"learning_rate": 2.871936486020021e-05,
"loss": 0.3378,
"step": 1555
},
{
"epoch": 1.4538676607642125,
"grad_norm": 0.3466122240719805,
"learning_rate": 2.8633068691750088e-05,
"loss": 0.3405,
"step": 1560
},
{
"epoch": 1.4585274930102516,
"grad_norm": 0.3575402768908414,
"learning_rate": 2.854677252329997e-05,
"loss": 0.3428,
"step": 1565
},
{
"epoch": 1.4631873252562908,
"grad_norm": 0.3096582186800248,
"learning_rate": 2.8460476354849848e-05,
"loss": 0.3497,
"step": 1570
},
{
"epoch": 1.46784715750233,
"grad_norm": 0.33013590252010105,
"learning_rate": 2.8374180186399723e-05,
"loss": 0.3496,
"step": 1575
},
{
"epoch": 1.4725069897483691,
"grad_norm": 0.3187151477116704,
"learning_rate": 2.8287884017949608e-05,
"loss": 0.3506,
"step": 1580
},
{
"epoch": 1.477166821994408,
"grad_norm": 0.34348421278060276,
"learning_rate": 2.8201587849499482e-05,
"loss": 0.3438,
"step": 1585
},
{
"epoch": 1.4818266542404475,
"grad_norm": 0.326900676939141,
"learning_rate": 2.811529168104936e-05,
"loss": 0.3538,
"step": 1590
},
{
"epoch": 1.4864864864864864,
"grad_norm": 0.3618370562846467,
"learning_rate": 2.8028995512599242e-05,
"loss": 0.345,
"step": 1595
},
{
"epoch": 1.4911463187325256,
"grad_norm": 0.41066760240240163,
"learning_rate": 2.794269934414912e-05,
"loss": 0.3465,
"step": 1600
},
{
"epoch": 1.4958061509785647,
"grad_norm": 0.29799959064598935,
"learning_rate": 2.7856403175699002e-05,
"loss": 0.3504,
"step": 1605
},
{
"epoch": 1.500465983224604,
"grad_norm": 0.32188730751398653,
"learning_rate": 2.777010700724888e-05,
"loss": 0.3459,
"step": 1610
},
{
"epoch": 1.505125815470643,
"grad_norm": 0.34489249948485085,
"learning_rate": 2.768381083879876e-05,
"loss": 0.3412,
"step": 1615
},
{
"epoch": 1.5097856477166822,
"grad_norm": 0.2743632271565497,
"learning_rate": 2.759751467034864e-05,
"loss": 0.3433,
"step": 1620
},
{
"epoch": 1.5144454799627214,
"grad_norm": 0.2632201162315543,
"learning_rate": 2.751121850189852e-05,
"loss": 0.3378,
"step": 1625
},
{
"epoch": 1.5191053122087603,
"grad_norm": 0.310113877848877,
"learning_rate": 2.7424922333448393e-05,
"loss": 0.331,
"step": 1630
},
{
"epoch": 1.5237651444547997,
"grad_norm": 0.36832951301654737,
"learning_rate": 2.7338626164998278e-05,
"loss": 0.3383,
"step": 1635
},
{
"epoch": 1.5284249767008387,
"grad_norm": 0.2870449857657397,
"learning_rate": 2.7252329996548153e-05,
"loss": 0.3412,
"step": 1640
},
{
"epoch": 1.533084808946878,
"grad_norm": 0.2952641137092537,
"learning_rate": 2.7166033828098038e-05,
"loss": 0.3563,
"step": 1645
},
{
"epoch": 1.537744641192917,
"grad_norm": 0.3299906919276126,
"learning_rate": 2.7079737659647913e-05,
"loss": 0.3452,
"step": 1650
},
{
"epoch": 1.5424044734389561,
"grad_norm": 0.3822117175450514,
"learning_rate": 2.699344149119779e-05,
"loss": 0.3524,
"step": 1655
},
{
"epoch": 1.5470643056849953,
"grad_norm": 0.41478336857871095,
"learning_rate": 2.6907145322747673e-05,
"loss": 0.3467,
"step": 1660
},
{
"epoch": 1.5517241379310345,
"grad_norm": 0.34501912306337074,
"learning_rate": 2.682084915429755e-05,
"loss": 0.3378,
"step": 1665
},
{
"epoch": 1.5563839701770736,
"grad_norm": 0.32970590684967527,
"learning_rate": 2.673455298584743e-05,
"loss": 0.3425,
"step": 1670
},
{
"epoch": 1.5610438024231128,
"grad_norm": 0.29535876114345505,
"learning_rate": 2.664825681739731e-05,
"loss": 0.3496,
"step": 1675
},
{
"epoch": 1.565703634669152,
"grad_norm": 0.28902437066340203,
"learning_rate": 2.6561960648947186e-05,
"loss": 0.3413,
"step": 1680
},
{
"epoch": 1.570363466915191,
"grad_norm": 0.28920609093403055,
"learning_rate": 2.647566448049707e-05,
"loss": 0.3376,
"step": 1685
},
{
"epoch": 1.5750232991612303,
"grad_norm": 0.300406772860295,
"learning_rate": 2.6389368312046945e-05,
"loss": 0.3496,
"step": 1690
},
{
"epoch": 1.5796831314072692,
"grad_norm": 0.2973719190935233,
"learning_rate": 2.6303072143596824e-05,
"loss": 0.3537,
"step": 1695
},
{
"epoch": 1.5843429636533086,
"grad_norm": 0.3559271770619745,
"learning_rate": 2.6216775975146705e-05,
"loss": 0.3463,
"step": 1700
},
{
"epoch": 1.5890027958993476,
"grad_norm": 0.2984734620990466,
"learning_rate": 2.6130479806696584e-05,
"loss": 0.3436,
"step": 1705
},
{
"epoch": 1.5936626281453867,
"grad_norm": 0.36976011646872764,
"learning_rate": 2.6044183638246462e-05,
"loss": 0.3492,
"step": 1710
},
{
"epoch": 1.598322460391426,
"grad_norm": 0.24901029674107222,
"learning_rate": 2.5957887469796343e-05,
"loss": 0.3385,
"step": 1715
},
{
"epoch": 1.602982292637465,
"grad_norm": 0.2934228784179318,
"learning_rate": 2.587159130134622e-05,
"loss": 0.3431,
"step": 1720
},
{
"epoch": 1.6076421248835042,
"grad_norm": 0.26723336889564814,
"learning_rate": 2.5785295132896096e-05,
"loss": 0.3424,
"step": 1725
},
{
"epoch": 1.6123019571295434,
"grad_norm": 0.298932772624824,
"learning_rate": 2.569899896444598e-05,
"loss": 0.3368,
"step": 1730
},
{
"epoch": 1.6169617893755825,
"grad_norm": 0.30311613432335305,
"learning_rate": 2.5612702795995856e-05,
"loss": 0.3472,
"step": 1735
},
{
"epoch": 1.6216216216216215,
"grad_norm": 0.29360494239139895,
"learning_rate": 2.552640662754574e-05,
"loss": 0.3419,
"step": 1740
},
{
"epoch": 1.6262814538676609,
"grad_norm": 0.2281942309825483,
"learning_rate": 2.5440110459095616e-05,
"loss": 0.3488,
"step": 1745
},
{
"epoch": 1.6309412861136998,
"grad_norm": 0.2975378743607759,
"learning_rate": 2.5353814290645494e-05,
"loss": 0.3475,
"step": 1750
},
{
"epoch": 1.6356011183597392,
"grad_norm": 0.2789041307290177,
"learning_rate": 2.5267518122195376e-05,
"loss": 0.3392,
"step": 1755
},
{
"epoch": 1.6402609506057781,
"grad_norm": 0.38202990153937605,
"learning_rate": 2.5181221953745254e-05,
"loss": 0.3479,
"step": 1760
},
{
"epoch": 1.6449207828518173,
"grad_norm": 0.3037667292862275,
"learning_rate": 2.5094925785295132e-05,
"loss": 0.3493,
"step": 1765
},
{
"epoch": 1.6495806150978565,
"grad_norm": 0.2660087977775579,
"learning_rate": 2.5008629616845014e-05,
"loss": 0.3425,
"step": 1770
},
{
"epoch": 1.6542404473438956,
"grad_norm": 0.3053784027438484,
"learning_rate": 2.4922333448394892e-05,
"loss": 0.3357,
"step": 1775
},
{
"epoch": 1.6589002795899348,
"grad_norm": 0.30314491817148603,
"learning_rate": 2.483603727994477e-05,
"loss": 0.3451,
"step": 1780
},
{
"epoch": 1.6635601118359737,
"grad_norm": 0.30640801376339205,
"learning_rate": 2.4749741111494652e-05,
"loss": 0.3425,
"step": 1785
},
{
"epoch": 1.6682199440820131,
"grad_norm": 0.28619788645869887,
"learning_rate": 2.466344494304453e-05,
"loss": 0.3416,
"step": 1790
},
{
"epoch": 1.672879776328052,
"grad_norm": 0.2739271771608511,
"learning_rate": 2.457714877459441e-05,
"loss": 0.3346,
"step": 1795
},
{
"epoch": 1.6775396085740915,
"grad_norm": 0.2951647529514162,
"learning_rate": 2.4490852606144287e-05,
"loss": 0.3348,
"step": 1800
},
{
"epoch": 1.6821994408201304,
"grad_norm": 0.29455068385598043,
"learning_rate": 2.440455643769417e-05,
"loss": 0.3378,
"step": 1805
},
{
"epoch": 1.6868592730661698,
"grad_norm": 0.26860247088817685,
"learning_rate": 2.4318260269244047e-05,
"loss": 0.341,
"step": 1810
},
{
"epoch": 1.6915191053122087,
"grad_norm": 0.2444734658271959,
"learning_rate": 2.4231964100793925e-05,
"loss": 0.3291,
"step": 1815
},
{
"epoch": 1.696178937558248,
"grad_norm": 0.26374792683498094,
"learning_rate": 2.4145667932343803e-05,
"loss": 0.3403,
"step": 1820
},
{
"epoch": 1.700838769804287,
"grad_norm": 0.24867933668825187,
"learning_rate": 2.4059371763893685e-05,
"loss": 0.3447,
"step": 1825
},
{
"epoch": 1.7054986020503262,
"grad_norm": 0.2794693850001193,
"learning_rate": 2.3973075595443563e-05,
"loss": 0.3499,
"step": 1830
},
{
"epoch": 1.7101584342963654,
"grad_norm": 0.3777593243260375,
"learning_rate": 2.388677942699344e-05,
"loss": 0.3448,
"step": 1835
},
{
"epoch": 1.7148182665424043,
"grad_norm": 0.27863573431474875,
"learning_rate": 2.3800483258543323e-05,
"loss": 0.3409,
"step": 1840
},
{
"epoch": 1.7194780987884437,
"grad_norm": 0.27358618998660655,
"learning_rate": 2.37141870900932e-05,
"loss": 0.3457,
"step": 1845
},
{
"epoch": 1.7241379310344827,
"grad_norm": 0.2455011014913662,
"learning_rate": 2.3627890921643083e-05,
"loss": 0.3288,
"step": 1850
},
{
"epoch": 1.728797763280522,
"grad_norm": 0.28836271838037164,
"learning_rate": 2.3541594753192957e-05,
"loss": 0.3384,
"step": 1855
},
{
"epoch": 1.733457595526561,
"grad_norm": 0.31542425547455827,
"learning_rate": 2.345529858474284e-05,
"loss": 0.3386,
"step": 1860
},
{
"epoch": 1.7381174277726001,
"grad_norm": 0.3449280835709219,
"learning_rate": 2.3369002416292717e-05,
"loss": 0.3381,
"step": 1865
},
{
"epoch": 1.7427772600186393,
"grad_norm": 0.3136685515414782,
"learning_rate": 2.32827062478426e-05,
"loss": 0.337,
"step": 1870
},
{
"epoch": 1.7474370922646785,
"grad_norm": 0.25085169672522356,
"learning_rate": 2.3196410079392474e-05,
"loss": 0.3454,
"step": 1875
},
{
"epoch": 1.7520969245107176,
"grad_norm": 0.32201426514790754,
"learning_rate": 2.3110113910942355e-05,
"loss": 0.34,
"step": 1880
},
{
"epoch": 1.7567567567567568,
"grad_norm": 0.28567210714179553,
"learning_rate": 2.3023817742492234e-05,
"loss": 0.3456,
"step": 1885
},
{
"epoch": 1.761416589002796,
"grad_norm": 0.27983348749444076,
"learning_rate": 2.2937521574042115e-05,
"loss": 0.3429,
"step": 1890
},
{
"epoch": 1.766076421248835,
"grad_norm": 0.3373872109864465,
"learning_rate": 2.285122540559199e-05,
"loss": 0.3409,
"step": 1895
},
{
"epoch": 1.7707362534948743,
"grad_norm": 0.3794270581387314,
"learning_rate": 2.276492923714187e-05,
"loss": 0.3331,
"step": 1900
},
{
"epoch": 1.7753960857409132,
"grad_norm": 0.2826762681367889,
"learning_rate": 2.267863306869175e-05,
"loss": 0.3325,
"step": 1905
},
{
"epoch": 1.7800559179869526,
"grad_norm": 0.26535676473539,
"learning_rate": 2.259233690024163e-05,
"loss": 0.3451,
"step": 1910
},
{
"epoch": 1.7847157502329916,
"grad_norm": 0.31037494100061985,
"learning_rate": 2.250604073179151e-05,
"loss": 0.3288,
"step": 1915
},
{
"epoch": 1.7893755824790307,
"grad_norm": 0.2976688611812692,
"learning_rate": 2.2419744563341388e-05,
"loss": 0.3446,
"step": 1920
},
{
"epoch": 1.7940354147250699,
"grad_norm": 0.24328341218076308,
"learning_rate": 2.233344839489127e-05,
"loss": 0.3412,
"step": 1925
},
{
"epoch": 1.798695246971109,
"grad_norm": 0.29166964439196236,
"learning_rate": 2.2247152226441148e-05,
"loss": 0.3379,
"step": 1930
},
{
"epoch": 1.8033550792171482,
"grad_norm": 0.28510438797289134,
"learning_rate": 2.2160856057991026e-05,
"loss": 0.343,
"step": 1935
},
{
"epoch": 1.8080149114631874,
"grad_norm": 0.26949563175148045,
"learning_rate": 2.2074559889540904e-05,
"loss": 0.3305,
"step": 1940
},
{
"epoch": 1.8126747437092265,
"grad_norm": 0.24779885849127598,
"learning_rate": 2.1988263721090786e-05,
"loss": 0.3471,
"step": 1945
},
{
"epoch": 1.8173345759552655,
"grad_norm": 0.2527580229109616,
"learning_rate": 2.1901967552640664e-05,
"loss": 0.3308,
"step": 1950
},
{
"epoch": 1.8219944082013049,
"grad_norm": 0.32138719398479965,
"learning_rate": 2.1815671384190542e-05,
"loss": 0.3386,
"step": 1955
},
{
"epoch": 1.8266542404473438,
"grad_norm": 0.3051026899981029,
"learning_rate": 2.172937521574042e-05,
"loss": 0.3424,
"step": 1960
},
{
"epoch": 1.8313140726933832,
"grad_norm": 0.2919317389332412,
"learning_rate": 2.1643079047290302e-05,
"loss": 0.3433,
"step": 1965
},
{
"epoch": 1.8359739049394221,
"grad_norm": 0.2896377748201226,
"learning_rate": 2.155678287884018e-05,
"loss": 0.3409,
"step": 1970
},
{
"epoch": 1.8406337371854613,
"grad_norm": 0.2558391030776698,
"learning_rate": 2.147048671039006e-05,
"loss": 0.349,
"step": 1975
},
{
"epoch": 1.8452935694315005,
"grad_norm": 0.26035521114289084,
"learning_rate": 2.1384190541939937e-05,
"loss": 0.3395,
"step": 1980
},
{
"epoch": 1.8499534016775396,
"grad_norm": 0.2609714474669437,
"learning_rate": 2.129789437348982e-05,
"loss": 0.336,
"step": 1985
},
{
"epoch": 1.8546132339235788,
"grad_norm": 0.2693923239510721,
"learning_rate": 2.1211598205039697e-05,
"loss": 0.3541,
"step": 1990
},
{
"epoch": 1.8592730661696177,
"grad_norm": 0.322687824919558,
"learning_rate": 2.1125302036589575e-05,
"loss": 0.3399,
"step": 1995
},
{
"epoch": 1.8639328984156571,
"grad_norm": 0.3197420712020152,
"learning_rate": 2.1039005868139457e-05,
"loss": 0.3388,
"step": 2000
},
{
"epoch": 1.868592730661696,
"grad_norm": 0.33199298239633046,
"learning_rate": 2.0952709699689335e-05,
"loss": 0.3324,
"step": 2005
},
{
"epoch": 1.8732525629077355,
"grad_norm": 0.27856296351857995,
"learning_rate": 2.0866413531239216e-05,
"loss": 0.3473,
"step": 2010
},
{
"epoch": 1.8779123951537744,
"grad_norm": 0.29304004571864817,
"learning_rate": 2.078011736278909e-05,
"loss": 0.334,
"step": 2015
},
{
"epoch": 1.8825722273998136,
"grad_norm": 0.24919415043627338,
"learning_rate": 2.0693821194338973e-05,
"loss": 0.3288,
"step": 2020
},
{
"epoch": 1.8872320596458527,
"grad_norm": 0.3180815683978963,
"learning_rate": 2.060752502588885e-05,
"loss": 0.3348,
"step": 2025
},
{
"epoch": 1.8918918918918919,
"grad_norm": 0.24445656460948836,
"learning_rate": 2.0521228857438733e-05,
"loss": 0.3419,
"step": 2030
},
{
"epoch": 1.896551724137931,
"grad_norm": 0.2498906892536783,
"learning_rate": 2.0434932688988608e-05,
"loss": 0.3351,
"step": 2035
},
{
"epoch": 1.9012115563839702,
"grad_norm": 0.2632214084281026,
"learning_rate": 2.034863652053849e-05,
"loss": 0.3381,
"step": 2040
},
{
"epoch": 1.9058713886300094,
"grad_norm": 0.24901758551275088,
"learning_rate": 2.0262340352088367e-05,
"loss": 0.353,
"step": 2045
},
{
"epoch": 1.9105312208760483,
"grad_norm": 0.2896812202839376,
"learning_rate": 2.017604418363825e-05,
"loss": 0.344,
"step": 2050
},
{
"epoch": 1.9151910531220877,
"grad_norm": 0.3204911440069581,
"learning_rate": 2.0089748015188127e-05,
"loss": 0.3382,
"step": 2055
},
{
"epoch": 1.9198508853681266,
"grad_norm": 0.2501138833900388,
"learning_rate": 2.0003451846738005e-05,
"loss": 0.3457,
"step": 2060
},
{
"epoch": 1.924510717614166,
"grad_norm": 0.27670902549239634,
"learning_rate": 1.9917155678287887e-05,
"loss": 0.3352,
"step": 2065
},
{
"epoch": 1.929170549860205,
"grad_norm": 0.261577148436467,
"learning_rate": 1.9830859509837765e-05,
"loss": 0.3427,
"step": 2070
},
{
"epoch": 1.9338303821062441,
"grad_norm": 0.2620237904497218,
"learning_rate": 1.9744563341387643e-05,
"loss": 0.3439,
"step": 2075
},
{
"epoch": 1.9384902143522833,
"grad_norm": 0.29404416747533085,
"learning_rate": 1.9658267172937522e-05,
"loss": 0.3392,
"step": 2080
},
{
"epoch": 1.9431500465983225,
"grad_norm": 0.3750096471173557,
"learning_rate": 1.9571971004487403e-05,
"loss": 0.3393,
"step": 2085
},
{
"epoch": 1.9478098788443616,
"grad_norm": 0.2926457559855875,
"learning_rate": 1.948567483603728e-05,
"loss": 0.3445,
"step": 2090
},
{
"epoch": 1.9524697110904008,
"grad_norm": 0.32683574545147326,
"learning_rate": 1.939937866758716e-05,
"loss": 0.3275,
"step": 2095
},
{
"epoch": 1.95712954333644,
"grad_norm": 0.2855923713407568,
"learning_rate": 1.9313082499137038e-05,
"loss": 0.3445,
"step": 2100
},
{
"epoch": 1.961789375582479,
"grad_norm": 0.2833543994022746,
"learning_rate": 1.922678633068692e-05,
"loss": 0.3388,
"step": 2105
},
{
"epoch": 1.9664492078285183,
"grad_norm": 0.2686665110389253,
"learning_rate": 1.9140490162236798e-05,
"loss": 0.3445,
"step": 2110
},
{
"epoch": 1.9711090400745572,
"grad_norm": 0.27211984977382214,
"learning_rate": 1.9054193993786676e-05,
"loss": 0.3514,
"step": 2115
},
{
"epoch": 1.9757688723205966,
"grad_norm": 0.2506426126864415,
"learning_rate": 1.8967897825336554e-05,
"loss": 0.3398,
"step": 2120
},
{
"epoch": 1.9804287045666356,
"grad_norm": 0.26027358549824325,
"learning_rate": 1.8881601656886436e-05,
"loss": 0.3367,
"step": 2125
},
{
"epoch": 1.9850885368126747,
"grad_norm": 0.2687840933765972,
"learning_rate": 1.8795305488436314e-05,
"loss": 0.3409,
"step": 2130
},
{
"epoch": 1.9897483690587139,
"grad_norm": 0.28597642998090084,
"learning_rate": 1.8709009319986192e-05,
"loss": 0.3294,
"step": 2135
},
{
"epoch": 1.994408201304753,
"grad_norm": 0.26520091936026574,
"learning_rate": 1.8622713151536074e-05,
"loss": 0.3363,
"step": 2140
},
{
"epoch": 1.9990680335507922,
"grad_norm": 0.2429965039838401,
"learning_rate": 1.8536416983085952e-05,
"loss": 0.3411,
"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": 1.83707713995486e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}