CocoRoF's picture
Training in progress, step 10000, checkpoint
45e087b verified
raw
history blame
167 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.7232619112196004,
"eval_steps": 2500,
"global_step": 10000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0007232619112196004,
"grad_norm": 16.890625,
"learning_rate": 9.99994349538668e-07,
"loss": 24.7314,
"step": 10
},
{
"epoch": 0.0014465238224392009,
"grad_norm": 16.96875,
"learning_rate": 9.999886990773362e-07,
"loss": 25.1718,
"step": 20
},
{
"epoch": 0.0021697857336588013,
"grad_norm": 15.609375,
"learning_rate": 9.999830486160042e-07,
"loss": 25.0494,
"step": 30
},
{
"epoch": 0.0028930476448784017,
"grad_norm": 15.890625,
"learning_rate": 9.999773981546724e-07,
"loss": 25.5317,
"step": 40
},
{
"epoch": 0.003616309556098002,
"grad_norm": 16.328125,
"learning_rate": 9.999717476933404e-07,
"loss": 24.9466,
"step": 50
},
{
"epoch": 0.0043395714673176026,
"grad_norm": 16.203125,
"learning_rate": 9.999660972320084e-07,
"loss": 25.3166,
"step": 60
},
{
"epoch": 0.0050628333785372025,
"grad_norm": 14.34375,
"learning_rate": 9.999604467706766e-07,
"loss": 24.963,
"step": 70
},
{
"epoch": 0.005786095289756803,
"grad_norm": 15.9609375,
"learning_rate": 9.999547963093446e-07,
"loss": 25.2256,
"step": 80
},
{
"epoch": 0.006509357200976403,
"grad_norm": 16.34375,
"learning_rate": 9.999491458480128e-07,
"loss": 25.5734,
"step": 90
},
{
"epoch": 0.007232619112196004,
"grad_norm": 16.65625,
"learning_rate": 9.999434953866808e-07,
"loss": 25.0955,
"step": 100
},
{
"epoch": 0.007955881023415604,
"grad_norm": 16.453125,
"learning_rate": 9.99937844925349e-07,
"loss": 25.173,
"step": 110
},
{
"epoch": 0.008679142934635205,
"grad_norm": 15.8125,
"learning_rate": 9.99932194464017e-07,
"loss": 24.9289,
"step": 120
},
{
"epoch": 0.009402404845854806,
"grad_norm": 16.3125,
"learning_rate": 9.99926544002685e-07,
"loss": 25.1639,
"step": 130
},
{
"epoch": 0.010125666757074405,
"grad_norm": 16.65625,
"learning_rate": 9.999208935413532e-07,
"loss": 25.2427,
"step": 140
},
{
"epoch": 0.010848928668294006,
"grad_norm": 16.46875,
"learning_rate": 9.999152430800212e-07,
"loss": 25.4152,
"step": 150
},
{
"epoch": 0.011572190579513607,
"grad_norm": 16.640625,
"learning_rate": 9.999095926186894e-07,
"loss": 25.3004,
"step": 160
},
{
"epoch": 0.012295452490733206,
"grad_norm": 16.1875,
"learning_rate": 9.999039421573574e-07,
"loss": 25.122,
"step": 170
},
{
"epoch": 0.013018714401952807,
"grad_norm": 16.96875,
"learning_rate": 9.998982916960254e-07,
"loss": 25.1076,
"step": 180
},
{
"epoch": 0.013741976313172408,
"grad_norm": 17.03125,
"learning_rate": 9.998926412346936e-07,
"loss": 25.7121,
"step": 190
},
{
"epoch": 0.014465238224392009,
"grad_norm": 17.203125,
"learning_rate": 9.998869907733616e-07,
"loss": 24.9972,
"step": 200
},
{
"epoch": 0.015188500135611608,
"grad_norm": 16.015625,
"learning_rate": 9.998813403120298e-07,
"loss": 25.1666,
"step": 210
},
{
"epoch": 0.01591176204683121,
"grad_norm": 16.328125,
"learning_rate": 9.998756898506978e-07,
"loss": 25.0473,
"step": 220
},
{
"epoch": 0.016635023958050808,
"grad_norm": 17.3125,
"learning_rate": 9.998700393893658e-07,
"loss": 25.1476,
"step": 230
},
{
"epoch": 0.01735828586927041,
"grad_norm": 16.125,
"learning_rate": 9.99864388928034e-07,
"loss": 25.5306,
"step": 240
},
{
"epoch": 0.01808154778049001,
"grad_norm": 15.53125,
"learning_rate": 9.99858738466702e-07,
"loss": 25.268,
"step": 250
},
{
"epoch": 0.018804809691709612,
"grad_norm": 15.9609375,
"learning_rate": 9.998530880053702e-07,
"loss": 25.2015,
"step": 260
},
{
"epoch": 0.01952807160292921,
"grad_norm": 16.3125,
"learning_rate": 9.998474375440382e-07,
"loss": 25.0597,
"step": 270
},
{
"epoch": 0.02025133351414881,
"grad_norm": 15.7578125,
"learning_rate": 9.998417870827062e-07,
"loss": 25.3037,
"step": 280
},
{
"epoch": 0.020974595425368413,
"grad_norm": 15.6484375,
"learning_rate": 9.998361366213744e-07,
"loss": 25.0957,
"step": 290
},
{
"epoch": 0.021697857336588012,
"grad_norm": 16.609375,
"learning_rate": 9.998304861600424e-07,
"loss": 25.1532,
"step": 300
},
{
"epoch": 0.02242111924780761,
"grad_norm": 16.75,
"learning_rate": 9.998248356987106e-07,
"loss": 25.0866,
"step": 310
},
{
"epoch": 0.023144381159027214,
"grad_norm": 16.125,
"learning_rate": 9.998191852373786e-07,
"loss": 25.4455,
"step": 320
},
{
"epoch": 0.023867643070246813,
"grad_norm": 16.140625,
"learning_rate": 9.998135347760468e-07,
"loss": 25.4253,
"step": 330
},
{
"epoch": 0.024590904981466412,
"grad_norm": 16.671875,
"learning_rate": 9.998078843147148e-07,
"loss": 24.9265,
"step": 340
},
{
"epoch": 0.025314166892686014,
"grad_norm": 16.71875,
"learning_rate": 9.998022338533828e-07,
"loss": 25.2683,
"step": 350
},
{
"epoch": 0.026037428803905614,
"grad_norm": 16.703125,
"learning_rate": 9.99796583392051e-07,
"loss": 25.2497,
"step": 360
},
{
"epoch": 0.026760690715125216,
"grad_norm": 15.8515625,
"learning_rate": 9.99790932930719e-07,
"loss": 25.099,
"step": 370
},
{
"epoch": 0.027483952626344815,
"grad_norm": 16.28125,
"learning_rate": 9.997852824693872e-07,
"loss": 25.7413,
"step": 380
},
{
"epoch": 0.028207214537564414,
"grad_norm": 15.7109375,
"learning_rate": 9.997796320080552e-07,
"loss": 24.8872,
"step": 390
},
{
"epoch": 0.028930476448784017,
"grad_norm": 16.421875,
"learning_rate": 9.997739815467232e-07,
"loss": 25.2099,
"step": 400
},
{
"epoch": 0.029653738360003616,
"grad_norm": 16.0,
"learning_rate": 9.997683310853914e-07,
"loss": 25.1945,
"step": 410
},
{
"epoch": 0.030377000271223215,
"grad_norm": 15.390625,
"learning_rate": 9.997626806240594e-07,
"loss": 24.9576,
"step": 420
},
{
"epoch": 0.031100262182442818,
"grad_norm": 15.3359375,
"learning_rate": 9.997570301627277e-07,
"loss": 24.7866,
"step": 430
},
{
"epoch": 0.03182352409366242,
"grad_norm": 16.296875,
"learning_rate": 9.997513797013957e-07,
"loss": 25.3049,
"step": 440
},
{
"epoch": 0.03254678600488202,
"grad_norm": 16.765625,
"learning_rate": 9.997457292400636e-07,
"loss": 25.151,
"step": 450
},
{
"epoch": 0.033270047916101615,
"grad_norm": 15.3515625,
"learning_rate": 9.997400787787319e-07,
"loss": 25.0841,
"step": 460
},
{
"epoch": 0.03399330982732122,
"grad_norm": 16.453125,
"learning_rate": 9.997344283173999e-07,
"loss": 25.2712,
"step": 470
},
{
"epoch": 0.03471657173854082,
"grad_norm": 14.78125,
"learning_rate": 9.99728777856068e-07,
"loss": 24.8674,
"step": 480
},
{
"epoch": 0.035439833649760416,
"grad_norm": 16.171875,
"learning_rate": 9.99723127394736e-07,
"loss": 25.2065,
"step": 490
},
{
"epoch": 0.03616309556098002,
"grad_norm": 16.625,
"learning_rate": 9.99717476933404e-07,
"loss": 25.0951,
"step": 500
},
{
"epoch": 0.03688635747219962,
"grad_norm": 15.8828125,
"learning_rate": 9.997118264720723e-07,
"loss": 25.1455,
"step": 510
},
{
"epoch": 0.037609619383419224,
"grad_norm": 15.6953125,
"learning_rate": 9.997061760107403e-07,
"loss": 25.313,
"step": 520
},
{
"epoch": 0.03833288129463882,
"grad_norm": 16.46875,
"learning_rate": 9.997005255494085e-07,
"loss": 24.8546,
"step": 530
},
{
"epoch": 0.03905614320585842,
"grad_norm": 16.0625,
"learning_rate": 9.996948750880765e-07,
"loss": 25.1006,
"step": 540
},
{
"epoch": 0.039779405117078025,
"grad_norm": 16.34375,
"learning_rate": 9.996892246267447e-07,
"loss": 25.1469,
"step": 550
},
{
"epoch": 0.04050266702829762,
"grad_norm": 15.6328125,
"learning_rate": 9.996835741654127e-07,
"loss": 25.0277,
"step": 560
},
{
"epoch": 0.04122592893951722,
"grad_norm": 15.328125,
"learning_rate": 9.996779237040807e-07,
"loss": 25.2235,
"step": 570
},
{
"epoch": 0.041949190850736826,
"grad_norm": 15.7265625,
"learning_rate": 9.996722732427489e-07,
"loss": 25.3082,
"step": 580
},
{
"epoch": 0.04267245276195642,
"grad_norm": 16.328125,
"learning_rate": 9.996666227814169e-07,
"loss": 25.329,
"step": 590
},
{
"epoch": 0.043395714673176024,
"grad_norm": 17.1875,
"learning_rate": 9.99660972320085e-07,
"loss": 25.3133,
"step": 600
},
{
"epoch": 0.044118976584395626,
"grad_norm": 16.578125,
"learning_rate": 9.99655321858753e-07,
"loss": 25.2397,
"step": 610
},
{
"epoch": 0.04484223849561522,
"grad_norm": 16.46875,
"learning_rate": 9.99649671397421e-07,
"loss": 25.2219,
"step": 620
},
{
"epoch": 0.045565500406834825,
"grad_norm": 16.484375,
"learning_rate": 9.996440209360893e-07,
"loss": 25.2115,
"step": 630
},
{
"epoch": 0.04628876231805443,
"grad_norm": 16.296875,
"learning_rate": 9.996383704747575e-07,
"loss": 24.9097,
"step": 640
},
{
"epoch": 0.04701202422927402,
"grad_norm": 16.171875,
"learning_rate": 9.996327200134255e-07,
"loss": 25.3739,
"step": 650
},
{
"epoch": 0.047735286140493625,
"grad_norm": 16.453125,
"learning_rate": 9.996270695520935e-07,
"loss": 25.4334,
"step": 660
},
{
"epoch": 0.04845854805171323,
"grad_norm": 16.625,
"learning_rate": 9.996214190907615e-07,
"loss": 25.3735,
"step": 670
},
{
"epoch": 0.049181809962932824,
"grad_norm": 16.546875,
"learning_rate": 9.996157686294297e-07,
"loss": 25.2253,
"step": 680
},
{
"epoch": 0.049905071874152426,
"grad_norm": 15.953125,
"learning_rate": 9.99610118168098e-07,
"loss": 25.3197,
"step": 690
},
{
"epoch": 0.05062833378537203,
"grad_norm": 16.046875,
"learning_rate": 9.99604467706766e-07,
"loss": 25.0905,
"step": 700
},
{
"epoch": 0.05135159569659163,
"grad_norm": 16.90625,
"learning_rate": 9.99598817245434e-07,
"loss": 25.5284,
"step": 710
},
{
"epoch": 0.05207485760781123,
"grad_norm": 15.7421875,
"learning_rate": 9.99593166784102e-07,
"loss": 25.1216,
"step": 720
},
{
"epoch": 0.05279811951903083,
"grad_norm": 16.5625,
"learning_rate": 9.995875163227701e-07,
"loss": 25.3238,
"step": 730
},
{
"epoch": 0.05352138143025043,
"grad_norm": 16.046875,
"learning_rate": 9.995818658614383e-07,
"loss": 25.0432,
"step": 740
},
{
"epoch": 0.05424464334147003,
"grad_norm": 16.0625,
"learning_rate": 9.995762154001063e-07,
"loss": 24.9606,
"step": 750
},
{
"epoch": 0.05496790525268963,
"grad_norm": 17.4375,
"learning_rate": 9.995705649387743e-07,
"loss": 25.1405,
"step": 760
},
{
"epoch": 0.05569116716390923,
"grad_norm": 16.3125,
"learning_rate": 9.995649144774425e-07,
"loss": 24.8765,
"step": 770
},
{
"epoch": 0.05641442907512883,
"grad_norm": 17.390625,
"learning_rate": 9.995592640161105e-07,
"loss": 24.962,
"step": 780
},
{
"epoch": 0.05713769098634843,
"grad_norm": 17.109375,
"learning_rate": 9.995536135547785e-07,
"loss": 25.0881,
"step": 790
},
{
"epoch": 0.057860952897568034,
"grad_norm": 16.4375,
"learning_rate": 9.995479630934467e-07,
"loss": 25.4495,
"step": 800
},
{
"epoch": 0.05858421480878763,
"grad_norm": 16.46875,
"learning_rate": 9.99542312632115e-07,
"loss": 25.3541,
"step": 810
},
{
"epoch": 0.05930747672000723,
"grad_norm": 16.125,
"learning_rate": 9.99536662170783e-07,
"loss": 25.2303,
"step": 820
},
{
"epoch": 0.060030738631226835,
"grad_norm": 16.625,
"learning_rate": 9.99531011709451e-07,
"loss": 24.929,
"step": 830
},
{
"epoch": 0.06075400054244643,
"grad_norm": 15.59375,
"learning_rate": 9.99525361248119e-07,
"loss": 25.0334,
"step": 840
},
{
"epoch": 0.06147726245366603,
"grad_norm": 16.5625,
"learning_rate": 9.995197107867871e-07,
"loss": 25.4058,
"step": 850
},
{
"epoch": 0.062200524364885636,
"grad_norm": 16.390625,
"learning_rate": 9.995140603254554e-07,
"loss": 24.9176,
"step": 860
},
{
"epoch": 0.06292378627610523,
"grad_norm": 15.96875,
"learning_rate": 9.995084098641234e-07,
"loss": 25.4821,
"step": 870
},
{
"epoch": 0.06364704818732483,
"grad_norm": 15.5625,
"learning_rate": 9.995027594027914e-07,
"loss": 25.2232,
"step": 880
},
{
"epoch": 0.06437031009854444,
"grad_norm": 16.140625,
"learning_rate": 9.994971089414593e-07,
"loss": 25.2073,
"step": 890
},
{
"epoch": 0.06509357200976404,
"grad_norm": 16.96875,
"learning_rate": 9.994914584801276e-07,
"loss": 24.9354,
"step": 900
},
{
"epoch": 0.06581683392098364,
"grad_norm": 15.890625,
"learning_rate": 9.994858080187958e-07,
"loss": 25.2106,
"step": 910
},
{
"epoch": 0.06654009583220323,
"grad_norm": 15.875,
"learning_rate": 9.994801575574638e-07,
"loss": 25.3214,
"step": 920
},
{
"epoch": 0.06726335774342283,
"grad_norm": 17.21875,
"learning_rate": 9.994745070961318e-07,
"loss": 25.1286,
"step": 930
},
{
"epoch": 0.06798661965464244,
"grad_norm": 15.75,
"learning_rate": 9.994688566347998e-07,
"loss": 25.0482,
"step": 940
},
{
"epoch": 0.06870988156586204,
"grad_norm": 15.71875,
"learning_rate": 9.99463206173468e-07,
"loss": 24.9044,
"step": 950
},
{
"epoch": 0.06943314347708164,
"grad_norm": 16.796875,
"learning_rate": 9.994575557121362e-07,
"loss": 25.3312,
"step": 960
},
{
"epoch": 0.07015640538830124,
"grad_norm": 16.28125,
"learning_rate": 9.994519052508042e-07,
"loss": 25.1329,
"step": 970
},
{
"epoch": 0.07087966729952083,
"grad_norm": 15.5546875,
"learning_rate": 9.994462547894722e-07,
"loss": 25.5721,
"step": 980
},
{
"epoch": 0.07160292921074043,
"grad_norm": 17.359375,
"learning_rate": 9.994406043281404e-07,
"loss": 25.4597,
"step": 990
},
{
"epoch": 0.07232619112196004,
"grad_norm": 16.03125,
"learning_rate": 9.994349538668084e-07,
"loss": 25.1876,
"step": 1000
},
{
"epoch": 0.07304945303317964,
"grad_norm": 15.4921875,
"learning_rate": 9.994293034054766e-07,
"loss": 24.6817,
"step": 1010
},
{
"epoch": 0.07377271494439924,
"grad_norm": 16.96875,
"learning_rate": 9.994236529441446e-07,
"loss": 25.1883,
"step": 1020
},
{
"epoch": 0.07449597685561885,
"grad_norm": 16.984375,
"learning_rate": 9.994180024828128e-07,
"loss": 24.8092,
"step": 1030
},
{
"epoch": 0.07521923876683845,
"grad_norm": 16.078125,
"learning_rate": 9.994123520214808e-07,
"loss": 25.0968,
"step": 1040
},
{
"epoch": 0.07594250067805804,
"grad_norm": 16.296875,
"learning_rate": 9.994067015601488e-07,
"loss": 25.0758,
"step": 1050
},
{
"epoch": 0.07666576258927764,
"grad_norm": 15.4375,
"learning_rate": 9.99401051098817e-07,
"loss": 25.3673,
"step": 1060
},
{
"epoch": 0.07738902450049724,
"grad_norm": 16.8125,
"learning_rate": 9.99395400637485e-07,
"loss": 25.293,
"step": 1070
},
{
"epoch": 0.07811228641171684,
"grad_norm": 15.3828125,
"learning_rate": 9.993897501761532e-07,
"loss": 25.3697,
"step": 1080
},
{
"epoch": 0.07883554832293645,
"grad_norm": 15.3125,
"learning_rate": 9.993840997148212e-07,
"loss": 25.3813,
"step": 1090
},
{
"epoch": 0.07955881023415605,
"grad_norm": 15.7890625,
"learning_rate": 9.993784492534892e-07,
"loss": 25.0103,
"step": 1100
},
{
"epoch": 0.08028207214537564,
"grad_norm": 17.609375,
"learning_rate": 9.993727987921572e-07,
"loss": 25.3168,
"step": 1110
},
{
"epoch": 0.08100533405659524,
"grad_norm": 17.765625,
"learning_rate": 9.993671483308254e-07,
"loss": 24.9627,
"step": 1120
},
{
"epoch": 0.08172859596781484,
"grad_norm": 17.265625,
"learning_rate": 9.993614978694936e-07,
"loss": 25.0904,
"step": 1130
},
{
"epoch": 0.08245185787903445,
"grad_norm": 15.6015625,
"learning_rate": 9.993558474081616e-07,
"loss": 25.0091,
"step": 1140
},
{
"epoch": 0.08317511979025405,
"grad_norm": 16.78125,
"learning_rate": 9.993501969468296e-07,
"loss": 25.1542,
"step": 1150
},
{
"epoch": 0.08389838170147365,
"grad_norm": 17.421875,
"learning_rate": 9.993445464854976e-07,
"loss": 25.0688,
"step": 1160
},
{
"epoch": 0.08462164361269324,
"grad_norm": 16.046875,
"learning_rate": 9.993388960241658e-07,
"loss": 24.9462,
"step": 1170
},
{
"epoch": 0.08534490552391284,
"grad_norm": 15.8125,
"learning_rate": 9.99333245562834e-07,
"loss": 25.2823,
"step": 1180
},
{
"epoch": 0.08606816743513244,
"grad_norm": 16.828125,
"learning_rate": 9.99327595101502e-07,
"loss": 25.2765,
"step": 1190
},
{
"epoch": 0.08679142934635205,
"grad_norm": 16.125,
"learning_rate": 9.9932194464017e-07,
"loss": 24.7678,
"step": 1200
},
{
"epoch": 0.08751469125757165,
"grad_norm": 16.28125,
"learning_rate": 9.993162941788382e-07,
"loss": 25.295,
"step": 1210
},
{
"epoch": 0.08823795316879125,
"grad_norm": 17.890625,
"learning_rate": 9.993106437175062e-07,
"loss": 25.3306,
"step": 1220
},
{
"epoch": 0.08896121508001086,
"grad_norm": 17.21875,
"learning_rate": 9.993049932561744e-07,
"loss": 24.9049,
"step": 1230
},
{
"epoch": 0.08968447699123044,
"grad_norm": 17.390625,
"learning_rate": 9.992993427948424e-07,
"loss": 24.8557,
"step": 1240
},
{
"epoch": 0.09040773890245005,
"grad_norm": 17.171875,
"learning_rate": 9.992936923335106e-07,
"loss": 24.8911,
"step": 1250
},
{
"epoch": 0.09113100081366965,
"grad_norm": 15.7890625,
"learning_rate": 9.992880418721786e-07,
"loss": 25.025,
"step": 1260
},
{
"epoch": 0.09185426272488925,
"grad_norm": 16.140625,
"learning_rate": 9.992823914108466e-07,
"loss": 25.0636,
"step": 1270
},
{
"epoch": 0.09257752463610885,
"grad_norm": 16.4375,
"learning_rate": 9.992767409495148e-07,
"loss": 25.3049,
"step": 1280
},
{
"epoch": 0.09330078654732846,
"grad_norm": 16.90625,
"learning_rate": 9.992710904881828e-07,
"loss": 25.2396,
"step": 1290
},
{
"epoch": 0.09402404845854805,
"grad_norm": 17.046875,
"learning_rate": 9.99265440026851e-07,
"loss": 25.1375,
"step": 1300
},
{
"epoch": 0.09474731036976765,
"grad_norm": 16.1875,
"learning_rate": 9.99259789565519e-07,
"loss": 25.2109,
"step": 1310
},
{
"epoch": 0.09547057228098725,
"grad_norm": 17.171875,
"learning_rate": 9.99254139104187e-07,
"loss": 25.2515,
"step": 1320
},
{
"epoch": 0.09619383419220685,
"grad_norm": 16.515625,
"learning_rate": 9.992484886428553e-07,
"loss": 25.175,
"step": 1330
},
{
"epoch": 0.09691709610342646,
"grad_norm": 16.84375,
"learning_rate": 9.992428381815233e-07,
"loss": 24.7771,
"step": 1340
},
{
"epoch": 0.09764035801464606,
"grad_norm": 17.390625,
"learning_rate": 9.992371877201915e-07,
"loss": 25.2527,
"step": 1350
},
{
"epoch": 0.09836361992586565,
"grad_norm": 15.9375,
"learning_rate": 9.992315372588595e-07,
"loss": 25.0993,
"step": 1360
},
{
"epoch": 0.09908688183708525,
"grad_norm": 16.359375,
"learning_rate": 9.992258867975275e-07,
"loss": 25.0632,
"step": 1370
},
{
"epoch": 0.09981014374830485,
"grad_norm": 15.578125,
"learning_rate": 9.992202363361957e-07,
"loss": 25.05,
"step": 1380
},
{
"epoch": 0.10053340565952446,
"grad_norm": 15.5234375,
"learning_rate": 9.992145858748637e-07,
"loss": 25.2352,
"step": 1390
},
{
"epoch": 0.10125666757074406,
"grad_norm": 15.8203125,
"learning_rate": 9.992089354135319e-07,
"loss": 25.1623,
"step": 1400
},
{
"epoch": 0.10197992948196366,
"grad_norm": 15.9140625,
"learning_rate": 9.992032849521999e-07,
"loss": 25.0429,
"step": 1410
},
{
"epoch": 0.10270319139318326,
"grad_norm": 16.40625,
"learning_rate": 9.991976344908679e-07,
"loss": 25.2809,
"step": 1420
},
{
"epoch": 0.10342645330440285,
"grad_norm": 15.5703125,
"learning_rate": 9.99191984029536e-07,
"loss": 25.3832,
"step": 1430
},
{
"epoch": 0.10414971521562245,
"grad_norm": 16.71875,
"learning_rate": 9.99186333568204e-07,
"loss": 25.2036,
"step": 1440
},
{
"epoch": 0.10487297712684206,
"grad_norm": 16.421875,
"learning_rate": 9.991806831068723e-07,
"loss": 25.1681,
"step": 1450
},
{
"epoch": 0.10559623903806166,
"grad_norm": 16.484375,
"learning_rate": 9.991750326455403e-07,
"loss": 25.2005,
"step": 1460
},
{
"epoch": 0.10631950094928126,
"grad_norm": 16.359375,
"learning_rate": 9.991693821842085e-07,
"loss": 24.7779,
"step": 1470
},
{
"epoch": 0.10704276286050086,
"grad_norm": 15.9140625,
"learning_rate": 9.991637317228765e-07,
"loss": 24.9989,
"step": 1480
},
{
"epoch": 0.10776602477172045,
"grad_norm": 15.984375,
"learning_rate": 9.991580812615445e-07,
"loss": 25.0596,
"step": 1490
},
{
"epoch": 0.10848928668294006,
"grad_norm": 15.3046875,
"learning_rate": 9.991524308002127e-07,
"loss": 25.118,
"step": 1500
},
{
"epoch": 0.10921254859415966,
"grad_norm": 17.5625,
"learning_rate": 9.991467803388807e-07,
"loss": 25.5828,
"step": 1510
},
{
"epoch": 0.10993581050537926,
"grad_norm": 15.6328125,
"learning_rate": 9.99141129877549e-07,
"loss": 24.9299,
"step": 1520
},
{
"epoch": 0.11065907241659886,
"grad_norm": 16.46875,
"learning_rate": 9.99135479416217e-07,
"loss": 25.4944,
"step": 1530
},
{
"epoch": 0.11138233432781847,
"grad_norm": 15.84375,
"learning_rate": 9.99129828954885e-07,
"loss": 25.2725,
"step": 1540
},
{
"epoch": 0.11210559623903806,
"grad_norm": 15.9453125,
"learning_rate": 9.99124178493553e-07,
"loss": 25.2005,
"step": 1550
},
{
"epoch": 0.11282885815025766,
"grad_norm": 16.40625,
"learning_rate": 9.99118528032221e-07,
"loss": 25.009,
"step": 1560
},
{
"epoch": 0.11355212006147726,
"grad_norm": 16.0,
"learning_rate": 9.991128775708893e-07,
"loss": 24.7517,
"step": 1570
},
{
"epoch": 0.11427538197269686,
"grad_norm": 16.09375,
"learning_rate": 9.991072271095573e-07,
"loss": 25.5721,
"step": 1580
},
{
"epoch": 0.11499864388391647,
"grad_norm": 16.84375,
"learning_rate": 9.991015766482253e-07,
"loss": 24.9795,
"step": 1590
},
{
"epoch": 0.11572190579513607,
"grad_norm": 16.671875,
"learning_rate": 9.990959261868935e-07,
"loss": 25.0254,
"step": 1600
},
{
"epoch": 0.11644516770635567,
"grad_norm": 15.4296875,
"learning_rate": 9.990902757255615e-07,
"loss": 24.8741,
"step": 1610
},
{
"epoch": 0.11716842961757526,
"grad_norm": 15.1875,
"learning_rate": 9.990846252642297e-07,
"loss": 25.0707,
"step": 1620
},
{
"epoch": 0.11789169152879486,
"grad_norm": 15.765625,
"learning_rate": 9.990789748028977e-07,
"loss": 25.1894,
"step": 1630
},
{
"epoch": 0.11861495344001446,
"grad_norm": 16.640625,
"learning_rate": 9.990733243415657e-07,
"loss": 25.275,
"step": 1640
},
{
"epoch": 0.11933821535123407,
"grad_norm": 15.9765625,
"learning_rate": 9.99067673880234e-07,
"loss": 25.5003,
"step": 1650
},
{
"epoch": 0.12006147726245367,
"grad_norm": 15.390625,
"learning_rate": 9.99062023418902e-07,
"loss": 25.217,
"step": 1660
},
{
"epoch": 0.12078473917367327,
"grad_norm": 15.6015625,
"learning_rate": 9.990563729575701e-07,
"loss": 24.9766,
"step": 1670
},
{
"epoch": 0.12150800108489286,
"grad_norm": 16.28125,
"learning_rate": 9.990507224962381e-07,
"loss": 24.9016,
"step": 1680
},
{
"epoch": 0.12223126299611246,
"grad_norm": 28.390625,
"learning_rate": 9.990450720349063e-07,
"loss": 25.5052,
"step": 1690
},
{
"epoch": 0.12295452490733207,
"grad_norm": 16.71875,
"learning_rate": 9.990394215735743e-07,
"loss": 25.113,
"step": 1700
},
{
"epoch": 0.12367778681855167,
"grad_norm": 15.9609375,
"learning_rate": 9.990337711122423e-07,
"loss": 25.0131,
"step": 1710
},
{
"epoch": 0.12440104872977127,
"grad_norm": 16.328125,
"learning_rate": 9.990281206509105e-07,
"loss": 25.1244,
"step": 1720
},
{
"epoch": 0.12512431064099086,
"grad_norm": 16.84375,
"learning_rate": 9.990224701895785e-07,
"loss": 25.1582,
"step": 1730
},
{
"epoch": 0.12584757255221046,
"grad_norm": 16.625,
"learning_rate": 9.990168197282468e-07,
"loss": 25.0834,
"step": 1740
},
{
"epoch": 0.12657083446343007,
"grad_norm": 15.8828125,
"learning_rate": 9.990111692669147e-07,
"loss": 25.0038,
"step": 1750
},
{
"epoch": 0.12729409637464967,
"grad_norm": 16.609375,
"learning_rate": 9.990055188055827e-07,
"loss": 25.0728,
"step": 1760
},
{
"epoch": 0.12801735828586927,
"grad_norm": 16.421875,
"learning_rate": 9.98999868344251e-07,
"loss": 25.295,
"step": 1770
},
{
"epoch": 0.12874062019708887,
"grad_norm": 15.1015625,
"learning_rate": 9.98994217882919e-07,
"loss": 25.2353,
"step": 1780
},
{
"epoch": 0.12946388210830848,
"grad_norm": 16.21875,
"learning_rate": 9.989885674215872e-07,
"loss": 25.2932,
"step": 1790
},
{
"epoch": 0.13018714401952808,
"grad_norm": 15.7890625,
"learning_rate": 9.989829169602552e-07,
"loss": 25.425,
"step": 1800
},
{
"epoch": 0.13091040593074768,
"grad_norm": 17.296875,
"learning_rate": 9.989772664989232e-07,
"loss": 24.8049,
"step": 1810
},
{
"epoch": 0.13163366784196728,
"grad_norm": 16.1875,
"learning_rate": 9.989716160375914e-07,
"loss": 25.0513,
"step": 1820
},
{
"epoch": 0.13235692975318689,
"grad_norm": 15.8828125,
"learning_rate": 9.989659655762594e-07,
"loss": 24.6162,
"step": 1830
},
{
"epoch": 0.13308019166440646,
"grad_norm": 15.953125,
"learning_rate": 9.989603151149276e-07,
"loss": 25.2815,
"step": 1840
},
{
"epoch": 0.13380345357562606,
"grad_norm": 15.8203125,
"learning_rate": 9.989546646535956e-07,
"loss": 24.9768,
"step": 1850
},
{
"epoch": 0.13452671548684567,
"grad_norm": 16.6875,
"learning_rate": 9.989490141922636e-07,
"loss": 24.9111,
"step": 1860
},
{
"epoch": 0.13524997739806527,
"grad_norm": 16.234375,
"learning_rate": 9.989433637309318e-07,
"loss": 25.027,
"step": 1870
},
{
"epoch": 0.13597323930928487,
"grad_norm": 16.609375,
"learning_rate": 9.989377132695998e-07,
"loss": 24.7659,
"step": 1880
},
{
"epoch": 0.13669650122050447,
"grad_norm": 15.546875,
"learning_rate": 9.98932062808268e-07,
"loss": 24.7565,
"step": 1890
},
{
"epoch": 0.13741976313172408,
"grad_norm": 16.28125,
"learning_rate": 9.98926412346936e-07,
"loss": 25.4091,
"step": 1900
},
{
"epoch": 0.13814302504294368,
"grad_norm": 17.0,
"learning_rate": 9.989207618856042e-07,
"loss": 25.1971,
"step": 1910
},
{
"epoch": 0.13886628695416328,
"grad_norm": 16.140625,
"learning_rate": 9.989151114242722e-07,
"loss": 25.1719,
"step": 1920
},
{
"epoch": 0.13958954886538288,
"grad_norm": 15.640625,
"learning_rate": 9.989094609629402e-07,
"loss": 25.2595,
"step": 1930
},
{
"epoch": 0.1403128107766025,
"grad_norm": 17.03125,
"learning_rate": 9.989038105016084e-07,
"loss": 25.1428,
"step": 1940
},
{
"epoch": 0.1410360726878221,
"grad_norm": 16.03125,
"learning_rate": 9.988981600402764e-07,
"loss": 24.9586,
"step": 1950
},
{
"epoch": 0.14175933459904166,
"grad_norm": 17.109375,
"learning_rate": 9.988925095789446e-07,
"loss": 25.2662,
"step": 1960
},
{
"epoch": 0.14248259651026127,
"grad_norm": 17.03125,
"learning_rate": 9.988868591176126e-07,
"loss": 25.16,
"step": 1970
},
{
"epoch": 0.14320585842148087,
"grad_norm": 16.5,
"learning_rate": 9.988812086562806e-07,
"loss": 25.2651,
"step": 1980
},
{
"epoch": 0.14392912033270047,
"grad_norm": 16.421875,
"learning_rate": 9.988755581949488e-07,
"loss": 25.2441,
"step": 1990
},
{
"epoch": 0.14465238224392007,
"grad_norm": 15.96875,
"learning_rate": 9.988699077336168e-07,
"loss": 25.0556,
"step": 2000
},
{
"epoch": 0.14537564415513968,
"grad_norm": 16.53125,
"learning_rate": 9.98864257272285e-07,
"loss": 25.4167,
"step": 2010
},
{
"epoch": 0.14609890606635928,
"grad_norm": 16.03125,
"learning_rate": 9.98858606810953e-07,
"loss": 25.1329,
"step": 2020
},
{
"epoch": 0.14682216797757888,
"grad_norm": 16.109375,
"learning_rate": 9.98852956349621e-07,
"loss": 25.4823,
"step": 2030
},
{
"epoch": 0.14754542988879848,
"grad_norm": 16.265625,
"learning_rate": 9.988473058882892e-07,
"loss": 25.0906,
"step": 2040
},
{
"epoch": 0.1482686918000181,
"grad_norm": 16.53125,
"learning_rate": 9.988416554269572e-07,
"loss": 24.8343,
"step": 2050
},
{
"epoch": 0.1489919537112377,
"grad_norm": 15.609375,
"learning_rate": 9.988360049656254e-07,
"loss": 25.3048,
"step": 2060
},
{
"epoch": 0.1497152156224573,
"grad_norm": 17.4375,
"learning_rate": 9.988303545042934e-07,
"loss": 25.2774,
"step": 2070
},
{
"epoch": 0.1504384775336769,
"grad_norm": 15.828125,
"learning_rate": 9.988247040429616e-07,
"loss": 25.2854,
"step": 2080
},
{
"epoch": 0.15116173944489647,
"grad_norm": 16.390625,
"learning_rate": 9.988190535816296e-07,
"loss": 25.0215,
"step": 2090
},
{
"epoch": 0.15188500135611607,
"grad_norm": 16.171875,
"learning_rate": 9.988134031202976e-07,
"loss": 25.2348,
"step": 2100
},
{
"epoch": 0.15260826326733568,
"grad_norm": 15.34375,
"learning_rate": 9.988077526589658e-07,
"loss": 24.6356,
"step": 2110
},
{
"epoch": 0.15333152517855528,
"grad_norm": 16.25,
"learning_rate": 9.988021021976338e-07,
"loss": 25.1007,
"step": 2120
},
{
"epoch": 0.15405478708977488,
"grad_norm": 16.03125,
"learning_rate": 9.98796451736302e-07,
"loss": 24.9828,
"step": 2130
},
{
"epoch": 0.15477804900099448,
"grad_norm": 16.5,
"learning_rate": 9.9879080127497e-07,
"loss": 25.0345,
"step": 2140
},
{
"epoch": 0.15550131091221409,
"grad_norm": 17.15625,
"learning_rate": 9.98785150813638e-07,
"loss": 25.5867,
"step": 2150
},
{
"epoch": 0.1562245728234337,
"grad_norm": 16.71875,
"learning_rate": 9.987795003523062e-07,
"loss": 25.3682,
"step": 2160
},
{
"epoch": 0.1569478347346533,
"grad_norm": 17.015625,
"learning_rate": 9.987738498909745e-07,
"loss": 25.4203,
"step": 2170
},
{
"epoch": 0.1576710966458729,
"grad_norm": 16.296875,
"learning_rate": 9.987681994296424e-07,
"loss": 24.7439,
"step": 2180
},
{
"epoch": 0.1583943585570925,
"grad_norm": 16.203125,
"learning_rate": 9.987625489683104e-07,
"loss": 25.2626,
"step": 2190
},
{
"epoch": 0.1591176204683121,
"grad_norm": 16.046875,
"learning_rate": 9.987568985069784e-07,
"loss": 25.3073,
"step": 2200
},
{
"epoch": 0.1598408823795317,
"grad_norm": 15.953125,
"learning_rate": 9.987512480456467e-07,
"loss": 25.1684,
"step": 2210
},
{
"epoch": 0.16056414429075128,
"grad_norm": 15.75,
"learning_rate": 9.987455975843149e-07,
"loss": 24.7415,
"step": 2220
},
{
"epoch": 0.16128740620197088,
"grad_norm": 16.5,
"learning_rate": 9.987399471229829e-07,
"loss": 25.3286,
"step": 2230
},
{
"epoch": 0.16201066811319048,
"grad_norm": 15.796875,
"learning_rate": 9.987342966616509e-07,
"loss": 24.9328,
"step": 2240
},
{
"epoch": 0.16273393002441008,
"grad_norm": 15.90625,
"learning_rate": 9.987286462003189e-07,
"loss": 25.232,
"step": 2250
},
{
"epoch": 0.1634571919356297,
"grad_norm": 15.65625,
"learning_rate": 9.98722995738987e-07,
"loss": 25.3422,
"step": 2260
},
{
"epoch": 0.1641804538468493,
"grad_norm": 16.359375,
"learning_rate": 9.98717345277655e-07,
"loss": 25.1399,
"step": 2270
},
{
"epoch": 0.1649037157580689,
"grad_norm": 15.375,
"learning_rate": 9.987116948163233e-07,
"loss": 25.268,
"step": 2280
},
{
"epoch": 0.1656269776692885,
"grad_norm": 15.9765625,
"learning_rate": 9.987060443549913e-07,
"loss": 25.0909,
"step": 2290
},
{
"epoch": 0.1663502395805081,
"grad_norm": 15.78125,
"learning_rate": 9.987003938936595e-07,
"loss": 24.9881,
"step": 2300
},
{
"epoch": 0.1670735014917277,
"grad_norm": 15.7109375,
"learning_rate": 9.986947434323275e-07,
"loss": 25.1584,
"step": 2310
},
{
"epoch": 0.1677967634029473,
"grad_norm": 16.6875,
"learning_rate": 9.986890929709955e-07,
"loss": 25.3915,
"step": 2320
},
{
"epoch": 0.1685200253141669,
"grad_norm": 16.03125,
"learning_rate": 9.986834425096637e-07,
"loss": 25.0778,
"step": 2330
},
{
"epoch": 0.16924328722538648,
"grad_norm": 16.5625,
"learning_rate": 9.986777920483317e-07,
"loss": 25.1175,
"step": 2340
},
{
"epoch": 0.16996654913660608,
"grad_norm": 16.484375,
"learning_rate": 9.986721415869999e-07,
"loss": 25.0781,
"step": 2350
},
{
"epoch": 0.17068981104782568,
"grad_norm": 16.359375,
"learning_rate": 9.986664911256679e-07,
"loss": 24.9664,
"step": 2360
},
{
"epoch": 0.1714130729590453,
"grad_norm": 16.359375,
"learning_rate": 9.986608406643359e-07,
"loss": 25.2058,
"step": 2370
},
{
"epoch": 0.1721363348702649,
"grad_norm": 16.765625,
"learning_rate": 9.98655190203004e-07,
"loss": 25.2812,
"step": 2380
},
{
"epoch": 0.1728595967814845,
"grad_norm": 15.3046875,
"learning_rate": 9.986495397416723e-07,
"loss": 24.9085,
"step": 2390
},
{
"epoch": 0.1735828586927041,
"grad_norm": 16.21875,
"learning_rate": 9.986438892803403e-07,
"loss": 25.5753,
"step": 2400
},
{
"epoch": 0.1743061206039237,
"grad_norm": 15.8984375,
"learning_rate": 9.986382388190083e-07,
"loss": 24.8124,
"step": 2410
},
{
"epoch": 0.1750293825151433,
"grad_norm": 16.53125,
"learning_rate": 9.986325883576763e-07,
"loss": 25.1711,
"step": 2420
},
{
"epoch": 0.1757526444263629,
"grad_norm": 15.953125,
"learning_rate": 9.986269378963445e-07,
"loss": 25.014,
"step": 2430
},
{
"epoch": 0.1764759063375825,
"grad_norm": 16.28125,
"learning_rate": 9.986212874350127e-07,
"loss": 25.099,
"step": 2440
},
{
"epoch": 0.1771991682488021,
"grad_norm": 17.46875,
"learning_rate": 9.986156369736807e-07,
"loss": 25.3132,
"step": 2450
},
{
"epoch": 0.1779224301600217,
"grad_norm": 16.96875,
"learning_rate": 9.986099865123487e-07,
"loss": 25.126,
"step": 2460
},
{
"epoch": 0.17864569207124129,
"grad_norm": 16.796875,
"learning_rate": 9.986043360510167e-07,
"loss": 25.1885,
"step": 2470
},
{
"epoch": 0.1793689539824609,
"grad_norm": 15.390625,
"learning_rate": 9.98598685589685e-07,
"loss": 25.0434,
"step": 2480
},
{
"epoch": 0.1800922158936805,
"grad_norm": 15.8203125,
"learning_rate": 9.985930351283531e-07,
"loss": 25.4111,
"step": 2490
},
{
"epoch": 0.1808154778049001,
"grad_norm": 15.4140625,
"learning_rate": 9.985873846670211e-07,
"loss": 24.7393,
"step": 2500
},
{
"epoch": 0.1808154778049001,
"eval_loss": 1.5891753435134888,
"eval_runtime": 361.4111,
"eval_samples_per_second": 2008.956,
"eval_steps_per_second": 31.391,
"step": 2500
},
{
"epoch": 0.1815387397161197,
"grad_norm": 15.7265625,
"learning_rate": 9.985817342056891e-07,
"loss": 24.7492,
"step": 2510
},
{
"epoch": 0.1822620016273393,
"grad_norm": 16.046875,
"learning_rate": 9.985760837443573e-07,
"loss": 24.7608,
"step": 2520
},
{
"epoch": 0.1829852635385589,
"grad_norm": 16.90625,
"learning_rate": 9.985704332830253e-07,
"loss": 24.9355,
"step": 2530
},
{
"epoch": 0.1837085254497785,
"grad_norm": 15.5625,
"learning_rate": 9.985647828216935e-07,
"loss": 25.3485,
"step": 2540
},
{
"epoch": 0.1844317873609981,
"grad_norm": 15.8671875,
"learning_rate": 9.985591323603615e-07,
"loss": 25.1266,
"step": 2550
},
{
"epoch": 0.1851550492722177,
"grad_norm": 17.3125,
"learning_rate": 9.985534818990295e-07,
"loss": 25.1881,
"step": 2560
},
{
"epoch": 0.1858783111834373,
"grad_norm": 16.53125,
"learning_rate": 9.985478314376977e-07,
"loss": 24.9183,
"step": 2570
},
{
"epoch": 0.18660157309465691,
"grad_norm": 15.8359375,
"learning_rate": 9.985421809763657e-07,
"loss": 25.1384,
"step": 2580
},
{
"epoch": 0.18732483500587652,
"grad_norm": 17.84375,
"learning_rate": 9.98536530515034e-07,
"loss": 24.6692,
"step": 2590
},
{
"epoch": 0.1880480969170961,
"grad_norm": 16.8125,
"learning_rate": 9.98530880053702e-07,
"loss": 24.7989,
"step": 2600
},
{
"epoch": 0.1887713588283157,
"grad_norm": 16.078125,
"learning_rate": 9.985252295923701e-07,
"loss": 24.9476,
"step": 2610
},
{
"epoch": 0.1894946207395353,
"grad_norm": 15.9765625,
"learning_rate": 9.985195791310381e-07,
"loss": 25.065,
"step": 2620
},
{
"epoch": 0.1902178826507549,
"grad_norm": 16.171875,
"learning_rate": 9.985139286697061e-07,
"loss": 24.9492,
"step": 2630
},
{
"epoch": 0.1909411445619745,
"grad_norm": 16.5,
"learning_rate": 9.985082782083741e-07,
"loss": 25.2619,
"step": 2640
},
{
"epoch": 0.1916644064731941,
"grad_norm": 16.03125,
"learning_rate": 9.985026277470424e-07,
"loss": 25.2593,
"step": 2650
},
{
"epoch": 0.1923876683844137,
"grad_norm": 15.625,
"learning_rate": 9.984969772857106e-07,
"loss": 24.9816,
"step": 2660
},
{
"epoch": 0.1931109302956333,
"grad_norm": 15.890625,
"learning_rate": 9.984913268243786e-07,
"loss": 25.0609,
"step": 2670
},
{
"epoch": 0.1938341922068529,
"grad_norm": 16.015625,
"learning_rate": 9.984856763630466e-07,
"loss": 25.2305,
"step": 2680
},
{
"epoch": 0.19455745411807251,
"grad_norm": 19.59375,
"learning_rate": 9.984800259017146e-07,
"loss": 24.9294,
"step": 2690
},
{
"epoch": 0.19528071602929212,
"grad_norm": 17.140625,
"learning_rate": 9.984743754403828e-07,
"loss": 25.3927,
"step": 2700
},
{
"epoch": 0.19600397794051172,
"grad_norm": 17.21875,
"learning_rate": 9.98468724979051e-07,
"loss": 25.1831,
"step": 2710
},
{
"epoch": 0.1967272398517313,
"grad_norm": 16.4375,
"learning_rate": 9.98463074517719e-07,
"loss": 25.128,
"step": 2720
},
{
"epoch": 0.1974505017629509,
"grad_norm": 16.953125,
"learning_rate": 9.98457424056387e-07,
"loss": 25.0455,
"step": 2730
},
{
"epoch": 0.1981737636741705,
"grad_norm": 16.5,
"learning_rate": 9.984517735950552e-07,
"loss": 24.7651,
"step": 2740
},
{
"epoch": 0.1988970255853901,
"grad_norm": 16.40625,
"learning_rate": 9.984461231337232e-07,
"loss": 25.3457,
"step": 2750
},
{
"epoch": 0.1996202874966097,
"grad_norm": 16.171875,
"learning_rate": 9.984404726723914e-07,
"loss": 25.3497,
"step": 2760
},
{
"epoch": 0.2003435494078293,
"grad_norm": 16.125,
"learning_rate": 9.984348222110594e-07,
"loss": 25.1117,
"step": 2770
},
{
"epoch": 0.2010668113190489,
"grad_norm": 16.09375,
"learning_rate": 9.984291717497276e-07,
"loss": 25.2813,
"step": 2780
},
{
"epoch": 0.2017900732302685,
"grad_norm": 15.5078125,
"learning_rate": 9.984235212883956e-07,
"loss": 25.4148,
"step": 2790
},
{
"epoch": 0.20251333514148812,
"grad_norm": 16.125,
"learning_rate": 9.984178708270636e-07,
"loss": 25.2476,
"step": 2800
},
{
"epoch": 0.20323659705270772,
"grad_norm": 16.375,
"learning_rate": 9.984122203657318e-07,
"loss": 25.1569,
"step": 2810
},
{
"epoch": 0.20395985896392732,
"grad_norm": 15.3984375,
"learning_rate": 9.984065699043998e-07,
"loss": 24.9938,
"step": 2820
},
{
"epoch": 0.20468312087514692,
"grad_norm": 15.8984375,
"learning_rate": 9.98400919443068e-07,
"loss": 25.0097,
"step": 2830
},
{
"epoch": 0.20540638278636653,
"grad_norm": 16.640625,
"learning_rate": 9.98395268981736e-07,
"loss": 25.2562,
"step": 2840
},
{
"epoch": 0.2061296446975861,
"grad_norm": 15.3828125,
"learning_rate": 9.98389618520404e-07,
"loss": 24.8865,
"step": 2850
},
{
"epoch": 0.2068529066088057,
"grad_norm": 15.5703125,
"learning_rate": 9.983839680590722e-07,
"loss": 25.2523,
"step": 2860
},
{
"epoch": 0.2075761685200253,
"grad_norm": 15.90625,
"learning_rate": 9.983783175977402e-07,
"loss": 25.2208,
"step": 2870
},
{
"epoch": 0.2082994304312449,
"grad_norm": 15.53125,
"learning_rate": 9.983726671364084e-07,
"loss": 25.2277,
"step": 2880
},
{
"epoch": 0.2090226923424645,
"grad_norm": 15.375,
"learning_rate": 9.983670166750764e-07,
"loss": 24.9516,
"step": 2890
},
{
"epoch": 0.2097459542536841,
"grad_norm": 17.4375,
"learning_rate": 9.983613662137444e-07,
"loss": 25.5087,
"step": 2900
},
{
"epoch": 0.21046921616490372,
"grad_norm": 16.28125,
"learning_rate": 9.983557157524126e-07,
"loss": 25.1188,
"step": 2910
},
{
"epoch": 0.21119247807612332,
"grad_norm": 16.390625,
"learning_rate": 9.983500652910806e-07,
"loss": 25.2662,
"step": 2920
},
{
"epoch": 0.21191573998734292,
"grad_norm": 16.03125,
"learning_rate": 9.983444148297488e-07,
"loss": 24.9432,
"step": 2930
},
{
"epoch": 0.21263900189856252,
"grad_norm": 17.234375,
"learning_rate": 9.983387643684168e-07,
"loss": 25.2459,
"step": 2940
},
{
"epoch": 0.21336226380978213,
"grad_norm": 15.28125,
"learning_rate": 9.983331139070848e-07,
"loss": 24.9326,
"step": 2950
},
{
"epoch": 0.21408552572100173,
"grad_norm": 18.0,
"learning_rate": 9.98327463445753e-07,
"loss": 25.2388,
"step": 2960
},
{
"epoch": 0.21480878763222133,
"grad_norm": 17.1875,
"learning_rate": 9.98321812984421e-07,
"loss": 25.6366,
"step": 2970
},
{
"epoch": 0.2155320495434409,
"grad_norm": 16.203125,
"learning_rate": 9.983161625230892e-07,
"loss": 25.1225,
"step": 2980
},
{
"epoch": 0.2162553114546605,
"grad_norm": 16.828125,
"learning_rate": 9.983105120617572e-07,
"loss": 25.1165,
"step": 2990
},
{
"epoch": 0.2169785733658801,
"grad_norm": 16.625,
"learning_rate": 9.983048616004254e-07,
"loss": 25.2762,
"step": 3000
},
{
"epoch": 0.21770183527709971,
"grad_norm": 16.1875,
"learning_rate": 9.982992111390934e-07,
"loss": 24.9826,
"step": 3010
},
{
"epoch": 0.21842509718831932,
"grad_norm": 16.890625,
"learning_rate": 9.982935606777614e-07,
"loss": 24.7336,
"step": 3020
},
{
"epoch": 0.21914835909953892,
"grad_norm": 15.765625,
"learning_rate": 9.982879102164296e-07,
"loss": 25.1037,
"step": 3030
},
{
"epoch": 0.21987162101075852,
"grad_norm": 16.734375,
"learning_rate": 9.982822597550976e-07,
"loss": 24.8429,
"step": 3040
},
{
"epoch": 0.22059488292197812,
"grad_norm": 15.96875,
"learning_rate": 9.982766092937658e-07,
"loss": 25.369,
"step": 3050
},
{
"epoch": 0.22131814483319773,
"grad_norm": 15.46875,
"learning_rate": 9.982709588324338e-07,
"loss": 25.6081,
"step": 3060
},
{
"epoch": 0.22204140674441733,
"grad_norm": 16.59375,
"learning_rate": 9.982653083711018e-07,
"loss": 25.4421,
"step": 3070
},
{
"epoch": 0.22276466865563693,
"grad_norm": 15.734375,
"learning_rate": 9.9825965790977e-07,
"loss": 24.9267,
"step": 3080
},
{
"epoch": 0.22348793056685654,
"grad_norm": 15.7421875,
"learning_rate": 9.98254007448438e-07,
"loss": 25.1702,
"step": 3090
},
{
"epoch": 0.2242111924780761,
"grad_norm": 16.34375,
"learning_rate": 9.982483569871063e-07,
"loss": 24.97,
"step": 3100
},
{
"epoch": 0.2249344543892957,
"grad_norm": 16.328125,
"learning_rate": 9.982427065257743e-07,
"loss": 25.1489,
"step": 3110
},
{
"epoch": 0.22565771630051532,
"grad_norm": 16.46875,
"learning_rate": 9.982370560644423e-07,
"loss": 25.291,
"step": 3120
},
{
"epoch": 0.22638097821173492,
"grad_norm": 16.859375,
"learning_rate": 9.982314056031105e-07,
"loss": 25.1147,
"step": 3130
},
{
"epoch": 0.22710424012295452,
"grad_norm": 16.1875,
"learning_rate": 9.982257551417785e-07,
"loss": 25.0294,
"step": 3140
},
{
"epoch": 0.22782750203417412,
"grad_norm": 16.609375,
"learning_rate": 9.982201046804467e-07,
"loss": 24.9066,
"step": 3150
},
{
"epoch": 0.22855076394539373,
"grad_norm": 16.296875,
"learning_rate": 9.982144542191147e-07,
"loss": 25.5522,
"step": 3160
},
{
"epoch": 0.22927402585661333,
"grad_norm": 16.875,
"learning_rate": 9.982088037577827e-07,
"loss": 25.046,
"step": 3170
},
{
"epoch": 0.22999728776783293,
"grad_norm": 17.34375,
"learning_rate": 9.982031532964509e-07,
"loss": 25.0408,
"step": 3180
},
{
"epoch": 0.23072054967905253,
"grad_norm": 16.078125,
"learning_rate": 9.981975028351189e-07,
"loss": 25.6239,
"step": 3190
},
{
"epoch": 0.23144381159027214,
"grad_norm": 14.75,
"learning_rate": 9.98191852373787e-07,
"loss": 24.7332,
"step": 3200
},
{
"epoch": 0.23216707350149174,
"grad_norm": 16.265625,
"learning_rate": 9.98186201912455e-07,
"loss": 25.0991,
"step": 3210
},
{
"epoch": 0.23289033541271134,
"grad_norm": 16.609375,
"learning_rate": 9.981805514511233e-07,
"loss": 24.9936,
"step": 3220
},
{
"epoch": 0.23361359732393092,
"grad_norm": 17.03125,
"learning_rate": 9.981749009897913e-07,
"loss": 25.4223,
"step": 3230
},
{
"epoch": 0.23433685923515052,
"grad_norm": 16.625,
"learning_rate": 9.981692505284593e-07,
"loss": 25.1375,
"step": 3240
},
{
"epoch": 0.23506012114637012,
"grad_norm": 16.71875,
"learning_rate": 9.981636000671275e-07,
"loss": 25.1016,
"step": 3250
},
{
"epoch": 0.23578338305758972,
"grad_norm": 16.15625,
"learning_rate": 9.981579496057955e-07,
"loss": 24.9417,
"step": 3260
},
{
"epoch": 0.23650664496880933,
"grad_norm": 16.84375,
"learning_rate": 9.981522991444637e-07,
"loss": 25.2421,
"step": 3270
},
{
"epoch": 0.23722990688002893,
"grad_norm": 15.8359375,
"learning_rate": 9.981466486831317e-07,
"loss": 25.1771,
"step": 3280
},
{
"epoch": 0.23795316879124853,
"grad_norm": 16.65625,
"learning_rate": 9.981409982217997e-07,
"loss": 25.3145,
"step": 3290
},
{
"epoch": 0.23867643070246813,
"grad_norm": 16.984375,
"learning_rate": 9.98135347760468e-07,
"loss": 24.93,
"step": 3300
},
{
"epoch": 0.23939969261368774,
"grad_norm": 16.6875,
"learning_rate": 9.98129697299136e-07,
"loss": 25.5047,
"step": 3310
},
{
"epoch": 0.24012295452490734,
"grad_norm": 16.28125,
"learning_rate": 9.98124046837804e-07,
"loss": 25.0701,
"step": 3320
},
{
"epoch": 0.24084621643612694,
"grad_norm": 15.265625,
"learning_rate": 9.98118396376472e-07,
"loss": 25.1109,
"step": 3330
},
{
"epoch": 0.24156947834734654,
"grad_norm": 16.5625,
"learning_rate": 9.9811274591514e-07,
"loss": 25.2658,
"step": 3340
},
{
"epoch": 0.24229274025856612,
"grad_norm": 16.140625,
"learning_rate": 9.981070954538083e-07,
"loss": 24.9092,
"step": 3350
},
{
"epoch": 0.24301600216978572,
"grad_norm": 16.71875,
"learning_rate": 9.981014449924763e-07,
"loss": 25.0737,
"step": 3360
},
{
"epoch": 0.24373926408100532,
"grad_norm": 17.578125,
"learning_rate": 9.980957945311445e-07,
"loss": 24.9847,
"step": 3370
},
{
"epoch": 0.24446252599222493,
"grad_norm": 16.125,
"learning_rate": 9.980901440698125e-07,
"loss": 25.1988,
"step": 3380
},
{
"epoch": 0.24518578790344453,
"grad_norm": 17.296875,
"learning_rate": 9.980844936084805e-07,
"loss": 25.0352,
"step": 3390
},
{
"epoch": 0.24590904981466413,
"grad_norm": 15.0078125,
"learning_rate": 9.980788431471487e-07,
"loss": 25.0363,
"step": 3400
},
{
"epoch": 0.24663231172588373,
"grad_norm": 17.265625,
"learning_rate": 9.980731926858167e-07,
"loss": 24.8165,
"step": 3410
},
{
"epoch": 0.24735557363710334,
"grad_norm": 18.09375,
"learning_rate": 9.98067542224485e-07,
"loss": 25.1422,
"step": 3420
},
{
"epoch": 0.24807883554832294,
"grad_norm": 16.203125,
"learning_rate": 9.98061891763153e-07,
"loss": 24.8054,
"step": 3430
},
{
"epoch": 0.24880209745954254,
"grad_norm": 16.703125,
"learning_rate": 9.980562413018211e-07,
"loss": 25.1454,
"step": 3440
},
{
"epoch": 0.24952535937076215,
"grad_norm": 15.359375,
"learning_rate": 9.980505908404891e-07,
"loss": 25.3514,
"step": 3450
},
{
"epoch": 0.2502486212819817,
"grad_norm": 15.3984375,
"learning_rate": 9.980449403791571e-07,
"loss": 25.141,
"step": 3460
},
{
"epoch": 0.25097188319320135,
"grad_norm": 16.078125,
"learning_rate": 9.980392899178253e-07,
"loss": 24.8799,
"step": 3470
},
{
"epoch": 0.2516951451044209,
"grad_norm": 16.515625,
"learning_rate": 9.980336394564933e-07,
"loss": 24.9967,
"step": 3480
},
{
"epoch": 0.25241840701564056,
"grad_norm": 16.46875,
"learning_rate": 9.980279889951615e-07,
"loss": 25.1666,
"step": 3490
},
{
"epoch": 0.25314166892686013,
"grad_norm": 16.25,
"learning_rate": 9.980223385338295e-07,
"loss": 25.3913,
"step": 3500
},
{
"epoch": 0.25386493083807976,
"grad_norm": 15.6640625,
"learning_rate": 9.980166880724975e-07,
"loss": 25.0079,
"step": 3510
},
{
"epoch": 0.25458819274929934,
"grad_norm": 16.015625,
"learning_rate": 9.980110376111657e-07,
"loss": 25.0063,
"step": 3520
},
{
"epoch": 0.25531145466051897,
"grad_norm": 15.8515625,
"learning_rate": 9.980053871498337e-07,
"loss": 25.5363,
"step": 3530
},
{
"epoch": 0.25603471657173854,
"grad_norm": 16.984375,
"learning_rate": 9.97999736688502e-07,
"loss": 25.1561,
"step": 3540
},
{
"epoch": 0.2567579784829581,
"grad_norm": 16.640625,
"learning_rate": 9.9799408622717e-07,
"loss": 25.1949,
"step": 3550
},
{
"epoch": 0.25748124039417775,
"grad_norm": 16.0,
"learning_rate": 9.97988435765838e-07,
"loss": 24.992,
"step": 3560
},
{
"epoch": 0.2582045023053973,
"grad_norm": 16.375,
"learning_rate": 9.979827853045062e-07,
"loss": 25.1319,
"step": 3570
},
{
"epoch": 0.25892776421661695,
"grad_norm": 15.9296875,
"learning_rate": 9.979771348431742e-07,
"loss": 25.0064,
"step": 3580
},
{
"epoch": 0.2596510261278365,
"grad_norm": 16.265625,
"learning_rate": 9.979714843818424e-07,
"loss": 24.9227,
"step": 3590
},
{
"epoch": 0.26037428803905616,
"grad_norm": 16.578125,
"learning_rate": 9.979658339205104e-07,
"loss": 25.1148,
"step": 3600
},
{
"epoch": 0.26109754995027573,
"grad_norm": 18.4375,
"learning_rate": 9.979601834591784e-07,
"loss": 24.9109,
"step": 3610
},
{
"epoch": 0.26182081186149536,
"grad_norm": 17.640625,
"learning_rate": 9.979545329978466e-07,
"loss": 25.1381,
"step": 3620
},
{
"epoch": 0.26254407377271494,
"grad_norm": 15.78125,
"learning_rate": 9.979488825365146e-07,
"loss": 25.0138,
"step": 3630
},
{
"epoch": 0.26326733568393457,
"grad_norm": 16.65625,
"learning_rate": 9.979432320751828e-07,
"loss": 25.0708,
"step": 3640
},
{
"epoch": 0.26399059759515414,
"grad_norm": 15.875,
"learning_rate": 9.979375816138508e-07,
"loss": 25.0401,
"step": 3650
},
{
"epoch": 0.26471385950637377,
"grad_norm": 16.359375,
"learning_rate": 9.97931931152519e-07,
"loss": 25.1882,
"step": 3660
},
{
"epoch": 0.26543712141759335,
"grad_norm": 15.5625,
"learning_rate": 9.97926280691187e-07,
"loss": 25.2344,
"step": 3670
},
{
"epoch": 0.2661603833288129,
"grad_norm": 15.921875,
"learning_rate": 9.97920630229855e-07,
"loss": 25.0361,
"step": 3680
},
{
"epoch": 0.26688364524003255,
"grad_norm": 15.7890625,
"learning_rate": 9.979149797685232e-07,
"loss": 25.2095,
"step": 3690
},
{
"epoch": 0.2676069071512521,
"grad_norm": 16.25,
"learning_rate": 9.979093293071914e-07,
"loss": 25.4166,
"step": 3700
},
{
"epoch": 0.26833016906247176,
"grad_norm": 17.03125,
"learning_rate": 9.979036788458594e-07,
"loss": 25.1349,
"step": 3710
},
{
"epoch": 0.26905343097369133,
"grad_norm": 15.703125,
"learning_rate": 9.978980283845274e-07,
"loss": 24.6449,
"step": 3720
},
{
"epoch": 0.26977669288491096,
"grad_norm": 15.625,
"learning_rate": 9.978923779231954e-07,
"loss": 25.4002,
"step": 3730
},
{
"epoch": 0.27049995479613054,
"grad_norm": 15.8671875,
"learning_rate": 9.978867274618636e-07,
"loss": 24.9894,
"step": 3740
},
{
"epoch": 0.27122321670735017,
"grad_norm": 17.234375,
"learning_rate": 9.978810770005318e-07,
"loss": 25.3439,
"step": 3750
},
{
"epoch": 0.27194647861856974,
"grad_norm": 16.171875,
"learning_rate": 9.978754265391998e-07,
"loss": 25.1633,
"step": 3760
},
{
"epoch": 0.2726697405297894,
"grad_norm": 16.875,
"learning_rate": 9.978697760778678e-07,
"loss": 25.0982,
"step": 3770
},
{
"epoch": 0.27339300244100895,
"grad_norm": 16.71875,
"learning_rate": 9.978641256165358e-07,
"loss": 25.2635,
"step": 3780
},
{
"epoch": 0.2741162643522286,
"grad_norm": 16.59375,
"learning_rate": 9.97858475155204e-07,
"loss": 25.4958,
"step": 3790
},
{
"epoch": 0.27483952626344815,
"grad_norm": 15.9921875,
"learning_rate": 9.97852824693872e-07,
"loss": 24.7305,
"step": 3800
},
{
"epoch": 0.2755627881746677,
"grad_norm": 16.46875,
"learning_rate": 9.978471742325402e-07,
"loss": 25.187,
"step": 3810
},
{
"epoch": 0.27628605008588736,
"grad_norm": 15.8984375,
"learning_rate": 9.978415237712082e-07,
"loss": 25.1958,
"step": 3820
},
{
"epoch": 0.27700931199710693,
"grad_norm": 16.109375,
"learning_rate": 9.978358733098762e-07,
"loss": 25.5358,
"step": 3830
},
{
"epoch": 0.27773257390832656,
"grad_norm": 15.5,
"learning_rate": 9.978302228485444e-07,
"loss": 24.9236,
"step": 3840
},
{
"epoch": 0.27845583581954614,
"grad_norm": 16.375,
"learning_rate": 9.978245723872124e-07,
"loss": 25.3469,
"step": 3850
},
{
"epoch": 0.27917909773076577,
"grad_norm": 16.15625,
"learning_rate": 9.978189219258806e-07,
"loss": 24.9141,
"step": 3860
},
{
"epoch": 0.27990235964198534,
"grad_norm": 17.125,
"learning_rate": 9.978132714645486e-07,
"loss": 25.0336,
"step": 3870
},
{
"epoch": 0.280625621553205,
"grad_norm": 15.640625,
"learning_rate": 9.978076210032168e-07,
"loss": 25.2943,
"step": 3880
},
{
"epoch": 0.28134888346442455,
"grad_norm": 15.8984375,
"learning_rate": 9.978019705418848e-07,
"loss": 24.9711,
"step": 3890
},
{
"epoch": 0.2820721453756442,
"grad_norm": 16.09375,
"learning_rate": 9.977963200805528e-07,
"loss": 25.2311,
"step": 3900
},
{
"epoch": 0.28279540728686375,
"grad_norm": 16.578125,
"learning_rate": 9.97790669619221e-07,
"loss": 25.4184,
"step": 3910
},
{
"epoch": 0.28351866919808333,
"grad_norm": 15.890625,
"learning_rate": 9.977850191578892e-07,
"loss": 25.2003,
"step": 3920
},
{
"epoch": 0.28424193110930296,
"grad_norm": 17.75,
"learning_rate": 9.977793686965572e-07,
"loss": 25.3082,
"step": 3930
},
{
"epoch": 0.28496519302052253,
"grad_norm": 16.03125,
"learning_rate": 9.977737182352252e-07,
"loss": 25.265,
"step": 3940
},
{
"epoch": 0.28568845493174216,
"grad_norm": 16.65625,
"learning_rate": 9.977680677738932e-07,
"loss": 25.6086,
"step": 3950
},
{
"epoch": 0.28641171684296174,
"grad_norm": 15.8671875,
"learning_rate": 9.977624173125614e-07,
"loss": 25.6748,
"step": 3960
},
{
"epoch": 0.28713497875418137,
"grad_norm": 16.5625,
"learning_rate": 9.977567668512297e-07,
"loss": 25.3439,
"step": 3970
},
{
"epoch": 0.28785824066540094,
"grad_norm": 16.78125,
"learning_rate": 9.977511163898977e-07,
"loss": 24.9551,
"step": 3980
},
{
"epoch": 0.2885815025766206,
"grad_norm": 16.546875,
"learning_rate": 9.977454659285656e-07,
"loss": 25.1372,
"step": 3990
},
{
"epoch": 0.28930476448784015,
"grad_norm": 15.8515625,
"learning_rate": 9.977398154672336e-07,
"loss": 25.5988,
"step": 4000
},
{
"epoch": 0.2900280263990598,
"grad_norm": 17.25,
"learning_rate": 9.977341650059019e-07,
"loss": 25.5525,
"step": 4010
},
{
"epoch": 0.29075128831027935,
"grad_norm": 16.734375,
"learning_rate": 9.9772851454457e-07,
"loss": 24.7127,
"step": 4020
},
{
"epoch": 0.291474550221499,
"grad_norm": 15.8125,
"learning_rate": 9.97722864083238e-07,
"loss": 24.8364,
"step": 4030
},
{
"epoch": 0.29219781213271856,
"grad_norm": 17.75,
"learning_rate": 9.97717213621906e-07,
"loss": 25.1328,
"step": 4040
},
{
"epoch": 0.29292107404393813,
"grad_norm": 15.7890625,
"learning_rate": 9.977115631605743e-07,
"loss": 24.932,
"step": 4050
},
{
"epoch": 0.29364433595515776,
"grad_norm": 16.71875,
"learning_rate": 9.977059126992423e-07,
"loss": 25.0942,
"step": 4060
},
{
"epoch": 0.29436759786637734,
"grad_norm": 17.0625,
"learning_rate": 9.977002622379105e-07,
"loss": 25.0096,
"step": 4070
},
{
"epoch": 0.29509085977759697,
"grad_norm": 15.59375,
"learning_rate": 9.976946117765785e-07,
"loss": 24.9936,
"step": 4080
},
{
"epoch": 0.29581412168881654,
"grad_norm": 16.890625,
"learning_rate": 9.976889613152465e-07,
"loss": 25.1561,
"step": 4090
},
{
"epoch": 0.2965373836000362,
"grad_norm": 16.5625,
"learning_rate": 9.976833108539147e-07,
"loss": 25.2981,
"step": 4100
},
{
"epoch": 0.29726064551125575,
"grad_norm": 16.03125,
"learning_rate": 9.976776603925827e-07,
"loss": 24.9605,
"step": 4110
},
{
"epoch": 0.2979839074224754,
"grad_norm": 16.09375,
"learning_rate": 9.976720099312507e-07,
"loss": 25.254,
"step": 4120
},
{
"epoch": 0.29870716933369496,
"grad_norm": 15.390625,
"learning_rate": 9.976663594699189e-07,
"loss": 25.2773,
"step": 4130
},
{
"epoch": 0.2994304312449146,
"grad_norm": 16.96875,
"learning_rate": 9.97660709008587e-07,
"loss": 25.3654,
"step": 4140
},
{
"epoch": 0.30015369315613416,
"grad_norm": 16.328125,
"learning_rate": 9.97655058547255e-07,
"loss": 24.5768,
"step": 4150
},
{
"epoch": 0.3008769550673538,
"grad_norm": 16.640625,
"learning_rate": 9.97649408085923e-07,
"loss": 25.0535,
"step": 4160
},
{
"epoch": 0.30160021697857337,
"grad_norm": 15.484375,
"learning_rate": 9.97643757624591e-07,
"loss": 25.2813,
"step": 4170
},
{
"epoch": 0.30232347888979294,
"grad_norm": 16.734375,
"learning_rate": 9.976381071632593e-07,
"loss": 25.172,
"step": 4180
},
{
"epoch": 0.30304674080101257,
"grad_norm": 16.140625,
"learning_rate": 9.976324567019275e-07,
"loss": 25.2808,
"step": 4190
},
{
"epoch": 0.30377000271223215,
"grad_norm": 14.953125,
"learning_rate": 9.976268062405955e-07,
"loss": 25.1555,
"step": 4200
},
{
"epoch": 0.3044932646234518,
"grad_norm": 16.0,
"learning_rate": 9.976211557792635e-07,
"loss": 25.4662,
"step": 4210
},
{
"epoch": 0.30521652653467135,
"grad_norm": 16.640625,
"learning_rate": 9.976155053179315e-07,
"loss": 25.3153,
"step": 4220
},
{
"epoch": 0.305939788445891,
"grad_norm": 15.9609375,
"learning_rate": 9.976098548565997e-07,
"loss": 25.1885,
"step": 4230
},
{
"epoch": 0.30666305035711056,
"grad_norm": 15.90625,
"learning_rate": 9.97604204395268e-07,
"loss": 24.9472,
"step": 4240
},
{
"epoch": 0.3073863122683302,
"grad_norm": 15.515625,
"learning_rate": 9.97598553933936e-07,
"loss": 25.0862,
"step": 4250
},
{
"epoch": 0.30810957417954976,
"grad_norm": 16.421875,
"learning_rate": 9.97592903472604e-07,
"loss": 24.8722,
"step": 4260
},
{
"epoch": 0.3088328360907694,
"grad_norm": 16.15625,
"learning_rate": 9.975872530112721e-07,
"loss": 25.3074,
"step": 4270
},
{
"epoch": 0.30955609800198897,
"grad_norm": 15.7109375,
"learning_rate": 9.975816025499401e-07,
"loss": 25.0129,
"step": 4280
},
{
"epoch": 0.3102793599132086,
"grad_norm": 15.859375,
"learning_rate": 9.975759520886083e-07,
"loss": 25.239,
"step": 4290
},
{
"epoch": 0.31100262182442817,
"grad_norm": 15.859375,
"learning_rate": 9.975703016272763e-07,
"loss": 24.9693,
"step": 4300
},
{
"epoch": 0.31172588373564775,
"grad_norm": 16.0625,
"learning_rate": 9.975646511659443e-07,
"loss": 25.1788,
"step": 4310
},
{
"epoch": 0.3124491456468674,
"grad_norm": 16.1875,
"learning_rate": 9.975590007046125e-07,
"loss": 25.3888,
"step": 4320
},
{
"epoch": 0.31317240755808695,
"grad_norm": 16.359375,
"learning_rate": 9.975533502432805e-07,
"loss": 25.269,
"step": 4330
},
{
"epoch": 0.3138956694693066,
"grad_norm": 16.796875,
"learning_rate": 9.975476997819487e-07,
"loss": 25.0877,
"step": 4340
},
{
"epoch": 0.31461893138052616,
"grad_norm": 16.453125,
"learning_rate": 9.975420493206167e-07,
"loss": 24.8608,
"step": 4350
},
{
"epoch": 0.3153421932917458,
"grad_norm": 16.71875,
"learning_rate": 9.97536398859285e-07,
"loss": 25.3031,
"step": 4360
},
{
"epoch": 0.31606545520296536,
"grad_norm": 16.03125,
"learning_rate": 9.97530748397953e-07,
"loss": 25.1261,
"step": 4370
},
{
"epoch": 0.316788717114185,
"grad_norm": 16.25,
"learning_rate": 9.97525097936621e-07,
"loss": 25.1505,
"step": 4380
},
{
"epoch": 0.31751197902540457,
"grad_norm": 15.640625,
"learning_rate": 9.975194474752891e-07,
"loss": 25.0906,
"step": 4390
},
{
"epoch": 0.3182352409366242,
"grad_norm": 16.109375,
"learning_rate": 9.975137970139571e-07,
"loss": 25.1493,
"step": 4400
},
{
"epoch": 0.3189585028478438,
"grad_norm": 16.25,
"learning_rate": 9.975081465526254e-07,
"loss": 25.2401,
"step": 4410
},
{
"epoch": 0.3196817647590634,
"grad_norm": 16.921875,
"learning_rate": 9.975024960912934e-07,
"loss": 25.1871,
"step": 4420
},
{
"epoch": 0.320405026670283,
"grad_norm": 16.328125,
"learning_rate": 9.974968456299613e-07,
"loss": 25.03,
"step": 4430
},
{
"epoch": 0.32112828858150255,
"grad_norm": 16.546875,
"learning_rate": 9.974911951686296e-07,
"loss": 25.1426,
"step": 4440
},
{
"epoch": 0.3218515504927222,
"grad_norm": 16.328125,
"learning_rate": 9.974855447072976e-07,
"loss": 25.0376,
"step": 4450
},
{
"epoch": 0.32257481240394176,
"grad_norm": 15.9375,
"learning_rate": 9.974798942459658e-07,
"loss": 24.6728,
"step": 4460
},
{
"epoch": 0.3232980743151614,
"grad_norm": 16.546875,
"learning_rate": 9.974742437846338e-07,
"loss": 25.2818,
"step": 4470
},
{
"epoch": 0.32402133622638096,
"grad_norm": 15.46875,
"learning_rate": 9.974685933233018e-07,
"loss": 25.5436,
"step": 4480
},
{
"epoch": 0.3247445981376006,
"grad_norm": 16.703125,
"learning_rate": 9.9746294286197e-07,
"loss": 25.0338,
"step": 4490
},
{
"epoch": 0.32546786004882017,
"grad_norm": 16.546875,
"learning_rate": 9.97457292400638e-07,
"loss": 24.9645,
"step": 4500
},
{
"epoch": 0.3261911219600398,
"grad_norm": 15.9296875,
"learning_rate": 9.974516419393062e-07,
"loss": 25.2643,
"step": 4510
},
{
"epoch": 0.3269143838712594,
"grad_norm": 15.984375,
"learning_rate": 9.974459914779742e-07,
"loss": 25.1678,
"step": 4520
},
{
"epoch": 0.327637645782479,
"grad_norm": 16.046875,
"learning_rate": 9.974403410166422e-07,
"loss": 25.2445,
"step": 4530
},
{
"epoch": 0.3283609076936986,
"grad_norm": 16.96875,
"learning_rate": 9.974346905553104e-07,
"loss": 24.8622,
"step": 4540
},
{
"epoch": 0.3290841696049182,
"grad_norm": 16.84375,
"learning_rate": 9.974290400939784e-07,
"loss": 25.0903,
"step": 4550
},
{
"epoch": 0.3298074315161378,
"grad_norm": 16.71875,
"learning_rate": 9.974233896326466e-07,
"loss": 24.9463,
"step": 4560
},
{
"epoch": 0.33053069342735736,
"grad_norm": 16.765625,
"learning_rate": 9.974177391713146e-07,
"loss": 25.1533,
"step": 4570
},
{
"epoch": 0.331253955338577,
"grad_norm": 16.828125,
"learning_rate": 9.974120887099828e-07,
"loss": 25.1094,
"step": 4580
},
{
"epoch": 0.33197721724979656,
"grad_norm": 16.8125,
"learning_rate": 9.974064382486508e-07,
"loss": 25.1725,
"step": 4590
},
{
"epoch": 0.3327004791610162,
"grad_norm": 16.65625,
"learning_rate": 9.974007877873188e-07,
"loss": 24.9989,
"step": 4600
},
{
"epoch": 0.33342374107223577,
"grad_norm": 15.7109375,
"learning_rate": 9.97395137325987e-07,
"loss": 24.916,
"step": 4610
},
{
"epoch": 0.3341470029834554,
"grad_norm": 16.140625,
"learning_rate": 9.97389486864655e-07,
"loss": 24.9188,
"step": 4620
},
{
"epoch": 0.334870264894675,
"grad_norm": 17.515625,
"learning_rate": 9.973838364033232e-07,
"loss": 24.9345,
"step": 4630
},
{
"epoch": 0.3355935268058946,
"grad_norm": 16.4375,
"learning_rate": 9.973781859419912e-07,
"loss": 25.1931,
"step": 4640
},
{
"epoch": 0.3363167887171142,
"grad_norm": 17.125,
"learning_rate": 9.973725354806592e-07,
"loss": 25.1182,
"step": 4650
},
{
"epoch": 0.3370400506283338,
"grad_norm": 16.515625,
"learning_rate": 9.973668850193274e-07,
"loss": 25.5517,
"step": 4660
},
{
"epoch": 0.3377633125395534,
"grad_norm": 16.8125,
"learning_rate": 9.973612345579954e-07,
"loss": 24.8154,
"step": 4670
},
{
"epoch": 0.33848657445077296,
"grad_norm": 16.40625,
"learning_rate": 9.973555840966636e-07,
"loss": 25.2684,
"step": 4680
},
{
"epoch": 0.3392098363619926,
"grad_norm": 16.109375,
"learning_rate": 9.973499336353316e-07,
"loss": 25.024,
"step": 4690
},
{
"epoch": 0.33993309827321216,
"grad_norm": 16.296875,
"learning_rate": 9.973442831739996e-07,
"loss": 25.1598,
"step": 4700
},
{
"epoch": 0.3406563601844318,
"grad_norm": 16.46875,
"learning_rate": 9.973386327126678e-07,
"loss": 25.0369,
"step": 4710
},
{
"epoch": 0.34137962209565137,
"grad_norm": 15.53125,
"learning_rate": 9.973329822513358e-07,
"loss": 25.4247,
"step": 4720
},
{
"epoch": 0.342102884006871,
"grad_norm": 16.75,
"learning_rate": 9.97327331790004e-07,
"loss": 25.0935,
"step": 4730
},
{
"epoch": 0.3428261459180906,
"grad_norm": 15.84375,
"learning_rate": 9.97321681328672e-07,
"loss": 24.8433,
"step": 4740
},
{
"epoch": 0.3435494078293102,
"grad_norm": 16.953125,
"learning_rate": 9.9731603086734e-07,
"loss": 25.3383,
"step": 4750
},
{
"epoch": 0.3442726697405298,
"grad_norm": 16.15625,
"learning_rate": 9.973103804060082e-07,
"loss": 24.9644,
"step": 4760
},
{
"epoch": 0.3449959316517494,
"grad_norm": 17.34375,
"learning_rate": 9.973047299446762e-07,
"loss": 25.2677,
"step": 4770
},
{
"epoch": 0.345719193562969,
"grad_norm": 17.125,
"learning_rate": 9.972990794833444e-07,
"loss": 25.1205,
"step": 4780
},
{
"epoch": 0.3464424554741886,
"grad_norm": 16.015625,
"learning_rate": 9.972934290220124e-07,
"loss": 24.9001,
"step": 4790
},
{
"epoch": 0.3471657173854082,
"grad_norm": 16.984375,
"learning_rate": 9.972877785606806e-07,
"loss": 24.6669,
"step": 4800
},
{
"epoch": 0.34788897929662776,
"grad_norm": 15.9453125,
"learning_rate": 9.972821280993486e-07,
"loss": 24.8667,
"step": 4810
},
{
"epoch": 0.3486122412078474,
"grad_norm": 15.609375,
"learning_rate": 9.972764776380166e-07,
"loss": 24.8402,
"step": 4820
},
{
"epoch": 0.34933550311906697,
"grad_norm": 15.3203125,
"learning_rate": 9.972708271766848e-07,
"loss": 25.0204,
"step": 4830
},
{
"epoch": 0.3500587650302866,
"grad_norm": 16.0625,
"learning_rate": 9.972651767153528e-07,
"loss": 25.1096,
"step": 4840
},
{
"epoch": 0.3507820269415062,
"grad_norm": 15.2421875,
"learning_rate": 9.97259526254021e-07,
"loss": 25.4209,
"step": 4850
},
{
"epoch": 0.3515052888527258,
"grad_norm": 18.0,
"learning_rate": 9.97253875792689e-07,
"loss": 25.3704,
"step": 4860
},
{
"epoch": 0.3522285507639454,
"grad_norm": 15.8984375,
"learning_rate": 9.97248225331357e-07,
"loss": 25.0614,
"step": 4870
},
{
"epoch": 0.352951812675165,
"grad_norm": 16.5,
"learning_rate": 9.972425748700253e-07,
"loss": 24.972,
"step": 4880
},
{
"epoch": 0.3536750745863846,
"grad_norm": 16.765625,
"learning_rate": 9.972369244086933e-07,
"loss": 25.0881,
"step": 4890
},
{
"epoch": 0.3543983364976042,
"grad_norm": 16.546875,
"learning_rate": 9.972312739473615e-07,
"loss": 25.3568,
"step": 4900
},
{
"epoch": 0.3551215984088238,
"grad_norm": 17.03125,
"learning_rate": 9.972256234860295e-07,
"loss": 24.9578,
"step": 4910
},
{
"epoch": 0.3558448603200434,
"grad_norm": 17.015625,
"learning_rate": 9.972199730246975e-07,
"loss": 25.1697,
"step": 4920
},
{
"epoch": 0.356568122231263,
"grad_norm": 17.28125,
"learning_rate": 9.972143225633657e-07,
"loss": 25.3285,
"step": 4930
},
{
"epoch": 0.35729138414248257,
"grad_norm": 16.671875,
"learning_rate": 9.972086721020337e-07,
"loss": 25.0862,
"step": 4940
},
{
"epoch": 0.3580146460537022,
"grad_norm": 15.0703125,
"learning_rate": 9.972030216407019e-07,
"loss": 25.1828,
"step": 4950
},
{
"epoch": 0.3587379079649218,
"grad_norm": 15.6484375,
"learning_rate": 9.971973711793699e-07,
"loss": 25.1832,
"step": 4960
},
{
"epoch": 0.3594611698761414,
"grad_norm": 16.46875,
"learning_rate": 9.97191720718038e-07,
"loss": 25.3505,
"step": 4970
},
{
"epoch": 0.360184431787361,
"grad_norm": 16.421875,
"learning_rate": 9.97186070256706e-07,
"loss": 24.9362,
"step": 4980
},
{
"epoch": 0.3609076936985806,
"grad_norm": 16.6875,
"learning_rate": 9.97180419795374e-07,
"loss": 25.5243,
"step": 4990
},
{
"epoch": 0.3616309556098002,
"grad_norm": 16.234375,
"learning_rate": 9.971747693340423e-07,
"loss": 25.2643,
"step": 5000
},
{
"epoch": 0.3616309556098002,
"eval_loss": 1.5911972522735596,
"eval_runtime": 376.056,
"eval_samples_per_second": 1930.72,
"eval_steps_per_second": 30.168,
"step": 5000
},
{
"epoch": 0.3623542175210198,
"grad_norm": 16.78125,
"learning_rate": 9.971691188727103e-07,
"loss": 25.2169,
"step": 5010
},
{
"epoch": 0.3630774794322394,
"grad_norm": 15.9765625,
"learning_rate": 9.971634684113785e-07,
"loss": 25.2699,
"step": 5020
},
{
"epoch": 0.363800741343459,
"grad_norm": 16.3125,
"learning_rate": 9.971578179500465e-07,
"loss": 25.0431,
"step": 5030
},
{
"epoch": 0.3645240032546786,
"grad_norm": 16.609375,
"learning_rate": 9.971521674887145e-07,
"loss": 25.1181,
"step": 5040
},
{
"epoch": 0.3652472651658982,
"grad_norm": 16.421875,
"learning_rate": 9.971465170273827e-07,
"loss": 25.0253,
"step": 5050
},
{
"epoch": 0.3659705270771178,
"grad_norm": 16.171875,
"learning_rate": 9.971408665660507e-07,
"loss": 25.1289,
"step": 5060
},
{
"epoch": 0.3666937889883374,
"grad_norm": 16.484375,
"learning_rate": 9.97135216104719e-07,
"loss": 25.298,
"step": 5070
},
{
"epoch": 0.367417050899557,
"grad_norm": 15.8984375,
"learning_rate": 9.97129565643387e-07,
"loss": 24.8407,
"step": 5080
},
{
"epoch": 0.3681403128107766,
"grad_norm": 16.421875,
"learning_rate": 9.971239151820549e-07,
"loss": 24.7991,
"step": 5090
},
{
"epoch": 0.3688635747219962,
"grad_norm": 17.015625,
"learning_rate": 9.97118264720723e-07,
"loss": 25.4768,
"step": 5100
},
{
"epoch": 0.3695868366332158,
"grad_norm": 16.15625,
"learning_rate": 9.97112614259391e-07,
"loss": 25.0624,
"step": 5110
},
{
"epoch": 0.3703100985444354,
"grad_norm": 15.9765625,
"learning_rate": 9.971069637980593e-07,
"loss": 24.9802,
"step": 5120
},
{
"epoch": 0.371033360455655,
"grad_norm": 16.21875,
"learning_rate": 9.971013133367273e-07,
"loss": 25.2557,
"step": 5130
},
{
"epoch": 0.3717566223668746,
"grad_norm": 15.625,
"learning_rate": 9.970956628753953e-07,
"loss": 25.0409,
"step": 5140
},
{
"epoch": 0.3724798842780942,
"grad_norm": 16.0625,
"learning_rate": 9.970900124140635e-07,
"loss": 24.8528,
"step": 5150
},
{
"epoch": 0.37320314618931383,
"grad_norm": 17.171875,
"learning_rate": 9.970843619527315e-07,
"loss": 25.2391,
"step": 5160
},
{
"epoch": 0.3739264081005334,
"grad_norm": 15.7265625,
"learning_rate": 9.970787114913997e-07,
"loss": 25.1315,
"step": 5170
},
{
"epoch": 0.37464967001175303,
"grad_norm": 15.890625,
"learning_rate": 9.970730610300677e-07,
"loss": 24.908,
"step": 5180
},
{
"epoch": 0.3753729319229726,
"grad_norm": 16.203125,
"learning_rate": 9.97067410568736e-07,
"loss": 24.8777,
"step": 5190
},
{
"epoch": 0.3760961938341922,
"grad_norm": 16.265625,
"learning_rate": 9.97061760107404e-07,
"loss": 24.7149,
"step": 5200
},
{
"epoch": 0.3768194557454118,
"grad_norm": 16.140625,
"learning_rate": 9.97056109646072e-07,
"loss": 25.073,
"step": 5210
},
{
"epoch": 0.3775427176566314,
"grad_norm": 16.078125,
"learning_rate": 9.970504591847401e-07,
"loss": 25.1022,
"step": 5220
},
{
"epoch": 0.378265979567851,
"grad_norm": 15.765625,
"learning_rate": 9.970448087234081e-07,
"loss": 25.1426,
"step": 5230
},
{
"epoch": 0.3789892414790706,
"grad_norm": 15.296875,
"learning_rate": 9.970391582620763e-07,
"loss": 25.0163,
"step": 5240
},
{
"epoch": 0.3797125033902902,
"grad_norm": 17.453125,
"learning_rate": 9.970335078007443e-07,
"loss": 25.334,
"step": 5250
},
{
"epoch": 0.3804357653015098,
"grad_norm": 15.5703125,
"learning_rate": 9.970278573394123e-07,
"loss": 24.9564,
"step": 5260
},
{
"epoch": 0.38115902721272943,
"grad_norm": 15.78125,
"learning_rate": 9.970222068780805e-07,
"loss": 25.264,
"step": 5270
},
{
"epoch": 0.381882289123949,
"grad_norm": 16.109375,
"learning_rate": 9.970165564167485e-07,
"loss": 25.3679,
"step": 5280
},
{
"epoch": 0.38260555103516863,
"grad_norm": 16.3125,
"learning_rate": 9.970109059554167e-07,
"loss": 25.1622,
"step": 5290
},
{
"epoch": 0.3833288129463882,
"grad_norm": 15.5703125,
"learning_rate": 9.970052554940847e-07,
"loss": 25.0249,
"step": 5300
},
{
"epoch": 0.3840520748576078,
"grad_norm": 16.453125,
"learning_rate": 9.969996050327527e-07,
"loss": 25.1322,
"step": 5310
},
{
"epoch": 0.3847753367688274,
"grad_norm": 17.53125,
"learning_rate": 9.96993954571421e-07,
"loss": 25.4187,
"step": 5320
},
{
"epoch": 0.385498598680047,
"grad_norm": 16.96875,
"learning_rate": 9.96988304110089e-07,
"loss": 24.8692,
"step": 5330
},
{
"epoch": 0.3862218605912666,
"grad_norm": 14.84375,
"learning_rate": 9.969826536487572e-07,
"loss": 25.1561,
"step": 5340
},
{
"epoch": 0.3869451225024862,
"grad_norm": 16.34375,
"learning_rate": 9.969770031874252e-07,
"loss": 25.4448,
"step": 5350
},
{
"epoch": 0.3876683844137058,
"grad_norm": 17.515625,
"learning_rate": 9.969713527260932e-07,
"loss": 25.3175,
"step": 5360
},
{
"epoch": 0.3883916463249254,
"grad_norm": 16.421875,
"learning_rate": 9.969657022647614e-07,
"loss": 24.7974,
"step": 5370
},
{
"epoch": 0.38911490823614503,
"grad_norm": 17.34375,
"learning_rate": 9.969600518034294e-07,
"loss": 24.831,
"step": 5380
},
{
"epoch": 0.3898381701473646,
"grad_norm": 16.34375,
"learning_rate": 9.969544013420976e-07,
"loss": 25.1299,
"step": 5390
},
{
"epoch": 0.39056143205858423,
"grad_norm": 17.765625,
"learning_rate": 9.969487508807656e-07,
"loss": 24.9919,
"step": 5400
},
{
"epoch": 0.3912846939698038,
"grad_norm": 15.75,
"learning_rate": 9.969431004194338e-07,
"loss": 25.1571,
"step": 5410
},
{
"epoch": 0.39200795588102344,
"grad_norm": 15.3671875,
"learning_rate": 9.969374499581018e-07,
"loss": 25.1051,
"step": 5420
},
{
"epoch": 0.392731217792243,
"grad_norm": 15.71875,
"learning_rate": 9.969317994967698e-07,
"loss": 24.9415,
"step": 5430
},
{
"epoch": 0.3934544797034626,
"grad_norm": 16.109375,
"learning_rate": 9.96926149035438e-07,
"loss": 25.4558,
"step": 5440
},
{
"epoch": 0.3941777416146822,
"grad_norm": 16.5,
"learning_rate": 9.96920498574106e-07,
"loss": 25.084,
"step": 5450
},
{
"epoch": 0.3949010035259018,
"grad_norm": 17.625,
"learning_rate": 9.969148481127742e-07,
"loss": 24.7463,
"step": 5460
},
{
"epoch": 0.3956242654371214,
"grad_norm": 16.234375,
"learning_rate": 9.969091976514422e-07,
"loss": 24.7555,
"step": 5470
},
{
"epoch": 0.396347527348341,
"grad_norm": 15.7421875,
"learning_rate": 9.969035471901102e-07,
"loss": 25.408,
"step": 5480
},
{
"epoch": 0.39707078925956063,
"grad_norm": 16.859375,
"learning_rate": 9.968978967287784e-07,
"loss": 25.1242,
"step": 5490
},
{
"epoch": 0.3977940511707802,
"grad_norm": 16.78125,
"learning_rate": 9.968922462674466e-07,
"loss": 25.2741,
"step": 5500
},
{
"epoch": 0.39851731308199984,
"grad_norm": 15.078125,
"learning_rate": 9.968865958061146e-07,
"loss": 24.884,
"step": 5510
},
{
"epoch": 0.3992405749932194,
"grad_norm": 16.421875,
"learning_rate": 9.968809453447826e-07,
"loss": 25.2885,
"step": 5520
},
{
"epoch": 0.39996383690443904,
"grad_norm": 16.40625,
"learning_rate": 9.968752948834506e-07,
"loss": 25.4121,
"step": 5530
},
{
"epoch": 0.4006870988156586,
"grad_norm": 16.78125,
"learning_rate": 9.968696444221188e-07,
"loss": 25.0091,
"step": 5540
},
{
"epoch": 0.40141036072687825,
"grad_norm": 16.140625,
"learning_rate": 9.96863993960787e-07,
"loss": 25.1445,
"step": 5550
},
{
"epoch": 0.4021336226380978,
"grad_norm": 15.875,
"learning_rate": 9.96858343499455e-07,
"loss": 24.9004,
"step": 5560
},
{
"epoch": 0.4028568845493174,
"grad_norm": 15.65625,
"learning_rate": 9.96852693038123e-07,
"loss": 25.0035,
"step": 5570
},
{
"epoch": 0.403580146460537,
"grad_norm": 15.796875,
"learning_rate": 9.96847042576791e-07,
"loss": 25.2146,
"step": 5580
},
{
"epoch": 0.4043034083717566,
"grad_norm": 15.6875,
"learning_rate": 9.968413921154592e-07,
"loss": 24.7197,
"step": 5590
},
{
"epoch": 0.40502667028297623,
"grad_norm": 16.6875,
"learning_rate": 9.968357416541274e-07,
"loss": 25.1168,
"step": 5600
},
{
"epoch": 0.4057499321941958,
"grad_norm": 16.109375,
"learning_rate": 9.968300911927954e-07,
"loss": 25.379,
"step": 5610
},
{
"epoch": 0.40647319410541544,
"grad_norm": 16.578125,
"learning_rate": 9.968244407314634e-07,
"loss": 24.966,
"step": 5620
},
{
"epoch": 0.407196456016635,
"grad_norm": 15.5390625,
"learning_rate": 9.968187902701316e-07,
"loss": 24.6402,
"step": 5630
},
{
"epoch": 0.40791971792785464,
"grad_norm": 16.46875,
"learning_rate": 9.968131398087996e-07,
"loss": 25.3287,
"step": 5640
},
{
"epoch": 0.4086429798390742,
"grad_norm": 16.34375,
"learning_rate": 9.968074893474676e-07,
"loss": 24.8552,
"step": 5650
},
{
"epoch": 0.40936624175029385,
"grad_norm": 16.78125,
"learning_rate": 9.968018388861358e-07,
"loss": 25.2226,
"step": 5660
},
{
"epoch": 0.4100895036615134,
"grad_norm": 16.3125,
"learning_rate": 9.96796188424804e-07,
"loss": 25.3465,
"step": 5670
},
{
"epoch": 0.41081276557273305,
"grad_norm": 16.28125,
"learning_rate": 9.96790537963472e-07,
"loss": 25.0835,
"step": 5680
},
{
"epoch": 0.4115360274839526,
"grad_norm": 16.4375,
"learning_rate": 9.9678488750214e-07,
"loss": 25.0594,
"step": 5690
},
{
"epoch": 0.4122592893951722,
"grad_norm": 16.046875,
"learning_rate": 9.96779237040808e-07,
"loss": 25.1546,
"step": 5700
},
{
"epoch": 0.41298255130639183,
"grad_norm": 16.53125,
"learning_rate": 9.967735865794762e-07,
"loss": 25.1157,
"step": 5710
},
{
"epoch": 0.4137058132176114,
"grad_norm": 16.265625,
"learning_rate": 9.967679361181444e-07,
"loss": 25.1408,
"step": 5720
},
{
"epoch": 0.41442907512883104,
"grad_norm": 15.8515625,
"learning_rate": 9.967622856568124e-07,
"loss": 24.9652,
"step": 5730
},
{
"epoch": 0.4151523370400506,
"grad_norm": 17.25,
"learning_rate": 9.967566351954804e-07,
"loss": 25.4921,
"step": 5740
},
{
"epoch": 0.41587559895127024,
"grad_norm": 16.765625,
"learning_rate": 9.967509847341484e-07,
"loss": 24.9062,
"step": 5750
},
{
"epoch": 0.4165988608624898,
"grad_norm": 16.59375,
"learning_rate": 9.967453342728166e-07,
"loss": 25.5228,
"step": 5760
},
{
"epoch": 0.41732212277370945,
"grad_norm": 16.3125,
"learning_rate": 9.967396838114849e-07,
"loss": 24.9683,
"step": 5770
},
{
"epoch": 0.418045384684929,
"grad_norm": 16.15625,
"learning_rate": 9.967340333501529e-07,
"loss": 24.8785,
"step": 5780
},
{
"epoch": 0.41876864659614865,
"grad_norm": 16.234375,
"learning_rate": 9.967283828888209e-07,
"loss": 25.188,
"step": 5790
},
{
"epoch": 0.4194919085073682,
"grad_norm": 16.453125,
"learning_rate": 9.967227324274889e-07,
"loss": 25.6373,
"step": 5800
},
{
"epoch": 0.42021517041858786,
"grad_norm": 16.8125,
"learning_rate": 9.96717081966157e-07,
"loss": 25.1054,
"step": 5810
},
{
"epoch": 0.42093843232980743,
"grad_norm": 16.921875,
"learning_rate": 9.967114315048253e-07,
"loss": 25.075,
"step": 5820
},
{
"epoch": 0.421661694241027,
"grad_norm": 16.28125,
"learning_rate": 9.967057810434933e-07,
"loss": 24.8702,
"step": 5830
},
{
"epoch": 0.42238495615224664,
"grad_norm": 15.7890625,
"learning_rate": 9.967001305821613e-07,
"loss": 25.1268,
"step": 5840
},
{
"epoch": 0.4231082180634662,
"grad_norm": 16.53125,
"learning_rate": 9.966944801208295e-07,
"loss": 25.222,
"step": 5850
},
{
"epoch": 0.42383147997468584,
"grad_norm": 15.78125,
"learning_rate": 9.966888296594975e-07,
"loss": 24.7222,
"step": 5860
},
{
"epoch": 0.4245547418859054,
"grad_norm": 17.375,
"learning_rate": 9.966831791981657e-07,
"loss": 24.8381,
"step": 5870
},
{
"epoch": 0.42527800379712505,
"grad_norm": 15.6953125,
"learning_rate": 9.966775287368337e-07,
"loss": 25.3032,
"step": 5880
},
{
"epoch": 0.4260012657083446,
"grad_norm": 16.484375,
"learning_rate": 9.966718782755019e-07,
"loss": 25.0616,
"step": 5890
},
{
"epoch": 0.42672452761956425,
"grad_norm": 16.296875,
"learning_rate": 9.966662278141699e-07,
"loss": 25.1365,
"step": 5900
},
{
"epoch": 0.42744778953078383,
"grad_norm": 16.296875,
"learning_rate": 9.966605773528379e-07,
"loss": 24.87,
"step": 5910
},
{
"epoch": 0.42817105144200346,
"grad_norm": 16.421875,
"learning_rate": 9.96654926891506e-07,
"loss": 25.0335,
"step": 5920
},
{
"epoch": 0.42889431335322303,
"grad_norm": 15.359375,
"learning_rate": 9.96649276430174e-07,
"loss": 25.2336,
"step": 5930
},
{
"epoch": 0.42961757526444266,
"grad_norm": 15.484375,
"learning_rate": 9.966436259688423e-07,
"loss": 24.7188,
"step": 5940
},
{
"epoch": 0.43034083717566224,
"grad_norm": 16.296875,
"learning_rate": 9.966379755075103e-07,
"loss": 25.243,
"step": 5950
},
{
"epoch": 0.4310640990868818,
"grad_norm": 15.8359375,
"learning_rate": 9.966323250461783e-07,
"loss": 25.4129,
"step": 5960
},
{
"epoch": 0.43178736099810144,
"grad_norm": 16.828125,
"learning_rate": 9.966266745848463e-07,
"loss": 25.113,
"step": 5970
},
{
"epoch": 0.432510622909321,
"grad_norm": 16.6875,
"learning_rate": 9.966210241235145e-07,
"loss": 25.1597,
"step": 5980
},
{
"epoch": 0.43323388482054065,
"grad_norm": 17.65625,
"learning_rate": 9.966153736621827e-07,
"loss": 25.085,
"step": 5990
},
{
"epoch": 0.4339571467317602,
"grad_norm": 17.078125,
"learning_rate": 9.966097232008507e-07,
"loss": 25.0908,
"step": 6000
},
{
"epoch": 0.43468040864297985,
"grad_norm": 16.578125,
"learning_rate": 9.966040727395187e-07,
"loss": 25.1842,
"step": 6010
},
{
"epoch": 0.43540367055419943,
"grad_norm": 16.890625,
"learning_rate": 9.96598422278187e-07,
"loss": 25.157,
"step": 6020
},
{
"epoch": 0.43612693246541906,
"grad_norm": 15.8984375,
"learning_rate": 9.96592771816855e-07,
"loss": 25.04,
"step": 6030
},
{
"epoch": 0.43685019437663863,
"grad_norm": 17.546875,
"learning_rate": 9.965871213555231e-07,
"loss": 25.1817,
"step": 6040
},
{
"epoch": 0.43757345628785826,
"grad_norm": 17.375,
"learning_rate": 9.965814708941911e-07,
"loss": 25.4064,
"step": 6050
},
{
"epoch": 0.43829671819907784,
"grad_norm": 16.703125,
"learning_rate": 9.965758204328591e-07,
"loss": 25.0107,
"step": 6060
},
{
"epoch": 0.4390199801102974,
"grad_norm": 16.828125,
"learning_rate": 9.965701699715273e-07,
"loss": 25.1997,
"step": 6070
},
{
"epoch": 0.43974324202151704,
"grad_norm": 16.171875,
"learning_rate": 9.965645195101953e-07,
"loss": 25.2761,
"step": 6080
},
{
"epoch": 0.4404665039327366,
"grad_norm": 15.578125,
"learning_rate": 9.965588690488635e-07,
"loss": 24.9903,
"step": 6090
},
{
"epoch": 0.44118976584395625,
"grad_norm": 15.625,
"learning_rate": 9.965532185875315e-07,
"loss": 24.8739,
"step": 6100
},
{
"epoch": 0.4419130277551758,
"grad_norm": 15.7734375,
"learning_rate": 9.965475681261997e-07,
"loss": 25.2024,
"step": 6110
},
{
"epoch": 0.44263628966639545,
"grad_norm": 15.9296875,
"learning_rate": 9.965419176648677e-07,
"loss": 25.122,
"step": 6120
},
{
"epoch": 0.44335955157761503,
"grad_norm": 17.40625,
"learning_rate": 9.965362672035357e-07,
"loss": 25.1236,
"step": 6130
},
{
"epoch": 0.44408281348883466,
"grad_norm": 16.234375,
"learning_rate": 9.96530616742204e-07,
"loss": 24.9833,
"step": 6140
},
{
"epoch": 0.44480607540005423,
"grad_norm": 16.28125,
"learning_rate": 9.96524966280872e-07,
"loss": 25.1377,
"step": 6150
},
{
"epoch": 0.44552933731127387,
"grad_norm": 16.078125,
"learning_rate": 9.965193158195401e-07,
"loss": 25.0597,
"step": 6160
},
{
"epoch": 0.44625259922249344,
"grad_norm": 15.96875,
"learning_rate": 9.965136653582081e-07,
"loss": 25.0991,
"step": 6170
},
{
"epoch": 0.44697586113371307,
"grad_norm": 15.8984375,
"learning_rate": 9.965080148968761e-07,
"loss": 25.0012,
"step": 6180
},
{
"epoch": 0.44769912304493265,
"grad_norm": 15.984375,
"learning_rate": 9.965023644355444e-07,
"loss": 24.924,
"step": 6190
},
{
"epoch": 0.4484223849561522,
"grad_norm": 16.21875,
"learning_rate": 9.964967139742123e-07,
"loss": 25.307,
"step": 6200
},
{
"epoch": 0.44914564686737185,
"grad_norm": 16.3125,
"learning_rate": 9.964910635128806e-07,
"loss": 25.1692,
"step": 6210
},
{
"epoch": 0.4498689087785914,
"grad_norm": 17.046875,
"learning_rate": 9.964854130515486e-07,
"loss": 25.219,
"step": 6220
},
{
"epoch": 0.45059217068981106,
"grad_norm": 16.5625,
"learning_rate": 9.964797625902166e-07,
"loss": 24.9296,
"step": 6230
},
{
"epoch": 0.45131543260103063,
"grad_norm": 16.78125,
"learning_rate": 9.964741121288848e-07,
"loss": 25.1963,
"step": 6240
},
{
"epoch": 0.45203869451225026,
"grad_norm": 17.078125,
"learning_rate": 9.964684616675528e-07,
"loss": 24.9102,
"step": 6250
},
{
"epoch": 0.45276195642346984,
"grad_norm": 15.9765625,
"learning_rate": 9.96462811206221e-07,
"loss": 25.297,
"step": 6260
},
{
"epoch": 0.45348521833468947,
"grad_norm": 16.109375,
"learning_rate": 9.96457160744889e-07,
"loss": 24.8933,
"step": 6270
},
{
"epoch": 0.45420848024590904,
"grad_norm": 15.9609375,
"learning_rate": 9.96451510283557e-07,
"loss": 25.0007,
"step": 6280
},
{
"epoch": 0.45493174215712867,
"grad_norm": 16.28125,
"learning_rate": 9.964458598222252e-07,
"loss": 25.013,
"step": 6290
},
{
"epoch": 0.45565500406834825,
"grad_norm": 18.0625,
"learning_rate": 9.964402093608932e-07,
"loss": 25.5618,
"step": 6300
},
{
"epoch": 0.4563782659795679,
"grad_norm": 16.21875,
"learning_rate": 9.964345588995614e-07,
"loss": 25.056,
"step": 6310
},
{
"epoch": 0.45710152789078745,
"grad_norm": 16.59375,
"learning_rate": 9.964289084382294e-07,
"loss": 24.9934,
"step": 6320
},
{
"epoch": 0.457824789802007,
"grad_norm": 16.03125,
"learning_rate": 9.964232579768976e-07,
"loss": 25.0928,
"step": 6330
},
{
"epoch": 0.45854805171322666,
"grad_norm": 16.75,
"learning_rate": 9.964176075155656e-07,
"loss": 25.194,
"step": 6340
},
{
"epoch": 0.45927131362444623,
"grad_norm": 16.296875,
"learning_rate": 9.964119570542336e-07,
"loss": 24.94,
"step": 6350
},
{
"epoch": 0.45999457553566586,
"grad_norm": 17.671875,
"learning_rate": 9.964063065929018e-07,
"loss": 25.4812,
"step": 6360
},
{
"epoch": 0.46071783744688544,
"grad_norm": 16.578125,
"learning_rate": 9.964006561315698e-07,
"loss": 25.5154,
"step": 6370
},
{
"epoch": 0.46144109935810507,
"grad_norm": 15.703125,
"learning_rate": 9.96395005670238e-07,
"loss": 25.3724,
"step": 6380
},
{
"epoch": 0.46216436126932464,
"grad_norm": 16.890625,
"learning_rate": 9.96389355208906e-07,
"loss": 25.0829,
"step": 6390
},
{
"epoch": 0.46288762318054427,
"grad_norm": 15.7734375,
"learning_rate": 9.96383704747574e-07,
"loss": 25.0513,
"step": 6400
},
{
"epoch": 0.46361088509176385,
"grad_norm": 16.96875,
"learning_rate": 9.963780542862422e-07,
"loss": 25.1159,
"step": 6410
},
{
"epoch": 0.4643341470029835,
"grad_norm": 15.6484375,
"learning_rate": 9.963724038249102e-07,
"loss": 24.9635,
"step": 6420
},
{
"epoch": 0.46505740891420305,
"grad_norm": 17.078125,
"learning_rate": 9.963667533635784e-07,
"loss": 25.3004,
"step": 6430
},
{
"epoch": 0.4657806708254227,
"grad_norm": 16.453125,
"learning_rate": 9.963611029022464e-07,
"loss": 25.2261,
"step": 6440
},
{
"epoch": 0.46650393273664226,
"grad_norm": 16.875,
"learning_rate": 9.963554524409144e-07,
"loss": 25.2098,
"step": 6450
},
{
"epoch": 0.46722719464786183,
"grad_norm": 15.6640625,
"learning_rate": 9.963498019795826e-07,
"loss": 25.0369,
"step": 6460
},
{
"epoch": 0.46795045655908146,
"grad_norm": 16.203125,
"learning_rate": 9.963441515182506e-07,
"loss": 25.028,
"step": 6470
},
{
"epoch": 0.46867371847030104,
"grad_norm": 15.9140625,
"learning_rate": 9.963385010569188e-07,
"loss": 24.9715,
"step": 6480
},
{
"epoch": 0.46939698038152067,
"grad_norm": 17.828125,
"learning_rate": 9.963328505955868e-07,
"loss": 25.007,
"step": 6490
},
{
"epoch": 0.47012024229274024,
"grad_norm": 16.1875,
"learning_rate": 9.963272001342548e-07,
"loss": 25.0585,
"step": 6500
},
{
"epoch": 0.4708435042039599,
"grad_norm": 15.1484375,
"learning_rate": 9.96321549672923e-07,
"loss": 25.3392,
"step": 6510
},
{
"epoch": 0.47156676611517945,
"grad_norm": 16.25,
"learning_rate": 9.96315899211591e-07,
"loss": 25.2201,
"step": 6520
},
{
"epoch": 0.4722900280263991,
"grad_norm": 16.171875,
"learning_rate": 9.963102487502592e-07,
"loss": 24.9467,
"step": 6530
},
{
"epoch": 0.47301328993761865,
"grad_norm": 16.5625,
"learning_rate": 9.963045982889272e-07,
"loss": 25.1083,
"step": 6540
},
{
"epoch": 0.4737365518488383,
"grad_norm": 16.515625,
"learning_rate": 9.962989478275954e-07,
"loss": 25.3217,
"step": 6550
},
{
"epoch": 0.47445981376005786,
"grad_norm": 15.578125,
"learning_rate": 9.962932973662634e-07,
"loss": 25.1594,
"step": 6560
},
{
"epoch": 0.4751830756712775,
"grad_norm": 17.09375,
"learning_rate": 9.962876469049314e-07,
"loss": 25.0634,
"step": 6570
},
{
"epoch": 0.47590633758249706,
"grad_norm": 16.984375,
"learning_rate": 9.962819964435996e-07,
"loss": 25.3349,
"step": 6580
},
{
"epoch": 0.47662959949371664,
"grad_norm": 15.734375,
"learning_rate": 9.962763459822676e-07,
"loss": 25.1982,
"step": 6590
},
{
"epoch": 0.47735286140493627,
"grad_norm": 16.390625,
"learning_rate": 9.962706955209358e-07,
"loss": 25.1291,
"step": 6600
},
{
"epoch": 0.47807612331615584,
"grad_norm": 16.3125,
"learning_rate": 9.962650450596038e-07,
"loss": 25.1213,
"step": 6610
},
{
"epoch": 0.4787993852273755,
"grad_norm": 16.546875,
"learning_rate": 9.962593945982718e-07,
"loss": 24.9788,
"step": 6620
},
{
"epoch": 0.47952264713859505,
"grad_norm": 15.84375,
"learning_rate": 9.9625374413694e-07,
"loss": 24.6411,
"step": 6630
},
{
"epoch": 0.4802459090498147,
"grad_norm": 15.8359375,
"learning_rate": 9.96248093675608e-07,
"loss": 24.8477,
"step": 6640
},
{
"epoch": 0.48096917096103425,
"grad_norm": 16.234375,
"learning_rate": 9.962424432142763e-07,
"loss": 25.1838,
"step": 6650
},
{
"epoch": 0.4816924328722539,
"grad_norm": 15.03125,
"learning_rate": 9.962367927529443e-07,
"loss": 24.8121,
"step": 6660
},
{
"epoch": 0.48241569478347346,
"grad_norm": 16.28125,
"learning_rate": 9.962311422916122e-07,
"loss": 24.8409,
"step": 6670
},
{
"epoch": 0.4831389566946931,
"grad_norm": 16.6875,
"learning_rate": 9.962254918302805e-07,
"loss": 24.7554,
"step": 6680
},
{
"epoch": 0.48386221860591266,
"grad_norm": 17.90625,
"learning_rate": 9.962198413689485e-07,
"loss": 25.3045,
"step": 6690
},
{
"epoch": 0.48458548051713224,
"grad_norm": 16.859375,
"learning_rate": 9.962141909076167e-07,
"loss": 25.1306,
"step": 6700
},
{
"epoch": 0.48530874242835187,
"grad_norm": 16.65625,
"learning_rate": 9.962085404462847e-07,
"loss": 25.1144,
"step": 6710
},
{
"epoch": 0.48603200433957144,
"grad_norm": 16.15625,
"learning_rate": 9.962028899849527e-07,
"loss": 24.9379,
"step": 6720
},
{
"epoch": 0.4867552662507911,
"grad_norm": 17.609375,
"learning_rate": 9.961972395236209e-07,
"loss": 24.8681,
"step": 6730
},
{
"epoch": 0.48747852816201065,
"grad_norm": 16.359375,
"learning_rate": 9.961915890622889e-07,
"loss": 24.836,
"step": 6740
},
{
"epoch": 0.4882017900732303,
"grad_norm": 16.3125,
"learning_rate": 9.96185938600957e-07,
"loss": 24.7874,
"step": 6750
},
{
"epoch": 0.48892505198444985,
"grad_norm": 17.46875,
"learning_rate": 9.96180288139625e-07,
"loss": 25.4011,
"step": 6760
},
{
"epoch": 0.4896483138956695,
"grad_norm": 15.1484375,
"learning_rate": 9.961746376782933e-07,
"loss": 25.1523,
"step": 6770
},
{
"epoch": 0.49037157580688906,
"grad_norm": 17.4375,
"learning_rate": 9.961689872169613e-07,
"loss": 25.3958,
"step": 6780
},
{
"epoch": 0.4910948377181087,
"grad_norm": 15.6796875,
"learning_rate": 9.961633367556293e-07,
"loss": 25.3325,
"step": 6790
},
{
"epoch": 0.49181809962932826,
"grad_norm": 22.125,
"learning_rate": 9.961576862942975e-07,
"loss": 25.1442,
"step": 6800
},
{
"epoch": 0.4925413615405479,
"grad_norm": 16.765625,
"learning_rate": 9.961520358329655e-07,
"loss": 25.2003,
"step": 6810
},
{
"epoch": 0.49326462345176747,
"grad_norm": 15.6640625,
"learning_rate": 9.961463853716337e-07,
"loss": 25.221,
"step": 6820
},
{
"epoch": 0.49398788536298704,
"grad_norm": 17.015625,
"learning_rate": 9.961407349103017e-07,
"loss": 24.856,
"step": 6830
},
{
"epoch": 0.4947111472742067,
"grad_norm": 16.359375,
"learning_rate": 9.961350844489697e-07,
"loss": 25.1671,
"step": 6840
},
{
"epoch": 0.49543440918542625,
"grad_norm": 15.8828125,
"learning_rate": 9.96129433987638e-07,
"loss": 25.145,
"step": 6850
},
{
"epoch": 0.4961576710966459,
"grad_norm": 16.15625,
"learning_rate": 9.961237835263059e-07,
"loss": 25.2164,
"step": 6860
},
{
"epoch": 0.49688093300786546,
"grad_norm": 17.078125,
"learning_rate": 9.96118133064974e-07,
"loss": 25.113,
"step": 6870
},
{
"epoch": 0.4976041949190851,
"grad_norm": 16.640625,
"learning_rate": 9.96112482603642e-07,
"loss": 24.7216,
"step": 6880
},
{
"epoch": 0.49832745683030466,
"grad_norm": 16.5,
"learning_rate": 9.9610683214231e-07,
"loss": 24.9603,
"step": 6890
},
{
"epoch": 0.4990507187415243,
"grad_norm": 16.0,
"learning_rate": 9.961011816809783e-07,
"loss": 25.1272,
"step": 6900
},
{
"epoch": 0.49977398065274387,
"grad_norm": 16.296875,
"learning_rate": 9.960955312196463e-07,
"loss": 24.8981,
"step": 6910
},
{
"epoch": 0.5004972425639634,
"grad_norm": 16.03125,
"learning_rate": 9.960898807583145e-07,
"loss": 25.1166,
"step": 6920
},
{
"epoch": 0.5012205044751831,
"grad_norm": 16.734375,
"learning_rate": 9.960842302969825e-07,
"loss": 25.522,
"step": 6930
},
{
"epoch": 0.5019437663864027,
"grad_norm": 16.203125,
"learning_rate": 9.960785798356507e-07,
"loss": 24.7892,
"step": 6940
},
{
"epoch": 0.5026670282976223,
"grad_norm": 16.890625,
"learning_rate": 9.960729293743187e-07,
"loss": 24.6813,
"step": 6950
},
{
"epoch": 0.5033902902088419,
"grad_norm": 15.59375,
"learning_rate": 9.960672789129867e-07,
"loss": 24.9678,
"step": 6960
},
{
"epoch": 0.5041135521200615,
"grad_norm": 16.34375,
"learning_rate": 9.96061628451655e-07,
"loss": 25.2393,
"step": 6970
},
{
"epoch": 0.5048368140312811,
"grad_norm": 15.8046875,
"learning_rate": 9.96055977990323e-07,
"loss": 25.1625,
"step": 6980
},
{
"epoch": 0.5055600759425006,
"grad_norm": 15.3671875,
"learning_rate": 9.960503275289911e-07,
"loss": 25.1241,
"step": 6990
},
{
"epoch": 0.5062833378537203,
"grad_norm": 16.3125,
"learning_rate": 9.960446770676591e-07,
"loss": 25.0165,
"step": 7000
},
{
"epoch": 0.5070065997649399,
"grad_norm": 15.8515625,
"learning_rate": 9.960390266063271e-07,
"loss": 25.5291,
"step": 7010
},
{
"epoch": 0.5077298616761595,
"grad_norm": 16.46875,
"learning_rate": 9.960333761449953e-07,
"loss": 25.1357,
"step": 7020
},
{
"epoch": 0.508453123587379,
"grad_norm": 15.515625,
"learning_rate": 9.960277256836635e-07,
"loss": 25.2483,
"step": 7030
},
{
"epoch": 0.5091763854985987,
"grad_norm": 15.9609375,
"learning_rate": 9.960220752223315e-07,
"loss": 24.7912,
"step": 7040
},
{
"epoch": 0.5098996474098183,
"grad_norm": 16.515625,
"learning_rate": 9.960164247609995e-07,
"loss": 25.3561,
"step": 7050
},
{
"epoch": 0.5106229093210379,
"grad_norm": 16.25,
"learning_rate": 9.960107742996675e-07,
"loss": 25.3195,
"step": 7060
},
{
"epoch": 0.5113461712322575,
"grad_norm": 16.859375,
"learning_rate": 9.960051238383357e-07,
"loss": 25.1948,
"step": 7070
},
{
"epoch": 0.5120694331434771,
"grad_norm": 15.890625,
"learning_rate": 9.95999473377004e-07,
"loss": 25.1284,
"step": 7080
},
{
"epoch": 0.5127926950546967,
"grad_norm": 16.296875,
"learning_rate": 9.95993822915672e-07,
"loss": 25.0781,
"step": 7090
},
{
"epoch": 0.5135159569659162,
"grad_norm": 16.65625,
"learning_rate": 9.9598817245434e-07,
"loss": 25.2929,
"step": 7100
},
{
"epoch": 0.5142392188771359,
"grad_norm": 16.5,
"learning_rate": 9.95982521993008e-07,
"loss": 25.0102,
"step": 7110
},
{
"epoch": 0.5149624807883555,
"grad_norm": 16.46875,
"learning_rate": 9.959768715316762e-07,
"loss": 25.1572,
"step": 7120
},
{
"epoch": 0.5156857426995751,
"grad_norm": 15.265625,
"learning_rate": 9.959712210703442e-07,
"loss": 25.0643,
"step": 7130
},
{
"epoch": 0.5164090046107946,
"grad_norm": 16.296875,
"learning_rate": 9.959655706090124e-07,
"loss": 25.0202,
"step": 7140
},
{
"epoch": 0.5171322665220143,
"grad_norm": 15.6484375,
"learning_rate": 9.959599201476804e-07,
"loss": 24.9407,
"step": 7150
},
{
"epoch": 0.5178555284332339,
"grad_norm": 16.765625,
"learning_rate": 9.959542696863486e-07,
"loss": 25.3482,
"step": 7160
},
{
"epoch": 0.5185787903444535,
"grad_norm": 16.296875,
"learning_rate": 9.959486192250166e-07,
"loss": 24.7983,
"step": 7170
},
{
"epoch": 0.519302052255673,
"grad_norm": 15.140625,
"learning_rate": 9.959429687636846e-07,
"loss": 24.9558,
"step": 7180
},
{
"epoch": 0.5200253141668927,
"grad_norm": 16.09375,
"learning_rate": 9.959373183023528e-07,
"loss": 25.2437,
"step": 7190
},
{
"epoch": 0.5207485760781123,
"grad_norm": 17.4375,
"learning_rate": 9.959316678410208e-07,
"loss": 24.9945,
"step": 7200
},
{
"epoch": 0.5214718379893318,
"grad_norm": 16.78125,
"learning_rate": 9.95926017379689e-07,
"loss": 24.876,
"step": 7210
},
{
"epoch": 0.5221950999005515,
"grad_norm": 15.6640625,
"learning_rate": 9.95920366918357e-07,
"loss": 25.1469,
"step": 7220
},
{
"epoch": 0.5229183618117711,
"grad_norm": 16.34375,
"learning_rate": 9.95914716457025e-07,
"loss": 25.0436,
"step": 7230
},
{
"epoch": 0.5236416237229907,
"grad_norm": 15.25,
"learning_rate": 9.959090659956932e-07,
"loss": 25.0983,
"step": 7240
},
{
"epoch": 0.5243648856342102,
"grad_norm": 16.578125,
"learning_rate": 9.959034155343614e-07,
"loss": 25.1828,
"step": 7250
},
{
"epoch": 0.5250881475454299,
"grad_norm": 15.375,
"learning_rate": 9.958977650730294e-07,
"loss": 25.0138,
"step": 7260
},
{
"epoch": 0.5258114094566495,
"grad_norm": 15.515625,
"learning_rate": 9.958921146116974e-07,
"loss": 25.336,
"step": 7270
},
{
"epoch": 0.5265346713678691,
"grad_norm": 16.59375,
"learning_rate": 9.958864641503654e-07,
"loss": 25.1535,
"step": 7280
},
{
"epoch": 0.5272579332790887,
"grad_norm": 15.8828125,
"learning_rate": 9.958808136890336e-07,
"loss": 25.1138,
"step": 7290
},
{
"epoch": 0.5279811951903083,
"grad_norm": 16.203125,
"learning_rate": 9.958751632277018e-07,
"loss": 24.8106,
"step": 7300
},
{
"epoch": 0.5287044571015279,
"grad_norm": 16.296875,
"learning_rate": 9.958695127663698e-07,
"loss": 24.7198,
"step": 7310
},
{
"epoch": 0.5294277190127475,
"grad_norm": 16.140625,
"learning_rate": 9.958638623050378e-07,
"loss": 24.796,
"step": 7320
},
{
"epoch": 0.5301509809239671,
"grad_norm": 16.390625,
"learning_rate": 9.958582118437058e-07,
"loss": 24.8893,
"step": 7330
},
{
"epoch": 0.5308742428351867,
"grad_norm": 15.859375,
"learning_rate": 9.95852561382374e-07,
"loss": 25.1709,
"step": 7340
},
{
"epoch": 0.5315975047464063,
"grad_norm": 15.9453125,
"learning_rate": 9.958469109210422e-07,
"loss": 25.0314,
"step": 7350
},
{
"epoch": 0.5323207666576258,
"grad_norm": 16.296875,
"learning_rate": 9.958412604597102e-07,
"loss": 24.7342,
"step": 7360
},
{
"epoch": 0.5330440285688455,
"grad_norm": 15.9140625,
"learning_rate": 9.958356099983782e-07,
"loss": 24.7556,
"step": 7370
},
{
"epoch": 0.5337672904800651,
"grad_norm": 16.375,
"learning_rate": 9.958299595370464e-07,
"loss": 25.1313,
"step": 7380
},
{
"epoch": 0.5344905523912847,
"grad_norm": 16.75,
"learning_rate": 9.958243090757144e-07,
"loss": 25.1914,
"step": 7390
},
{
"epoch": 0.5352138143025043,
"grad_norm": 15.2109375,
"learning_rate": 9.958186586143826e-07,
"loss": 25.0681,
"step": 7400
},
{
"epoch": 0.5359370762137239,
"grad_norm": 16.40625,
"learning_rate": 9.958130081530506e-07,
"loss": 25.4192,
"step": 7410
},
{
"epoch": 0.5366603381249435,
"grad_norm": 15.515625,
"learning_rate": 9.958073576917186e-07,
"loss": 25.2562,
"step": 7420
},
{
"epoch": 0.5373836000361631,
"grad_norm": 15.7109375,
"learning_rate": 9.958017072303868e-07,
"loss": 25.1077,
"step": 7430
},
{
"epoch": 0.5381068619473827,
"grad_norm": 17.1875,
"learning_rate": 9.957960567690548e-07,
"loss": 25.1754,
"step": 7440
},
{
"epoch": 0.5388301238586023,
"grad_norm": 16.34375,
"learning_rate": 9.95790406307723e-07,
"loss": 25.0318,
"step": 7450
},
{
"epoch": 0.5395533857698219,
"grad_norm": 16.375,
"learning_rate": 9.95784755846391e-07,
"loss": 25.2289,
"step": 7460
},
{
"epoch": 0.5402766476810414,
"grad_norm": 15.21875,
"learning_rate": 9.957791053850592e-07,
"loss": 24.9497,
"step": 7470
},
{
"epoch": 0.5409999095922611,
"grad_norm": 15.671875,
"learning_rate": 9.957734549237272e-07,
"loss": 25.0775,
"step": 7480
},
{
"epoch": 0.5417231715034807,
"grad_norm": 16.5,
"learning_rate": 9.957678044623952e-07,
"loss": 25.3473,
"step": 7490
},
{
"epoch": 0.5424464334147003,
"grad_norm": 16.734375,
"learning_rate": 9.957621540010632e-07,
"loss": 25.1996,
"step": 7500
},
{
"epoch": 0.5424464334147003,
"eval_loss": 1.5896706581115723,
"eval_runtime": 371.382,
"eval_samples_per_second": 1955.019,
"eval_steps_per_second": 30.548,
"step": 7500
},
{
"epoch": 0.5431696953259199,
"grad_norm": 17.65625,
"learning_rate": 9.957565035397314e-07,
"loss": 25.4949,
"step": 7510
},
{
"epoch": 0.5438929572371395,
"grad_norm": 15.7265625,
"learning_rate": 9.957508530783997e-07,
"loss": 25.2191,
"step": 7520
},
{
"epoch": 0.5446162191483591,
"grad_norm": 16.109375,
"learning_rate": 9.957452026170676e-07,
"loss": 25.1094,
"step": 7530
},
{
"epoch": 0.5453394810595787,
"grad_norm": 16.703125,
"learning_rate": 9.957395521557356e-07,
"loss": 24.9443,
"step": 7540
},
{
"epoch": 0.5460627429707983,
"grad_norm": 15.3515625,
"learning_rate": 9.957339016944036e-07,
"loss": 25.173,
"step": 7550
},
{
"epoch": 0.5467860048820179,
"grad_norm": 16.21875,
"learning_rate": 9.957282512330719e-07,
"loss": 24.9756,
"step": 7560
},
{
"epoch": 0.5475092667932375,
"grad_norm": 16.21875,
"learning_rate": 9.9572260077174e-07,
"loss": 24.9386,
"step": 7570
},
{
"epoch": 0.5482325287044572,
"grad_norm": 16.109375,
"learning_rate": 9.95716950310408e-07,
"loss": 24.7594,
"step": 7580
},
{
"epoch": 0.5489557906156767,
"grad_norm": 16.75,
"learning_rate": 9.95711299849076e-07,
"loss": 25.0118,
"step": 7590
},
{
"epoch": 0.5496790525268963,
"grad_norm": 16.34375,
"learning_rate": 9.957056493877443e-07,
"loss": 24.9402,
"step": 7600
},
{
"epoch": 0.5504023144381159,
"grad_norm": 16.5,
"learning_rate": 9.956999989264123e-07,
"loss": 25.188,
"step": 7610
},
{
"epoch": 0.5511255763493355,
"grad_norm": 14.90625,
"learning_rate": 9.956943484650805e-07,
"loss": 25.4746,
"step": 7620
},
{
"epoch": 0.5518488382605551,
"grad_norm": 16.265625,
"learning_rate": 9.956886980037485e-07,
"loss": 25.1864,
"step": 7630
},
{
"epoch": 0.5525721001717747,
"grad_norm": 16.328125,
"learning_rate": 9.956830475424167e-07,
"loss": 25.0443,
"step": 7640
},
{
"epoch": 0.5532953620829943,
"grad_norm": 15.890625,
"learning_rate": 9.956773970810847e-07,
"loss": 24.9956,
"step": 7650
},
{
"epoch": 0.5540186239942139,
"grad_norm": 16.203125,
"learning_rate": 9.956717466197527e-07,
"loss": 25.133,
"step": 7660
},
{
"epoch": 0.5547418859054335,
"grad_norm": 15.78125,
"learning_rate": 9.956660961584209e-07,
"loss": 25.1481,
"step": 7670
},
{
"epoch": 0.5554651478166531,
"grad_norm": 16.015625,
"learning_rate": 9.956604456970889e-07,
"loss": 25.5133,
"step": 7680
},
{
"epoch": 0.5561884097278728,
"grad_norm": 15.7265625,
"learning_rate": 9.95654795235757e-07,
"loss": 25.4187,
"step": 7690
},
{
"epoch": 0.5569116716390923,
"grad_norm": 17.234375,
"learning_rate": 9.95649144774425e-07,
"loss": 24.9255,
"step": 7700
},
{
"epoch": 0.5576349335503119,
"grad_norm": 16.4375,
"learning_rate": 9.95643494313093e-07,
"loss": 25.2905,
"step": 7710
},
{
"epoch": 0.5583581954615315,
"grad_norm": 15.4296875,
"learning_rate": 9.956378438517613e-07,
"loss": 25.1232,
"step": 7720
},
{
"epoch": 0.5590814573727511,
"grad_norm": 16.328125,
"learning_rate": 9.956321933904293e-07,
"loss": 25.0427,
"step": 7730
},
{
"epoch": 0.5598047192839707,
"grad_norm": 16.203125,
"learning_rate": 9.956265429290975e-07,
"loss": 24.8122,
"step": 7740
},
{
"epoch": 0.5605279811951903,
"grad_norm": 16.265625,
"learning_rate": 9.956208924677655e-07,
"loss": 24.9768,
"step": 7750
},
{
"epoch": 0.56125124310641,
"grad_norm": 15.9765625,
"learning_rate": 9.956152420064335e-07,
"loss": 25.1609,
"step": 7760
},
{
"epoch": 0.5619745050176295,
"grad_norm": 15.4765625,
"learning_rate": 9.956095915451017e-07,
"loss": 24.8014,
"step": 7770
},
{
"epoch": 0.5626977669288491,
"grad_norm": 15.8125,
"learning_rate": 9.956039410837697e-07,
"loss": 25.1054,
"step": 7780
},
{
"epoch": 0.5634210288400687,
"grad_norm": 15.953125,
"learning_rate": 9.95598290622438e-07,
"loss": 25.3523,
"step": 7790
},
{
"epoch": 0.5641442907512884,
"grad_norm": 16.71875,
"learning_rate": 9.95592640161106e-07,
"loss": 25.1315,
"step": 7800
},
{
"epoch": 0.5648675526625079,
"grad_norm": 17.046875,
"learning_rate": 9.95586989699774e-07,
"loss": 25.2146,
"step": 7810
},
{
"epoch": 0.5655908145737275,
"grad_norm": 16.84375,
"learning_rate": 9.955813392384421e-07,
"loss": 24.9596,
"step": 7820
},
{
"epoch": 0.5663140764849471,
"grad_norm": 17.671875,
"learning_rate": 9.955756887771101e-07,
"loss": 25.124,
"step": 7830
},
{
"epoch": 0.5670373383961667,
"grad_norm": 16.28125,
"learning_rate": 9.955700383157783e-07,
"loss": 24.9963,
"step": 7840
},
{
"epoch": 0.5677606003073863,
"grad_norm": 15.890625,
"learning_rate": 9.955643878544463e-07,
"loss": 25.4564,
"step": 7850
},
{
"epoch": 0.5684838622186059,
"grad_norm": 17.0625,
"learning_rate": 9.955587373931145e-07,
"loss": 25.0387,
"step": 7860
},
{
"epoch": 0.5692071241298255,
"grad_norm": 16.625,
"learning_rate": 9.955530869317825e-07,
"loss": 25.2127,
"step": 7870
},
{
"epoch": 0.5699303860410451,
"grad_norm": 16.203125,
"learning_rate": 9.955474364704505e-07,
"loss": 25.1584,
"step": 7880
},
{
"epoch": 0.5706536479522647,
"grad_norm": 16.515625,
"learning_rate": 9.955417860091187e-07,
"loss": 25.2296,
"step": 7890
},
{
"epoch": 0.5713769098634843,
"grad_norm": 16.5,
"learning_rate": 9.955361355477867e-07,
"loss": 25.0436,
"step": 7900
},
{
"epoch": 0.572100171774704,
"grad_norm": 16.0625,
"learning_rate": 9.95530485086455e-07,
"loss": 25.4235,
"step": 7910
},
{
"epoch": 0.5728234336859235,
"grad_norm": 16.640625,
"learning_rate": 9.95524834625123e-07,
"loss": 25.1063,
"step": 7920
},
{
"epoch": 0.5735466955971431,
"grad_norm": 16.484375,
"learning_rate": 9.95519184163791e-07,
"loss": 24.9058,
"step": 7930
},
{
"epoch": 0.5742699575083627,
"grad_norm": 15.8671875,
"learning_rate": 9.955135337024591e-07,
"loss": 24.9924,
"step": 7940
},
{
"epoch": 0.5749932194195824,
"grad_norm": 16.296875,
"learning_rate": 9.955078832411271e-07,
"loss": 24.8562,
"step": 7950
},
{
"epoch": 0.5757164813308019,
"grad_norm": 16.390625,
"learning_rate": 9.955022327797954e-07,
"loss": 25.0083,
"step": 7960
},
{
"epoch": 0.5764397432420215,
"grad_norm": 17.484375,
"learning_rate": 9.954965823184633e-07,
"loss": 24.5593,
"step": 7970
},
{
"epoch": 0.5771630051532411,
"grad_norm": 17.234375,
"learning_rate": 9.954909318571313e-07,
"loss": 25.0094,
"step": 7980
},
{
"epoch": 0.5778862670644607,
"grad_norm": 16.34375,
"learning_rate": 9.954852813957996e-07,
"loss": 25.4218,
"step": 7990
},
{
"epoch": 0.5786095289756803,
"grad_norm": 16.515625,
"learning_rate": 9.954796309344676e-07,
"loss": 24.8672,
"step": 8000
},
{
"epoch": 0.5793327908868999,
"grad_norm": 17.0,
"learning_rate": 9.954739804731358e-07,
"loss": 25.1298,
"step": 8010
},
{
"epoch": 0.5800560527981196,
"grad_norm": 15.703125,
"learning_rate": 9.954683300118038e-07,
"loss": 24.8071,
"step": 8020
},
{
"epoch": 0.5807793147093391,
"grad_norm": 15.8125,
"learning_rate": 9.954626795504718e-07,
"loss": 25.4645,
"step": 8030
},
{
"epoch": 0.5815025766205587,
"grad_norm": 16.484375,
"learning_rate": 9.9545702908914e-07,
"loss": 25.2866,
"step": 8040
},
{
"epoch": 0.5822258385317783,
"grad_norm": 16.59375,
"learning_rate": 9.95451378627808e-07,
"loss": 25.0499,
"step": 8050
},
{
"epoch": 0.582949100442998,
"grad_norm": 15.9453125,
"learning_rate": 9.954457281664762e-07,
"loss": 25.265,
"step": 8060
},
{
"epoch": 0.5836723623542175,
"grad_norm": 16.75,
"learning_rate": 9.954400777051442e-07,
"loss": 25.1575,
"step": 8070
},
{
"epoch": 0.5843956242654371,
"grad_norm": 16.796875,
"learning_rate": 9.954344272438124e-07,
"loss": 25.0611,
"step": 8080
},
{
"epoch": 0.5851188861766567,
"grad_norm": 16.328125,
"learning_rate": 9.954287767824804e-07,
"loss": 25.374,
"step": 8090
},
{
"epoch": 0.5858421480878763,
"grad_norm": 15.5390625,
"learning_rate": 9.954231263211484e-07,
"loss": 24.4109,
"step": 8100
},
{
"epoch": 0.5865654099990959,
"grad_norm": 15.9453125,
"learning_rate": 9.954174758598166e-07,
"loss": 25.0325,
"step": 8110
},
{
"epoch": 0.5872886719103155,
"grad_norm": 16.796875,
"learning_rate": 9.954118253984846e-07,
"loss": 25.0507,
"step": 8120
},
{
"epoch": 0.5880119338215352,
"grad_norm": 16.390625,
"learning_rate": 9.954061749371528e-07,
"loss": 25.0707,
"step": 8130
},
{
"epoch": 0.5887351957327547,
"grad_norm": 16.546875,
"learning_rate": 9.954005244758208e-07,
"loss": 25.5378,
"step": 8140
},
{
"epoch": 0.5894584576439743,
"grad_norm": 16.0625,
"learning_rate": 9.953948740144888e-07,
"loss": 25.1591,
"step": 8150
},
{
"epoch": 0.5901817195551939,
"grad_norm": 17.171875,
"learning_rate": 9.95389223553157e-07,
"loss": 25.3384,
"step": 8160
},
{
"epoch": 0.5909049814664136,
"grad_norm": 16.5,
"learning_rate": 9.95383573091825e-07,
"loss": 25.1535,
"step": 8170
},
{
"epoch": 0.5916282433776331,
"grad_norm": 15.65625,
"learning_rate": 9.953779226304932e-07,
"loss": 24.9001,
"step": 8180
},
{
"epoch": 0.5923515052888527,
"grad_norm": 16.375,
"learning_rate": 9.953722721691612e-07,
"loss": 25.1227,
"step": 8190
},
{
"epoch": 0.5930747672000724,
"grad_norm": 16.078125,
"learning_rate": 9.953666217078292e-07,
"loss": 25.1242,
"step": 8200
},
{
"epoch": 0.593798029111292,
"grad_norm": 17.21875,
"learning_rate": 9.953609712464974e-07,
"loss": 25.037,
"step": 8210
},
{
"epoch": 0.5945212910225115,
"grad_norm": 17.265625,
"learning_rate": 9.953553207851654e-07,
"loss": 25.1524,
"step": 8220
},
{
"epoch": 0.5952445529337311,
"grad_norm": 17.125,
"learning_rate": 9.953496703238336e-07,
"loss": 25.221,
"step": 8230
},
{
"epoch": 0.5959678148449508,
"grad_norm": 17.109375,
"learning_rate": 9.953440198625016e-07,
"loss": 24.8512,
"step": 8240
},
{
"epoch": 0.5966910767561703,
"grad_norm": 16.03125,
"learning_rate": 9.953383694011696e-07,
"loss": 25.1353,
"step": 8250
},
{
"epoch": 0.5974143386673899,
"grad_norm": 15.09375,
"learning_rate": 9.953327189398378e-07,
"loss": 24.9567,
"step": 8260
},
{
"epoch": 0.5981376005786095,
"grad_norm": 17.640625,
"learning_rate": 9.953270684785058e-07,
"loss": 24.9228,
"step": 8270
},
{
"epoch": 0.5988608624898292,
"grad_norm": 15.8984375,
"learning_rate": 9.95321418017174e-07,
"loss": 25.0197,
"step": 8280
},
{
"epoch": 0.5995841244010487,
"grad_norm": 17.03125,
"learning_rate": 9.95315767555842e-07,
"loss": 25.3745,
"step": 8290
},
{
"epoch": 0.6003073863122683,
"grad_norm": 16.453125,
"learning_rate": 9.953101170945102e-07,
"loss": 25.306,
"step": 8300
},
{
"epoch": 0.601030648223488,
"grad_norm": 14.90625,
"learning_rate": 9.953044666331782e-07,
"loss": 24.8696,
"step": 8310
},
{
"epoch": 0.6017539101347076,
"grad_norm": 15.59375,
"learning_rate": 9.952988161718462e-07,
"loss": 25.2775,
"step": 8320
},
{
"epoch": 0.6024771720459271,
"grad_norm": 15.8515625,
"learning_rate": 9.952931657105144e-07,
"loss": 24.8399,
"step": 8330
},
{
"epoch": 0.6032004339571467,
"grad_norm": 16.484375,
"learning_rate": 9.952875152491824e-07,
"loss": 24.971,
"step": 8340
},
{
"epoch": 0.6039236958683664,
"grad_norm": 16.0625,
"learning_rate": 9.952818647878506e-07,
"loss": 25.0859,
"step": 8350
},
{
"epoch": 0.6046469577795859,
"grad_norm": 16.3125,
"learning_rate": 9.952762143265186e-07,
"loss": 25.063,
"step": 8360
},
{
"epoch": 0.6053702196908055,
"grad_norm": 15.828125,
"learning_rate": 9.952705638651866e-07,
"loss": 25.029,
"step": 8370
},
{
"epoch": 0.6060934816020251,
"grad_norm": 16.984375,
"learning_rate": 9.952649134038548e-07,
"loss": 24.7311,
"step": 8380
},
{
"epoch": 0.6068167435132448,
"grad_norm": 15.703125,
"learning_rate": 9.952592629425228e-07,
"loss": 24.7605,
"step": 8390
},
{
"epoch": 0.6075400054244643,
"grad_norm": 16.140625,
"learning_rate": 9.95253612481191e-07,
"loss": 25.1468,
"step": 8400
},
{
"epoch": 0.6082632673356839,
"grad_norm": 15.5546875,
"learning_rate": 9.95247962019859e-07,
"loss": 24.8656,
"step": 8410
},
{
"epoch": 0.6089865292469036,
"grad_norm": 16.1875,
"learning_rate": 9.95242311558527e-07,
"loss": 24.9105,
"step": 8420
},
{
"epoch": 0.6097097911581232,
"grad_norm": 16.03125,
"learning_rate": 9.952366610971953e-07,
"loss": 25.0424,
"step": 8430
},
{
"epoch": 0.6104330530693427,
"grad_norm": 16.28125,
"learning_rate": 9.952310106358632e-07,
"loss": 25.3547,
"step": 8440
},
{
"epoch": 0.6111563149805623,
"grad_norm": 15.5625,
"learning_rate": 9.952253601745315e-07,
"loss": 24.777,
"step": 8450
},
{
"epoch": 0.611879576891782,
"grad_norm": 17.03125,
"learning_rate": 9.952197097131995e-07,
"loss": 24.958,
"step": 8460
},
{
"epoch": 0.6126028388030015,
"grad_norm": 16.703125,
"learning_rate": 9.952140592518675e-07,
"loss": 25.1826,
"step": 8470
},
{
"epoch": 0.6133261007142211,
"grad_norm": 16.078125,
"learning_rate": 9.952084087905357e-07,
"loss": 25.2109,
"step": 8480
},
{
"epoch": 0.6140493626254407,
"grad_norm": 15.921875,
"learning_rate": 9.952027583292037e-07,
"loss": 25.3674,
"step": 8490
},
{
"epoch": 0.6147726245366604,
"grad_norm": 15.6875,
"learning_rate": 9.951971078678719e-07,
"loss": 25.07,
"step": 8500
},
{
"epoch": 0.6154958864478799,
"grad_norm": 16.0625,
"learning_rate": 9.951914574065399e-07,
"loss": 24.7706,
"step": 8510
},
{
"epoch": 0.6162191483590995,
"grad_norm": 16.25,
"learning_rate": 9.95185806945208e-07,
"loss": 25.121,
"step": 8520
},
{
"epoch": 0.6169424102703192,
"grad_norm": 16.875,
"learning_rate": 9.95180156483876e-07,
"loss": 24.9344,
"step": 8530
},
{
"epoch": 0.6176656721815388,
"grad_norm": 15.8671875,
"learning_rate": 9.95174506022544e-07,
"loss": 25.3272,
"step": 8540
},
{
"epoch": 0.6183889340927583,
"grad_norm": 16.484375,
"learning_rate": 9.951688555612123e-07,
"loss": 25.1814,
"step": 8550
},
{
"epoch": 0.6191121960039779,
"grad_norm": 15.4921875,
"learning_rate": 9.951632050998805e-07,
"loss": 24.6664,
"step": 8560
},
{
"epoch": 0.6198354579151976,
"grad_norm": 17.25,
"learning_rate": 9.951575546385485e-07,
"loss": 25.6137,
"step": 8570
},
{
"epoch": 0.6205587198264172,
"grad_norm": 16.984375,
"learning_rate": 9.951519041772165e-07,
"loss": 25.0584,
"step": 8580
},
{
"epoch": 0.6212819817376367,
"grad_norm": 15.4765625,
"learning_rate": 9.951462537158845e-07,
"loss": 25.0154,
"step": 8590
},
{
"epoch": 0.6220052436488563,
"grad_norm": 16.796875,
"learning_rate": 9.951406032545527e-07,
"loss": 24.9978,
"step": 8600
},
{
"epoch": 0.622728505560076,
"grad_norm": 15.6640625,
"learning_rate": 9.95134952793221e-07,
"loss": 25.5721,
"step": 8610
},
{
"epoch": 0.6234517674712955,
"grad_norm": 16.640625,
"learning_rate": 9.95129302331889e-07,
"loss": 25.3643,
"step": 8620
},
{
"epoch": 0.6241750293825151,
"grad_norm": 16.140625,
"learning_rate": 9.951236518705569e-07,
"loss": 25.4492,
"step": 8630
},
{
"epoch": 0.6248982912937348,
"grad_norm": 16.25,
"learning_rate": 9.951180014092249e-07,
"loss": 24.9718,
"step": 8640
},
{
"epoch": 0.6256215532049544,
"grad_norm": 16.65625,
"learning_rate": 9.95112350947893e-07,
"loss": 24.6082,
"step": 8650
},
{
"epoch": 0.6263448151161739,
"grad_norm": 15.8828125,
"learning_rate": 9.95106700486561e-07,
"loss": 24.9935,
"step": 8660
},
{
"epoch": 0.6270680770273935,
"grad_norm": 15.328125,
"learning_rate": 9.951010500252293e-07,
"loss": 25.0441,
"step": 8670
},
{
"epoch": 0.6277913389386132,
"grad_norm": 15.6015625,
"learning_rate": 9.950953995638973e-07,
"loss": 24.6859,
"step": 8680
},
{
"epoch": 0.6285146008498328,
"grad_norm": 16.03125,
"learning_rate": 9.950897491025653e-07,
"loss": 24.9053,
"step": 8690
},
{
"epoch": 0.6292378627610523,
"grad_norm": 16.1875,
"learning_rate": 9.950840986412335e-07,
"loss": 24.9122,
"step": 8700
},
{
"epoch": 0.6299611246722719,
"grad_norm": 16.984375,
"learning_rate": 9.950784481799015e-07,
"loss": 25.4552,
"step": 8710
},
{
"epoch": 0.6306843865834916,
"grad_norm": 15.515625,
"learning_rate": 9.950727977185697e-07,
"loss": 24.8516,
"step": 8720
},
{
"epoch": 0.6314076484947111,
"grad_norm": 15.8125,
"learning_rate": 9.950671472572377e-07,
"loss": 24.9679,
"step": 8730
},
{
"epoch": 0.6321309104059307,
"grad_norm": 16.375,
"learning_rate": 9.95061496795906e-07,
"loss": 25.5788,
"step": 8740
},
{
"epoch": 0.6328541723171504,
"grad_norm": 17.078125,
"learning_rate": 9.95055846334574e-07,
"loss": 25.3259,
"step": 8750
},
{
"epoch": 0.63357743422837,
"grad_norm": 15.7109375,
"learning_rate": 9.95050195873242e-07,
"loss": 24.7144,
"step": 8760
},
{
"epoch": 0.6343006961395895,
"grad_norm": 16.015625,
"learning_rate": 9.950445454119101e-07,
"loss": 24.7972,
"step": 8770
},
{
"epoch": 0.6350239580508091,
"grad_norm": 16.828125,
"learning_rate": 9.950388949505783e-07,
"loss": 25.0281,
"step": 8780
},
{
"epoch": 0.6357472199620288,
"grad_norm": 16.65625,
"learning_rate": 9.950332444892463e-07,
"loss": 24.8294,
"step": 8790
},
{
"epoch": 0.6364704818732484,
"grad_norm": 16.90625,
"learning_rate": 9.950275940279143e-07,
"loss": 25.2674,
"step": 8800
},
{
"epoch": 0.6371937437844679,
"grad_norm": 16.5,
"learning_rate": 9.950219435665823e-07,
"loss": 25.1726,
"step": 8810
},
{
"epoch": 0.6379170056956875,
"grad_norm": 17.265625,
"learning_rate": 9.950162931052505e-07,
"loss": 24.7559,
"step": 8820
},
{
"epoch": 0.6386402676069072,
"grad_norm": 15.828125,
"learning_rate": 9.950106426439187e-07,
"loss": 25.1427,
"step": 8830
},
{
"epoch": 0.6393635295181268,
"grad_norm": 16.796875,
"learning_rate": 9.950049921825867e-07,
"loss": 24.9646,
"step": 8840
},
{
"epoch": 0.6400867914293463,
"grad_norm": 15.2265625,
"learning_rate": 9.949993417212547e-07,
"loss": 25.1181,
"step": 8850
},
{
"epoch": 0.640810053340566,
"grad_norm": 15.84375,
"learning_rate": 9.949936912599227e-07,
"loss": 24.9643,
"step": 8860
},
{
"epoch": 0.6415333152517856,
"grad_norm": 15.3515625,
"learning_rate": 9.94988040798591e-07,
"loss": 25.2106,
"step": 8870
},
{
"epoch": 0.6422565771630051,
"grad_norm": 16.484375,
"learning_rate": 9.949823903372592e-07,
"loss": 25.0076,
"step": 8880
},
{
"epoch": 0.6429798390742247,
"grad_norm": 16.953125,
"learning_rate": 9.949767398759272e-07,
"loss": 25.3634,
"step": 8890
},
{
"epoch": 0.6437031009854444,
"grad_norm": 15.96875,
"learning_rate": 9.949710894145952e-07,
"loss": 25.1013,
"step": 8900
},
{
"epoch": 0.644426362896664,
"grad_norm": 15.453125,
"learning_rate": 9.949654389532634e-07,
"loss": 24.9565,
"step": 8910
},
{
"epoch": 0.6451496248078835,
"grad_norm": 15.7578125,
"learning_rate": 9.949597884919314e-07,
"loss": 24.6701,
"step": 8920
},
{
"epoch": 0.6458728867191031,
"grad_norm": 16.9375,
"learning_rate": 9.949541380305996e-07,
"loss": 25.131,
"step": 8930
},
{
"epoch": 0.6465961486303228,
"grad_norm": 16.625,
"learning_rate": 9.949484875692676e-07,
"loss": 25.0346,
"step": 8940
},
{
"epoch": 0.6473194105415424,
"grad_norm": 16.046875,
"learning_rate": 9.949428371079356e-07,
"loss": 25.1957,
"step": 8950
},
{
"epoch": 0.6480426724527619,
"grad_norm": 16.171875,
"learning_rate": 9.949371866466038e-07,
"loss": 25.2245,
"step": 8960
},
{
"epoch": 0.6487659343639816,
"grad_norm": 16.125,
"learning_rate": 9.949315361852718e-07,
"loss": 25.3402,
"step": 8970
},
{
"epoch": 0.6494891962752012,
"grad_norm": 16.046875,
"learning_rate": 9.949258857239398e-07,
"loss": 25.0492,
"step": 8980
},
{
"epoch": 0.6502124581864207,
"grad_norm": 16.171875,
"learning_rate": 9.94920235262608e-07,
"loss": 25.3217,
"step": 8990
},
{
"epoch": 0.6509357200976403,
"grad_norm": 15.8984375,
"learning_rate": 9.949145848012762e-07,
"loss": 25.2028,
"step": 9000
},
{
"epoch": 0.65165898200886,
"grad_norm": 16.9375,
"learning_rate": 9.949089343399442e-07,
"loss": 25.19,
"step": 9010
},
{
"epoch": 0.6523822439200796,
"grad_norm": 16.6875,
"learning_rate": 9.949032838786122e-07,
"loss": 25.143,
"step": 9020
},
{
"epoch": 0.6531055058312991,
"grad_norm": 16.828125,
"learning_rate": 9.948976334172802e-07,
"loss": 25.1178,
"step": 9030
},
{
"epoch": 0.6538287677425187,
"grad_norm": 16.9375,
"learning_rate": 9.948919829559484e-07,
"loss": 25.0101,
"step": 9040
},
{
"epoch": 0.6545520296537384,
"grad_norm": 16.15625,
"learning_rate": 9.948863324946166e-07,
"loss": 25.3933,
"step": 9050
},
{
"epoch": 0.655275291564958,
"grad_norm": 16.984375,
"learning_rate": 9.948806820332846e-07,
"loss": 24.9395,
"step": 9060
},
{
"epoch": 0.6559985534761775,
"grad_norm": 16.53125,
"learning_rate": 9.948750315719526e-07,
"loss": 24.8895,
"step": 9070
},
{
"epoch": 0.6567218153873972,
"grad_norm": 15.796875,
"learning_rate": 9.948693811106206e-07,
"loss": 24.8654,
"step": 9080
},
{
"epoch": 0.6574450772986168,
"grad_norm": 15.6953125,
"learning_rate": 9.948637306492888e-07,
"loss": 24.8138,
"step": 9090
},
{
"epoch": 0.6581683392098364,
"grad_norm": 16.171875,
"learning_rate": 9.94858080187957e-07,
"loss": 25.1842,
"step": 9100
},
{
"epoch": 0.6588916011210559,
"grad_norm": 16.109375,
"learning_rate": 9.94852429726625e-07,
"loss": 24.8857,
"step": 9110
},
{
"epoch": 0.6596148630322756,
"grad_norm": 16.234375,
"learning_rate": 9.94846779265293e-07,
"loss": 25.2401,
"step": 9120
},
{
"epoch": 0.6603381249434952,
"grad_norm": 16.28125,
"learning_rate": 9.948411288039612e-07,
"loss": 24.8999,
"step": 9130
},
{
"epoch": 0.6610613868547147,
"grad_norm": 16.40625,
"learning_rate": 9.948354783426292e-07,
"loss": 24.8383,
"step": 9140
},
{
"epoch": 0.6617846487659343,
"grad_norm": 16.53125,
"learning_rate": 9.948298278812974e-07,
"loss": 24.9992,
"step": 9150
},
{
"epoch": 0.662507910677154,
"grad_norm": 15.8828125,
"learning_rate": 9.948241774199654e-07,
"loss": 24.7903,
"step": 9160
},
{
"epoch": 0.6632311725883736,
"grad_norm": 16.265625,
"learning_rate": 9.948185269586334e-07,
"loss": 25.0166,
"step": 9170
},
{
"epoch": 0.6639544344995931,
"grad_norm": 16.109375,
"learning_rate": 9.948128764973016e-07,
"loss": 25.1261,
"step": 9180
},
{
"epoch": 0.6646776964108128,
"grad_norm": 15.8046875,
"learning_rate": 9.948072260359696e-07,
"loss": 25.2559,
"step": 9190
},
{
"epoch": 0.6654009583220324,
"grad_norm": 16.703125,
"learning_rate": 9.948015755746378e-07,
"loss": 24.8266,
"step": 9200
},
{
"epoch": 0.666124220233252,
"grad_norm": 16.203125,
"learning_rate": 9.947959251133058e-07,
"loss": 25.231,
"step": 9210
},
{
"epoch": 0.6668474821444715,
"grad_norm": 16.53125,
"learning_rate": 9.94790274651974e-07,
"loss": 25.3356,
"step": 9220
},
{
"epoch": 0.6675707440556912,
"grad_norm": 15.234375,
"learning_rate": 9.94784624190642e-07,
"loss": 25.1178,
"step": 9230
},
{
"epoch": 0.6682940059669108,
"grad_norm": 16.171875,
"learning_rate": 9.9477897372931e-07,
"loss": 25.1904,
"step": 9240
},
{
"epoch": 0.6690172678781303,
"grad_norm": 15.515625,
"learning_rate": 9.947733232679782e-07,
"loss": 25.057,
"step": 9250
},
{
"epoch": 0.66974052978935,
"grad_norm": 16.9375,
"learning_rate": 9.947676728066462e-07,
"loss": 25.0339,
"step": 9260
},
{
"epoch": 0.6704637917005696,
"grad_norm": 16.0625,
"learning_rate": 9.947620223453144e-07,
"loss": 25.0058,
"step": 9270
},
{
"epoch": 0.6711870536117892,
"grad_norm": 17.0625,
"learning_rate": 9.947563718839824e-07,
"loss": 24.9719,
"step": 9280
},
{
"epoch": 0.6719103155230087,
"grad_norm": 15.65625,
"learning_rate": 9.947507214226504e-07,
"loss": 25.0308,
"step": 9290
},
{
"epoch": 0.6726335774342284,
"grad_norm": 16.03125,
"learning_rate": 9.947450709613186e-07,
"loss": 25.2972,
"step": 9300
},
{
"epoch": 0.673356839345448,
"grad_norm": 15.671875,
"learning_rate": 9.947394204999866e-07,
"loss": 25.0732,
"step": 9310
},
{
"epoch": 0.6740801012566676,
"grad_norm": 16.234375,
"learning_rate": 9.947337700386549e-07,
"loss": 25.0754,
"step": 9320
},
{
"epoch": 0.6748033631678871,
"grad_norm": 15.2578125,
"learning_rate": 9.947281195773229e-07,
"loss": 25.5137,
"step": 9330
},
{
"epoch": 0.6755266250791068,
"grad_norm": 15.859375,
"learning_rate": 9.947224691159909e-07,
"loss": 25.1258,
"step": 9340
},
{
"epoch": 0.6762498869903264,
"grad_norm": 16.65625,
"learning_rate": 9.94716818654659e-07,
"loss": 25.1174,
"step": 9350
},
{
"epoch": 0.6769731489015459,
"grad_norm": 16.25,
"learning_rate": 9.94711168193327e-07,
"loss": 25.1364,
"step": 9360
},
{
"epoch": 0.6776964108127655,
"grad_norm": 15.8359375,
"learning_rate": 9.947055177319953e-07,
"loss": 25.13,
"step": 9370
},
{
"epoch": 0.6784196727239852,
"grad_norm": 17.828125,
"learning_rate": 9.946998672706633e-07,
"loss": 25.317,
"step": 9380
},
{
"epoch": 0.6791429346352048,
"grad_norm": 15.5703125,
"learning_rate": 9.946942168093313e-07,
"loss": 25.0771,
"step": 9390
},
{
"epoch": 0.6798661965464243,
"grad_norm": 16.3125,
"learning_rate": 9.946885663479995e-07,
"loss": 25.558,
"step": 9400
},
{
"epoch": 0.680589458457644,
"grad_norm": 15.7265625,
"learning_rate": 9.946829158866675e-07,
"loss": 24.8281,
"step": 9410
},
{
"epoch": 0.6813127203688636,
"grad_norm": 17.03125,
"learning_rate": 9.946772654253357e-07,
"loss": 24.8212,
"step": 9420
},
{
"epoch": 0.6820359822800832,
"grad_norm": 16.296875,
"learning_rate": 9.946716149640037e-07,
"loss": 25.2519,
"step": 9430
},
{
"epoch": 0.6827592441913027,
"grad_norm": 16.328125,
"learning_rate": 9.946659645026719e-07,
"loss": 25.086,
"step": 9440
},
{
"epoch": 0.6834825061025224,
"grad_norm": 16.34375,
"learning_rate": 9.946603140413399e-07,
"loss": 25.1416,
"step": 9450
},
{
"epoch": 0.684205768013742,
"grad_norm": 15.1328125,
"learning_rate": 9.946546635800079e-07,
"loss": 25.0502,
"step": 9460
},
{
"epoch": 0.6849290299249616,
"grad_norm": 16.140625,
"learning_rate": 9.94649013118676e-07,
"loss": 25.018,
"step": 9470
},
{
"epoch": 0.6856522918361811,
"grad_norm": 16.359375,
"learning_rate": 9.94643362657344e-07,
"loss": 24.6714,
"step": 9480
},
{
"epoch": 0.6863755537474008,
"grad_norm": 16.03125,
"learning_rate": 9.946377121960123e-07,
"loss": 25.0868,
"step": 9490
},
{
"epoch": 0.6870988156586204,
"grad_norm": 15.8515625,
"learning_rate": 9.946320617346803e-07,
"loss": 25.0213,
"step": 9500
},
{
"epoch": 0.6878220775698399,
"grad_norm": 17.421875,
"learning_rate": 9.946264112733483e-07,
"loss": 25.2668,
"step": 9510
},
{
"epoch": 0.6885453394810596,
"grad_norm": 16.703125,
"learning_rate": 9.946207608120165e-07,
"loss": 25.1576,
"step": 9520
},
{
"epoch": 0.6892686013922792,
"grad_norm": 16.453125,
"learning_rate": 9.946151103506845e-07,
"loss": 25.0762,
"step": 9530
},
{
"epoch": 0.6899918633034988,
"grad_norm": 15.5859375,
"learning_rate": 9.946094598893527e-07,
"loss": 24.9267,
"step": 9540
},
{
"epoch": 0.6907151252147183,
"grad_norm": 17.0625,
"learning_rate": 9.946038094280207e-07,
"loss": 25.3065,
"step": 9550
},
{
"epoch": 0.691438387125938,
"grad_norm": 17.46875,
"learning_rate": 9.945981589666887e-07,
"loss": 24.975,
"step": 9560
},
{
"epoch": 0.6921616490371576,
"grad_norm": 17.625,
"learning_rate": 9.94592508505357e-07,
"loss": 25.006,
"step": 9570
},
{
"epoch": 0.6928849109483772,
"grad_norm": 15.5546875,
"learning_rate": 9.94586858044025e-07,
"loss": 25.3168,
"step": 9580
},
{
"epoch": 0.6936081728595967,
"grad_norm": 17.15625,
"learning_rate": 9.945812075826931e-07,
"loss": 25.2658,
"step": 9590
},
{
"epoch": 0.6943314347708164,
"grad_norm": 16.90625,
"learning_rate": 9.945755571213611e-07,
"loss": 25.3412,
"step": 9600
},
{
"epoch": 0.695054696682036,
"grad_norm": 17.359375,
"learning_rate": 9.945699066600293e-07,
"loss": 24.7888,
"step": 9610
},
{
"epoch": 0.6957779585932555,
"grad_norm": 15.4296875,
"learning_rate": 9.945642561986973e-07,
"loss": 24.7344,
"step": 9620
},
{
"epoch": 0.6965012205044752,
"grad_norm": 15.84375,
"learning_rate": 9.945586057373653e-07,
"loss": 24.9273,
"step": 9630
},
{
"epoch": 0.6972244824156948,
"grad_norm": 15.859375,
"learning_rate": 9.945529552760335e-07,
"loss": 25.0604,
"step": 9640
},
{
"epoch": 0.6979477443269144,
"grad_norm": 15.5078125,
"learning_rate": 9.945473048147015e-07,
"loss": 25.0373,
"step": 9650
},
{
"epoch": 0.6986710062381339,
"grad_norm": 16.734375,
"learning_rate": 9.945416543533697e-07,
"loss": 24.89,
"step": 9660
},
{
"epoch": 0.6993942681493536,
"grad_norm": 16.984375,
"learning_rate": 9.945360038920377e-07,
"loss": 25.5452,
"step": 9670
},
{
"epoch": 0.7001175300605732,
"grad_norm": 15.609375,
"learning_rate": 9.945303534307057e-07,
"loss": 25.3505,
"step": 9680
},
{
"epoch": 0.7008407919717928,
"grad_norm": 16.140625,
"learning_rate": 9.94524702969374e-07,
"loss": 25.1811,
"step": 9690
},
{
"epoch": 0.7015640538830124,
"grad_norm": 17.21875,
"learning_rate": 9.94519052508042e-07,
"loss": 25.2871,
"step": 9700
},
{
"epoch": 0.702287315794232,
"grad_norm": 17.234375,
"learning_rate": 9.945134020467101e-07,
"loss": 25.123,
"step": 9710
},
{
"epoch": 0.7030105777054516,
"grad_norm": 16.578125,
"learning_rate": 9.945077515853781e-07,
"loss": 25.1046,
"step": 9720
},
{
"epoch": 0.7037338396166712,
"grad_norm": 17.0625,
"learning_rate": 9.945021011240461e-07,
"loss": 24.8252,
"step": 9730
},
{
"epoch": 0.7044571015278908,
"grad_norm": 15.671875,
"learning_rate": 9.944964506627143e-07,
"loss": 25.0734,
"step": 9740
},
{
"epoch": 0.7051803634391104,
"grad_norm": 16.546875,
"learning_rate": 9.944908002013823e-07,
"loss": 25.0185,
"step": 9750
},
{
"epoch": 0.70590362535033,
"grad_norm": 16.1875,
"learning_rate": 9.944851497400506e-07,
"loss": 25.2923,
"step": 9760
},
{
"epoch": 0.7066268872615495,
"grad_norm": 15.984375,
"learning_rate": 9.944794992787186e-07,
"loss": 24.9048,
"step": 9770
},
{
"epoch": 0.7073501491727692,
"grad_norm": 15.90625,
"learning_rate": 9.944738488173865e-07,
"loss": 25.2429,
"step": 9780
},
{
"epoch": 0.7080734110839888,
"grad_norm": 16.390625,
"learning_rate": 9.944681983560548e-07,
"loss": 24.9763,
"step": 9790
},
{
"epoch": 0.7087966729952084,
"grad_norm": 17.609375,
"learning_rate": 9.944625478947228e-07,
"loss": 24.7899,
"step": 9800
},
{
"epoch": 0.709519934906428,
"grad_norm": 15.5234375,
"learning_rate": 9.94456897433391e-07,
"loss": 25.1678,
"step": 9810
},
{
"epoch": 0.7102431968176476,
"grad_norm": 16.296875,
"learning_rate": 9.94451246972059e-07,
"loss": 24.7197,
"step": 9820
},
{
"epoch": 0.7109664587288672,
"grad_norm": 16.921875,
"learning_rate": 9.944455965107272e-07,
"loss": 25.1769,
"step": 9830
},
{
"epoch": 0.7116897206400868,
"grad_norm": 15.6796875,
"learning_rate": 9.944399460493952e-07,
"loss": 24.9675,
"step": 9840
},
{
"epoch": 0.7124129825513064,
"grad_norm": 16.5625,
"learning_rate": 9.944342955880632e-07,
"loss": 25.0993,
"step": 9850
},
{
"epoch": 0.713136244462526,
"grad_norm": 16.203125,
"learning_rate": 9.944286451267314e-07,
"loss": 24.9978,
"step": 9860
},
{
"epoch": 0.7138595063737456,
"grad_norm": 16.78125,
"learning_rate": 9.944229946653994e-07,
"loss": 24.8196,
"step": 9870
},
{
"epoch": 0.7145827682849651,
"grad_norm": 16.140625,
"learning_rate": 9.944173442040676e-07,
"loss": 25.2091,
"step": 9880
},
{
"epoch": 0.7153060301961848,
"grad_norm": 16.484375,
"learning_rate": 9.944116937427356e-07,
"loss": 25.1268,
"step": 9890
},
{
"epoch": 0.7160292921074044,
"grad_norm": 16.171875,
"learning_rate": 9.944060432814036e-07,
"loss": 25.0055,
"step": 9900
},
{
"epoch": 0.716752554018624,
"grad_norm": 16.203125,
"learning_rate": 9.944003928200718e-07,
"loss": 25.0594,
"step": 9910
},
{
"epoch": 0.7174758159298436,
"grad_norm": 16.78125,
"learning_rate": 9.943947423587398e-07,
"loss": 25.3482,
"step": 9920
},
{
"epoch": 0.7181990778410632,
"grad_norm": 16.03125,
"learning_rate": 9.94389091897408e-07,
"loss": 25.1836,
"step": 9930
},
{
"epoch": 0.7189223397522828,
"grad_norm": 16.609375,
"learning_rate": 9.94383441436076e-07,
"loss": 25.2864,
"step": 9940
},
{
"epoch": 0.7196456016635024,
"grad_norm": 16.046875,
"learning_rate": 9.94377790974744e-07,
"loss": 25.058,
"step": 9950
},
{
"epoch": 0.720368863574722,
"grad_norm": 17.21875,
"learning_rate": 9.943721405134122e-07,
"loss": 24.9827,
"step": 9960
},
{
"epoch": 0.7210921254859416,
"grad_norm": 17.34375,
"learning_rate": 9.943664900520802e-07,
"loss": 25.0163,
"step": 9970
},
{
"epoch": 0.7218153873971612,
"grad_norm": 16.109375,
"learning_rate": 9.943608395907484e-07,
"loss": 25.2643,
"step": 9980
},
{
"epoch": 0.7225386493083807,
"grad_norm": 16.390625,
"learning_rate": 9.943551891294164e-07,
"loss": 25.2327,
"step": 9990
},
{
"epoch": 0.7232619112196004,
"grad_norm": 16.8125,
"learning_rate": 9.943495386680844e-07,
"loss": 24.98,
"step": 10000
},
{
"epoch": 0.7232619112196004,
"eval_loss": 1.5873770713806152,
"eval_runtime": 366.777,
"eval_samples_per_second": 1979.565,
"eval_steps_per_second": 30.932,
"step": 10000
}
],
"logging_steps": 10,
"max_steps": 13826,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 2500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1.3969285567217664e+19,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}