gpt2-sft-port / trainer_state.json
pkarypis's picture
Model save
008f90f verified
raw
history blame
73.5 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0,
"eval_steps": 500,
"global_step": 2258,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 6.16668000797642,
"learning_rate": 8.849557522123894e-08,
"loss": 3.1048,
"step": 1
},
{
"epoch": 0.0,
"grad_norm": 6.5471896525435085,
"learning_rate": 4.4247787610619474e-07,
"loss": 3.1401,
"step": 5
},
{
"epoch": 0.01,
"grad_norm": 6.696338000754039,
"learning_rate": 8.849557522123895e-07,
"loss": 3.1347,
"step": 10
},
{
"epoch": 0.01,
"grad_norm": 6.564941869190695,
"learning_rate": 1.3274336283185843e-06,
"loss": 3.1075,
"step": 15
},
{
"epoch": 0.02,
"grad_norm": 6.564022310256844,
"learning_rate": 1.769911504424779e-06,
"loss": 3.1301,
"step": 20
},
{
"epoch": 0.02,
"grad_norm": 5.857459703567939,
"learning_rate": 2.212389380530974e-06,
"loss": 3.0924,
"step": 25
},
{
"epoch": 0.03,
"grad_norm": 5.070174288225833,
"learning_rate": 2.6548672566371687e-06,
"loss": 3.0522,
"step": 30
},
{
"epoch": 0.03,
"grad_norm": 4.855168020326285,
"learning_rate": 3.097345132743363e-06,
"loss": 2.9879,
"step": 35
},
{
"epoch": 0.04,
"grad_norm": 3.7053321557646437,
"learning_rate": 3.539823008849558e-06,
"loss": 2.9571,
"step": 40
},
{
"epoch": 0.04,
"grad_norm": 3.385179078690331,
"learning_rate": 3.982300884955752e-06,
"loss": 2.8408,
"step": 45
},
{
"epoch": 0.04,
"grad_norm": 2.7593797822420125,
"learning_rate": 4.424778761061948e-06,
"loss": 2.8395,
"step": 50
},
{
"epoch": 0.05,
"grad_norm": 2.4167880367832946,
"learning_rate": 4.867256637168142e-06,
"loss": 2.7639,
"step": 55
},
{
"epoch": 0.05,
"grad_norm": 1.90271914290754,
"learning_rate": 5.309734513274337e-06,
"loss": 2.7241,
"step": 60
},
{
"epoch": 0.06,
"grad_norm": 1.7113126693108835,
"learning_rate": 5.752212389380532e-06,
"loss": 2.6849,
"step": 65
},
{
"epoch": 0.06,
"grad_norm": 1.529840444833872,
"learning_rate": 6.194690265486726e-06,
"loss": 2.6677,
"step": 70
},
{
"epoch": 0.07,
"grad_norm": 1.330967989395003,
"learning_rate": 6.6371681415929215e-06,
"loss": 2.6535,
"step": 75
},
{
"epoch": 0.07,
"grad_norm": 1.2369520958493831,
"learning_rate": 7.079646017699116e-06,
"loss": 2.6052,
"step": 80
},
{
"epoch": 0.08,
"grad_norm": 1.1701918999480954,
"learning_rate": 7.5221238938053095e-06,
"loss": 2.5976,
"step": 85
},
{
"epoch": 0.08,
"grad_norm": 1.0431136754020183,
"learning_rate": 7.964601769911505e-06,
"loss": 2.5714,
"step": 90
},
{
"epoch": 0.08,
"grad_norm": 1.046491645305411,
"learning_rate": 8.4070796460177e-06,
"loss": 2.5294,
"step": 95
},
{
"epoch": 0.09,
"grad_norm": 1.002838976495586,
"learning_rate": 8.849557522123895e-06,
"loss": 2.5194,
"step": 100
},
{
"epoch": 0.09,
"grad_norm": 0.9044119407546927,
"learning_rate": 9.29203539823009e-06,
"loss": 2.4896,
"step": 105
},
{
"epoch": 0.1,
"grad_norm": 0.8334848336315438,
"learning_rate": 9.734513274336284e-06,
"loss": 2.4613,
"step": 110
},
{
"epoch": 0.1,
"grad_norm": 0.828439583027838,
"learning_rate": 1.0176991150442479e-05,
"loss": 2.4672,
"step": 115
},
{
"epoch": 0.11,
"grad_norm": 0.7767404696640728,
"learning_rate": 1.0619469026548675e-05,
"loss": 2.4404,
"step": 120
},
{
"epoch": 0.11,
"grad_norm": 0.711776929431372,
"learning_rate": 1.1061946902654867e-05,
"loss": 2.4359,
"step": 125
},
{
"epoch": 0.12,
"grad_norm": 0.6315041207215156,
"learning_rate": 1.1504424778761064e-05,
"loss": 2.4162,
"step": 130
},
{
"epoch": 0.12,
"grad_norm": 0.6324896043486851,
"learning_rate": 1.1946902654867258e-05,
"loss": 2.4092,
"step": 135
},
{
"epoch": 0.12,
"grad_norm": 0.6169604572085972,
"learning_rate": 1.2389380530973452e-05,
"loss": 2.3839,
"step": 140
},
{
"epoch": 0.13,
"grad_norm": 0.6226373424179094,
"learning_rate": 1.2831858407079647e-05,
"loss": 2.3835,
"step": 145
},
{
"epoch": 0.13,
"grad_norm": 0.5289987275716352,
"learning_rate": 1.3274336283185843e-05,
"loss": 2.3808,
"step": 150
},
{
"epoch": 0.14,
"grad_norm": 0.5453594016937603,
"learning_rate": 1.3716814159292036e-05,
"loss": 2.3643,
"step": 155
},
{
"epoch": 0.14,
"grad_norm": 0.516762242242035,
"learning_rate": 1.4159292035398232e-05,
"loss": 2.3454,
"step": 160
},
{
"epoch": 0.15,
"grad_norm": 0.5368880823545079,
"learning_rate": 1.4601769911504426e-05,
"loss": 2.3305,
"step": 165
},
{
"epoch": 0.15,
"grad_norm": 0.47438616375480386,
"learning_rate": 1.5044247787610619e-05,
"loss": 2.3065,
"step": 170
},
{
"epoch": 0.16,
"grad_norm": 0.5491458973903554,
"learning_rate": 1.5486725663716813e-05,
"loss": 2.3428,
"step": 175
},
{
"epoch": 0.16,
"grad_norm": 0.511377973091286,
"learning_rate": 1.592920353982301e-05,
"loss": 2.352,
"step": 180
},
{
"epoch": 0.16,
"grad_norm": 0.46313283368485314,
"learning_rate": 1.6371681415929206e-05,
"loss": 2.3497,
"step": 185
},
{
"epoch": 0.17,
"grad_norm": 0.5204938365133991,
"learning_rate": 1.68141592920354e-05,
"loss": 2.3224,
"step": 190
},
{
"epoch": 0.17,
"grad_norm": 0.49147593332523326,
"learning_rate": 1.7256637168141594e-05,
"loss": 2.32,
"step": 195
},
{
"epoch": 0.18,
"grad_norm": 0.4591039414766313,
"learning_rate": 1.769911504424779e-05,
"loss": 2.3187,
"step": 200
},
{
"epoch": 0.18,
"grad_norm": 0.5292810235066809,
"learning_rate": 1.8141592920353983e-05,
"loss": 2.318,
"step": 205
},
{
"epoch": 0.19,
"grad_norm": 0.4471389099066468,
"learning_rate": 1.858407079646018e-05,
"loss": 2.3148,
"step": 210
},
{
"epoch": 0.19,
"grad_norm": 0.4727315621186687,
"learning_rate": 1.9026548672566376e-05,
"loss": 2.2873,
"step": 215
},
{
"epoch": 0.19,
"grad_norm": 0.4474346990399113,
"learning_rate": 1.946902654867257e-05,
"loss": 2.2928,
"step": 220
},
{
"epoch": 0.2,
"grad_norm": 0.45811951151514696,
"learning_rate": 1.991150442477876e-05,
"loss": 2.2823,
"step": 225
},
{
"epoch": 0.2,
"grad_norm": 0.4758359215778834,
"learning_rate": 1.9999808776641724e-05,
"loss": 2.2906,
"step": 230
},
{
"epoch": 0.21,
"grad_norm": 0.4562250088331338,
"learning_rate": 1.999903194428269e-05,
"loss": 2.2843,
"step": 235
},
{
"epoch": 0.21,
"grad_norm": 0.45880244952022176,
"learning_rate": 1.999765759784862e-05,
"loss": 2.2903,
"step": 240
},
{
"epoch": 0.22,
"grad_norm": 0.446778077231725,
"learning_rate": 1.9995685819466593e-05,
"loss": 2.2752,
"step": 245
},
{
"epoch": 0.22,
"grad_norm": 0.48686346365157646,
"learning_rate": 1.9993116726964554e-05,
"loss": 2.2752,
"step": 250
},
{
"epoch": 0.23,
"grad_norm": 0.44639791563091075,
"learning_rate": 1.9989950473864254e-05,
"loss": 2.2766,
"step": 255
},
{
"epoch": 0.23,
"grad_norm": 0.4505549459373634,
"learning_rate": 1.998618724937209e-05,
"loss": 2.2672,
"step": 260
},
{
"epoch": 0.23,
"grad_norm": 0.43187262303167306,
"learning_rate": 1.9981827278367796e-05,
"loss": 2.2748,
"step": 265
},
{
"epoch": 0.24,
"grad_norm": 0.4460489393323405,
"learning_rate": 1.997687082139099e-05,
"loss": 2.2488,
"step": 270
},
{
"epoch": 0.24,
"grad_norm": 0.4735321696451869,
"learning_rate": 1.9971318174625633e-05,
"loss": 2.2535,
"step": 275
},
{
"epoch": 0.25,
"grad_norm": 0.4526230412160001,
"learning_rate": 1.9965169669882293e-05,
"loss": 2.2387,
"step": 280
},
{
"epoch": 0.25,
"grad_norm": 0.4788294815903439,
"learning_rate": 1.9958425674578364e-05,
"loss": 2.245,
"step": 285
},
{
"epoch": 0.26,
"grad_norm": 0.45377311149909166,
"learning_rate": 1.995108659171607e-05,
"loss": 2.2723,
"step": 290
},
{
"epoch": 0.26,
"grad_norm": 0.43094696529248516,
"learning_rate": 1.9943152859858386e-05,
"loss": 2.2374,
"step": 295
},
{
"epoch": 0.27,
"grad_norm": 0.45873287085170134,
"learning_rate": 1.9934624953102858e-05,
"loss": 2.2524,
"step": 300
},
{
"epoch": 0.27,
"grad_norm": 0.49321291345146306,
"learning_rate": 1.9925503381053258e-05,
"loss": 2.2438,
"step": 305
},
{
"epoch": 0.27,
"grad_norm": 0.4600493359269162,
"learning_rate": 1.9915788688789107e-05,
"loss": 2.2538,
"step": 310
},
{
"epoch": 0.28,
"grad_norm": 0.5028182925643601,
"learning_rate": 1.990548145683315e-05,
"loss": 2.2403,
"step": 315
},
{
"epoch": 0.28,
"grad_norm": 0.45532384433440837,
"learning_rate": 1.9894582301116633e-05,
"loss": 2.2373,
"step": 320
},
{
"epoch": 0.29,
"grad_norm": 0.4487282040168157,
"learning_rate": 1.9883091872942484e-05,
"loss": 2.2358,
"step": 325
},
{
"epoch": 0.29,
"grad_norm": 0.4764926656513436,
"learning_rate": 1.9871010858946443e-05,
"loss": 2.2351,
"step": 330
},
{
"epoch": 0.3,
"grad_norm": 0.4199971699008786,
"learning_rate": 1.985833998105598e-05,
"loss": 2.2175,
"step": 335
},
{
"epoch": 0.3,
"grad_norm": 0.4586423717481669,
"learning_rate": 1.984507999644719e-05,
"loss": 2.2244,
"step": 340
},
{
"epoch": 0.31,
"grad_norm": 0.4613425170988114,
"learning_rate": 1.9831231697499515e-05,
"loss": 2.2288,
"step": 345
},
{
"epoch": 0.31,
"grad_norm": 0.456218191386037,
"learning_rate": 1.9816795911748422e-05,
"loss": 2.2502,
"step": 350
},
{
"epoch": 0.31,
"grad_norm": 0.4239570753803919,
"learning_rate": 1.980177350183594e-05,
"loss": 2.2326,
"step": 355
},
{
"epoch": 0.32,
"grad_norm": 0.4290635987649535,
"learning_rate": 1.9786165365459102e-05,
"loss": 2.2368,
"step": 360
},
{
"epoch": 0.32,
"grad_norm": 0.4269016309566689,
"learning_rate": 1.976997243531632e-05,
"loss": 2.2194,
"step": 365
},
{
"epoch": 0.33,
"grad_norm": 0.44327099284543003,
"learning_rate": 1.975319567905163e-05,
"loss": 2.2236,
"step": 370
},
{
"epoch": 0.33,
"grad_norm": 0.4391241219476269,
"learning_rate": 1.9735836099196882e-05,
"loss": 2.2242,
"step": 375
},
{
"epoch": 0.34,
"grad_norm": 0.44135665059050083,
"learning_rate": 1.971789473311184e-05,
"loss": 2.2149,
"step": 380
},
{
"epoch": 0.34,
"grad_norm": 0.4454759372996008,
"learning_rate": 1.9699372652922154e-05,
"loss": 2.2342,
"step": 385
},
{
"epoch": 0.35,
"grad_norm": 0.4358200897154911,
"learning_rate": 1.9680270965455343e-05,
"loss": 2.2333,
"step": 390
},
{
"epoch": 0.35,
"grad_norm": 0.451674904618942,
"learning_rate": 1.966059081217461e-05,
"loss": 2.2198,
"step": 395
},
{
"epoch": 0.35,
"grad_norm": 0.4495411033694467,
"learning_rate": 1.9640333369110662e-05,
"loss": 2.2177,
"step": 400
},
{
"epoch": 0.36,
"grad_norm": 0.4557748974050081,
"learning_rate": 1.9619499846791426e-05,
"loss": 2.2251,
"step": 405
},
{
"epoch": 0.36,
"grad_norm": 0.46508006961836623,
"learning_rate": 1.9598091490169696e-05,
"loss": 2.2033,
"step": 410
},
{
"epoch": 0.37,
"grad_norm": 0.42183477184382523,
"learning_rate": 1.9576109578548757e-05,
"loss": 2.2156,
"step": 415
},
{
"epoch": 0.37,
"grad_norm": 0.4789216349529269,
"learning_rate": 1.9553555425505933e-05,
"loss": 2.187,
"step": 420
},
{
"epoch": 0.38,
"grad_norm": 0.44945247248582815,
"learning_rate": 1.953043037881408e-05,
"loss": 2.212,
"step": 425
},
{
"epoch": 0.38,
"grad_norm": 0.46496471170649234,
"learning_rate": 1.9506735820361065e-05,
"loss": 2.2215,
"step": 430
},
{
"epoch": 0.39,
"grad_norm": 0.4182435105418993,
"learning_rate": 1.9482473166067177e-05,
"loss": 2.2124,
"step": 435
},
{
"epoch": 0.39,
"grad_norm": 0.42509241432837735,
"learning_rate": 1.945764386580051e-05,
"loss": 2.2188,
"step": 440
},
{
"epoch": 0.39,
"grad_norm": 0.42561775297071197,
"learning_rate": 1.9432249403290337e-05,
"loss": 2.2159,
"step": 445
},
{
"epoch": 0.4,
"grad_norm": 0.4275982255504015,
"learning_rate": 1.940629129603844e-05,
"loss": 2.202,
"step": 450
},
{
"epoch": 0.4,
"grad_norm": 0.4625693629122074,
"learning_rate": 1.9379771095228426e-05,
"loss": 2.1969,
"step": 455
},
{
"epoch": 0.41,
"grad_norm": 0.44678800517788825,
"learning_rate": 1.935269038563303e-05,
"loss": 2.1957,
"step": 460
},
{
"epoch": 0.41,
"grad_norm": 0.4440605685011746,
"learning_rate": 1.9325050785519438e-05,
"loss": 2.2145,
"step": 465
},
{
"epoch": 0.42,
"grad_norm": 0.49426967537286975,
"learning_rate": 1.9296853946552532e-05,
"loss": 2.1951,
"step": 470
},
{
"epoch": 0.42,
"grad_norm": 0.4475032010354562,
"learning_rate": 1.9268101553696255e-05,
"loss": 2.2028,
"step": 475
},
{
"epoch": 0.43,
"grad_norm": 0.4369593068597004,
"learning_rate": 1.9238795325112867e-05,
"loss": 2.1919,
"step": 480
},
{
"epoch": 0.43,
"grad_norm": 0.45965332711713663,
"learning_rate": 1.9208937012060316e-05,
"loss": 2.1974,
"step": 485
},
{
"epoch": 0.43,
"grad_norm": 0.44148650224612423,
"learning_rate": 1.9178528398787553e-05,
"loss": 2.2052,
"step": 490
},
{
"epoch": 0.44,
"grad_norm": 0.43518697725913585,
"learning_rate": 1.9147571302427927e-05,
"loss": 2.1925,
"step": 495
},
{
"epoch": 0.44,
"grad_norm": 0.43050554279692543,
"learning_rate": 1.9116067572890603e-05,
"loss": 2.1824,
"step": 500
},
{
"epoch": 0.45,
"grad_norm": 0.4548569462037583,
"learning_rate": 1.9084019092750007e-05,
"loss": 2.1732,
"step": 505
},
{
"epoch": 0.45,
"grad_norm": 0.42197579298043214,
"learning_rate": 1.9051427777133328e-05,
"loss": 2.1803,
"step": 510
},
{
"epoch": 0.46,
"grad_norm": 0.4415090288064239,
"learning_rate": 1.901829557360608e-05,
"loss": 2.1803,
"step": 515
},
{
"epoch": 0.46,
"grad_norm": 0.43714364510293374,
"learning_rate": 1.8984624462055724e-05,
"loss": 2.1779,
"step": 520
},
{
"epoch": 0.47,
"grad_norm": 0.4432571257308879,
"learning_rate": 1.895041645457335e-05,
"loss": 2.1787,
"step": 525
},
{
"epoch": 0.47,
"grad_norm": 0.4783484138075745,
"learning_rate": 1.8915673595333443e-05,
"loss": 2.1894,
"step": 530
},
{
"epoch": 0.47,
"grad_norm": 0.42905150652939833,
"learning_rate": 1.8880397960471724e-05,
"loss": 2.1769,
"step": 535
},
{
"epoch": 0.48,
"grad_norm": 0.4211292593290474,
"learning_rate": 1.8844591657961083e-05,
"loss": 2.1678,
"step": 540
},
{
"epoch": 0.48,
"grad_norm": 0.4179846393433192,
"learning_rate": 1.880825682748563e-05,
"loss": 2.1944,
"step": 545
},
{
"epoch": 0.49,
"grad_norm": 0.4829098687563685,
"learning_rate": 1.877139564031282e-05,
"loss": 2.1905,
"step": 550
},
{
"epoch": 0.49,
"grad_norm": 0.44989057349646744,
"learning_rate": 1.87340102991637e-05,
"loss": 2.1568,
"step": 555
},
{
"epoch": 0.5,
"grad_norm": 0.4542696674908085,
"learning_rate": 1.8696103038081297e-05,
"loss": 2.1736,
"step": 560
},
{
"epoch": 0.5,
"grad_norm": 0.45797580492436063,
"learning_rate": 1.86576761222971e-05,
"loss": 2.1616,
"step": 565
},
{
"epoch": 0.5,
"grad_norm": 0.4552317221563487,
"learning_rate": 1.8618731848095706e-05,
"loss": 2.1693,
"step": 570
},
{
"epoch": 0.51,
"grad_norm": 0.41807581958189405,
"learning_rate": 1.8579272542677597e-05,
"loss": 2.1782,
"step": 575
},
{
"epoch": 0.51,
"grad_norm": 0.4327822652288263,
"learning_rate": 1.853930056402008e-05,
"loss": 2.18,
"step": 580
},
{
"epoch": 0.52,
"grad_norm": 0.4773654657191617,
"learning_rate": 1.849881830073637e-05,
"loss": 2.1631,
"step": 585
},
{
"epoch": 0.52,
"grad_norm": 0.42353740329128814,
"learning_rate": 1.845782817193286e-05,
"loss": 2.1567,
"step": 590
},
{
"epoch": 0.53,
"grad_norm": 0.4139807169693792,
"learning_rate": 1.841633262706456e-05,
"loss": 2.1756,
"step": 595
},
{
"epoch": 0.53,
"grad_norm": 0.43829753304650737,
"learning_rate": 1.8374334145788723e-05,
"loss": 2.1694,
"step": 600
},
{
"epoch": 0.54,
"grad_norm": 0.44749374993683705,
"learning_rate": 1.833183523781668e-05,
"loss": 2.1879,
"step": 605
},
{
"epoch": 0.54,
"grad_norm": 0.4238815840015181,
"learning_rate": 1.8288838442763838e-05,
"loss": 2.1538,
"step": 610
},
{
"epoch": 0.54,
"grad_norm": 0.5173013816010105,
"learning_rate": 1.824534632999796e-05,
"loss": 2.1667,
"step": 615
},
{
"epoch": 0.55,
"grad_norm": 0.4485521005267073,
"learning_rate": 1.820136149848559e-05,
"loss": 2.161,
"step": 620
},
{
"epoch": 0.55,
"grad_norm": 0.4682651494534772,
"learning_rate": 1.8156886576636758e-05,
"loss": 2.1816,
"step": 625
},
{
"epoch": 0.56,
"grad_norm": 0.4534252457964566,
"learning_rate": 1.8111924222147927e-05,
"loss": 2.1684,
"step": 630
},
{
"epoch": 0.56,
"grad_norm": 0.41174103733160045,
"learning_rate": 1.8066477121843163e-05,
"loss": 2.1557,
"step": 635
},
{
"epoch": 0.57,
"grad_norm": 0.47681214256967025,
"learning_rate": 1.8020547991513583e-05,
"loss": 2.1671,
"step": 640
},
{
"epoch": 0.57,
"grad_norm": 0.46865533356504113,
"learning_rate": 1.7974139575755055e-05,
"loss": 2.1622,
"step": 645
},
{
"epoch": 0.58,
"grad_norm": 0.4022540531757412,
"learning_rate": 1.792725464780421e-05,
"loss": 2.1715,
"step": 650
},
{
"epoch": 0.58,
"grad_norm": 0.42886171783708427,
"learning_rate": 1.7879896009372698e-05,
"loss": 2.1668,
"step": 655
},
{
"epoch": 0.58,
"grad_norm": 0.44665491622025383,
"learning_rate": 1.7832066490479797e-05,
"loss": 2.1821,
"step": 660
},
{
"epoch": 0.59,
"grad_norm": 0.4358947246825003,
"learning_rate": 1.7783768949283258e-05,
"loss": 2.1658,
"step": 665
},
{
"epoch": 0.59,
"grad_norm": 0.4448804405857869,
"learning_rate": 1.773500627190854e-05,
"loss": 2.172,
"step": 670
},
{
"epoch": 0.6,
"grad_norm": 0.43650990199089035,
"learning_rate": 1.7685781372276338e-05,
"loss": 2.1711,
"step": 675
},
{
"epoch": 0.6,
"grad_norm": 0.42379321592602226,
"learning_rate": 1.7636097191928437e-05,
"loss": 2.161,
"step": 680
},
{
"epoch": 0.61,
"grad_norm": 0.44716869007099813,
"learning_rate": 1.758595669985197e-05,
"loss": 2.1715,
"step": 685
},
{
"epoch": 0.61,
"grad_norm": 0.44116327473533345,
"learning_rate": 1.7535362892301953e-05,
"loss": 2.173,
"step": 690
},
{
"epoch": 0.62,
"grad_norm": 0.41266222988011164,
"learning_rate": 1.748431879262229e-05,
"loss": 2.167,
"step": 695
},
{
"epoch": 0.62,
"grad_norm": 0.4150965113412605,
"learning_rate": 1.7432827451065052e-05,
"loss": 2.174,
"step": 700
},
{
"epoch": 0.62,
"grad_norm": 0.4642657537042056,
"learning_rate": 1.7380891944608243e-05,
"loss": 2.1566,
"step": 705
},
{
"epoch": 0.63,
"grad_norm": 0.41399080032074204,
"learning_rate": 1.732851537677191e-05,
"loss": 2.1415,
"step": 710
},
{
"epoch": 0.63,
"grad_norm": 0.43184189620946783,
"learning_rate": 1.7275700877432693e-05,
"loss": 2.1382,
"step": 715
},
{
"epoch": 0.64,
"grad_norm": 0.4144508133370711,
"learning_rate": 1.7222451602636785e-05,
"loss": 2.1524,
"step": 720
},
{
"epoch": 0.64,
"grad_norm": 0.4437467129614119,
"learning_rate": 1.7168770734411344e-05,
"loss": 2.1498,
"step": 725
},
{
"epoch": 0.65,
"grad_norm": 0.42641976004801235,
"learning_rate": 1.711466148057433e-05,
"loss": 2.1498,
"step": 730
},
{
"epoch": 0.65,
"grad_norm": 0.40901436028900945,
"learning_rate": 1.7060127074542847e-05,
"loss": 2.1407,
"step": 735
},
{
"epoch": 0.66,
"grad_norm": 0.44229015311354253,
"learning_rate": 1.700517077513987e-05,
"loss": 2.1445,
"step": 740
},
{
"epoch": 0.66,
"grad_norm": 0.456420201073181,
"learning_rate": 1.6949795866399554e-05,
"loss": 2.1559,
"step": 745
},
{
"epoch": 0.66,
"grad_norm": 0.4012310766849398,
"learning_rate": 1.689400565737098e-05,
"loss": 2.1699,
"step": 750
},
{
"epoch": 0.67,
"grad_norm": 0.4533350615602428,
"learning_rate": 1.6837803481920393e-05,
"loss": 2.1655,
"step": 755
},
{
"epoch": 0.67,
"grad_norm": 0.43653673137315946,
"learning_rate": 1.6781192698532e-05,
"loss": 2.1475,
"step": 760
},
{
"epoch": 0.68,
"grad_norm": 0.45156685296683646,
"learning_rate": 1.6724176690107272e-05,
"loss": 2.1689,
"step": 765
},
{
"epoch": 0.68,
"grad_norm": 0.4330861856245563,
"learning_rate": 1.6666758863762796e-05,
"loss": 2.1566,
"step": 770
},
{
"epoch": 0.69,
"grad_norm": 0.47044619679799987,
"learning_rate": 1.6608942650626655e-05,
"loss": 2.1498,
"step": 775
},
{
"epoch": 0.69,
"grad_norm": 0.45478891368198726,
"learning_rate": 1.655073150563343e-05,
"loss": 2.1371,
"step": 780
},
{
"epoch": 0.7,
"grad_norm": 0.41308199064438017,
"learning_rate": 1.6492128907317696e-05,
"loss": 2.1449,
"step": 785
},
{
"epoch": 0.7,
"grad_norm": 0.42855671920472144,
"learning_rate": 1.6433138357606198e-05,
"loss": 2.1386,
"step": 790
},
{
"epoch": 0.7,
"grad_norm": 0.4826101298226364,
"learning_rate": 1.637376338160856e-05,
"loss": 2.1501,
"step": 795
},
{
"epoch": 0.71,
"grad_norm": 0.4166185210327446,
"learning_rate": 1.6314007527406643e-05,
"loss": 2.148,
"step": 800
},
{
"epoch": 0.71,
"grad_norm": 0.5022480623663947,
"learning_rate": 1.6253874365842518e-05,
"loss": 2.1617,
"step": 805
},
{
"epoch": 0.72,
"grad_norm": 0.39970334788657425,
"learning_rate": 1.619336749030509e-05,
"loss": 2.1512,
"step": 810
},
{
"epoch": 0.72,
"grad_norm": 0.4216001027958341,
"learning_rate": 1.613249051651535e-05,
"loss": 2.1392,
"step": 815
},
{
"epoch": 0.73,
"grad_norm": 0.4235687713801533,
"learning_rate": 1.6071247082310337e-05,
"loss": 2.1453,
"step": 820
},
{
"epoch": 0.73,
"grad_norm": 0.41566664668702064,
"learning_rate": 1.6009640847425726e-05,
"loss": 2.1523,
"step": 825
},
{
"epoch": 0.74,
"grad_norm": 0.41330453088636837,
"learning_rate": 1.594767549327714e-05,
"loss": 2.1364,
"step": 830
},
{
"epoch": 0.74,
"grad_norm": 0.41437265776610643,
"learning_rate": 1.588535472274017e-05,
"loss": 2.1413,
"step": 835
},
{
"epoch": 0.74,
"grad_norm": 0.4348718031197105,
"learning_rate": 1.5822682259929086e-05,
"loss": 2.1528,
"step": 840
},
{
"epoch": 0.75,
"grad_norm": 0.420542730016367,
"learning_rate": 1.57596618499743e-05,
"loss": 2.1338,
"step": 845
},
{
"epoch": 0.75,
"grad_norm": 0.40722769776451634,
"learning_rate": 1.5696297258798573e-05,
"loss": 2.133,
"step": 850
},
{
"epoch": 0.76,
"grad_norm": 0.40399518567705306,
"learning_rate": 1.5632592272891964e-05,
"loss": 2.1276,
"step": 855
},
{
"epoch": 0.76,
"grad_norm": 0.4369726580995414,
"learning_rate": 1.5568550699085574e-05,
"loss": 2.1346,
"step": 860
},
{
"epoch": 0.77,
"grad_norm": 0.4145493711186958,
"learning_rate": 1.550417636432404e-05,
"loss": 2.1318,
"step": 865
},
{
"epoch": 0.77,
"grad_norm": 0.4331703033873727,
"learning_rate": 1.5439473115436872e-05,
"loss": 2.1352,
"step": 870
},
{
"epoch": 0.78,
"grad_norm": 0.40919925545952135,
"learning_rate": 1.5374444818908553e-05,
"loss": 2.1155,
"step": 875
},
{
"epoch": 0.78,
"grad_norm": 0.42660353474918383,
"learning_rate": 1.5309095360647505e-05,
"loss": 2.1413,
"step": 880
},
{
"epoch": 0.78,
"grad_norm": 0.3974677057741376,
"learning_rate": 1.5243428645753877e-05,
"loss": 2.1379,
"step": 885
},
{
"epoch": 0.79,
"grad_norm": 0.4386116163244799,
"learning_rate": 1.5177448598286182e-05,
"loss": 2.152,
"step": 890
},
{
"epoch": 0.79,
"grad_norm": 0.43793476843921425,
"learning_rate": 1.5111159161026802e-05,
"loss": 2.126,
"step": 895
},
{
"epoch": 0.8,
"grad_norm": 0.411373483274045,
"learning_rate": 1.5044564295246395e-05,
"loss": 2.1289,
"step": 900
},
{
"epoch": 0.8,
"grad_norm": 0.4113083163266006,
"learning_rate": 1.4977667980467162e-05,
"loss": 2.1446,
"step": 905
},
{
"epoch": 0.81,
"grad_norm": 0.4085799608091074,
"learning_rate": 1.491047421422505e-05,
"loss": 2.1455,
"step": 910
},
{
"epoch": 0.81,
"grad_norm": 0.4283053890615202,
"learning_rate": 1.4842987011830871e-05,
"loss": 2.146,
"step": 915
},
{
"epoch": 0.81,
"grad_norm": 0.4440435185259168,
"learning_rate": 1.4775210406130358e-05,
"loss": 2.1247,
"step": 920
},
{
"epoch": 0.82,
"grad_norm": 0.4006570253639838,
"learning_rate": 1.4707148447263178e-05,
"loss": 2.1182,
"step": 925
},
{
"epoch": 0.82,
"grad_norm": 0.4460349616301842,
"learning_rate": 1.4638805202420896e-05,
"loss": 2.1547,
"step": 930
},
{
"epoch": 0.83,
"grad_norm": 0.43753878951126113,
"learning_rate": 1.4570184755603936e-05,
"loss": 2.1368,
"step": 935
},
{
"epoch": 0.83,
"grad_norm": 0.4329925050207014,
"learning_rate": 1.4501291207377537e-05,
"loss": 2.1362,
"step": 940
},
{
"epoch": 0.84,
"grad_norm": 0.4539296942720091,
"learning_rate": 1.4432128674626713e-05,
"loss": 2.1414,
"step": 945
},
{
"epoch": 0.84,
"grad_norm": 0.42917181768983026,
"learning_rate": 1.4362701290310234e-05,
"loss": 2.1348,
"step": 950
},
{
"epoch": 0.85,
"grad_norm": 0.4262885050230876,
"learning_rate": 1.4293013203213662e-05,
"loss": 2.1229,
"step": 955
},
{
"epoch": 0.85,
"grad_norm": 0.4243870677661146,
"learning_rate": 1.422306857770141e-05,
"loss": 2.1287,
"step": 960
},
{
"epoch": 0.85,
"grad_norm": 0.4003112170725217,
"learning_rate": 1.415287159346793e-05,
"loss": 2.1325,
"step": 965
},
{
"epoch": 0.86,
"grad_norm": 0.40432875292500364,
"learning_rate": 1.4082426445287904e-05,
"loss": 2.1254,
"step": 970
},
{
"epoch": 0.86,
"grad_norm": 0.41678222540401577,
"learning_rate": 1.4011737342765604e-05,
"loss": 2.1436,
"step": 975
},
{
"epoch": 0.87,
"grad_norm": 0.42907637988471764,
"learning_rate": 1.3940808510083321e-05,
"loss": 2.152,
"step": 980
},
{
"epoch": 0.87,
"grad_norm": 0.44791611745861026,
"learning_rate": 1.3869644185748954e-05,
"loss": 2.1294,
"step": 985
},
{
"epoch": 0.88,
"grad_norm": 0.4004395103339801,
"learning_rate": 1.3798248622342719e-05,
"loss": 2.1373,
"step": 990
},
{
"epoch": 0.88,
"grad_norm": 0.4202858585970897,
"learning_rate": 1.3726626086263029e-05,
"loss": 2.127,
"step": 995
},
{
"epoch": 0.89,
"grad_norm": 0.4271532881302461,
"learning_rate": 1.3654780857471548e-05,
"loss": 2.1204,
"step": 1000
},
{
"epoch": 0.89,
"grad_norm": 0.4458140514803481,
"learning_rate": 1.3582717229237434e-05,
"loss": 2.1385,
"step": 1005
},
{
"epoch": 0.89,
"grad_norm": 0.436341151336672,
"learning_rate": 1.3510439507880778e-05,
"loss": 2.1494,
"step": 1010
},
{
"epoch": 0.9,
"grad_norm": 0.4042069130122514,
"learning_rate": 1.3437952012515275e-05,
"loss": 2.1321,
"step": 1015
},
{
"epoch": 0.9,
"grad_norm": 0.4449385582483308,
"learning_rate": 1.336525907479013e-05,
"loss": 2.1362,
"step": 1020
},
{
"epoch": 0.91,
"grad_norm": 0.41131034280821704,
"learning_rate": 1.32923650386312e-05,
"loss": 2.1279,
"step": 1025
},
{
"epoch": 0.91,
"grad_norm": 0.41503686754884667,
"learning_rate": 1.321927425998143e-05,
"loss": 2.1422,
"step": 1030
},
{
"epoch": 0.92,
"grad_norm": 0.42913078973534735,
"learning_rate": 1.314599110654053e-05,
"loss": 2.1298,
"step": 1035
},
{
"epoch": 0.92,
"grad_norm": 0.4375522087703456,
"learning_rate": 1.3072519957504e-05,
"loss": 2.1327,
"step": 1040
},
{
"epoch": 0.93,
"grad_norm": 0.39777062205197167,
"learning_rate": 1.2998865203301424e-05,
"loss": 2.1233,
"step": 1045
},
{
"epoch": 0.93,
"grad_norm": 0.41930871323738256,
"learning_rate": 1.2925031245334112e-05,
"loss": 2.132,
"step": 1050
},
{
"epoch": 0.93,
"grad_norm": 0.4381474710886509,
"learning_rate": 1.2851022495712092e-05,
"loss": 2.1316,
"step": 1055
},
{
"epoch": 0.94,
"grad_norm": 0.42807628586234325,
"learning_rate": 1.2776843376990448e-05,
"loss": 2.12,
"step": 1060
},
{
"epoch": 0.94,
"grad_norm": 0.4338973499824451,
"learning_rate": 1.270249832190505e-05,
"loss": 2.1191,
"step": 1065
},
{
"epoch": 0.95,
"grad_norm": 0.43301978924071477,
"learning_rate": 1.2627991773107651e-05,
"loss": 2.1168,
"step": 1070
},
{
"epoch": 0.95,
"grad_norm": 0.4434445595390298,
"learning_rate": 1.2553328182900414e-05,
"loss": 2.1438,
"step": 1075
},
{
"epoch": 0.96,
"grad_norm": 0.44970886991124087,
"learning_rate": 1.2478512012969864e-05,
"loss": 2.1134,
"step": 1080
},
{
"epoch": 0.96,
"grad_norm": 0.429311171844366,
"learning_rate": 1.2403547734120253e-05,
"loss": 2.1256,
"step": 1085
},
{
"epoch": 0.97,
"grad_norm": 0.4097438629027672,
"learning_rate": 1.2328439826006415e-05,
"loss": 2.1301,
"step": 1090
},
{
"epoch": 0.97,
"grad_norm": 0.40558336586323696,
"learning_rate": 1.2253192776866059e-05,
"loss": 2.1218,
"step": 1095
},
{
"epoch": 0.97,
"grad_norm": 0.3969230786876914,
"learning_rate": 1.2177811083251572e-05,
"loss": 2.0971,
"step": 1100
},
{
"epoch": 0.98,
"grad_norm": 0.42264096644737814,
"learning_rate": 1.2102299249761315e-05,
"loss": 2.1216,
"step": 1105
},
{
"epoch": 0.98,
"grad_norm": 0.43041870396541304,
"learning_rate": 1.2026661788770453e-05,
"loss": 2.1199,
"step": 1110
},
{
"epoch": 0.99,
"grad_norm": 0.43792544078155893,
"learning_rate": 1.1950903220161286e-05,
"loss": 2.1193,
"step": 1115
},
{
"epoch": 0.99,
"grad_norm": 0.46024843969749113,
"learning_rate": 1.1875028071053165e-05,
"loss": 2.1384,
"step": 1120
},
{
"epoch": 1.0,
"grad_norm": 0.4262908584944093,
"learning_rate": 1.1799040875531975e-05,
"loss": 2.1213,
"step": 1125
},
{
"epoch": 1.0,
"eval_loss": 2.127272605895996,
"eval_runtime": 22.1739,
"eval_samples_per_second": 1442.192,
"eval_steps_per_second": 22.549,
"step": 1129
},
{
"epoch": 1.0,
"grad_norm": 0.40723005257539824,
"learning_rate": 1.1722946174379168e-05,
"loss": 2.1371,
"step": 1130
},
{
"epoch": 1.01,
"grad_norm": 0.4165006619376059,
"learning_rate": 1.1646748514800441e-05,
"loss": 2.1133,
"step": 1135
},
{
"epoch": 1.01,
"grad_norm": 0.4265534445969313,
"learning_rate": 1.1570452450153992e-05,
"loss": 2.1102,
"step": 1140
},
{
"epoch": 1.01,
"grad_norm": 0.41417530766953414,
"learning_rate": 1.149406253967843e-05,
"loss": 2.1215,
"step": 1145
},
{
"epoch": 1.02,
"grad_norm": 0.4200503893771155,
"learning_rate": 1.1417583348220322e-05,
"loss": 2.1132,
"step": 1150
},
{
"epoch": 1.02,
"grad_norm": 0.3901369484912683,
"learning_rate": 1.134101944596143e-05,
"loss": 2.1223,
"step": 1155
},
{
"epoch": 1.03,
"grad_norm": 0.39095544350917155,
"learning_rate": 1.1264375408145582e-05,
"loss": 2.1277,
"step": 1160
},
{
"epoch": 1.03,
"grad_norm": 0.395664418651239,
"learning_rate": 1.118765581480529e-05,
"loss": 2.1014,
"step": 1165
},
{
"epoch": 1.04,
"grad_norm": 0.41528855354394845,
"learning_rate": 1.1110865250488047e-05,
"loss": 2.1277,
"step": 1170
},
{
"epoch": 1.04,
"grad_norm": 0.4220840220117444,
"learning_rate": 1.1034008303982373e-05,
"loss": 2.124,
"step": 1175
},
{
"epoch": 1.05,
"grad_norm": 0.4147919955179845,
"learning_rate": 1.0957089568043607e-05,
"loss": 2.1107,
"step": 1180
},
{
"epoch": 1.05,
"grad_norm": 0.41180982643238745,
"learning_rate": 1.088011363911944e-05,
"loss": 2.1053,
"step": 1185
},
{
"epoch": 1.05,
"grad_norm": 0.4226688550104511,
"learning_rate": 1.080308511707527e-05,
"loss": 2.118,
"step": 1190
},
{
"epoch": 1.06,
"grad_norm": 0.4127679361918531,
"learning_rate": 1.0726008604919296e-05,
"loss": 2.1071,
"step": 1195
},
{
"epoch": 1.06,
"grad_norm": 0.41422806192198935,
"learning_rate": 1.0648888708527481e-05,
"loss": 2.1217,
"step": 1200
},
{
"epoch": 1.07,
"grad_norm": 0.40793604660952315,
"learning_rate": 1.0571730036368308e-05,
"loss": 2.1057,
"step": 1205
},
{
"epoch": 1.07,
"grad_norm": 0.41928090055173245,
"learning_rate": 1.0494537199227393e-05,
"loss": 2.1029,
"step": 1210
},
{
"epoch": 1.08,
"grad_norm": 0.3963968527749634,
"learning_rate": 1.0417314809931945e-05,
"loss": 2.1043,
"step": 1215
},
{
"epoch": 1.08,
"grad_norm": 0.4333425677178677,
"learning_rate": 1.0340067483075135e-05,
"loss": 2.0963,
"step": 1220
},
{
"epoch": 1.09,
"grad_norm": 0.4129126276223174,
"learning_rate": 1.0262799834740334e-05,
"loss": 2.129,
"step": 1225
},
{
"epoch": 1.09,
"grad_norm": 0.40898715204006264,
"learning_rate": 1.0185516482225264e-05,
"loss": 2.1263,
"step": 1230
},
{
"epoch": 1.09,
"grad_norm": 0.4126097914010792,
"learning_rate": 1.0108222043766087e-05,
"loss": 2.12,
"step": 1235
},
{
"epoch": 1.1,
"grad_norm": 0.4012576574381173,
"learning_rate": 1.0030921138261422e-05,
"loss": 2.0842,
"step": 1240
},
{
"epoch": 1.1,
"grad_norm": 0.40240868081754594,
"learning_rate": 9.953618384996353e-06,
"loss": 2.1134,
"step": 1245
},
{
"epoch": 1.11,
"grad_norm": 0.40559837490056405,
"learning_rate": 9.876318403366371e-06,
"loss": 2.0966,
"step": 1250
},
{
"epoch": 1.11,
"grad_norm": 0.4013529363767149,
"learning_rate": 9.79902581260135e-06,
"loss": 2.1092,
"step": 1255
},
{
"epoch": 1.12,
"grad_norm": 0.4055440562747367,
"learning_rate": 9.721745231489499e-06,
"loss": 2.1149,
"step": 1260
},
{
"epoch": 1.12,
"grad_norm": 0.4035773407220577,
"learning_rate": 9.644481278101366e-06,
"loss": 2.1309,
"step": 1265
},
{
"epoch": 1.12,
"grad_norm": 0.41371656524676326,
"learning_rate": 9.567238569513872e-06,
"loss": 2.1018,
"step": 1270
},
{
"epoch": 1.13,
"grad_norm": 0.42252949212672125,
"learning_rate": 9.49002172153442e-06,
"loss": 2.1129,
"step": 1275
},
{
"epoch": 1.13,
"grad_norm": 0.44147408528722076,
"learning_rate": 9.412835348425038e-06,
"loss": 2.1136,
"step": 1280
},
{
"epoch": 1.14,
"grad_norm": 0.4373864316683171,
"learning_rate": 9.335684062626669e-06,
"loss": 2.112,
"step": 1285
},
{
"epoch": 1.14,
"grad_norm": 0.4186144161163772,
"learning_rate": 9.25857247448354e-06,
"loss": 2.1002,
"step": 1290
},
{
"epoch": 1.15,
"grad_norm": 0.42068101148699355,
"learning_rate": 9.181505191967656e-06,
"loss": 2.1033,
"step": 1295
},
{
"epoch": 1.15,
"grad_norm": 0.4068327137098232,
"learning_rate": 9.104486820403438e-06,
"loss": 2.1137,
"step": 1300
},
{
"epoch": 1.16,
"grad_norm": 0.3925407893472356,
"learning_rate": 9.027521962192532e-06,
"loss": 2.115,
"step": 1305
},
{
"epoch": 1.16,
"grad_norm": 0.43767243051487653,
"learning_rate": 8.950615216538765e-06,
"loss": 2.1281,
"step": 1310
},
{
"epoch": 1.16,
"grad_norm": 0.4341377754982897,
"learning_rate": 8.873771179173339e-06,
"loss": 2.1031,
"step": 1315
},
{
"epoch": 1.17,
"grad_norm": 0.4169360771557741,
"learning_rate": 8.796994442080167e-06,
"loss": 2.1161,
"step": 1320
},
{
"epoch": 1.17,
"grad_norm": 0.41508105421346786,
"learning_rate": 8.720289593221502e-06,
"loss": 2.1147,
"step": 1325
},
{
"epoch": 1.18,
"grad_norm": 0.4376766510575495,
"learning_rate": 8.643661216263744e-06,
"loss": 2.0824,
"step": 1330
},
{
"epoch": 1.18,
"grad_norm": 0.39554269033624057,
"learning_rate": 8.567113890303554e-06,
"loss": 2.1216,
"step": 1335
},
{
"epoch": 1.19,
"grad_norm": 0.4395238196101571,
"learning_rate": 8.490652189594212e-06,
"loss": 2.1071,
"step": 1340
},
{
"epoch": 1.19,
"grad_norm": 0.4010878903838548,
"learning_rate": 8.414280683272273e-06,
"loss": 2.1239,
"step": 1345
},
{
"epoch": 1.2,
"grad_norm": 0.3942671903374018,
"learning_rate": 8.338003935084531e-06,
"loss": 2.113,
"step": 1350
},
{
"epoch": 1.2,
"grad_norm": 0.4278130962487623,
"learning_rate": 8.2618265031153e-06,
"loss": 2.1125,
"step": 1355
},
{
"epoch": 1.2,
"grad_norm": 0.3944639027917202,
"learning_rate": 8.185752939514026e-06,
"loss": 2.0962,
"step": 1360
},
{
"epoch": 1.21,
"grad_norm": 0.40659734461588076,
"learning_rate": 8.109787790223285e-06,
"loss": 2.0982,
"step": 1365
},
{
"epoch": 1.21,
"grad_norm": 0.45381622699152685,
"learning_rate": 8.033935594707116e-06,
"loss": 2.1023,
"step": 1370
},
{
"epoch": 1.22,
"grad_norm": 0.42089854648317804,
"learning_rate": 7.958200885679752e-06,
"loss": 2.1216,
"step": 1375
},
{
"epoch": 1.22,
"grad_norm": 0.40474946249263904,
"learning_rate": 7.88258818883477e-06,
"loss": 2.1316,
"step": 1380
},
{
"epoch": 1.23,
"grad_norm": 0.40839298659360906,
"learning_rate": 7.807102022574631e-06,
"loss": 2.0998,
"step": 1385
},
{
"epoch": 1.23,
"grad_norm": 0.4108597211066138,
"learning_rate": 7.7317468977407e-06,
"loss": 2.1043,
"step": 1390
},
{
"epoch": 1.24,
"grad_norm": 0.4140704738815021,
"learning_rate": 7.65652731734366e-06,
"loss": 2.0905,
"step": 1395
},
{
"epoch": 1.24,
"grad_norm": 0.41776686260392426,
"learning_rate": 7.5814477762944435e-06,
"loss": 2.1261,
"step": 1400
},
{
"epoch": 1.24,
"grad_norm": 0.41530489559602735,
"learning_rate": 7.506512761135627e-06,
"loss": 2.1039,
"step": 1405
},
{
"epoch": 1.25,
"grad_norm": 0.40700432981592705,
"learning_rate": 7.431726749773322e-06,
"loss": 2.0818,
"step": 1410
},
{
"epoch": 1.25,
"grad_norm": 0.39564305744757305,
"learning_rate": 7.3570942112095955e-06,
"loss": 2.1178,
"step": 1415
},
{
"epoch": 1.26,
"grad_norm": 0.4178636304864647,
"learning_rate": 7.282619605275409e-06,
"loss": 2.1073,
"step": 1420
},
{
"epoch": 1.26,
"grad_norm": 0.42732536293687734,
"learning_rate": 7.208307382364111e-06,
"loss": 2.0965,
"step": 1425
},
{
"epoch": 1.27,
"grad_norm": 0.4174767550441002,
"learning_rate": 7.134161983165498e-06,
"loss": 2.1125,
"step": 1430
},
{
"epoch": 1.27,
"grad_norm": 0.3970977394089729,
"learning_rate": 7.060187838400451e-06,
"loss": 2.1136,
"step": 1435
},
{
"epoch": 1.28,
"grad_norm": 0.41936318103223125,
"learning_rate": 6.986389368556168e-06,
"loss": 2.1076,
"step": 1440
},
{
"epoch": 1.28,
"grad_norm": 0.3999968334112073,
"learning_rate": 6.912770983622008e-06,
"loss": 2.1171,
"step": 1445
},
{
"epoch": 1.28,
"grad_norm": 0.40546677393248054,
"learning_rate": 6.839337082825954e-06,
"loss": 2.1166,
"step": 1450
},
{
"epoch": 1.29,
"grad_norm": 0.4083040392461004,
"learning_rate": 6.766092054371744e-06,
"loss": 2.1066,
"step": 1455
},
{
"epoch": 1.29,
"grad_norm": 0.4258312102729685,
"learning_rate": 6.693040275176623e-06,
"loss": 2.1284,
"step": 1460
},
{
"epoch": 1.3,
"grad_norm": 0.4093311630959738,
"learning_rate": 6.62018611060982e-06,
"loss": 2.1022,
"step": 1465
},
{
"epoch": 1.3,
"grad_norm": 0.40040887955887994,
"learning_rate": 6.547533914231654e-06,
"loss": 2.1095,
"step": 1470
},
{
"epoch": 1.31,
"grad_norm": 0.42225943956721124,
"learning_rate": 6.475088027533399e-06,
"loss": 2.1251,
"step": 1475
},
{
"epoch": 1.31,
"grad_norm": 0.4140975992383474,
"learning_rate": 6.40285277967784e-06,
"loss": 2.1165,
"step": 1480
},
{
"epoch": 1.32,
"grad_norm": 0.4244163593071348,
"learning_rate": 6.330832487240573e-06,
"loss": 2.0902,
"step": 1485
},
{
"epoch": 1.32,
"grad_norm": 0.4097848769182914,
"learning_rate": 6.2590314539520695e-06,
"loss": 2.1172,
"step": 1490
},
{
"epoch": 1.32,
"grad_norm": 0.41141590708499054,
"learning_rate": 6.187453970440484e-06,
"loss": 2.0995,
"step": 1495
},
{
"epoch": 1.33,
"grad_norm": 0.4094485361390189,
"learning_rate": 6.116104313975267e-06,
"loss": 2.1215,
"step": 1500
},
{
"epoch": 1.33,
"grad_norm": 0.4038677821218221,
"learning_rate": 6.044986748211556e-06,
"loss": 2.0907,
"step": 1505
},
{
"epoch": 1.34,
"grad_norm": 0.39065768813151286,
"learning_rate": 5.974105522935416e-06,
"loss": 2.1003,
"step": 1510
},
{
"epoch": 1.34,
"grad_norm": 0.3962049547910227,
"learning_rate": 5.903464873809854e-06,
"loss": 2.0894,
"step": 1515
},
{
"epoch": 1.35,
"grad_norm": 0.4654900808214661,
"learning_rate": 5.833069022121727e-06,
"loss": 2.1079,
"step": 1520
},
{
"epoch": 1.35,
"grad_norm": 0.40121323056151253,
"learning_rate": 5.762922174529482e-06,
"loss": 2.0802,
"step": 1525
},
{
"epoch": 1.36,
"grad_norm": 0.4272967812226492,
"learning_rate": 5.693028522811783e-06,
"loss": 2.0999,
"step": 1530
},
{
"epoch": 1.36,
"grad_norm": 0.41351449563322684,
"learning_rate": 5.6233922436170205e-06,
"loss": 2.0991,
"step": 1535
},
{
"epoch": 1.36,
"grad_norm": 0.4218885369397254,
"learning_rate": 5.5540174982137185e-06,
"loss": 2.1105,
"step": 1540
},
{
"epoch": 1.37,
"grad_norm": 0.40918986819578146,
"learning_rate": 5.484908432241889e-06,
"loss": 2.1046,
"step": 1545
},
{
"epoch": 1.37,
"grad_norm": 0.4017822298125311,
"learning_rate": 5.416069175465274e-06,
"loss": 2.1033,
"step": 1550
},
{
"epoch": 1.38,
"grad_norm": 0.4100251647421733,
"learning_rate": 5.347503841524582e-06,
"loss": 2.1235,
"step": 1555
},
{
"epoch": 1.38,
"grad_norm": 0.4194869497973335,
"learning_rate": 5.279216527691657e-06,
"loss": 2.1115,
"step": 1560
},
{
"epoch": 1.39,
"grad_norm": 0.4173285243712288,
"learning_rate": 5.211211314624653e-06,
"loss": 2.1007,
"step": 1565
},
{
"epoch": 1.39,
"grad_norm": 0.4171312909053348,
"learning_rate": 5.143492266124164e-06,
"loss": 2.1085,
"step": 1570
},
{
"epoch": 1.4,
"grad_norm": 0.39448869145365084,
"learning_rate": 5.076063428890393e-06,
"loss": 2.1039,
"step": 1575
},
{
"epoch": 1.4,
"grad_norm": 0.4197610608758638,
"learning_rate": 5.008928832281339e-06,
"loss": 2.095,
"step": 1580
},
{
"epoch": 1.4,
"grad_norm": 0.4182941699698344,
"learning_rate": 4.942092488072e-06,
"loss": 2.1066,
"step": 1585
},
{
"epoch": 1.41,
"grad_norm": 0.41038377020595973,
"learning_rate": 4.875558390214652e-06,
"loss": 2.0944,
"step": 1590
},
{
"epoch": 1.41,
"grad_norm": 0.41931401936390567,
"learning_rate": 4.8093305146001815e-06,
"loss": 2.0941,
"step": 1595
},
{
"epoch": 1.42,
"grad_norm": 0.4142471239944317,
"learning_rate": 4.743412818820488e-06,
"loss": 2.1052,
"step": 1600
},
{
"epoch": 1.42,
"grad_norm": 0.3965863879905922,
"learning_rate": 4.677809241931994e-06,
"loss": 2.1039,
"step": 1605
},
{
"epoch": 1.43,
"grad_norm": 0.4077258460899842,
"learning_rate": 4.612523704220264e-06,
"loss": 2.1022,
"step": 1610
},
{
"epoch": 1.43,
"grad_norm": 0.4400824748890086,
"learning_rate": 4.5475601069657304e-06,
"loss": 2.1029,
"step": 1615
},
{
"epoch": 1.43,
"grad_norm": 0.4618350439187338,
"learning_rate": 4.482922332210569e-06,
"loss": 2.0943,
"step": 1620
},
{
"epoch": 1.44,
"grad_norm": 0.4536620607174402,
"learning_rate": 4.418614242526717e-06,
"loss": 2.0889,
"step": 1625
},
{
"epoch": 1.44,
"grad_norm": 0.395184232757659,
"learning_rate": 4.354639680785059e-06,
"loss": 2.1045,
"step": 1630
},
{
"epoch": 1.45,
"grad_norm": 0.3983844206518319,
"learning_rate": 4.291002469925782e-06,
"loss": 2.1184,
"step": 1635
},
{
"epoch": 1.45,
"grad_norm": 0.42341867611083195,
"learning_rate": 4.227706412729943e-06,
"loss": 2.1014,
"step": 1640
},
{
"epoch": 1.46,
"grad_norm": 0.38390499054508137,
"learning_rate": 4.1647552915922e-06,
"loss": 2.0999,
"step": 1645
},
{
"epoch": 1.46,
"grad_norm": 0.4020575992796089,
"learning_rate": 4.1021528682948064e-06,
"loss": 2.0864,
"step": 1650
},
{
"epoch": 1.47,
"grad_norm": 0.3893311251004592,
"learning_rate": 4.039902883782814e-06,
"loss": 2.1092,
"step": 1655
},
{
"epoch": 1.47,
"grad_norm": 0.4022507142375332,
"learning_rate": 3.978009057940518e-06,
"loss": 2.1063,
"step": 1660
},
{
"epoch": 1.47,
"grad_norm": 0.39132377038485955,
"learning_rate": 3.916475089369175e-06,
"loss": 2.099,
"step": 1665
},
{
"epoch": 1.48,
"grad_norm": 0.4054379326638264,
"learning_rate": 3.855304655165978e-06,
"loss": 2.111,
"step": 1670
},
{
"epoch": 1.48,
"grad_norm": 0.4153443947887743,
"learning_rate": 3.794501410704331e-06,
"loss": 2.1022,
"step": 1675
},
{
"epoch": 1.49,
"grad_norm": 0.3958658218820195,
"learning_rate": 3.7340689894154023e-06,
"loss": 2.0816,
"step": 1680
},
{
"epoch": 1.49,
"grad_norm": 0.3935656557523995,
"learning_rate": 3.674011002571022e-06,
"loss": 2.0942,
"step": 1685
},
{
"epoch": 1.5,
"grad_norm": 0.39101679788894395,
"learning_rate": 3.6143310390678544e-06,
"loss": 2.0923,
"step": 1690
},
{
"epoch": 1.5,
"grad_norm": 0.4385258009544127,
"learning_rate": 3.555032665212964e-06,
"loss": 2.1226,
"step": 1695
},
{
"epoch": 1.51,
"grad_norm": 0.4177391551580632,
"learning_rate": 3.496119424510678e-06,
"loss": 2.0929,
"step": 1700
},
{
"epoch": 1.51,
"grad_norm": 0.41620389629735616,
"learning_rate": 3.4375948374508516e-06,
"loss": 2.0741,
"step": 1705
},
{
"epoch": 1.51,
"grad_norm": 0.4114783396501206,
"learning_rate": 3.3794624012984913e-06,
"loss": 2.0994,
"step": 1710
},
{
"epoch": 1.52,
"grad_norm": 0.4409745190643104,
"learning_rate": 3.3217255898847635e-06,
"loss": 2.0996,
"step": 1715
},
{
"epoch": 1.52,
"grad_norm": 0.40264327807277855,
"learning_rate": 3.2643878533994145e-06,
"loss": 2.1051,
"step": 1720
},
{
"epoch": 1.53,
"grad_norm": 0.39595811156963595,
"learning_rate": 3.20745261818459e-06,
"loss": 2.0883,
"step": 1725
},
{
"epoch": 1.53,
"grad_norm": 0.3924141706237974,
"learning_rate": 3.1509232865300886e-06,
"loss": 2.0967,
"step": 1730
},
{
"epoch": 1.54,
"grad_norm": 0.38763293693962486,
"learning_rate": 3.09480323647006e-06,
"loss": 2.0907,
"step": 1735
},
{
"epoch": 1.54,
"grad_norm": 0.3979048685211274,
"learning_rate": 3.039095821581127e-06,
"loss": 2.1031,
"step": 1740
},
{
"epoch": 1.55,
"grad_norm": 0.39294112678579723,
"learning_rate": 2.983804370781996e-06,
"loss": 2.1007,
"step": 1745
},
{
"epoch": 1.55,
"grad_norm": 0.39182233954069495,
"learning_rate": 2.9289321881345257e-06,
"loss": 2.1112,
"step": 1750
},
{
"epoch": 1.55,
"grad_norm": 0.4191507971226216,
"learning_rate": 2.8744825526462882e-06,
"loss": 2.1027,
"step": 1755
},
{
"epoch": 1.56,
"grad_norm": 0.38632711327880076,
"learning_rate": 2.8204587180746256e-06,
"loss": 2.1024,
"step": 1760
},
{
"epoch": 1.56,
"grad_norm": 0.401153479741774,
"learning_rate": 2.7668639127322084e-06,
"loss": 2.1158,
"step": 1765
},
{
"epoch": 1.57,
"grad_norm": 0.3825496267553849,
"learning_rate": 2.713701339294129e-06,
"loss": 2.0938,
"step": 1770
},
{
"epoch": 1.57,
"grad_norm": 0.40987757351301446,
"learning_rate": 2.66097417460651e-06,
"loss": 2.0973,
"step": 1775
},
{
"epoch": 1.58,
"grad_norm": 0.3822408562218838,
"learning_rate": 2.6086855694966795e-06,
"loss": 2.0823,
"step": 1780
},
{
"epoch": 1.58,
"grad_norm": 0.4023814082684278,
"learning_rate": 2.5568386485848663e-06,
"loss": 2.1023,
"step": 1785
},
{
"epoch": 1.59,
"grad_norm": 0.38995290656891923,
"learning_rate": 2.505436510097494e-06,
"loss": 2.1128,
"step": 1790
},
{
"epoch": 1.59,
"grad_norm": 0.4192274003350971,
"learning_rate": 2.45448222568204e-06,
"loss": 2.108,
"step": 1795
},
{
"epoch": 1.59,
"grad_norm": 0.3958146568701287,
"learning_rate": 2.4039788402234787e-06,
"loss": 2.0815,
"step": 1800
},
{
"epoch": 1.6,
"grad_norm": 0.3899345487225284,
"learning_rate": 2.3539293716623268e-06,
"loss": 2.1155,
"step": 1805
},
{
"epoch": 1.6,
"grad_norm": 0.4117980185924797,
"learning_rate": 2.304336810814305e-06,
"loss": 2.0988,
"step": 1810
},
{
"epoch": 1.61,
"grad_norm": 0.3929442131747748,
"learning_rate": 2.2552041211916052e-06,
"loss": 2.0972,
"step": 1815
},
{
"epoch": 1.61,
"grad_norm": 0.4192037048524669,
"learning_rate": 2.2065342388258193e-06,
"loss": 2.0875,
"step": 1820
},
{
"epoch": 1.62,
"grad_norm": 0.40277691417023526,
"learning_rate": 2.1583300720924604e-06,
"loss": 2.0743,
"step": 1825
},
{
"epoch": 1.62,
"grad_norm": 0.4115665165694226,
"learning_rate": 2.1105945015371985e-06,
"loss": 2.092,
"step": 1830
},
{
"epoch": 1.63,
"grad_norm": 0.40792864223683667,
"learning_rate": 2.063330379703702e-06,
"loss": 2.1053,
"step": 1835
},
{
"epoch": 1.63,
"grad_norm": 0.3960421374782312,
"learning_rate": 2.016540530963188e-06,
"loss": 2.0908,
"step": 1840
},
{
"epoch": 1.63,
"grad_norm": 0.39468768468781434,
"learning_rate": 1.9702277513456493e-06,
"loss": 2.1015,
"step": 1845
},
{
"epoch": 1.64,
"grad_norm": 0.3951629256392003,
"learning_rate": 1.9243948083727626e-06,
"loss": 2.1066,
"step": 1850
},
{
"epoch": 1.64,
"grad_norm": 0.37908230643141794,
"learning_rate": 1.879044440892517e-06,
"loss": 2.105,
"step": 1855
},
{
"epoch": 1.65,
"grad_norm": 0.4043119825923411,
"learning_rate": 1.8341793589155444e-06,
"loss": 2.0941,
"step": 1860
},
{
"epoch": 1.65,
"grad_norm": 0.3992125291402619,
"learning_rate": 1.789802243453178e-06,
"loss": 2.0958,
"step": 1865
},
{
"epoch": 1.66,
"grad_norm": 0.44824510154999736,
"learning_rate": 1.7459157463572396e-06,
"loss": 2.1184,
"step": 1870
},
{
"epoch": 1.66,
"grad_norm": 0.40290066426181237,
"learning_rate": 1.7025224901615811e-06,
"loss": 2.087,
"step": 1875
},
{
"epoch": 1.67,
"grad_norm": 0.40329042437474516,
"learning_rate": 1.6596250679253568e-06,
"loss": 2.1043,
"step": 1880
},
{
"epoch": 1.67,
"grad_norm": 0.4026581756741885,
"learning_rate": 1.6172260430780772e-06,
"loss": 2.0896,
"step": 1885
},
{
"epoch": 1.67,
"grad_norm": 0.40186470883700653,
"learning_rate": 1.5753279492664264e-06,
"loss": 2.0809,
"step": 1890
},
{
"epoch": 1.68,
"grad_norm": 0.40243244247873866,
"learning_rate": 1.5339332902028537e-06,
"loss": 2.1114,
"step": 1895
},
{
"epoch": 1.68,
"grad_norm": 0.4017594562983069,
"learning_rate": 1.493044539515961e-06,
"loss": 2.1031,
"step": 1900
},
{
"epoch": 1.69,
"grad_norm": 0.3981128121055056,
"learning_rate": 1.4526641406026898e-06,
"loss": 2.1044,
"step": 1905
},
{
"epoch": 1.69,
"grad_norm": 0.392491209687315,
"learning_rate": 1.4127945064823023e-06,
"loss": 2.0905,
"step": 1910
},
{
"epoch": 1.7,
"grad_norm": 0.38637746157683445,
"learning_rate": 1.3734380196521923e-06,
"loss": 2.0806,
"step": 1915
},
{
"epoch": 1.7,
"grad_norm": 0.3860383350327143,
"learning_rate": 1.334597031945517e-06,
"loss": 2.0961,
"step": 1920
},
{
"epoch": 1.71,
"grad_norm": 0.39899628934706927,
"learning_rate": 1.296273864390646e-06,
"loss": 2.0921,
"step": 1925
},
{
"epoch": 1.71,
"grad_norm": 0.4015920544044468,
"learning_rate": 1.2584708070724738e-06,
"loss": 2.0863,
"step": 1930
},
{
"epoch": 1.71,
"grad_norm": 0.40852618728142004,
"learning_rate": 1.2211901189955689e-06,
"loss": 2.0844,
"step": 1935
},
{
"epoch": 1.72,
"grad_norm": 0.41728571657256014,
"learning_rate": 1.1844340279491772e-06,
"loss": 2.091,
"step": 1940
},
{
"epoch": 1.72,
"grad_norm": 0.3756106716332952,
"learning_rate": 1.1482047303740996e-06,
"loss": 2.1058,
"step": 1945
},
{
"epoch": 1.73,
"grad_norm": 0.3787679724558361,
"learning_rate": 1.1125043912314438e-06,
"loss": 2.0792,
"step": 1950
},
{
"epoch": 1.73,
"grad_norm": 0.39416480322765324,
"learning_rate": 1.0773351438732392e-06,
"loss": 2.0941,
"step": 1955
},
{
"epoch": 1.74,
"grad_norm": 0.39660635274260575,
"learning_rate": 1.0426990899149658e-06,
"loss": 2.1108,
"step": 1960
},
{
"epoch": 1.74,
"grad_norm": 0.41574515299547316,
"learning_rate": 1.0085982991099585e-06,
"loss": 2.0842,
"step": 1965
},
{
"epoch": 1.74,
"grad_norm": 0.4015546675146476,
"learning_rate": 9.750348092257368e-07,
"loss": 2.1133,
"step": 1970
},
{
"epoch": 1.75,
"grad_norm": 0.4011902676802935,
"learning_rate": 9.420106259222184e-07,
"loss": 2.0977,
"step": 1975
},
{
"epoch": 1.75,
"grad_norm": 0.38641725135656807,
"learning_rate": 9.095277226318766e-07,
"loss": 2.1061,
"step": 1980
},
{
"epoch": 1.76,
"grad_norm": 0.3919138060745633,
"learning_rate": 8.775880404418113e-07,
"loss": 2.1073,
"step": 1985
},
{
"epoch": 1.76,
"grad_norm": 0.3921573507167864,
"learning_rate": 8.461934879777545e-07,
"loss": 2.107,
"step": 1990
},
{
"epoch": 1.77,
"grad_norm": 0.3927198106647622,
"learning_rate": 8.153459412900156e-07,
"loss": 2.1048,
"step": 1995
},
{
"epoch": 1.77,
"grad_norm": 0.3889128398277548,
"learning_rate": 7.850472437413748e-07,
"loss": 2.0887,
"step": 2000
},
{
"epoch": 1.78,
"grad_norm": 0.40798311368353046,
"learning_rate": 7.552992058969299e-07,
"loss": 2.0947,
"step": 2005
},
{
"epoch": 1.78,
"grad_norm": 0.40181448732733616,
"learning_rate": 7.261036054158965e-07,
"loss": 2.0936,
"step": 2010
},
{
"epoch": 1.78,
"grad_norm": 0.38883853844853333,
"learning_rate": 6.974621869453924e-07,
"loss": 2.1059,
"step": 2015
},
{
"epoch": 1.79,
"grad_norm": 0.40388469626182455,
"learning_rate": 6.693766620161691e-07,
"loss": 2.0738,
"step": 2020
},
{
"epoch": 1.79,
"grad_norm": 0.39957045639742295,
"learning_rate": 6.418487089403392e-07,
"loss": 2.0968,
"step": 2025
},
{
"epoch": 1.8,
"grad_norm": 0.41286689792246,
"learning_rate": 6.148799727110911e-07,
"loss": 2.113,
"step": 2030
},
{
"epoch": 1.8,
"grad_norm": 0.40197726001637857,
"learning_rate": 5.884720649043807e-07,
"loss": 2.1007,
"step": 2035
},
{
"epoch": 1.81,
"grad_norm": 0.39910390504267484,
"learning_rate": 5.626265635826367e-07,
"loss": 2.0865,
"step": 2040
},
{
"epoch": 1.81,
"grad_norm": 0.41231928670539864,
"learning_rate": 5.373450132004499e-07,
"loss": 2.0913,
"step": 2045
},
{
"epoch": 1.82,
"grad_norm": 0.3990823019488567,
"learning_rate": 5.126289245122906e-07,
"loss": 2.0932,
"step": 2050
},
{
"epoch": 1.82,
"grad_norm": 0.38839408989932034,
"learning_rate": 4.884797744822212e-07,
"loss": 2.0978,
"step": 2055
},
{
"epoch": 1.82,
"grad_norm": 0.3806673308803765,
"learning_rate": 4.648990061956493e-07,
"loss": 2.07,
"step": 2060
},
{
"epoch": 1.83,
"grad_norm": 0.4074456595154459,
"learning_rate": 4.418880287730798e-07,
"loss": 2.1086,
"step": 2065
},
{
"epoch": 1.83,
"grad_norm": 0.39196159709559636,
"learning_rate": 4.194482172859127e-07,
"loss": 2.1096,
"step": 2070
},
{
"epoch": 1.84,
"grad_norm": 0.38968873697963935,
"learning_rate": 3.9758091267428245e-07,
"loss": 2.1058,
"step": 2075
},
{
"epoch": 1.84,
"grad_norm": 0.4035061362528785,
"learning_rate": 3.762874216669166e-07,
"loss": 2.0968,
"step": 2080
},
{
"epoch": 1.85,
"grad_norm": 0.3844793064672678,
"learning_rate": 3.555690167030512e-07,
"loss": 2.0942,
"step": 2085
},
{
"epoch": 1.85,
"grad_norm": 0.38877399092787523,
"learning_rate": 3.354269358563966e-07,
"loss": 2.0752,
"step": 2090
},
{
"epoch": 1.86,
"grad_norm": 0.39974848670931834,
"learning_rate": 3.158623827611529e-07,
"loss": 2.082,
"step": 2095
},
{
"epoch": 1.86,
"grad_norm": 0.39175445568930795,
"learning_rate": 2.968765265400808e-07,
"loss": 2.1093,
"step": 2100
},
{
"epoch": 1.86,
"grad_norm": 0.38984444652162176,
"learning_rate": 2.784705017346423e-07,
"loss": 2.1011,
"step": 2105
},
{
"epoch": 1.87,
"grad_norm": 0.4158962064409584,
"learning_rate": 2.606454082372045e-07,
"loss": 2.118,
"step": 2110
},
{
"epoch": 1.87,
"grad_norm": 0.3937202740943716,
"learning_rate": 2.4340231122530477e-07,
"loss": 2.091,
"step": 2115
},
{
"epoch": 1.88,
"grad_norm": 0.39757180612367116,
"learning_rate": 2.2674224109800913e-07,
"loss": 2.1105,
"step": 2120
},
{
"epoch": 1.88,
"grad_norm": 0.4090245098298795,
"learning_rate": 2.106661934143317e-07,
"loss": 2.0776,
"step": 2125
},
{
"epoch": 1.89,
"grad_norm": 0.4012268787820084,
"learning_rate": 1.9517512883374667e-07,
"loss": 2.0766,
"step": 2130
},
{
"epoch": 1.89,
"grad_norm": 0.3943402628984385,
"learning_rate": 1.802699730587798e-07,
"loss": 2.0959,
"step": 2135
},
{
"epoch": 1.9,
"grad_norm": 0.38257656512904975,
"learning_rate": 1.659516167796904e-07,
"loss": 2.0748,
"step": 2140
},
{
"epoch": 1.9,
"grad_norm": 0.3902633313179273,
"learning_rate": 1.522209156212484e-07,
"loss": 2.0951,
"step": 2145
},
{
"epoch": 1.9,
"grad_norm": 0.3883365627039246,
"learning_rate": 1.3907869009160525e-07,
"loss": 2.0935,
"step": 2150
},
{
"epoch": 1.91,
"grad_norm": 0.4028488464833583,
"learning_rate": 1.265257255332586e-07,
"loss": 2.0797,
"step": 2155
},
{
"epoch": 1.91,
"grad_norm": 0.4008149442131381,
"learning_rate": 1.1456277207612554e-07,
"loss": 2.1012,
"step": 2160
},
{
"epoch": 1.92,
"grad_norm": 0.3983604404045211,
"learning_rate": 1.0319054459271837e-07,
"loss": 2.1072,
"step": 2165
},
{
"epoch": 1.92,
"grad_norm": 0.3944471181564264,
"learning_rate": 9.240972265541992e-08,
"loss": 2.1013,
"step": 2170
},
{
"epoch": 1.93,
"grad_norm": 0.39386145535861466,
"learning_rate": 8.222095049588264e-08,
"loss": 2.0907,
"step": 2175
},
{
"epoch": 1.93,
"grad_norm": 0.38893023580488223,
"learning_rate": 7.262483696652167e-08,
"loss": 2.1053,
"step": 2180
},
{
"epoch": 1.94,
"grad_norm": 0.38978073750688746,
"learning_rate": 6.362195550413953e-08,
"loss": 2.0855,
"step": 2185
},
{
"epoch": 1.94,
"grad_norm": 0.3849427663072533,
"learning_rate": 5.521284409565675e-08,
"loss": 2.0939,
"step": 2190
},
{
"epoch": 1.94,
"grad_norm": 0.4098717539733243,
"learning_rate": 4.739800524595884e-08,
"loss": 2.1072,
"step": 2195
},
{
"epoch": 1.95,
"grad_norm": 0.3951183361579776,
"learning_rate": 4.017790594787574e-08,
"loss": 2.0963,
"step": 2200
},
{
"epoch": 1.95,
"grad_norm": 0.39882271458674357,
"learning_rate": 3.355297765426868e-08,
"loss": 2.0959,
"step": 2205
},
{
"epoch": 1.96,
"grad_norm": 0.38698652997579813,
"learning_rate": 2.7523616252252972e-08,
"loss": 2.0801,
"step": 2210
},
{
"epoch": 1.96,
"grad_norm": 0.3875845337094545,
"learning_rate": 2.2090182039538055e-08,
"loss": 2.0809,
"step": 2215
},
{
"epoch": 1.97,
"grad_norm": 0.3815217398442563,
"learning_rate": 1.7252999702894736e-08,
"loss": 2.089,
"step": 2220
},
{
"epoch": 1.97,
"grad_norm": 0.38137458714435063,
"learning_rate": 1.3012358298760686e-08,
"loss": 2.0984,
"step": 2225
},
{
"epoch": 1.98,
"grad_norm": 0.41804458050954507,
"learning_rate": 9.368511235958722e-09,
"loss": 2.1075,
"step": 2230
},
{
"epoch": 1.98,
"grad_norm": 0.38803597828465214,
"learning_rate": 6.3216762605589064e-09,
"loss": 2.0877,
"step": 2235
},
{
"epoch": 1.98,
"grad_norm": 0.3794346361216384,
"learning_rate": 3.87203544286563e-09,
"loss": 2.1056,
"step": 2240
},
{
"epoch": 1.99,
"grad_norm": 0.39366177580375067,
"learning_rate": 2.019735166534087e-09,
"loss": 2.0838,
"step": 2245
},
{
"epoch": 1.99,
"grad_norm": 0.4204884785544385,
"learning_rate": 7.648861198306101e-10,
"loss": 2.1125,
"step": 2250
},
{
"epoch": 2.0,
"grad_norm": 0.4047224809246568,
"learning_rate": 1.0756328901018188e-10,
"loss": 2.0929,
"step": 2255
},
{
"epoch": 2.0,
"eval_loss": 2.106668472290039,
"eval_runtime": 22.144,
"eval_samples_per_second": 1444.14,
"eval_steps_per_second": 22.58,
"step": 2258
},
{
"epoch": 2.0,
"step": 2258,
"total_flos": 13637863342080.0,
"train_loss": 2.17739642542796,
"train_runtime": 641.6533,
"train_samples_per_second": 900.574,
"train_steps_per_second": 3.519
}
],
"logging_steps": 5,
"max_steps": 2258,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 1000000,
"total_flos": 13637863342080.0,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}