OpenR1-Qwen-7B-Sparse-P40 / trainer_state.json
ZMC2019's picture
Model save
8869d30 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": 2.9358379585521024,
"learning_rate": 4.651162790697675e-07,
"loss": 0.4889,
"step": 5
},
{
"epoch": 0.009319664492078284,
"grad_norm": 3.0095855100891242,
"learning_rate": 9.30232558139535e-07,
"loss": 0.4777,
"step": 10
},
{
"epoch": 0.013979496738117428,
"grad_norm": 2.3588796393229203,
"learning_rate": 1.3953488372093025e-06,
"loss": 0.4643,
"step": 15
},
{
"epoch": 0.01863932898415657,
"grad_norm": 1.3763840278489534,
"learning_rate": 1.86046511627907e-06,
"loss": 0.4556,
"step": 20
},
{
"epoch": 0.023299161230195712,
"grad_norm": 0.6192077367736554,
"learning_rate": 2.3255813953488376e-06,
"loss": 0.3961,
"step": 25
},
{
"epoch": 0.027958993476234855,
"grad_norm": 0.617990002942175,
"learning_rate": 2.790697674418605e-06,
"loss": 0.3762,
"step": 30
},
{
"epoch": 0.032618825722273995,
"grad_norm": 0.5925450142970052,
"learning_rate": 3.2558139534883724e-06,
"loss": 0.374,
"step": 35
},
{
"epoch": 0.03727865796831314,
"grad_norm": 0.3561624832181387,
"learning_rate": 3.72093023255814e-06,
"loss": 0.3725,
"step": 40
},
{
"epoch": 0.04193849021435228,
"grad_norm": 0.25081650759652796,
"learning_rate": 4.186046511627907e-06,
"loss": 0.3645,
"step": 45
},
{
"epoch": 0.046598322460391424,
"grad_norm": 0.20558496554294722,
"learning_rate": 4.651162790697675e-06,
"loss": 0.3667,
"step": 50
},
{
"epoch": 0.05125815470643057,
"grad_norm": 0.24486074932144072,
"learning_rate": 5.116279069767442e-06,
"loss": 0.3571,
"step": 55
},
{
"epoch": 0.05591798695246971,
"grad_norm": 0.19060108583147642,
"learning_rate": 5.58139534883721e-06,
"loss": 0.3505,
"step": 60
},
{
"epoch": 0.06057781919850885,
"grad_norm": 0.18154982231055428,
"learning_rate": 6.046511627906977e-06,
"loss": 0.3434,
"step": 65
},
{
"epoch": 0.06523765144454799,
"grad_norm": 0.1779776294534931,
"learning_rate": 6.511627906976745e-06,
"loss": 0.3492,
"step": 70
},
{
"epoch": 0.06989748369058714,
"grad_norm": 0.17606219612014878,
"learning_rate": 6.976744186046513e-06,
"loss": 0.3416,
"step": 75
},
{
"epoch": 0.07455731593662628,
"grad_norm": 0.1814825848846307,
"learning_rate": 7.44186046511628e-06,
"loss": 0.3377,
"step": 80
},
{
"epoch": 0.07921714818266543,
"grad_norm": 0.17880816384452355,
"learning_rate": 7.906976744186048e-06,
"loss": 0.3393,
"step": 85
},
{
"epoch": 0.08387698042870456,
"grad_norm": 0.1568045157823065,
"learning_rate": 8.372093023255815e-06,
"loss": 0.3461,
"step": 90
},
{
"epoch": 0.08853681267474371,
"grad_norm": 0.16443817784114392,
"learning_rate": 8.837209302325582e-06,
"loss": 0.336,
"step": 95
},
{
"epoch": 0.09319664492078285,
"grad_norm": 0.16992057839542438,
"learning_rate": 9.30232558139535e-06,
"loss": 0.3316,
"step": 100
},
{
"epoch": 0.097856477166822,
"grad_norm": 0.1605564738554441,
"learning_rate": 9.767441860465117e-06,
"loss": 0.3311,
"step": 105
},
{
"epoch": 0.10251630941286113,
"grad_norm": 0.16236919155376553,
"learning_rate": 1.0232558139534884e-05,
"loss": 0.3316,
"step": 110
},
{
"epoch": 0.10717614165890028,
"grad_norm": 0.1698286305625274,
"learning_rate": 1.0697674418604651e-05,
"loss": 0.3328,
"step": 115
},
{
"epoch": 0.11183597390493942,
"grad_norm": 0.16840606723491117,
"learning_rate": 1.116279069767442e-05,
"loss": 0.3289,
"step": 120
},
{
"epoch": 0.11649580615097857,
"grad_norm": 0.16484123295216122,
"learning_rate": 1.1627906976744187e-05,
"loss": 0.325,
"step": 125
},
{
"epoch": 0.1211556383970177,
"grad_norm": 0.18951085550690486,
"learning_rate": 1.2093023255813954e-05,
"loss": 0.3324,
"step": 130
},
{
"epoch": 0.12581547064305684,
"grad_norm": 0.16528197993781607,
"learning_rate": 1.2558139534883723e-05,
"loss": 0.3282,
"step": 135
},
{
"epoch": 0.13047530288909598,
"grad_norm": 0.18137576101744035,
"learning_rate": 1.302325581395349e-05,
"loss": 0.3248,
"step": 140
},
{
"epoch": 0.13513513513513514,
"grad_norm": 0.18673134625349136,
"learning_rate": 1.3488372093023257e-05,
"loss": 0.3413,
"step": 145
},
{
"epoch": 0.13979496738117428,
"grad_norm": 0.19883071868742613,
"learning_rate": 1.3953488372093025e-05,
"loss": 0.3275,
"step": 150
},
{
"epoch": 0.14445479962721341,
"grad_norm": 0.19063978229188172,
"learning_rate": 1.441860465116279e-05,
"loss": 0.3396,
"step": 155
},
{
"epoch": 0.14911463187325255,
"grad_norm": 0.20126727074239933,
"learning_rate": 1.488372093023256e-05,
"loss": 0.3283,
"step": 160
},
{
"epoch": 0.15377446411929171,
"grad_norm": 0.17583289349910128,
"learning_rate": 1.5348837209302328e-05,
"loss": 0.3295,
"step": 165
},
{
"epoch": 0.15843429636533085,
"grad_norm": 0.17444177904401117,
"learning_rate": 1.5813953488372095e-05,
"loss": 0.3114,
"step": 170
},
{
"epoch": 0.16309412861137,
"grad_norm": 0.18551430775271738,
"learning_rate": 1.6279069767441862e-05,
"loss": 0.3283,
"step": 175
},
{
"epoch": 0.16775396085740912,
"grad_norm": 0.1858093588140544,
"learning_rate": 1.674418604651163e-05,
"loss": 0.3256,
"step": 180
},
{
"epoch": 0.1724137931034483,
"grad_norm": 0.18567509691449366,
"learning_rate": 1.7209302325581396e-05,
"loss": 0.3237,
"step": 185
},
{
"epoch": 0.17707362534948742,
"grad_norm": 0.20208132114348742,
"learning_rate": 1.7674418604651163e-05,
"loss": 0.3309,
"step": 190
},
{
"epoch": 0.18173345759552656,
"grad_norm": 0.22338210289869026,
"learning_rate": 1.813953488372093e-05,
"loss": 0.3275,
"step": 195
},
{
"epoch": 0.1863932898415657,
"grad_norm": 0.2287598111489109,
"learning_rate": 1.86046511627907e-05,
"loss": 0.3299,
"step": 200
},
{
"epoch": 0.19105312208760486,
"grad_norm": 0.19421590377562648,
"learning_rate": 1.9069767441860468e-05,
"loss": 0.3249,
"step": 205
},
{
"epoch": 0.195712954333644,
"grad_norm": 0.22779060574090246,
"learning_rate": 1.9534883720930235e-05,
"loss": 0.3307,
"step": 210
},
{
"epoch": 0.20037278657968313,
"grad_norm": 0.19697826843895966,
"learning_rate": 2e-05,
"loss": 0.3229,
"step": 215
},
{
"epoch": 0.20503261882572227,
"grad_norm": 0.1971661281754431,
"learning_rate": 1.9948213360952874e-05,
"loss": 0.3286,
"step": 220
},
{
"epoch": 0.2096924510717614,
"grad_norm": 0.19606725204483122,
"learning_rate": 1.989642672190575e-05,
"loss": 0.3177,
"step": 225
},
{
"epoch": 0.21435228331780057,
"grad_norm": 0.194751313295783,
"learning_rate": 1.9844640082858625e-05,
"loss": 0.3264,
"step": 230
},
{
"epoch": 0.2190121155638397,
"grad_norm": 0.19401867383810653,
"learning_rate": 1.9792853443811498e-05,
"loss": 0.3236,
"step": 235
},
{
"epoch": 0.22367194780987884,
"grad_norm": 0.2061487676182812,
"learning_rate": 1.9741066804764374e-05,
"loss": 0.3262,
"step": 240
},
{
"epoch": 0.22833178005591798,
"grad_norm": 0.22633953405241564,
"learning_rate": 1.9689280165717246e-05,
"loss": 0.3278,
"step": 245
},
{
"epoch": 0.23299161230195714,
"grad_norm": 0.19231837200784246,
"learning_rate": 1.963749352667012e-05,
"loss": 0.3239,
"step": 250
},
{
"epoch": 0.23765144454799628,
"grad_norm": 0.20567774474325756,
"learning_rate": 1.9585706887622994e-05,
"loss": 0.3262,
"step": 255
},
{
"epoch": 0.2423112767940354,
"grad_norm": 0.22721757172361462,
"learning_rate": 1.953392024857587e-05,
"loss": 0.3295,
"step": 260
},
{
"epoch": 0.24697110904007455,
"grad_norm": 0.22305696098164374,
"learning_rate": 1.9482133609528746e-05,
"loss": 0.3279,
"step": 265
},
{
"epoch": 0.2516309412861137,
"grad_norm": 0.1869588017000237,
"learning_rate": 1.9430346970481618e-05,
"loss": 0.3249,
"step": 270
},
{
"epoch": 0.25629077353215285,
"grad_norm": 0.2172985514276064,
"learning_rate": 1.937856033143449e-05,
"loss": 0.3262,
"step": 275
},
{
"epoch": 0.26095060577819196,
"grad_norm": 0.18457639562753428,
"learning_rate": 1.9326773692387366e-05,
"loss": 0.3229,
"step": 280
},
{
"epoch": 0.2656104380242311,
"grad_norm": 0.2085704847556598,
"learning_rate": 1.927498705334024e-05,
"loss": 0.3254,
"step": 285
},
{
"epoch": 0.2702702702702703,
"grad_norm": 0.2127987275321348,
"learning_rate": 1.9223200414293114e-05,
"loss": 0.3313,
"step": 290
},
{
"epoch": 0.2749301025163094,
"grad_norm": 0.22872842226387602,
"learning_rate": 1.917141377524599e-05,
"loss": 0.341,
"step": 295
},
{
"epoch": 0.27958993476234856,
"grad_norm": 0.20656421805333372,
"learning_rate": 1.9119627136198862e-05,
"loss": 0.3367,
"step": 300
},
{
"epoch": 0.2842497670083877,
"grad_norm": 0.21983247009596246,
"learning_rate": 1.9067840497151738e-05,
"loss": 0.3267,
"step": 305
},
{
"epoch": 0.28890959925442683,
"grad_norm": 0.23918660845251774,
"learning_rate": 1.901605385810461e-05,
"loss": 0.3236,
"step": 310
},
{
"epoch": 0.293569431500466,
"grad_norm": 0.19505454122387392,
"learning_rate": 1.8964267219057483e-05,
"loss": 0.3296,
"step": 315
},
{
"epoch": 0.2982292637465051,
"grad_norm": 0.2131522282361315,
"learning_rate": 1.891248058001036e-05,
"loss": 0.3154,
"step": 320
},
{
"epoch": 0.30288909599254427,
"grad_norm": 0.20650244568080883,
"learning_rate": 1.8860693940963234e-05,
"loss": 0.3172,
"step": 325
},
{
"epoch": 0.30754892823858343,
"grad_norm": 0.19264072341187666,
"learning_rate": 1.8808907301916107e-05,
"loss": 0.3263,
"step": 330
},
{
"epoch": 0.31220876048462254,
"grad_norm": 0.19575618641019582,
"learning_rate": 1.8757120662868982e-05,
"loss": 0.3246,
"step": 335
},
{
"epoch": 0.3168685927306617,
"grad_norm": 0.19258229616398992,
"learning_rate": 1.8705334023821855e-05,
"loss": 0.3238,
"step": 340
},
{
"epoch": 0.32152842497670087,
"grad_norm": 0.2051192074004652,
"learning_rate": 1.865354738477473e-05,
"loss": 0.3243,
"step": 345
},
{
"epoch": 0.32618825722274,
"grad_norm": 0.21124173698060286,
"learning_rate": 1.8601760745727603e-05,
"loss": 0.3351,
"step": 350
},
{
"epoch": 0.33084808946877914,
"grad_norm": 0.22780074322712987,
"learning_rate": 1.854997410668048e-05,
"loss": 0.3307,
"step": 355
},
{
"epoch": 0.33550792171481825,
"grad_norm": 0.20050843718250042,
"learning_rate": 1.8498187467633354e-05,
"loss": 0.3231,
"step": 360
},
{
"epoch": 0.3401677539608574,
"grad_norm": 0.20951533429996613,
"learning_rate": 1.8446400828586227e-05,
"loss": 0.3212,
"step": 365
},
{
"epoch": 0.3448275862068966,
"grad_norm": 0.21171260508207002,
"learning_rate": 1.83946141895391e-05,
"loss": 0.3292,
"step": 370
},
{
"epoch": 0.3494874184529357,
"grad_norm": 0.21633188290889002,
"learning_rate": 1.8342827550491975e-05,
"loss": 0.3224,
"step": 375
},
{
"epoch": 0.35414725069897485,
"grad_norm": 0.20122697603683964,
"learning_rate": 1.8291040911444847e-05,
"loss": 0.3327,
"step": 380
},
{
"epoch": 0.35880708294501396,
"grad_norm": 0.24727101450140662,
"learning_rate": 1.8239254272397723e-05,
"loss": 0.3294,
"step": 385
},
{
"epoch": 0.3634669151910531,
"grad_norm": 0.20467658801471714,
"learning_rate": 1.81874676333506e-05,
"loss": 0.3251,
"step": 390
},
{
"epoch": 0.3681267474370923,
"grad_norm": 0.21341519432060765,
"learning_rate": 1.813568099430347e-05,
"loss": 0.3273,
"step": 395
},
{
"epoch": 0.3727865796831314,
"grad_norm": 0.19729258558856816,
"learning_rate": 1.8083894355256347e-05,
"loss": 0.3174,
"step": 400
},
{
"epoch": 0.37744641192917056,
"grad_norm": 0.23364055194092742,
"learning_rate": 1.803210771620922e-05,
"loss": 0.3371,
"step": 405
},
{
"epoch": 0.3821062441752097,
"grad_norm": 0.2188644833600693,
"learning_rate": 1.798032107716209e-05,
"loss": 0.3293,
"step": 410
},
{
"epoch": 0.38676607642124883,
"grad_norm": 0.21936389966299402,
"learning_rate": 1.7928534438114967e-05,
"loss": 0.3216,
"step": 415
},
{
"epoch": 0.391425908667288,
"grad_norm": 0.2192776499326504,
"learning_rate": 1.7876747799067843e-05,
"loss": 0.3329,
"step": 420
},
{
"epoch": 0.3960857409133271,
"grad_norm": 0.22918816226113867,
"learning_rate": 1.7824961160020715e-05,
"loss": 0.3355,
"step": 425
},
{
"epoch": 0.40074557315936626,
"grad_norm": 0.20292598837602438,
"learning_rate": 1.777317452097359e-05,
"loss": 0.3296,
"step": 430
},
{
"epoch": 0.40540540540540543,
"grad_norm": 0.21448495438266224,
"learning_rate": 1.7721387881926463e-05,
"loss": 0.3361,
"step": 435
},
{
"epoch": 0.41006523765144454,
"grad_norm": 0.22652364723732418,
"learning_rate": 1.766960124287934e-05,
"loss": 0.3276,
"step": 440
},
{
"epoch": 0.4147250698974837,
"grad_norm": 0.23415606339172687,
"learning_rate": 1.761781460383221e-05,
"loss": 0.3259,
"step": 445
},
{
"epoch": 0.4193849021435228,
"grad_norm": 0.20093495271086492,
"learning_rate": 1.7566027964785087e-05,
"loss": 0.3187,
"step": 450
},
{
"epoch": 0.424044734389562,
"grad_norm": 0.2158827654398668,
"learning_rate": 1.7514241325737963e-05,
"loss": 0.3262,
"step": 455
},
{
"epoch": 0.42870456663560114,
"grad_norm": 0.20249378577929067,
"learning_rate": 1.7462454686690835e-05,
"loss": 0.3242,
"step": 460
},
{
"epoch": 0.43336439888164024,
"grad_norm": 0.2478822044298927,
"learning_rate": 1.7410668047643708e-05,
"loss": 0.3211,
"step": 465
},
{
"epoch": 0.4380242311276794,
"grad_norm": 0.21466364366586702,
"learning_rate": 1.7358881408596583e-05,
"loss": 0.3315,
"step": 470
},
{
"epoch": 0.4426840633737186,
"grad_norm": 0.21583782535840262,
"learning_rate": 1.7307094769549456e-05,
"loss": 0.3225,
"step": 475
},
{
"epoch": 0.4473438956197577,
"grad_norm": 0.2277116814012742,
"learning_rate": 1.725530813050233e-05,
"loss": 0.3373,
"step": 480
},
{
"epoch": 0.45200372786579684,
"grad_norm": 0.2200145220713097,
"learning_rate": 1.7203521491455207e-05,
"loss": 0.328,
"step": 485
},
{
"epoch": 0.45666356011183595,
"grad_norm": 0.20597194480299022,
"learning_rate": 1.715173485240808e-05,
"loss": 0.3227,
"step": 490
},
{
"epoch": 0.4613233923578751,
"grad_norm": 0.24111938352420786,
"learning_rate": 1.7099948213360955e-05,
"loss": 0.3262,
"step": 495
},
{
"epoch": 0.4659832246039143,
"grad_norm": 0.21117173274751344,
"learning_rate": 1.7048161574313828e-05,
"loss": 0.336,
"step": 500
},
{
"epoch": 0.4706430568499534,
"grad_norm": 0.2109477726961095,
"learning_rate": 1.6996374935266703e-05,
"loss": 0.3249,
"step": 505
},
{
"epoch": 0.47530288909599255,
"grad_norm": 0.19834985237292366,
"learning_rate": 1.6944588296219576e-05,
"loss": 0.3265,
"step": 510
},
{
"epoch": 0.47996272134203166,
"grad_norm": 0.19919064505532363,
"learning_rate": 1.689280165717245e-05,
"loss": 0.3258,
"step": 515
},
{
"epoch": 0.4846225535880708,
"grad_norm": 0.21189208212338323,
"learning_rate": 1.6841015018125324e-05,
"loss": 0.3305,
"step": 520
},
{
"epoch": 0.48928238583411,
"grad_norm": 0.22407539115980474,
"learning_rate": 1.67892283790782e-05,
"loss": 0.3305,
"step": 525
},
{
"epoch": 0.4939422180801491,
"grad_norm": 0.20719461369420158,
"learning_rate": 1.6737441740031072e-05,
"loss": 0.3256,
"step": 530
},
{
"epoch": 0.49860205032618826,
"grad_norm": 0.21338336824874424,
"learning_rate": 1.6685655100983948e-05,
"loss": 0.3301,
"step": 535
},
{
"epoch": 0.5032618825722274,
"grad_norm": 0.21755987080911546,
"learning_rate": 1.6633868461936823e-05,
"loss": 0.3173,
"step": 540
},
{
"epoch": 0.5079217148182665,
"grad_norm": 0.23805239729872102,
"learning_rate": 1.6582081822889696e-05,
"loss": 0.3242,
"step": 545
},
{
"epoch": 0.5125815470643057,
"grad_norm": 0.2178591347578284,
"learning_rate": 1.653029518384257e-05,
"loss": 0.3294,
"step": 550
},
{
"epoch": 0.5172413793103449,
"grad_norm": 0.22861501679158425,
"learning_rate": 1.6478508544795444e-05,
"loss": 0.3273,
"step": 555
},
{
"epoch": 0.5219012115563839,
"grad_norm": 0.22391633980488124,
"learning_rate": 1.6426721905748316e-05,
"loss": 0.3382,
"step": 560
},
{
"epoch": 0.5265610438024231,
"grad_norm": 0.19479709922612634,
"learning_rate": 1.6374935266701192e-05,
"loss": 0.3157,
"step": 565
},
{
"epoch": 0.5312208760484622,
"grad_norm": 0.24570860862235,
"learning_rate": 1.6323148627654068e-05,
"loss": 0.3197,
"step": 570
},
{
"epoch": 0.5358807082945014,
"grad_norm": 0.22193001265554263,
"learning_rate": 1.6271361988606944e-05,
"loss": 0.3238,
"step": 575
},
{
"epoch": 0.5405405405405406,
"grad_norm": 0.23961884549409623,
"learning_rate": 1.6219575349559816e-05,
"loss": 0.3232,
"step": 580
},
{
"epoch": 0.5452003727865797,
"grad_norm": 0.19137567045296802,
"learning_rate": 1.6167788710512688e-05,
"loss": 0.3265,
"step": 585
},
{
"epoch": 0.5498602050326188,
"grad_norm": 0.20115755434956437,
"learning_rate": 1.6116002071465564e-05,
"loss": 0.3193,
"step": 590
},
{
"epoch": 0.554520037278658,
"grad_norm": 0.2124470228159018,
"learning_rate": 1.6064215432418436e-05,
"loss": 0.3233,
"step": 595
},
{
"epoch": 0.5591798695246971,
"grad_norm": 0.2180160408149773,
"learning_rate": 1.6012428793371312e-05,
"loss": 0.3244,
"step": 600
},
{
"epoch": 0.5638397017707363,
"grad_norm": 0.2160242094251449,
"learning_rate": 1.5960642154324188e-05,
"loss": 0.3234,
"step": 605
},
{
"epoch": 0.5684995340167754,
"grad_norm": 0.24806655464152233,
"learning_rate": 1.590885551527706e-05,
"loss": 0.332,
"step": 610
},
{
"epoch": 0.5731593662628145,
"grad_norm": 0.21806755157780733,
"learning_rate": 1.5857068876229933e-05,
"loss": 0.3229,
"step": 615
},
{
"epoch": 0.5778191985088537,
"grad_norm": 1.2870679394584568,
"learning_rate": 1.5805282237182808e-05,
"loss": 0.3168,
"step": 620
},
{
"epoch": 0.5824790307548928,
"grad_norm": 0.2124373058232124,
"learning_rate": 1.575349559813568e-05,
"loss": 0.3322,
"step": 625
},
{
"epoch": 0.587138863000932,
"grad_norm": 0.5434431989836286,
"learning_rate": 1.5701708959088556e-05,
"loss": 0.3294,
"step": 630
},
{
"epoch": 0.5917986952469712,
"grad_norm": 0.22747818610994888,
"learning_rate": 1.5649922320041432e-05,
"loss": 0.3307,
"step": 635
},
{
"epoch": 0.5964585274930102,
"grad_norm": 0.19699104995801411,
"learning_rate": 1.5598135680994304e-05,
"loss": 0.329,
"step": 640
},
{
"epoch": 0.6011183597390494,
"grad_norm": 0.19502045494995346,
"learning_rate": 1.554634904194718e-05,
"loss": 0.3267,
"step": 645
},
{
"epoch": 0.6057781919850885,
"grad_norm": 0.2192638389835302,
"learning_rate": 1.5494562402900053e-05,
"loss": 0.3294,
"step": 650
},
{
"epoch": 0.6104380242311277,
"grad_norm": 0.21933984732022657,
"learning_rate": 1.5442775763852925e-05,
"loss": 0.332,
"step": 655
},
{
"epoch": 0.6150978564771669,
"grad_norm": 0.21041459913026447,
"learning_rate": 1.53909891248058e-05,
"loss": 0.3156,
"step": 660
},
{
"epoch": 0.6197576887232059,
"grad_norm": 0.23530573949874736,
"learning_rate": 1.5339202485758676e-05,
"loss": 0.3184,
"step": 665
},
{
"epoch": 0.6244175209692451,
"grad_norm": 0.21850565712919703,
"learning_rate": 1.5287415846711552e-05,
"loss": 0.3188,
"step": 670
},
{
"epoch": 0.6290773532152842,
"grad_norm": 0.2361213653169026,
"learning_rate": 1.5235629207664425e-05,
"loss": 0.32,
"step": 675
},
{
"epoch": 0.6337371854613234,
"grad_norm": 0.20552105027007925,
"learning_rate": 1.5183842568617297e-05,
"loss": 0.3282,
"step": 680
},
{
"epoch": 0.6383970177073626,
"grad_norm": 0.24722706453350193,
"learning_rate": 1.5132055929570173e-05,
"loss": 0.3312,
"step": 685
},
{
"epoch": 0.6430568499534017,
"grad_norm": 0.267746674789141,
"learning_rate": 1.5080269290523047e-05,
"loss": 0.3215,
"step": 690
},
{
"epoch": 0.6477166821994408,
"grad_norm": 0.20969336261995508,
"learning_rate": 1.5028482651475919e-05,
"loss": 0.3279,
"step": 695
},
{
"epoch": 0.65237651444548,
"grad_norm": 0.2257138045484241,
"learning_rate": 1.4976696012428795e-05,
"loss": 0.3337,
"step": 700
},
{
"epoch": 0.6570363466915191,
"grad_norm": 0.17961148295092744,
"learning_rate": 1.4924909373381669e-05,
"loss": 0.3224,
"step": 705
},
{
"epoch": 0.6616961789375583,
"grad_norm": 0.23789960713870506,
"learning_rate": 1.4873122734334541e-05,
"loss": 0.3237,
"step": 710
},
{
"epoch": 0.6663560111835974,
"grad_norm": 0.2208025857214297,
"learning_rate": 1.4821336095287417e-05,
"loss": 0.3175,
"step": 715
},
{
"epoch": 0.6710158434296365,
"grad_norm": 0.19989756193207112,
"learning_rate": 1.4769549456240291e-05,
"loss": 0.3227,
"step": 720
},
{
"epoch": 0.6756756756756757,
"grad_norm": 0.26463761665216684,
"learning_rate": 1.4717762817193167e-05,
"loss": 0.3266,
"step": 725
},
{
"epoch": 0.6803355079217148,
"grad_norm": 0.21434087964970538,
"learning_rate": 1.4665976178146039e-05,
"loss": 0.3226,
"step": 730
},
{
"epoch": 0.684995340167754,
"grad_norm": 0.2065276035839868,
"learning_rate": 1.4614189539098913e-05,
"loss": 0.3235,
"step": 735
},
{
"epoch": 0.6896551724137931,
"grad_norm": 0.20649033808104794,
"learning_rate": 1.4562402900051789e-05,
"loss": 0.3187,
"step": 740
},
{
"epoch": 0.6943150046598322,
"grad_norm": 0.20751516672171208,
"learning_rate": 1.4510616261004661e-05,
"loss": 0.319,
"step": 745
},
{
"epoch": 0.6989748369058714,
"grad_norm": 0.19697106412326107,
"learning_rate": 1.4458829621957535e-05,
"loss": 0.3195,
"step": 750
},
{
"epoch": 0.7036346691519105,
"grad_norm": 0.22267160213111145,
"learning_rate": 1.4407042982910411e-05,
"loss": 0.3234,
"step": 755
},
{
"epoch": 0.7082945013979497,
"grad_norm": 0.20795208077111424,
"learning_rate": 1.4355256343863283e-05,
"loss": 0.3239,
"step": 760
},
{
"epoch": 0.7129543336439889,
"grad_norm": 0.20931413370029262,
"learning_rate": 1.430346970481616e-05,
"loss": 0.3157,
"step": 765
},
{
"epoch": 0.7176141658900279,
"grad_norm": 0.2279031584837272,
"learning_rate": 1.4251683065769033e-05,
"loss": 0.3229,
"step": 770
},
{
"epoch": 0.7222739981360671,
"grad_norm": 0.28081041443039917,
"learning_rate": 1.4199896426721906e-05,
"loss": 0.33,
"step": 775
},
{
"epoch": 0.7269338303821062,
"grad_norm": 0.21402684688996287,
"learning_rate": 1.4148109787674781e-05,
"loss": 0.3176,
"step": 780
},
{
"epoch": 0.7315936626281454,
"grad_norm": 0.2170900802684808,
"learning_rate": 1.4096323148627655e-05,
"loss": 0.3217,
"step": 785
},
{
"epoch": 0.7362534948741846,
"grad_norm": 0.24860030155760016,
"learning_rate": 1.404453650958053e-05,
"loss": 0.3265,
"step": 790
},
{
"epoch": 0.7409133271202236,
"grad_norm": 0.19118776157108477,
"learning_rate": 1.3992749870533403e-05,
"loss": 0.3247,
"step": 795
},
{
"epoch": 0.7455731593662628,
"grad_norm": 0.2260662140886731,
"learning_rate": 1.3940963231486278e-05,
"loss": 0.3231,
"step": 800
},
{
"epoch": 0.750232991612302,
"grad_norm": 0.20631237485654458,
"learning_rate": 1.3889176592439152e-05,
"loss": 0.3232,
"step": 805
},
{
"epoch": 0.7548928238583411,
"grad_norm": 0.2504046456069529,
"learning_rate": 1.3837389953392026e-05,
"loss": 0.3332,
"step": 810
},
{
"epoch": 0.7595526561043803,
"grad_norm": 0.22716466517888942,
"learning_rate": 1.37856033143449e-05,
"loss": 0.3202,
"step": 815
},
{
"epoch": 0.7642124883504194,
"grad_norm": 0.2609982780137847,
"learning_rate": 1.3733816675297775e-05,
"loss": 0.3285,
"step": 820
},
{
"epoch": 0.7688723205964585,
"grad_norm": 0.269151275613793,
"learning_rate": 1.3682030036250648e-05,
"loss": 0.3206,
"step": 825
},
{
"epoch": 0.7735321528424977,
"grad_norm": 0.21925839142695047,
"learning_rate": 1.3630243397203522e-05,
"loss": 0.3192,
"step": 830
},
{
"epoch": 0.7781919850885368,
"grad_norm": 0.20386131918941994,
"learning_rate": 1.3578456758156398e-05,
"loss": 0.321,
"step": 835
},
{
"epoch": 0.782851817334576,
"grad_norm": 0.2532163945055847,
"learning_rate": 1.3526670119109272e-05,
"loss": 0.3275,
"step": 840
},
{
"epoch": 0.7875116495806151,
"grad_norm": 0.21472966630917703,
"learning_rate": 1.3474883480062144e-05,
"loss": 0.3349,
"step": 845
},
{
"epoch": 0.7921714818266542,
"grad_norm": 0.24312133461634136,
"learning_rate": 1.342309684101502e-05,
"loss": 0.321,
"step": 850
},
{
"epoch": 0.7968313140726934,
"grad_norm": 0.2229944694531994,
"learning_rate": 1.3371310201967894e-05,
"loss": 0.3278,
"step": 855
},
{
"epoch": 0.8014911463187325,
"grad_norm": 0.19371137005852632,
"learning_rate": 1.3319523562920766e-05,
"loss": 0.3338,
"step": 860
},
{
"epoch": 0.8061509785647717,
"grad_norm": 0.22950989752366102,
"learning_rate": 1.3267736923873642e-05,
"loss": 0.3208,
"step": 865
},
{
"epoch": 0.8108108108108109,
"grad_norm": 0.22496120833511551,
"learning_rate": 1.3215950284826516e-05,
"loss": 0.328,
"step": 870
},
{
"epoch": 0.8154706430568499,
"grad_norm": 0.20551584337966472,
"learning_rate": 1.3164163645779392e-05,
"loss": 0.3257,
"step": 875
},
{
"epoch": 0.8201304753028891,
"grad_norm": 0.21692857406595228,
"learning_rate": 1.3112377006732264e-05,
"loss": 0.3226,
"step": 880
},
{
"epoch": 0.8247903075489282,
"grad_norm": 0.2171027152401885,
"learning_rate": 1.3060590367685138e-05,
"loss": 0.3278,
"step": 885
},
{
"epoch": 0.8294501397949674,
"grad_norm": 0.22927464515991447,
"learning_rate": 1.3008803728638014e-05,
"loss": 0.3212,
"step": 890
},
{
"epoch": 0.8341099720410066,
"grad_norm": 0.19864068045743993,
"learning_rate": 1.2957017089590886e-05,
"loss": 0.337,
"step": 895
},
{
"epoch": 0.8387698042870456,
"grad_norm": 0.19961778149739876,
"learning_rate": 1.290523045054376e-05,
"loss": 0.3241,
"step": 900
},
{
"epoch": 0.8434296365330848,
"grad_norm": 0.18416032414845795,
"learning_rate": 1.2853443811496636e-05,
"loss": 0.3222,
"step": 905
},
{
"epoch": 0.848089468779124,
"grad_norm": 0.2110298314090597,
"learning_rate": 1.2801657172449508e-05,
"loss": 0.3262,
"step": 910
},
{
"epoch": 0.8527493010251631,
"grad_norm": 0.2281982369217672,
"learning_rate": 1.2749870533402384e-05,
"loss": 0.327,
"step": 915
},
{
"epoch": 0.8574091332712023,
"grad_norm": 0.22872935927067808,
"learning_rate": 1.2698083894355258e-05,
"loss": 0.323,
"step": 920
},
{
"epoch": 0.8620689655172413,
"grad_norm": 0.21165791707242024,
"learning_rate": 1.264629725530813e-05,
"loss": 0.3169,
"step": 925
},
{
"epoch": 0.8667287977632805,
"grad_norm": 0.18628554545791506,
"learning_rate": 1.2594510616261006e-05,
"loss": 0.3164,
"step": 930
},
{
"epoch": 0.8713886300093197,
"grad_norm": 0.23083205226589923,
"learning_rate": 1.254272397721388e-05,
"loss": 0.3222,
"step": 935
},
{
"epoch": 0.8760484622553588,
"grad_norm": 0.2192954299935438,
"learning_rate": 1.2490937338166753e-05,
"loss": 0.3244,
"step": 940
},
{
"epoch": 0.880708294501398,
"grad_norm": 0.2248357209056571,
"learning_rate": 1.2439150699119628e-05,
"loss": 0.3219,
"step": 945
},
{
"epoch": 0.8853681267474371,
"grad_norm": 0.20471170755414897,
"learning_rate": 1.2387364060072502e-05,
"loss": 0.3333,
"step": 950
},
{
"epoch": 0.8900279589934762,
"grad_norm": 0.2076140935278383,
"learning_rate": 1.2335577421025375e-05,
"loss": 0.3269,
"step": 955
},
{
"epoch": 0.8946877912395154,
"grad_norm": 0.1767922723082159,
"learning_rate": 1.228379078197825e-05,
"loss": 0.3296,
"step": 960
},
{
"epoch": 0.8993476234855545,
"grad_norm": 0.24187201186511662,
"learning_rate": 1.2232004142931125e-05,
"loss": 0.3251,
"step": 965
},
{
"epoch": 0.9040074557315937,
"grad_norm": 0.2309707614632519,
"learning_rate": 1.2180217503884e-05,
"loss": 0.3166,
"step": 970
},
{
"epoch": 0.9086672879776329,
"grad_norm": 0.20032479610884243,
"learning_rate": 1.2128430864836873e-05,
"loss": 0.332,
"step": 975
},
{
"epoch": 0.9133271202236719,
"grad_norm": 0.20908939058088777,
"learning_rate": 1.2076644225789747e-05,
"loss": 0.3231,
"step": 980
},
{
"epoch": 0.9179869524697111,
"grad_norm": 0.21029655462480856,
"learning_rate": 1.2024857586742623e-05,
"loss": 0.3219,
"step": 985
},
{
"epoch": 0.9226467847157502,
"grad_norm": 0.21845134201809333,
"learning_rate": 1.1973070947695495e-05,
"loss": 0.3203,
"step": 990
},
{
"epoch": 0.9273066169617894,
"grad_norm": 0.20388120839780344,
"learning_rate": 1.1921284308648369e-05,
"loss": 0.3224,
"step": 995
},
{
"epoch": 0.9319664492078286,
"grad_norm": 0.19143754175819644,
"learning_rate": 1.1869497669601245e-05,
"loss": 0.3221,
"step": 1000
},
{
"epoch": 0.9366262814538676,
"grad_norm": 0.19860377602780624,
"learning_rate": 1.1817711030554117e-05,
"loss": 0.326,
"step": 1005
},
{
"epoch": 0.9412861136999068,
"grad_norm": 0.22191428976312808,
"learning_rate": 1.1765924391506993e-05,
"loss": 0.3241,
"step": 1010
},
{
"epoch": 0.9459459459459459,
"grad_norm": 0.2264899602702503,
"learning_rate": 1.1714137752459867e-05,
"loss": 0.328,
"step": 1015
},
{
"epoch": 0.9506057781919851,
"grad_norm": 0.19315587082731886,
"learning_rate": 1.166235111341274e-05,
"loss": 0.3229,
"step": 1020
},
{
"epoch": 0.9552656104380243,
"grad_norm": 0.19951522522778153,
"learning_rate": 1.1610564474365615e-05,
"loss": 0.3311,
"step": 1025
},
{
"epoch": 0.9599254426840633,
"grad_norm": 0.18858346428525566,
"learning_rate": 1.1558777835318489e-05,
"loss": 0.3249,
"step": 1030
},
{
"epoch": 0.9645852749301025,
"grad_norm": 0.19792425530271554,
"learning_rate": 1.1506991196271361e-05,
"loss": 0.3259,
"step": 1035
},
{
"epoch": 0.9692451071761417,
"grad_norm": 0.20030469631792833,
"learning_rate": 1.1455204557224237e-05,
"loss": 0.3226,
"step": 1040
},
{
"epoch": 0.9739049394221808,
"grad_norm": 0.20770307462171755,
"learning_rate": 1.1403417918177111e-05,
"loss": 0.3171,
"step": 1045
},
{
"epoch": 0.97856477166822,
"grad_norm": 0.28646869241469997,
"learning_rate": 1.1351631279129983e-05,
"loss": 0.3392,
"step": 1050
},
{
"epoch": 0.983224603914259,
"grad_norm": 0.21788488981566065,
"learning_rate": 1.129984464008286e-05,
"loss": 0.3191,
"step": 1055
},
{
"epoch": 0.9878844361602982,
"grad_norm": 0.20606943186110416,
"learning_rate": 1.1248058001035733e-05,
"loss": 0.3217,
"step": 1060
},
{
"epoch": 0.9925442684063374,
"grad_norm": 0.21528787214750336,
"learning_rate": 1.1196271361988609e-05,
"loss": 0.325,
"step": 1065
},
{
"epoch": 0.9972041006523765,
"grad_norm": 0.21376540320281065,
"learning_rate": 1.1144484722941481e-05,
"loss": 0.3225,
"step": 1070
},
{
"epoch": 1.0018639328984156,
"grad_norm": 0.2259025613794045,
"learning_rate": 1.1092698083894355e-05,
"loss": 0.2983,
"step": 1075
},
{
"epoch": 1.0065237651444547,
"grad_norm": 0.2195605783591442,
"learning_rate": 1.1040911444847231e-05,
"loss": 0.2827,
"step": 1080
},
{
"epoch": 1.011183597390494,
"grad_norm": 0.23344991666250092,
"learning_rate": 1.0989124805800104e-05,
"loss": 0.2878,
"step": 1085
},
{
"epoch": 1.015843429636533,
"grad_norm": 0.2134504165256278,
"learning_rate": 1.0937338166752978e-05,
"loss": 0.2924,
"step": 1090
},
{
"epoch": 1.0205032618825722,
"grad_norm": 0.20631302270411553,
"learning_rate": 1.0885551527705853e-05,
"loss": 0.2823,
"step": 1095
},
{
"epoch": 1.0251630941286114,
"grad_norm": 0.19643085882318784,
"learning_rate": 1.0833764888658726e-05,
"loss": 0.2736,
"step": 1100
},
{
"epoch": 1.0298229263746506,
"grad_norm": 0.22146197945390989,
"learning_rate": 1.0781978249611601e-05,
"loss": 0.2815,
"step": 1105
},
{
"epoch": 1.0344827586206897,
"grad_norm": 0.18822022387097448,
"learning_rate": 1.0730191610564476e-05,
"loss": 0.2798,
"step": 1110
},
{
"epoch": 1.0391425908667289,
"grad_norm": 0.2048702940946389,
"learning_rate": 1.067840497151735e-05,
"loss": 0.2827,
"step": 1115
},
{
"epoch": 1.0438024231127678,
"grad_norm": 0.21862482584858495,
"learning_rate": 1.0626618332470224e-05,
"loss": 0.2774,
"step": 1120
},
{
"epoch": 1.048462255358807,
"grad_norm": 0.17556754940127575,
"learning_rate": 1.0574831693423098e-05,
"loss": 0.281,
"step": 1125
},
{
"epoch": 1.0531220876048462,
"grad_norm": 0.17905762828840907,
"learning_rate": 1.0523045054375972e-05,
"loss": 0.2827,
"step": 1130
},
{
"epoch": 1.0577819198508853,
"grad_norm": 0.19929098240833457,
"learning_rate": 1.0471258415328846e-05,
"loss": 0.2908,
"step": 1135
},
{
"epoch": 1.0624417520969245,
"grad_norm": 0.1911168370013747,
"learning_rate": 1.041947177628172e-05,
"loss": 0.2865,
"step": 1140
},
{
"epoch": 1.0671015843429636,
"grad_norm": 0.19469722744535697,
"learning_rate": 1.0367685137234594e-05,
"loss": 0.281,
"step": 1145
},
{
"epoch": 1.0717614165890028,
"grad_norm": 0.1915246964885849,
"learning_rate": 1.031589849818747e-05,
"loss": 0.2738,
"step": 1150
},
{
"epoch": 1.076421248835042,
"grad_norm": 0.20821547972324014,
"learning_rate": 1.0264111859140342e-05,
"loss": 0.2797,
"step": 1155
},
{
"epoch": 1.0810810810810811,
"grad_norm": 0.1784617175285205,
"learning_rate": 1.0212325220093218e-05,
"loss": 0.2804,
"step": 1160
},
{
"epoch": 1.0857409133271203,
"grad_norm": 0.21890203338152653,
"learning_rate": 1.0160538581046092e-05,
"loss": 0.2829,
"step": 1165
},
{
"epoch": 1.0904007455731595,
"grad_norm": 0.19918388826900157,
"learning_rate": 1.0108751941998964e-05,
"loss": 0.289,
"step": 1170
},
{
"epoch": 1.0950605778191984,
"grad_norm": 0.19026021585624292,
"learning_rate": 1.005696530295184e-05,
"loss": 0.2805,
"step": 1175
},
{
"epoch": 1.0997204100652376,
"grad_norm": 0.21456689205757995,
"learning_rate": 1.0005178663904714e-05,
"loss": 0.2838,
"step": 1180
},
{
"epoch": 1.1043802423112767,
"grad_norm": 0.20700020922949647,
"learning_rate": 9.953392024857588e-06,
"loss": 0.2802,
"step": 1185
},
{
"epoch": 1.109040074557316,
"grad_norm": 0.18236027964263013,
"learning_rate": 9.901605385810462e-06,
"loss": 0.2852,
"step": 1190
},
{
"epoch": 1.113699906803355,
"grad_norm": 0.21077637095731513,
"learning_rate": 9.849818746763336e-06,
"loss": 0.2896,
"step": 1195
},
{
"epoch": 1.1183597390493942,
"grad_norm": 0.18470733903435116,
"learning_rate": 9.79803210771621e-06,
"loss": 0.2844,
"step": 1200
},
{
"epoch": 1.1230195712954334,
"grad_norm": 0.20723918029516097,
"learning_rate": 9.746245468669084e-06,
"loss": 0.2851,
"step": 1205
},
{
"epoch": 1.1276794035414726,
"grad_norm": 0.1898177888255008,
"learning_rate": 9.694458829621958e-06,
"loss": 0.2854,
"step": 1210
},
{
"epoch": 1.1323392357875117,
"grad_norm": 0.18458308742885182,
"learning_rate": 9.642672190574832e-06,
"loss": 0.2833,
"step": 1215
},
{
"epoch": 1.1369990680335509,
"grad_norm": 0.19195564599792803,
"learning_rate": 9.590885551527706e-06,
"loss": 0.2842,
"step": 1220
},
{
"epoch": 1.14165890027959,
"grad_norm": 0.2042846743518091,
"learning_rate": 9.53909891248058e-06,
"loss": 0.2824,
"step": 1225
},
{
"epoch": 1.146318732525629,
"grad_norm": 0.19117917471615137,
"learning_rate": 9.487312273433454e-06,
"loss": 0.2844,
"step": 1230
},
{
"epoch": 1.1509785647716682,
"grad_norm": 0.1889691345807206,
"learning_rate": 9.435525634386328e-06,
"loss": 0.2881,
"step": 1235
},
{
"epoch": 1.1556383970177073,
"grad_norm": 0.1859542475312062,
"learning_rate": 9.383738995339203e-06,
"loss": 0.2802,
"step": 1240
},
{
"epoch": 1.1602982292637465,
"grad_norm": 0.18878674736359274,
"learning_rate": 9.331952356292078e-06,
"loss": 0.287,
"step": 1245
},
{
"epoch": 1.1649580615097856,
"grad_norm": 0.18804677940623368,
"learning_rate": 9.28016571724495e-06,
"loss": 0.2828,
"step": 1250
},
{
"epoch": 1.1696178937558248,
"grad_norm": 0.18920448811635315,
"learning_rate": 9.228379078197825e-06,
"loss": 0.2688,
"step": 1255
},
{
"epoch": 1.174277726001864,
"grad_norm": 0.18953887671939224,
"learning_rate": 9.1765924391507e-06,
"loss": 0.2848,
"step": 1260
},
{
"epoch": 1.1789375582479031,
"grad_norm": 0.20452563159049228,
"learning_rate": 9.124805800103574e-06,
"loss": 0.2807,
"step": 1265
},
{
"epoch": 1.1835973904939423,
"grad_norm": 0.19366714572234173,
"learning_rate": 9.073019161056449e-06,
"loss": 0.2905,
"step": 1270
},
{
"epoch": 1.1882572227399812,
"grad_norm": 0.18697810604500456,
"learning_rate": 9.021232522009323e-06,
"loss": 0.2843,
"step": 1275
},
{
"epoch": 1.1929170549860204,
"grad_norm": 0.20123824295274137,
"learning_rate": 8.969445882962197e-06,
"loss": 0.2857,
"step": 1280
},
{
"epoch": 1.1975768872320596,
"grad_norm": 0.1915692749101002,
"learning_rate": 8.91765924391507e-06,
"loss": 0.2898,
"step": 1285
},
{
"epoch": 1.2022367194780987,
"grad_norm": 0.18916488913676208,
"learning_rate": 8.865872604867945e-06,
"loss": 0.2804,
"step": 1290
},
{
"epoch": 1.206896551724138,
"grad_norm": 0.18617028635469698,
"learning_rate": 8.814085965820819e-06,
"loss": 0.2817,
"step": 1295
},
{
"epoch": 1.211556383970177,
"grad_norm": 0.22609923703108467,
"learning_rate": 8.762299326773693e-06,
"loss": 0.2905,
"step": 1300
},
{
"epoch": 1.2162162162162162,
"grad_norm": 0.2126857111729568,
"learning_rate": 8.710512687726569e-06,
"loss": 0.2879,
"step": 1305
},
{
"epoch": 1.2208760484622554,
"grad_norm": 0.18126697001357067,
"learning_rate": 8.658726048679441e-06,
"loss": 0.2777,
"step": 1310
},
{
"epoch": 1.2255358807082946,
"grad_norm": 0.19570909476945747,
"learning_rate": 8.606939409632315e-06,
"loss": 0.2941,
"step": 1315
},
{
"epoch": 1.2301957129543337,
"grad_norm": 0.19971390800204433,
"learning_rate": 8.55515277058519e-06,
"loss": 0.285,
"step": 1320
},
{
"epoch": 1.2348555452003729,
"grad_norm": 0.18248980133165246,
"learning_rate": 8.503366131538063e-06,
"loss": 0.2888,
"step": 1325
},
{
"epoch": 1.2395153774464118,
"grad_norm": 0.18297875521481516,
"learning_rate": 8.451579492490937e-06,
"loss": 0.2829,
"step": 1330
},
{
"epoch": 1.244175209692451,
"grad_norm": 0.19520992573003834,
"learning_rate": 8.399792853443813e-06,
"loss": 0.2826,
"step": 1335
},
{
"epoch": 1.2488350419384902,
"grad_norm": 0.19384120322676446,
"learning_rate": 8.348006214396687e-06,
"loss": 0.2868,
"step": 1340
},
{
"epoch": 1.2534948741845293,
"grad_norm": 0.1783920464815875,
"learning_rate": 8.29621957534956e-06,
"loss": 0.2751,
"step": 1345
},
{
"epoch": 1.2581547064305685,
"grad_norm": 0.2021231519848707,
"learning_rate": 8.244432936302435e-06,
"loss": 0.2821,
"step": 1350
},
{
"epoch": 1.2628145386766076,
"grad_norm": 0.19093463948979375,
"learning_rate": 8.192646297255309e-06,
"loss": 0.2938,
"step": 1355
},
{
"epoch": 1.2674743709226468,
"grad_norm": 0.199915448813584,
"learning_rate": 8.140859658208183e-06,
"loss": 0.2858,
"step": 1360
},
{
"epoch": 1.272134203168686,
"grad_norm": 0.17486575781096428,
"learning_rate": 8.089073019161057e-06,
"loss": 0.2837,
"step": 1365
},
{
"epoch": 1.2767940354147251,
"grad_norm": 0.20131105656692982,
"learning_rate": 8.037286380113931e-06,
"loss": 0.2827,
"step": 1370
},
{
"epoch": 1.281453867660764,
"grad_norm": 0.19114388849396574,
"learning_rate": 7.985499741066805e-06,
"loss": 0.2764,
"step": 1375
},
{
"epoch": 1.2861136999068035,
"grad_norm": 0.18459191813016385,
"learning_rate": 7.93371310201968e-06,
"loss": 0.2869,
"step": 1380
},
{
"epoch": 1.2907735321528424,
"grad_norm": 0.18749447304277989,
"learning_rate": 7.881926462972553e-06,
"loss": 0.2892,
"step": 1385
},
{
"epoch": 1.2954333643988816,
"grad_norm": 0.19126389742101615,
"learning_rate": 7.830139823925427e-06,
"loss": 0.2821,
"step": 1390
},
{
"epoch": 1.3000931966449207,
"grad_norm": 0.18893656425511854,
"learning_rate": 7.778353184878302e-06,
"loss": 0.288,
"step": 1395
},
{
"epoch": 1.30475302889096,
"grad_norm": 0.1749441012261624,
"learning_rate": 7.726566545831177e-06,
"loss": 0.2782,
"step": 1400
},
{
"epoch": 1.309412861136999,
"grad_norm": 0.17984937281141697,
"learning_rate": 7.67477990678405e-06,
"loss": 0.2877,
"step": 1405
},
{
"epoch": 1.3140726933830382,
"grad_norm": 0.18676954296910422,
"learning_rate": 7.6229932677369245e-06,
"loss": 0.2828,
"step": 1410
},
{
"epoch": 1.3187325256290774,
"grad_norm": 0.18513926663515104,
"learning_rate": 7.571206628689799e-06,
"loss": 0.2896,
"step": 1415
},
{
"epoch": 1.3233923578751166,
"grad_norm": 0.18724138325257705,
"learning_rate": 7.519419989642673e-06,
"loss": 0.2734,
"step": 1420
},
{
"epoch": 1.3280521901211557,
"grad_norm": 0.18319826116355747,
"learning_rate": 7.467633350595547e-06,
"loss": 0.2851,
"step": 1425
},
{
"epoch": 1.3327120223671947,
"grad_norm": 0.18428323561659832,
"learning_rate": 7.415846711548422e-06,
"loss": 0.2874,
"step": 1430
},
{
"epoch": 1.337371854613234,
"grad_norm": 0.17000821424536014,
"learning_rate": 7.364060072501296e-06,
"loss": 0.2909,
"step": 1435
},
{
"epoch": 1.342031686859273,
"grad_norm": 0.1841048828984189,
"learning_rate": 7.312273433454169e-06,
"loss": 0.2861,
"step": 1440
},
{
"epoch": 1.3466915191053122,
"grad_norm": 0.17345256964148933,
"learning_rate": 7.260486794407044e-06,
"loss": 0.2882,
"step": 1445
},
{
"epoch": 1.3513513513513513,
"grad_norm": 0.18258453132832908,
"learning_rate": 7.208700155359918e-06,
"loss": 0.2806,
"step": 1450
},
{
"epoch": 1.3560111835973905,
"grad_norm": 0.17378696156161202,
"learning_rate": 7.156913516312793e-06,
"loss": 0.2841,
"step": 1455
},
{
"epoch": 1.3606710158434296,
"grad_norm": 0.17509727665850616,
"learning_rate": 7.105126877265666e-06,
"loss": 0.2921,
"step": 1460
},
{
"epoch": 1.3653308480894688,
"grad_norm": 0.20267254775086155,
"learning_rate": 7.05334023821854e-06,
"loss": 0.2861,
"step": 1465
},
{
"epoch": 1.369990680335508,
"grad_norm": 0.1826780882718024,
"learning_rate": 7.001553599171415e-06,
"loss": 0.2835,
"step": 1470
},
{
"epoch": 1.3746505125815471,
"grad_norm": 0.17275106110892652,
"learning_rate": 6.949766960124289e-06,
"loss": 0.2822,
"step": 1475
},
{
"epoch": 1.3793103448275863,
"grad_norm": 0.1743364966907928,
"learning_rate": 6.897980321077162e-06,
"loss": 0.2717,
"step": 1480
},
{
"epoch": 1.3839701770736252,
"grad_norm": 0.17305807049397479,
"learning_rate": 6.846193682030037e-06,
"loss": 0.2828,
"step": 1485
},
{
"epoch": 1.3886300093196646,
"grad_norm": 0.18234220662171363,
"learning_rate": 6.794407042982911e-06,
"loss": 0.2928,
"step": 1490
},
{
"epoch": 1.3932898415657036,
"grad_norm": 0.19901527937775287,
"learning_rate": 6.742620403935786e-06,
"loss": 0.2963,
"step": 1495
},
{
"epoch": 1.3979496738117427,
"grad_norm": 0.20010097090257092,
"learning_rate": 6.690833764888659e-06,
"loss": 0.2918,
"step": 1500
},
{
"epoch": 1.402609506057782,
"grad_norm": 0.20029238519412934,
"learning_rate": 6.639047125841533e-06,
"loss": 0.2855,
"step": 1505
},
{
"epoch": 1.407269338303821,
"grad_norm": 0.1936425086687196,
"learning_rate": 6.587260486794408e-06,
"loss": 0.2785,
"step": 1510
},
{
"epoch": 1.4119291705498602,
"grad_norm": 0.1991532365434793,
"learning_rate": 6.535473847747281e-06,
"loss": 0.2881,
"step": 1515
},
{
"epoch": 1.4165890027958994,
"grad_norm": 0.17075404785787288,
"learning_rate": 6.483687208700155e-06,
"loss": 0.2875,
"step": 1520
},
{
"epoch": 1.4212488350419386,
"grad_norm": 0.1840441188498844,
"learning_rate": 6.43190056965303e-06,
"loss": 0.2846,
"step": 1525
},
{
"epoch": 1.4259086672879777,
"grad_norm": 0.1812038546431852,
"learning_rate": 6.380113930605904e-06,
"loss": 0.2799,
"step": 1530
},
{
"epoch": 1.4305684995340169,
"grad_norm": 0.19072904209724684,
"learning_rate": 6.3283272915587775e-06,
"loss": 0.2847,
"step": 1535
},
{
"epoch": 1.4352283317800558,
"grad_norm": 0.1704872579719148,
"learning_rate": 6.276540652511652e-06,
"loss": 0.2825,
"step": 1540
},
{
"epoch": 1.439888164026095,
"grad_norm": 0.19536704137537816,
"learning_rate": 6.2247540134645265e-06,
"loss": 0.2838,
"step": 1545
},
{
"epoch": 1.4445479962721341,
"grad_norm": 0.17789121544016315,
"learning_rate": 6.172967374417401e-06,
"loss": 0.2847,
"step": 1550
},
{
"epoch": 1.4492078285181733,
"grad_norm": 0.19130871528617927,
"learning_rate": 6.1211807353702746e-06,
"loss": 0.28,
"step": 1555
},
{
"epoch": 1.4538676607642125,
"grad_norm": 0.17749331170718344,
"learning_rate": 6.069394096323149e-06,
"loss": 0.2817,
"step": 1560
},
{
"epoch": 1.4585274930102516,
"grad_norm": 0.17125040626587648,
"learning_rate": 6.0176074572760235e-06,
"loss": 0.2842,
"step": 1565
},
{
"epoch": 1.4631873252562908,
"grad_norm": 0.1717733923202594,
"learning_rate": 5.9658208182288976e-06,
"loss": 0.2852,
"step": 1570
},
{
"epoch": 1.46784715750233,
"grad_norm": 0.18029835185575757,
"learning_rate": 5.914034179181772e-06,
"loss": 0.2868,
"step": 1575
},
{
"epoch": 1.4725069897483691,
"grad_norm": 0.17321438940936698,
"learning_rate": 5.862247540134646e-06,
"loss": 0.289,
"step": 1580
},
{
"epoch": 1.477166821994408,
"grad_norm": 0.18465932185370307,
"learning_rate": 5.81046090108752e-06,
"loss": 0.2868,
"step": 1585
},
{
"epoch": 1.4818266542404475,
"grad_norm": 0.17130471704458686,
"learning_rate": 5.758674262040395e-06,
"loss": 0.2862,
"step": 1590
},
{
"epoch": 1.4864864864864864,
"grad_norm": 0.18491783556891386,
"learning_rate": 5.706887622993268e-06,
"loss": 0.2823,
"step": 1595
},
{
"epoch": 1.4911463187325256,
"grad_norm": 0.18312451894682794,
"learning_rate": 5.655100983946143e-06,
"loss": 0.2834,
"step": 1600
},
{
"epoch": 1.4958061509785647,
"grad_norm": 0.20865879661440087,
"learning_rate": 5.603314344899017e-06,
"loss": 0.2924,
"step": 1605
},
{
"epoch": 1.500465983224604,
"grad_norm": 0.19218279954004253,
"learning_rate": 5.55152770585189e-06,
"loss": 0.2881,
"step": 1610
},
{
"epoch": 1.505125815470643,
"grad_norm": 0.20044041867998658,
"learning_rate": 5.499741066804765e-06,
"loss": 0.2833,
"step": 1615
},
{
"epoch": 1.5097856477166822,
"grad_norm": 0.17165212019189693,
"learning_rate": 5.447954427757639e-06,
"loss": 0.2831,
"step": 1620
},
{
"epoch": 1.5144454799627214,
"grad_norm": 0.16485996786783458,
"learning_rate": 5.396167788710514e-06,
"loss": 0.2785,
"step": 1625
},
{
"epoch": 1.5191053122087603,
"grad_norm": 0.1869210240796828,
"learning_rate": 5.344381149663387e-06,
"loss": 0.2797,
"step": 1630
},
{
"epoch": 1.5237651444547997,
"grad_norm": 0.19194565166760805,
"learning_rate": 5.292594510616261e-06,
"loss": 0.2784,
"step": 1635
},
{
"epoch": 1.5284249767008387,
"grad_norm": 0.1863087676239078,
"learning_rate": 5.240807871569136e-06,
"loss": 0.2758,
"step": 1640
},
{
"epoch": 1.533084808946878,
"grad_norm": 0.17516911308691216,
"learning_rate": 5.18902123252201e-06,
"loss": 0.2922,
"step": 1645
},
{
"epoch": 1.537744641192917,
"grad_norm": 0.19150134236928595,
"learning_rate": 5.137234593474883e-06,
"loss": 0.288,
"step": 1650
},
{
"epoch": 1.5424044734389561,
"grad_norm": 0.17534568446933935,
"learning_rate": 5.085447954427758e-06,
"loss": 0.2904,
"step": 1655
},
{
"epoch": 1.5470643056849953,
"grad_norm": 0.18924120071576864,
"learning_rate": 5.033661315380632e-06,
"loss": 0.293,
"step": 1660
},
{
"epoch": 1.5517241379310345,
"grad_norm": 0.19954078997223346,
"learning_rate": 4.981874676333506e-06,
"loss": 0.2797,
"step": 1665
},
{
"epoch": 1.5563839701770736,
"grad_norm": 0.18388310486629303,
"learning_rate": 4.930088037286381e-06,
"loss": 0.2875,
"step": 1670
},
{
"epoch": 1.5610438024231128,
"grad_norm": 0.19490866092310646,
"learning_rate": 4.878301398239254e-06,
"loss": 0.2931,
"step": 1675
},
{
"epoch": 1.565703634669152,
"grad_norm": 0.19751354733303017,
"learning_rate": 4.826514759192129e-06,
"loss": 0.289,
"step": 1680
},
{
"epoch": 1.570363466915191,
"grad_norm": 0.18733868260649666,
"learning_rate": 4.774728120145003e-06,
"loss": 0.2785,
"step": 1685
},
{
"epoch": 1.5750232991612303,
"grad_norm": 0.19221116473809075,
"learning_rate": 4.7229414810978765e-06,
"loss": 0.2925,
"step": 1690
},
{
"epoch": 1.5796831314072692,
"grad_norm": 0.18085900113347148,
"learning_rate": 4.671154842050751e-06,
"loss": 0.2911,
"step": 1695
},
{
"epoch": 1.5843429636533086,
"grad_norm": 0.1973140421662704,
"learning_rate": 4.6193682030036255e-06,
"loss": 0.2803,
"step": 1700
},
{
"epoch": 1.5890027958993476,
"grad_norm": 0.18709409447385142,
"learning_rate": 4.5675815639564995e-06,
"loss": 0.2845,
"step": 1705
},
{
"epoch": 1.5936626281453867,
"grad_norm": 0.17665797548222434,
"learning_rate": 4.5157949249093736e-06,
"loss": 0.2869,
"step": 1710
},
{
"epoch": 1.598322460391426,
"grad_norm": 0.17996538581217197,
"learning_rate": 4.464008285862248e-06,
"loss": 0.2795,
"step": 1715
},
{
"epoch": 1.602982292637465,
"grad_norm": 0.19445653575092758,
"learning_rate": 4.412221646815122e-06,
"loss": 0.2885,
"step": 1720
},
{
"epoch": 1.6076421248835042,
"grad_norm": 0.17825241400226646,
"learning_rate": 4.3604350077679966e-06,
"loss": 0.2855,
"step": 1725
},
{
"epoch": 1.6123019571295434,
"grad_norm": 0.18474656142781384,
"learning_rate": 4.308648368720871e-06,
"loss": 0.282,
"step": 1730
},
{
"epoch": 1.6169617893755825,
"grad_norm": 0.1950234628172142,
"learning_rate": 4.256861729673745e-06,
"loss": 0.2872,
"step": 1735
},
{
"epoch": 1.6216216216216215,
"grad_norm": 0.18770953389353345,
"learning_rate": 4.205075090626619e-06,
"loss": 0.2867,
"step": 1740
},
{
"epoch": 1.6262814538676609,
"grad_norm": 0.178276907246367,
"learning_rate": 4.153288451579493e-06,
"loss": 0.2837,
"step": 1745
},
{
"epoch": 1.6309412861136998,
"grad_norm": 0.20173726657394422,
"learning_rate": 4.101501812532367e-06,
"loss": 0.2878,
"step": 1750
},
{
"epoch": 1.6356011183597392,
"grad_norm": 0.18457623482800892,
"learning_rate": 4.049715173485242e-06,
"loss": 0.2849,
"step": 1755
},
{
"epoch": 1.6402609506057781,
"grad_norm": 0.19451630359937264,
"learning_rate": 3.997928534438115e-06,
"loss": 0.2911,
"step": 1760
},
{
"epoch": 1.6449207828518173,
"grad_norm": 0.182302636059145,
"learning_rate": 3.94614189539099e-06,
"loss": 0.2873,
"step": 1765
},
{
"epoch": 1.6495806150978565,
"grad_norm": 0.17227161525874077,
"learning_rate": 3.894355256343864e-06,
"loss": 0.2813,
"step": 1770
},
{
"epoch": 1.6542404473438956,
"grad_norm": 0.18388975389617138,
"learning_rate": 3.842568617296738e-06,
"loss": 0.2778,
"step": 1775
},
{
"epoch": 1.6589002795899348,
"grad_norm": 0.18763197465021858,
"learning_rate": 3.790781978249612e-06,
"loss": 0.2874,
"step": 1780
},
{
"epoch": 1.6635601118359737,
"grad_norm": 0.18648811108535399,
"learning_rate": 3.7389953392024856e-06,
"loss": 0.287,
"step": 1785
},
{
"epoch": 1.6682199440820131,
"grad_norm": 0.16990216601005864,
"learning_rate": 3.68720870015536e-06,
"loss": 0.2844,
"step": 1790
},
{
"epoch": 1.672879776328052,
"grad_norm": 0.19315242332028645,
"learning_rate": 3.635422061108234e-06,
"loss": 0.2778,
"step": 1795
},
{
"epoch": 1.6775396085740915,
"grad_norm": 0.18107706144773053,
"learning_rate": 3.5836354220611086e-06,
"loss": 0.2746,
"step": 1800
},
{
"epoch": 1.6821994408201304,
"grad_norm": 0.17626787448033834,
"learning_rate": 3.5318487830139827e-06,
"loss": 0.2759,
"step": 1805
},
{
"epoch": 1.6868592730661698,
"grad_norm": 0.17699842425781145,
"learning_rate": 3.480062143966857e-06,
"loss": 0.2812,
"step": 1810
},
{
"epoch": 1.6915191053122087,
"grad_norm": 0.1755900253752614,
"learning_rate": 3.4282755049197308e-06,
"loss": 0.2798,
"step": 1815
},
{
"epoch": 1.696178937558248,
"grad_norm": 0.16931795318247234,
"learning_rate": 3.3764888658726052e-06,
"loss": 0.2795,
"step": 1820
},
{
"epoch": 1.700838769804287,
"grad_norm": 0.20489778400137415,
"learning_rate": 3.3247022268254793e-06,
"loss": 0.2908,
"step": 1825
},
{
"epoch": 1.7054986020503262,
"grad_norm": 0.17146060928109105,
"learning_rate": 3.2729155877783538e-06,
"loss": 0.2874,
"step": 1830
},
{
"epoch": 1.7101584342963654,
"grad_norm": 0.20743117222876825,
"learning_rate": 3.2211289487312274e-06,
"loss": 0.2899,
"step": 1835
},
{
"epoch": 1.7148182665424043,
"grad_norm": 0.17532744243024526,
"learning_rate": 3.169342309684102e-06,
"loss": 0.2835,
"step": 1840
},
{
"epoch": 1.7194780987884437,
"grad_norm": 0.17155149443847173,
"learning_rate": 3.117555670636976e-06,
"loss": 0.2843,
"step": 1845
},
{
"epoch": 1.7241379310344827,
"grad_norm": 0.17510709024655646,
"learning_rate": 3.0657690315898504e-06,
"loss": 0.2709,
"step": 1850
},
{
"epoch": 1.728797763280522,
"grad_norm": 0.1936769403956301,
"learning_rate": 3.013982392542724e-06,
"loss": 0.2795,
"step": 1855
},
{
"epoch": 1.733457595526561,
"grad_norm": 0.18362587824643453,
"learning_rate": 2.9621957534955985e-06,
"loss": 0.2793,
"step": 1860
},
{
"epoch": 1.7381174277726001,
"grad_norm": 0.1733400456233926,
"learning_rate": 2.9104091144484725e-06,
"loss": 0.2833,
"step": 1865
},
{
"epoch": 1.7427772600186393,
"grad_norm": 0.1707366786490628,
"learning_rate": 2.858622475401347e-06,
"loss": 0.2787,
"step": 1870
},
{
"epoch": 1.7474370922646785,
"grad_norm": 0.1621088673034961,
"learning_rate": 2.8068358363542206e-06,
"loss": 0.2863,
"step": 1875
},
{
"epoch": 1.7520969245107176,
"grad_norm": 0.18628674415468172,
"learning_rate": 2.7550491973070947e-06,
"loss": 0.2831,
"step": 1880
},
{
"epoch": 1.7567567567567568,
"grad_norm": 0.17521662090290133,
"learning_rate": 2.703262558259969e-06,
"loss": 0.2888,
"step": 1885
},
{
"epoch": 1.761416589002796,
"grad_norm": 0.17377765263252967,
"learning_rate": 2.6514759192128432e-06,
"loss": 0.2818,
"step": 1890
},
{
"epoch": 1.766076421248835,
"grad_norm": 0.17845993065674873,
"learning_rate": 2.5996892801657177e-06,
"loss": 0.2818,
"step": 1895
},
{
"epoch": 1.7707362534948743,
"grad_norm": 0.17969442761482135,
"learning_rate": 2.5479026411185913e-06,
"loss": 0.2778,
"step": 1900
},
{
"epoch": 1.7753960857409132,
"grad_norm": 0.1694996391077242,
"learning_rate": 2.496116002071466e-06,
"loss": 0.2781,
"step": 1905
},
{
"epoch": 1.7800559179869526,
"grad_norm": 0.1657595379031282,
"learning_rate": 2.44432936302434e-06,
"loss": 0.2838,
"step": 1910
},
{
"epoch": 1.7847157502329916,
"grad_norm": 0.17428612808942404,
"learning_rate": 2.3925427239772143e-06,
"loss": 0.2724,
"step": 1915
},
{
"epoch": 1.7893755824790307,
"grad_norm": 0.17198875130716848,
"learning_rate": 2.340756084930088e-06,
"loss": 0.2878,
"step": 1920
},
{
"epoch": 1.7940354147250699,
"grad_norm": 0.16955861871882308,
"learning_rate": 2.2889694458829624e-06,
"loss": 0.2807,
"step": 1925
},
{
"epoch": 1.798695246971109,
"grad_norm": 0.17174832677033502,
"learning_rate": 2.2371828068358365e-06,
"loss": 0.2786,
"step": 1930
},
{
"epoch": 1.8033550792171482,
"grad_norm": 0.1751748502595037,
"learning_rate": 2.1853961677887105e-06,
"loss": 0.2843,
"step": 1935
},
{
"epoch": 1.8080149114631874,
"grad_norm": 0.1600435428137458,
"learning_rate": 2.1336095287415846e-06,
"loss": 0.2782,
"step": 1940
},
{
"epoch": 1.8126747437092265,
"grad_norm": 0.16951263851585602,
"learning_rate": 2.081822889694459e-06,
"loss": 0.2897,
"step": 1945
},
{
"epoch": 1.8173345759552655,
"grad_norm": 0.17043335019815078,
"learning_rate": 2.030036250647333e-06,
"loss": 0.276,
"step": 1950
},
{
"epoch": 1.8219944082013049,
"grad_norm": 0.18796171228162317,
"learning_rate": 1.978249611600207e-06,
"loss": 0.2821,
"step": 1955
},
{
"epoch": 1.8266542404473438,
"grad_norm": 0.16512159795603745,
"learning_rate": 1.9264629725530816e-06,
"loss": 0.284,
"step": 1960
},
{
"epoch": 1.8313140726933832,
"grad_norm": 0.17527046387428025,
"learning_rate": 1.8746763335059557e-06,
"loss": 0.2845,
"step": 1965
},
{
"epoch": 1.8359739049394221,
"grad_norm": 0.17075350188159297,
"learning_rate": 1.8228896944588297e-06,
"loss": 0.2811,
"step": 1970
},
{
"epoch": 1.8406337371854613,
"grad_norm": 0.1611238780162702,
"learning_rate": 1.771103055411704e-06,
"loss": 0.2874,
"step": 1975
},
{
"epoch": 1.8452935694315005,
"grad_norm": 0.16686965231642037,
"learning_rate": 1.719316416364578e-06,
"loss": 0.2808,
"step": 1980
},
{
"epoch": 1.8499534016775396,
"grad_norm": 0.18363600453053663,
"learning_rate": 1.6675297773174523e-06,
"loss": 0.2757,
"step": 1985
},
{
"epoch": 1.8546132339235788,
"grad_norm": 0.17346986089912622,
"learning_rate": 1.6157431382703264e-06,
"loss": 0.296,
"step": 1990
},
{
"epoch": 1.8592730661696177,
"grad_norm": 0.16302686401969246,
"learning_rate": 1.5639564992232006e-06,
"loss": 0.2831,
"step": 1995
},
{
"epoch": 1.8639328984156571,
"grad_norm": 0.17714019680547557,
"learning_rate": 1.512169860176075e-06,
"loss": 0.2801,
"step": 2000
},
{
"epoch": 1.868592730661696,
"grad_norm": 0.17573560819468692,
"learning_rate": 1.460383221128949e-06,
"loss": 0.2844,
"step": 2005
},
{
"epoch": 1.8732525629077355,
"grad_norm": 0.16964855694079678,
"learning_rate": 1.4085965820818228e-06,
"loss": 0.2854,
"step": 2010
},
{
"epoch": 1.8779123951537744,
"grad_norm": 0.1747943844382359,
"learning_rate": 1.356809943034697e-06,
"loss": 0.2772,
"step": 2015
},
{
"epoch": 1.8825722273998136,
"grad_norm": 0.17766856004277642,
"learning_rate": 1.305023303987571e-06,
"loss": 0.2765,
"step": 2020
},
{
"epoch": 1.8872320596458527,
"grad_norm": 0.18210315847664407,
"learning_rate": 1.2532366649404454e-06,
"loss": 0.2801,
"step": 2025
},
{
"epoch": 1.8918918918918919,
"grad_norm": 0.1702245037263823,
"learning_rate": 1.2014500258933196e-06,
"loss": 0.2804,
"step": 2030
},
{
"epoch": 1.896551724137931,
"grad_norm": 0.16687796494139084,
"learning_rate": 1.1496633868461937e-06,
"loss": 0.2833,
"step": 2035
},
{
"epoch": 1.9012115563839702,
"grad_norm": 0.16709036530571203,
"learning_rate": 1.097876747799068e-06,
"loss": 0.2828,
"step": 2040
},
{
"epoch": 1.9058713886300094,
"grad_norm": 0.17632942060794365,
"learning_rate": 1.046090108751942e-06,
"loss": 0.2916,
"step": 2045
},
{
"epoch": 1.9105312208760483,
"grad_norm": 0.1659570265982279,
"learning_rate": 9.943034697048163e-07,
"loss": 0.2873,
"step": 2050
},
{
"epoch": 1.9151910531220877,
"grad_norm": 0.16259315421842638,
"learning_rate": 9.425168306576904e-07,
"loss": 0.2772,
"step": 2055
},
{
"epoch": 1.9198508853681266,
"grad_norm": 0.1704776417406957,
"learning_rate": 8.907301916105646e-07,
"loss": 0.2865,
"step": 2060
},
{
"epoch": 1.924510717614166,
"grad_norm": 0.18010747503070362,
"learning_rate": 8.389435525634387e-07,
"loss": 0.2828,
"step": 2065
},
{
"epoch": 1.929170549860205,
"grad_norm": 0.16347013947993091,
"learning_rate": 7.871569135163129e-07,
"loss": 0.2847,
"step": 2070
},
{
"epoch": 1.9338303821062441,
"grad_norm": 0.17518692368495511,
"learning_rate": 7.35370274469187e-07,
"loss": 0.2933,
"step": 2075
},
{
"epoch": 1.9384902143522833,
"grad_norm": 0.16956394794756754,
"learning_rate": 6.835836354220611e-07,
"loss": 0.2788,
"step": 2080
},
{
"epoch": 1.9431500465983225,
"grad_norm": 0.18735277361498398,
"learning_rate": 6.317969963749353e-07,
"loss": 0.2809,
"step": 2085
},
{
"epoch": 1.9478098788443616,
"grad_norm": 0.1730004551971713,
"learning_rate": 5.800103573278094e-07,
"loss": 0.2824,
"step": 2090
},
{
"epoch": 1.9524697110904008,
"grad_norm": 0.1739877181954761,
"learning_rate": 5.282237182806836e-07,
"loss": 0.2709,
"step": 2095
},
{
"epoch": 1.95712954333644,
"grad_norm": 0.18343227110963495,
"learning_rate": 4.764370792335578e-07,
"loss": 0.2855,
"step": 2100
},
{
"epoch": 1.961789375582479,
"grad_norm": 0.1618328590005831,
"learning_rate": 4.2465044018643194e-07,
"loss": 0.2852,
"step": 2105
},
{
"epoch": 1.9664492078285183,
"grad_norm": 0.1627906424832234,
"learning_rate": 3.728638011393061e-07,
"loss": 0.2859,
"step": 2110
},
{
"epoch": 1.9711090400745572,
"grad_norm": 0.16365417745617075,
"learning_rate": 3.210771620921802e-07,
"loss": 0.2891,
"step": 2115
},
{
"epoch": 1.9757688723205966,
"grad_norm": 0.15612849284974395,
"learning_rate": 2.692905230450544e-07,
"loss": 0.2848,
"step": 2120
},
{
"epoch": 1.9804287045666356,
"grad_norm": 0.163025042449839,
"learning_rate": 2.1750388399792857e-07,
"loss": 0.2789,
"step": 2125
},
{
"epoch": 1.9850885368126747,
"grad_norm": 0.16808874057585674,
"learning_rate": 1.657172449508027e-07,
"loss": 0.2836,
"step": 2130
},
{
"epoch": 1.9897483690587139,
"grad_norm": 0.1605707022006799,
"learning_rate": 1.1393060590367686e-07,
"loss": 0.2766,
"step": 2135
},
{
"epoch": 1.994408201304753,
"grad_norm": 0.16194115983227303,
"learning_rate": 6.214396685655102e-08,
"loss": 0.282,
"step": 2140
},
{
"epoch": 1.9990680335507922,
"grad_norm": 0.17713436476094685,
"learning_rate": 1.0357327809425169e-08,
"loss": 0.2862,
"step": 2145
},
{
"epoch": 2.0,
"step": 2146,
"total_flos": 1.8367600516462019e+18,
"train_loss": 0.3072341699830195,
"train_runtime": 72164.3066,
"train_samples_per_second": 0.476,
"train_steps_per_second": 0.03
}
],
"logging_steps": 5,
"max_steps": 2146,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.8367600516462019e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}