DouglasBraga's picture
End of training
41d3861 verified
raw
history blame
151 kB
{
"best_metric": 0.77775,
"best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-leukemia.v2.2/checkpoint-4687",
"epoch": 29.951999999999998,
"eval_steps": 500,
"global_step": 9360,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.032,
"grad_norm": 301599.375,
"learning_rate": 1.0683760683760684e-05,
"loss": 0.2217,
"step": 10
},
{
"epoch": 0.064,
"grad_norm": 364430.96875,
"learning_rate": 2.1367521367521368e-05,
"loss": 0.2072,
"step": 20
},
{
"epoch": 0.096,
"grad_norm": 279609.3125,
"learning_rate": 3.205128205128205e-05,
"loss": 0.215,
"step": 30
},
{
"epoch": 0.128,
"grad_norm": 206238.953125,
"learning_rate": 4.2735042735042735e-05,
"loss": 0.1963,
"step": 40
},
{
"epoch": 0.16,
"grad_norm": 380894.78125,
"learning_rate": 5.341880341880342e-05,
"loss": 0.2141,
"step": 50
},
{
"epoch": 0.192,
"grad_norm": 397403.78125,
"learning_rate": 6.41025641025641e-05,
"loss": 0.2216,
"step": 60
},
{
"epoch": 0.224,
"grad_norm": 208578.28125,
"learning_rate": 7.478632478632479e-05,
"loss": 0.1989,
"step": 70
},
{
"epoch": 0.256,
"grad_norm": 252869.15625,
"learning_rate": 8.547008547008547e-05,
"loss": 0.2031,
"step": 80
},
{
"epoch": 0.288,
"grad_norm": 229751.796875,
"learning_rate": 9.615384615384617e-05,
"loss": 0.1923,
"step": 90
},
{
"epoch": 0.32,
"grad_norm": 236082.515625,
"learning_rate": 0.00010683760683760684,
"loss": 0.1989,
"step": 100
},
{
"epoch": 0.352,
"grad_norm": 146456.484375,
"learning_rate": 0.00011752136752136752,
"loss": 0.2219,
"step": 110
},
{
"epoch": 0.384,
"grad_norm": 298750.875,
"learning_rate": 0.0001282051282051282,
"loss": 0.2128,
"step": 120
},
{
"epoch": 0.416,
"grad_norm": 274198.6875,
"learning_rate": 0.0001388888888888889,
"loss": 0.2079,
"step": 130
},
{
"epoch": 0.448,
"grad_norm": 337207.125,
"learning_rate": 0.00014957264957264957,
"loss": 0.2041,
"step": 140
},
{
"epoch": 0.48,
"grad_norm": 292073.5625,
"learning_rate": 0.00016025641025641028,
"loss": 0.2339,
"step": 150
},
{
"epoch": 0.512,
"grad_norm": 227215.015625,
"learning_rate": 0.00017094017094017094,
"loss": 0.2227,
"step": 160
},
{
"epoch": 0.544,
"grad_norm": 314843.59375,
"learning_rate": 0.00018162393162393162,
"loss": 0.2069,
"step": 170
},
{
"epoch": 0.576,
"grad_norm": 412868.6875,
"learning_rate": 0.00019230769230769233,
"loss": 0.2338,
"step": 180
},
{
"epoch": 0.608,
"grad_norm": 200428.625,
"learning_rate": 0.000202991452991453,
"loss": 0.2106,
"step": 190
},
{
"epoch": 0.64,
"grad_norm": 215299.359375,
"learning_rate": 0.00021367521367521368,
"loss": 0.2043,
"step": 200
},
{
"epoch": 0.672,
"grad_norm": 457425.53125,
"learning_rate": 0.00022435897435897436,
"loss": 0.2194,
"step": 210
},
{
"epoch": 0.704,
"grad_norm": 263560.4375,
"learning_rate": 0.00023504273504273504,
"loss": 0.2015,
"step": 220
},
{
"epoch": 0.736,
"grad_norm": 348618.3125,
"learning_rate": 0.0002457264957264957,
"loss": 0.2029,
"step": 230
},
{
"epoch": 0.768,
"grad_norm": 275649.71875,
"learning_rate": 0.0002564102564102564,
"loss": 0.2159,
"step": 240
},
{
"epoch": 0.8,
"grad_norm": 204672.875,
"learning_rate": 0.0002670940170940171,
"loss": 0.2032,
"step": 250
},
{
"epoch": 0.832,
"grad_norm": 173324.40625,
"learning_rate": 0.0002777777777777778,
"loss": 0.1903,
"step": 260
},
{
"epoch": 0.864,
"grad_norm": 282196.0,
"learning_rate": 0.00028846153846153843,
"loss": 0.2278,
"step": 270
},
{
"epoch": 0.896,
"grad_norm": 327001.0,
"learning_rate": 0.00029914529914529915,
"loss": 0.2043,
"step": 280
},
{
"epoch": 0.928,
"grad_norm": 218572.71875,
"learning_rate": 0.00030982905982905986,
"loss": 0.205,
"step": 290
},
{
"epoch": 0.96,
"grad_norm": 234121.171875,
"learning_rate": 0.00032051282051282057,
"loss": 0.2175,
"step": 300
},
{
"epoch": 0.992,
"grad_norm": 285625.0625,
"learning_rate": 0.00033119658119658117,
"loss": 0.2349,
"step": 310
},
{
"epoch": 0.9984,
"eval_accuracy": 0.76975,
"eval_loss": 0.557490348815918,
"eval_runtime": 58.1095,
"eval_samples_per_second": 68.836,
"eval_steps_per_second": 2.151,
"step": 312
},
{
"epoch": 1.024,
"grad_norm": 275641.71875,
"learning_rate": 0.0003418803418803419,
"loss": 0.2282,
"step": 320
},
{
"epoch": 1.056,
"grad_norm": 360517.9375,
"learning_rate": 0.0003525641025641026,
"loss": 0.2088,
"step": 330
},
{
"epoch": 1.088,
"grad_norm": 243124.90625,
"learning_rate": 0.00036324786324786325,
"loss": 0.2142,
"step": 340
},
{
"epoch": 1.12,
"grad_norm": 348227.40625,
"learning_rate": 0.00037393162393162396,
"loss": 0.2073,
"step": 350
},
{
"epoch": 1.152,
"grad_norm": 269863.5625,
"learning_rate": 0.00038461538461538467,
"loss": 0.2141,
"step": 360
},
{
"epoch": 1.184,
"grad_norm": 244875.484375,
"learning_rate": 0.00039529914529914527,
"loss": 0.2124,
"step": 370
},
{
"epoch": 1.216,
"grad_norm": 233065.015625,
"learning_rate": 0.000405982905982906,
"loss": 0.2315,
"step": 380
},
{
"epoch": 1.248,
"grad_norm": 285089.71875,
"learning_rate": 0.0004166666666666667,
"loss": 0.2135,
"step": 390
},
{
"epoch": 1.28,
"grad_norm": 242519.3125,
"learning_rate": 0.00042735042735042735,
"loss": 0.2131,
"step": 400
},
{
"epoch": 1.312,
"grad_norm": 337405.78125,
"learning_rate": 0.00043803418803418806,
"loss": 0.1909,
"step": 410
},
{
"epoch": 1.3439999999999999,
"grad_norm": 197278.3125,
"learning_rate": 0.0004487179487179487,
"loss": 0.196,
"step": 420
},
{
"epoch": 1.376,
"grad_norm": 295350.5625,
"learning_rate": 0.0004594017094017094,
"loss": 0.215,
"step": 430
},
{
"epoch": 1.408,
"grad_norm": 200897.625,
"learning_rate": 0.0004700854700854701,
"loss": 0.2103,
"step": 440
},
{
"epoch": 1.44,
"grad_norm": 304716.84375,
"learning_rate": 0.0004807692307692308,
"loss": 0.1853,
"step": 450
},
{
"epoch": 1.472,
"grad_norm": 337455.09375,
"learning_rate": 0.0004914529914529914,
"loss": 0.2001,
"step": 460
},
{
"epoch": 1.504,
"grad_norm": 292779.3125,
"learning_rate": 0.0005021367521367521,
"loss": 0.1941,
"step": 470
},
{
"epoch": 1.536,
"grad_norm": 280025.1875,
"learning_rate": 0.0005128205128205128,
"loss": 0.2208,
"step": 480
},
{
"epoch": 1.568,
"grad_norm": 309745.53125,
"learning_rate": 0.0005235042735042735,
"loss": 0.2267,
"step": 490
},
{
"epoch": 1.6,
"grad_norm": 176447.296875,
"learning_rate": 0.0005341880341880342,
"loss": 0.2137,
"step": 500
},
{
"epoch": 1.6320000000000001,
"grad_norm": 258145.578125,
"learning_rate": 0.0005448717948717948,
"loss": 0.2039,
"step": 510
},
{
"epoch": 1.6640000000000001,
"grad_norm": 279034.375,
"learning_rate": 0.0005555555555555556,
"loss": 0.2057,
"step": 520
},
{
"epoch": 1.696,
"grad_norm": 194719.9375,
"learning_rate": 0.0005662393162393163,
"loss": 0.23,
"step": 530
},
{
"epoch": 1.728,
"grad_norm": 232231.15625,
"learning_rate": 0.0005769230769230769,
"loss": 0.2142,
"step": 540
},
{
"epoch": 1.76,
"grad_norm": 227828.96875,
"learning_rate": 0.0005876068376068377,
"loss": 0.2012,
"step": 550
},
{
"epoch": 1.792,
"grad_norm": 260916.8125,
"learning_rate": 0.0005982905982905983,
"loss": 0.2154,
"step": 560
},
{
"epoch": 1.8239999999999998,
"grad_norm": 332572.875,
"learning_rate": 0.0006089743589743589,
"loss": 0.2104,
"step": 570
},
{
"epoch": 1.8559999999999999,
"grad_norm": 274341.21875,
"learning_rate": 0.0006196581196581197,
"loss": 0.1927,
"step": 580
},
{
"epoch": 1.888,
"grad_norm": 221814.578125,
"learning_rate": 0.0006303418803418803,
"loss": 0.2181,
"step": 590
},
{
"epoch": 1.92,
"grad_norm": 266745.46875,
"learning_rate": 0.0006410256410256411,
"loss": 0.1901,
"step": 600
},
{
"epoch": 1.952,
"grad_norm": 245638.390625,
"learning_rate": 0.0006517094017094017,
"loss": 0.2292,
"step": 610
},
{
"epoch": 1.984,
"grad_norm": 305467.09375,
"learning_rate": 0.0006623931623931623,
"loss": 0.2191,
"step": 620
},
{
"epoch": 2.0,
"eval_accuracy": 0.76175,
"eval_loss": 0.5572408437728882,
"eval_runtime": 57.8321,
"eval_samples_per_second": 69.166,
"eval_steps_per_second": 2.161,
"step": 625
},
{
"epoch": 2.016,
"grad_norm": 145681.390625,
"learning_rate": 0.0006730769230769232,
"loss": 0.1823,
"step": 630
},
{
"epoch": 2.048,
"grad_norm": 380796.4375,
"learning_rate": 0.0006837606837606838,
"loss": 0.2308,
"step": 640
},
{
"epoch": 2.08,
"grad_norm": 256831.84375,
"learning_rate": 0.0006944444444444445,
"loss": 0.2223,
"step": 650
},
{
"epoch": 2.112,
"grad_norm": 241387.109375,
"learning_rate": 0.0007051282051282052,
"loss": 0.2134,
"step": 660
},
{
"epoch": 2.144,
"grad_norm": 361426.84375,
"learning_rate": 0.0007158119658119658,
"loss": 0.2055,
"step": 670
},
{
"epoch": 2.176,
"grad_norm": 213368.515625,
"learning_rate": 0.0007264957264957265,
"loss": 0.2068,
"step": 680
},
{
"epoch": 2.208,
"grad_norm": 351293.4375,
"learning_rate": 0.0007371794871794872,
"loss": 0.2278,
"step": 690
},
{
"epoch": 2.24,
"grad_norm": 351138.28125,
"learning_rate": 0.0007478632478632479,
"loss": 0.216,
"step": 700
},
{
"epoch": 2.2720000000000002,
"grad_norm": 281240.5,
"learning_rate": 0.0007585470085470085,
"loss": 0.2064,
"step": 710
},
{
"epoch": 2.304,
"grad_norm": 273138.125,
"learning_rate": 0.0007692307692307693,
"loss": 0.2021,
"step": 720
},
{
"epoch": 2.336,
"grad_norm": 198611.640625,
"learning_rate": 0.0007799145299145299,
"loss": 0.218,
"step": 730
},
{
"epoch": 2.368,
"grad_norm": 334765.21875,
"learning_rate": 0.0007905982905982905,
"loss": 0.1988,
"step": 740
},
{
"epoch": 2.4,
"grad_norm": 321548.25,
"learning_rate": 0.0008012820512820514,
"loss": 0.2446,
"step": 750
},
{
"epoch": 2.432,
"grad_norm": 230110.046875,
"learning_rate": 0.000811965811965812,
"loss": 0.213,
"step": 760
},
{
"epoch": 2.464,
"grad_norm": 216522.890625,
"learning_rate": 0.0008226495726495727,
"loss": 0.2114,
"step": 770
},
{
"epoch": 2.496,
"grad_norm": 290017.125,
"learning_rate": 0.0008333333333333334,
"loss": 0.2253,
"step": 780
},
{
"epoch": 2.528,
"grad_norm": 307569.59375,
"learning_rate": 0.000844017094017094,
"loss": 0.1961,
"step": 790
},
{
"epoch": 2.56,
"grad_norm": 258996.3125,
"learning_rate": 0.0008547008547008547,
"loss": 0.2226,
"step": 800
},
{
"epoch": 2.592,
"grad_norm": 187987.453125,
"learning_rate": 0.0008653846153846154,
"loss": 0.2018,
"step": 810
},
{
"epoch": 2.624,
"grad_norm": 249946.1875,
"learning_rate": 0.0008760683760683761,
"loss": 0.1907,
"step": 820
},
{
"epoch": 2.656,
"grad_norm": 265943.46875,
"learning_rate": 0.0008867521367521367,
"loss": 0.2025,
"step": 830
},
{
"epoch": 2.6879999999999997,
"grad_norm": 232717.453125,
"learning_rate": 0.0008974358974358974,
"loss": 0.2252,
"step": 840
},
{
"epoch": 2.7199999999999998,
"grad_norm": 212033.515625,
"learning_rate": 0.0009081196581196581,
"loss": 0.2126,
"step": 850
},
{
"epoch": 2.752,
"grad_norm": 233050.203125,
"learning_rate": 0.0009188034188034187,
"loss": 0.2281,
"step": 860
},
{
"epoch": 2.784,
"grad_norm": 359475.75,
"learning_rate": 0.0009294871794871796,
"loss": 0.188,
"step": 870
},
{
"epoch": 2.816,
"grad_norm": 343301.15625,
"learning_rate": 0.0009401709401709402,
"loss": 0.2316,
"step": 880
},
{
"epoch": 2.848,
"grad_norm": 316784.375,
"learning_rate": 0.0009508547008547009,
"loss": 0.2014,
"step": 890
},
{
"epoch": 2.88,
"grad_norm": 375741.0,
"learning_rate": 0.0009615384615384616,
"loss": 0.2196,
"step": 900
},
{
"epoch": 2.912,
"grad_norm": 320915.21875,
"learning_rate": 0.0009722222222222222,
"loss": 0.2092,
"step": 910
},
{
"epoch": 2.944,
"grad_norm": 254477.09375,
"learning_rate": 0.0009829059829059828,
"loss": 0.2147,
"step": 920
},
{
"epoch": 2.976,
"grad_norm": 256587.71875,
"learning_rate": 0.0009935897435897436,
"loss": 0.2124,
"step": 930
},
{
"epoch": 2.9984,
"eval_accuracy": 0.769,
"eval_loss": 0.5579845309257507,
"eval_runtime": 58.375,
"eval_samples_per_second": 68.523,
"eval_steps_per_second": 2.141,
"step": 937
},
{
"epoch": 3.008,
"grad_norm": 284885.21875,
"learning_rate": 0.0009995251661918329,
"loss": 0.2,
"step": 940
},
{
"epoch": 3.04,
"grad_norm": 280128.65625,
"learning_rate": 0.000998338081671415,
"loss": 0.2062,
"step": 950
},
{
"epoch": 3.072,
"grad_norm": 327251.5,
"learning_rate": 0.0009971509971509972,
"loss": 0.2285,
"step": 960
},
{
"epoch": 3.104,
"grad_norm": 246022.15625,
"learning_rate": 0.0009959639126305793,
"loss": 0.2138,
"step": 970
},
{
"epoch": 3.136,
"grad_norm": 149789.59375,
"learning_rate": 0.0009947768281101616,
"loss": 0.2108,
"step": 980
},
{
"epoch": 3.168,
"grad_norm": 171794.5625,
"learning_rate": 0.0009935897435897436,
"loss": 0.2167,
"step": 990
},
{
"epoch": 3.2,
"grad_norm": 227827.90625,
"learning_rate": 0.000992402659069326,
"loss": 0.2043,
"step": 1000
},
{
"epoch": 3.232,
"grad_norm": 196560.046875,
"learning_rate": 0.000991215574548908,
"loss": 0.2247,
"step": 1010
},
{
"epoch": 3.2640000000000002,
"grad_norm": 198754.34375,
"learning_rate": 0.00099002849002849,
"loss": 0.2033,
"step": 1020
},
{
"epoch": 3.296,
"grad_norm": 277303.875,
"learning_rate": 0.0009888414055080723,
"loss": 0.209,
"step": 1030
},
{
"epoch": 3.328,
"grad_norm": 291565.90625,
"learning_rate": 0.0009876543209876543,
"loss": 0.2248,
"step": 1040
},
{
"epoch": 3.36,
"grad_norm": 231893.078125,
"learning_rate": 0.0009864672364672364,
"loss": 0.1879,
"step": 1050
},
{
"epoch": 3.392,
"grad_norm": 160630.546875,
"learning_rate": 0.0009852801519468187,
"loss": 0.2316,
"step": 1060
},
{
"epoch": 3.424,
"grad_norm": 274359.40625,
"learning_rate": 0.0009840930674264007,
"loss": 0.2191,
"step": 1070
},
{
"epoch": 3.456,
"grad_norm": 287355.15625,
"learning_rate": 0.0009829059829059828,
"loss": 0.1981,
"step": 1080
},
{
"epoch": 3.488,
"grad_norm": 286551.75,
"learning_rate": 0.000981718898385565,
"loss": 0.2253,
"step": 1090
},
{
"epoch": 3.52,
"grad_norm": 293656.21875,
"learning_rate": 0.0009805318138651471,
"loss": 0.2113,
"step": 1100
},
{
"epoch": 3.552,
"grad_norm": 226414.3125,
"learning_rate": 0.0009793447293447294,
"loss": 0.2366,
"step": 1110
},
{
"epoch": 3.584,
"grad_norm": 304422.28125,
"learning_rate": 0.0009781576448243115,
"loss": 0.2296,
"step": 1120
},
{
"epoch": 3.616,
"grad_norm": 200736.75,
"learning_rate": 0.0009769705603038937,
"loss": 0.2085,
"step": 1130
},
{
"epoch": 3.648,
"grad_norm": 191285.28125,
"learning_rate": 0.0009757834757834758,
"loss": 0.2168,
"step": 1140
},
{
"epoch": 3.68,
"grad_norm": 265826.28125,
"learning_rate": 0.000974596391263058,
"loss": 0.2242,
"step": 1150
},
{
"epoch": 3.7119999999999997,
"grad_norm": 170613.078125,
"learning_rate": 0.00097340930674264,
"loss": 0.2015,
"step": 1160
},
{
"epoch": 3.7439999999999998,
"grad_norm": 272692.40625,
"learning_rate": 0.0009722222222222222,
"loss": 0.2162,
"step": 1170
},
{
"epoch": 3.776,
"grad_norm": 245685.46875,
"learning_rate": 0.0009710351377018044,
"loss": 0.2166,
"step": 1180
},
{
"epoch": 3.808,
"grad_norm": 359136.125,
"learning_rate": 0.0009698480531813865,
"loss": 0.2023,
"step": 1190
},
{
"epoch": 3.84,
"grad_norm": 273622.90625,
"learning_rate": 0.0009686609686609687,
"loss": 0.1966,
"step": 1200
},
{
"epoch": 3.872,
"grad_norm": 291427.96875,
"learning_rate": 0.0009674738841405509,
"loss": 0.216,
"step": 1210
},
{
"epoch": 3.904,
"grad_norm": 328024.90625,
"learning_rate": 0.000966286799620133,
"loss": 0.1891,
"step": 1220
},
{
"epoch": 3.936,
"grad_norm": 263930.625,
"learning_rate": 0.0009650997150997152,
"loss": 0.2155,
"step": 1230
},
{
"epoch": 3.968,
"grad_norm": 214059.265625,
"learning_rate": 0.0009639126305792973,
"loss": 0.232,
"step": 1240
},
{
"epoch": 4.0,
"grad_norm": 204246.984375,
"learning_rate": 0.0009627255460588794,
"loss": 0.2207,
"step": 1250
},
{
"epoch": 4.0,
"eval_accuracy": 0.763,
"eval_loss": 0.5500344038009644,
"eval_runtime": 57.7064,
"eval_samples_per_second": 69.316,
"eval_steps_per_second": 2.166,
"step": 1250
},
{
"epoch": 4.032,
"grad_norm": 295677.8125,
"learning_rate": 0.0009615384615384616,
"loss": 0.2247,
"step": 1260
},
{
"epoch": 4.064,
"grad_norm": 197746.8125,
"learning_rate": 0.0009603513770180438,
"loss": 0.2203,
"step": 1270
},
{
"epoch": 4.096,
"grad_norm": 287209.65625,
"learning_rate": 0.0009591642924976258,
"loss": 0.2201,
"step": 1280
},
{
"epoch": 4.128,
"grad_norm": 279235.9375,
"learning_rate": 0.000957977207977208,
"loss": 0.2344,
"step": 1290
},
{
"epoch": 4.16,
"grad_norm": 293263.59375,
"learning_rate": 0.0009567901234567902,
"loss": 0.2069,
"step": 1300
},
{
"epoch": 4.192,
"grad_norm": 212506.359375,
"learning_rate": 0.0009556030389363722,
"loss": 0.1787,
"step": 1310
},
{
"epoch": 4.224,
"grad_norm": 274108.71875,
"learning_rate": 0.0009544159544159544,
"loss": 0.2311,
"step": 1320
},
{
"epoch": 4.256,
"grad_norm": 319637.59375,
"learning_rate": 0.0009532288698955365,
"loss": 0.2136,
"step": 1330
},
{
"epoch": 4.288,
"grad_norm": 229458.78125,
"learning_rate": 0.0009520417853751187,
"loss": 0.2019,
"step": 1340
},
{
"epoch": 4.32,
"grad_norm": 307182.09375,
"learning_rate": 0.0009508547008547009,
"loss": 0.2162,
"step": 1350
},
{
"epoch": 4.352,
"grad_norm": 288039.09375,
"learning_rate": 0.000949667616334283,
"loss": 0.1945,
"step": 1360
},
{
"epoch": 4.384,
"grad_norm": 253276.65625,
"learning_rate": 0.0009484805318138652,
"loss": 0.2089,
"step": 1370
},
{
"epoch": 4.416,
"grad_norm": 250874.515625,
"learning_rate": 0.0009472934472934474,
"loss": 0.2135,
"step": 1380
},
{
"epoch": 4.448,
"grad_norm": 173051.5625,
"learning_rate": 0.0009461063627730294,
"loss": 0.2005,
"step": 1390
},
{
"epoch": 4.48,
"grad_norm": 198301.75,
"learning_rate": 0.0009449192782526116,
"loss": 0.2227,
"step": 1400
},
{
"epoch": 4.5120000000000005,
"grad_norm": 252200.671875,
"learning_rate": 0.0009437321937321938,
"loss": 0.2233,
"step": 1410
},
{
"epoch": 4.5440000000000005,
"grad_norm": 243128.0,
"learning_rate": 0.0009425451092117759,
"loss": 0.2291,
"step": 1420
},
{
"epoch": 4.576,
"grad_norm": 353883.34375,
"learning_rate": 0.000941358024691358,
"loss": 0.2207,
"step": 1430
},
{
"epoch": 4.608,
"grad_norm": 223227.578125,
"learning_rate": 0.0009401709401709402,
"loss": 0.2076,
"step": 1440
},
{
"epoch": 4.64,
"grad_norm": 283902.90625,
"learning_rate": 0.0009389838556505223,
"loss": 0.1916,
"step": 1450
},
{
"epoch": 4.672,
"grad_norm": 301072.40625,
"learning_rate": 0.0009377967711301045,
"loss": 0.1951,
"step": 1460
},
{
"epoch": 4.704,
"grad_norm": 300642.03125,
"learning_rate": 0.0009366096866096866,
"loss": 0.2129,
"step": 1470
},
{
"epoch": 4.736,
"grad_norm": 254449.90625,
"learning_rate": 0.0009354226020892687,
"loss": 0.2043,
"step": 1480
},
{
"epoch": 4.768,
"grad_norm": 278776.9375,
"learning_rate": 0.0009342355175688509,
"loss": 0.2139,
"step": 1490
},
{
"epoch": 4.8,
"grad_norm": 169466.421875,
"learning_rate": 0.0009330484330484332,
"loss": 0.1944,
"step": 1500
},
{
"epoch": 4.832,
"grad_norm": 378268.8125,
"learning_rate": 0.0009318613485280152,
"loss": 0.1903,
"step": 1510
},
{
"epoch": 4.864,
"grad_norm": 232065.859375,
"learning_rate": 0.0009306742640075974,
"loss": 0.2188,
"step": 1520
},
{
"epoch": 4.896,
"grad_norm": 222731.53125,
"learning_rate": 0.0009294871794871796,
"loss": 0.2239,
"step": 1530
},
{
"epoch": 4.928,
"grad_norm": 250957.84375,
"learning_rate": 0.0009283000949667616,
"loss": 0.2051,
"step": 1540
},
{
"epoch": 4.96,
"grad_norm": 265056.0,
"learning_rate": 0.0009271130104463438,
"loss": 0.1948,
"step": 1550
},
{
"epoch": 4.992,
"grad_norm": 263161.90625,
"learning_rate": 0.000925925925925926,
"loss": 0.2143,
"step": 1560
},
{
"epoch": 4.9984,
"eval_accuracy": 0.76525,
"eval_loss": 0.5575093030929565,
"eval_runtime": 58.0778,
"eval_samples_per_second": 68.873,
"eval_steps_per_second": 2.152,
"step": 1562
},
{
"epoch": 5.024,
"grad_norm": 242776.59375,
"learning_rate": 0.0009247388414055081,
"loss": 0.2211,
"step": 1570
},
{
"epoch": 5.056,
"grad_norm": 233410.25,
"learning_rate": 0.0009235517568850902,
"loss": 0.2181,
"step": 1580
},
{
"epoch": 5.088,
"grad_norm": 293507.25,
"learning_rate": 0.0009223646723646724,
"loss": 0.1975,
"step": 1590
},
{
"epoch": 5.12,
"grad_norm": 228211.375,
"learning_rate": 0.0009211775878442545,
"loss": 0.2207,
"step": 1600
},
{
"epoch": 5.152,
"grad_norm": 219766.359375,
"learning_rate": 0.0009199905033238367,
"loss": 0.185,
"step": 1610
},
{
"epoch": 5.184,
"grad_norm": 243226.15625,
"learning_rate": 0.0009188034188034187,
"loss": 0.2023,
"step": 1620
},
{
"epoch": 5.216,
"grad_norm": 227722.0625,
"learning_rate": 0.0009176163342830009,
"loss": 0.2233,
"step": 1630
},
{
"epoch": 5.248,
"grad_norm": 182398.234375,
"learning_rate": 0.0009164292497625831,
"loss": 0.2204,
"step": 1640
},
{
"epoch": 5.28,
"grad_norm": 227450.109375,
"learning_rate": 0.0009152421652421653,
"loss": 0.2206,
"step": 1650
},
{
"epoch": 5.312,
"grad_norm": 416922.8125,
"learning_rate": 0.0009140550807217474,
"loss": 0.2024,
"step": 1660
},
{
"epoch": 5.344,
"grad_norm": 174068.0,
"learning_rate": 0.0009128679962013296,
"loss": 0.2166,
"step": 1670
},
{
"epoch": 5.376,
"grad_norm": 295677.71875,
"learning_rate": 0.0009116809116809118,
"loss": 0.2201,
"step": 1680
},
{
"epoch": 5.408,
"grad_norm": 219667.796875,
"learning_rate": 0.0009104938271604939,
"loss": 0.1898,
"step": 1690
},
{
"epoch": 5.44,
"grad_norm": 196588.1875,
"learning_rate": 0.000909306742640076,
"loss": 0.1963,
"step": 1700
},
{
"epoch": 5.4719999999999995,
"grad_norm": 198457.703125,
"learning_rate": 0.0009081196581196581,
"loss": 0.1988,
"step": 1710
},
{
"epoch": 5.504,
"grad_norm": 177612.46875,
"learning_rate": 0.0009069325735992403,
"loss": 0.1918,
"step": 1720
},
{
"epoch": 5.536,
"grad_norm": 231351.9375,
"learning_rate": 0.0009057454890788225,
"loss": 0.2098,
"step": 1730
},
{
"epoch": 5.568,
"grad_norm": 281818.4375,
"learning_rate": 0.0009045584045584045,
"loss": 0.205,
"step": 1740
},
{
"epoch": 5.6,
"grad_norm": 249959.890625,
"learning_rate": 0.0009033713200379867,
"loss": 0.1928,
"step": 1750
},
{
"epoch": 5.632,
"grad_norm": 192714.796875,
"learning_rate": 0.0009021842355175689,
"loss": 0.2063,
"step": 1760
},
{
"epoch": 5.664,
"grad_norm": 244175.8125,
"learning_rate": 0.0009009971509971509,
"loss": 0.2161,
"step": 1770
},
{
"epoch": 5.696,
"grad_norm": 238826.9375,
"learning_rate": 0.0008998100664767331,
"loss": 0.2042,
"step": 1780
},
{
"epoch": 5.728,
"grad_norm": 174162.5625,
"learning_rate": 0.0008986229819563153,
"loss": 0.205,
"step": 1790
},
{
"epoch": 5.76,
"grad_norm": 359398.59375,
"learning_rate": 0.0008974358974358974,
"loss": 0.2137,
"step": 1800
},
{
"epoch": 5.792,
"grad_norm": 324436.6875,
"learning_rate": 0.0008962488129154796,
"loss": 0.2223,
"step": 1810
},
{
"epoch": 5.824,
"grad_norm": 274255.4375,
"learning_rate": 0.0008950617283950618,
"loss": 0.2043,
"step": 1820
},
{
"epoch": 5.856,
"grad_norm": 212538.953125,
"learning_rate": 0.0008938746438746439,
"loss": 0.2054,
"step": 1830
},
{
"epoch": 5.888,
"grad_norm": 170735.9375,
"learning_rate": 0.0008926875593542261,
"loss": 0.1999,
"step": 1840
},
{
"epoch": 5.92,
"grad_norm": 241050.0625,
"learning_rate": 0.0008915004748338082,
"loss": 0.2154,
"step": 1850
},
{
"epoch": 5.952,
"grad_norm": 213356.40625,
"learning_rate": 0.0008903133903133903,
"loss": 0.2068,
"step": 1860
},
{
"epoch": 5.984,
"grad_norm": 252578.8125,
"learning_rate": 0.0008891263057929725,
"loss": 0.2191,
"step": 1870
},
{
"epoch": 6.0,
"eval_accuracy": 0.77275,
"eval_loss": 0.5486457347869873,
"eval_runtime": 58.401,
"eval_samples_per_second": 68.492,
"eval_steps_per_second": 2.14,
"step": 1875
},
{
"epoch": 6.016,
"grad_norm": 195587.765625,
"learning_rate": 0.0008879392212725547,
"loss": 0.2163,
"step": 1880
},
{
"epoch": 6.048,
"grad_norm": 298880.3125,
"learning_rate": 0.0008867521367521367,
"loss": 0.2318,
"step": 1890
},
{
"epoch": 6.08,
"grad_norm": 254082.453125,
"learning_rate": 0.0008855650522317189,
"loss": 0.2296,
"step": 1900
},
{
"epoch": 6.112,
"grad_norm": 209368.6875,
"learning_rate": 0.0008843779677113011,
"loss": 0.2059,
"step": 1910
},
{
"epoch": 6.144,
"grad_norm": 296605.71875,
"learning_rate": 0.0008831908831908832,
"loss": 0.1965,
"step": 1920
},
{
"epoch": 6.176,
"grad_norm": 232373.09375,
"learning_rate": 0.0008820037986704653,
"loss": 0.2062,
"step": 1930
},
{
"epoch": 6.208,
"grad_norm": 212144.140625,
"learning_rate": 0.0008808167141500475,
"loss": 0.2242,
"step": 1940
},
{
"epoch": 6.24,
"grad_norm": 223206.578125,
"learning_rate": 0.0008796296296296296,
"loss": 0.2109,
"step": 1950
},
{
"epoch": 6.272,
"grad_norm": 237154.78125,
"learning_rate": 0.0008784425451092119,
"loss": 0.2167,
"step": 1960
},
{
"epoch": 6.304,
"grad_norm": 185542.953125,
"learning_rate": 0.000877255460588794,
"loss": 0.2052,
"step": 1970
},
{
"epoch": 6.336,
"grad_norm": 203563.640625,
"learning_rate": 0.0008760683760683761,
"loss": 0.2096,
"step": 1980
},
{
"epoch": 6.368,
"grad_norm": 285480.0625,
"learning_rate": 0.0008748812915479583,
"loss": 0.2037,
"step": 1990
},
{
"epoch": 6.4,
"grad_norm": 303778.9375,
"learning_rate": 0.0008736942070275405,
"loss": 0.1981,
"step": 2000
},
{
"epoch": 6.432,
"grad_norm": 462646.96875,
"learning_rate": 0.0008725071225071225,
"loss": 0.2092,
"step": 2010
},
{
"epoch": 6.464,
"grad_norm": 461746.4375,
"learning_rate": 0.0008713200379867047,
"loss": 0.2294,
"step": 2020
},
{
"epoch": 6.496,
"grad_norm": 533448.4375,
"learning_rate": 0.0008701329534662869,
"loss": 0.2055,
"step": 2030
},
{
"epoch": 6.5280000000000005,
"grad_norm": 548531.5625,
"learning_rate": 0.0008689458689458689,
"loss": 0.2184,
"step": 2040
},
{
"epoch": 6.5600000000000005,
"grad_norm": 495748.65625,
"learning_rate": 0.0008677587844254511,
"loss": 0.2184,
"step": 2050
},
{
"epoch": 6.592,
"grad_norm": 543121.0625,
"learning_rate": 0.0008665716999050332,
"loss": 0.1997,
"step": 2060
},
{
"epoch": 6.624,
"grad_norm": 461117.59375,
"learning_rate": 0.0008653846153846154,
"loss": 0.2019,
"step": 2070
},
{
"epoch": 6.656,
"grad_norm": 765304.125,
"learning_rate": 0.0008641975308641975,
"loss": 0.1989,
"step": 2080
},
{
"epoch": 6.688,
"grad_norm": 499025.09375,
"learning_rate": 0.0008630104463437796,
"loss": 0.2139,
"step": 2090
},
{
"epoch": 6.72,
"grad_norm": 443804.0625,
"learning_rate": 0.0008618233618233618,
"loss": 0.2212,
"step": 2100
},
{
"epoch": 6.752,
"grad_norm": 513447.84375,
"learning_rate": 0.0008606362773029441,
"loss": 0.2014,
"step": 2110
},
{
"epoch": 6.784,
"grad_norm": 417790.125,
"learning_rate": 0.0008594491927825261,
"loss": 0.2092,
"step": 2120
},
{
"epoch": 6.816,
"grad_norm": 580280.5625,
"learning_rate": 0.0008582621082621083,
"loss": 0.2185,
"step": 2130
},
{
"epoch": 6.848,
"grad_norm": 553312.1875,
"learning_rate": 0.0008570750237416905,
"loss": 0.2094,
"step": 2140
},
{
"epoch": 6.88,
"grad_norm": 498577.25,
"learning_rate": 0.0008558879392212726,
"loss": 0.214,
"step": 2150
},
{
"epoch": 6.912,
"grad_norm": 698254.125,
"learning_rate": 0.0008547008547008547,
"loss": 0.2407,
"step": 2160
},
{
"epoch": 6.944,
"grad_norm": 423589.03125,
"learning_rate": 0.0008535137701804369,
"loss": 0.2203,
"step": 2170
},
{
"epoch": 6.976,
"grad_norm": 755190.0625,
"learning_rate": 0.000852326685660019,
"loss": 0.2063,
"step": 2180
},
{
"epoch": 6.9984,
"eval_accuracy": 0.7615,
"eval_loss": 0.5593643188476562,
"eval_runtime": 58.3849,
"eval_samples_per_second": 68.511,
"eval_steps_per_second": 2.141,
"step": 2187
},
{
"epoch": 7.008,
"grad_norm": 672768.3125,
"learning_rate": 0.0008511396011396012,
"loss": 0.2376,
"step": 2190
},
{
"epoch": 7.04,
"grad_norm": 590710.6875,
"learning_rate": 0.0008499525166191833,
"loss": 0.2167,
"step": 2200
},
{
"epoch": 7.072,
"grad_norm": 451509.625,
"learning_rate": 0.0008487654320987654,
"loss": 0.1995,
"step": 2210
},
{
"epoch": 7.104,
"grad_norm": 438460.8125,
"learning_rate": 0.0008475783475783476,
"loss": 0.2262,
"step": 2220
},
{
"epoch": 7.136,
"grad_norm": 538988.8125,
"learning_rate": 0.0008463912630579298,
"loss": 0.2178,
"step": 2230
},
{
"epoch": 7.168,
"grad_norm": 525871.25,
"learning_rate": 0.0008452041785375118,
"loss": 0.1961,
"step": 2240
},
{
"epoch": 7.2,
"grad_norm": 420527.3125,
"learning_rate": 0.000844017094017094,
"loss": 0.2175,
"step": 2250
},
{
"epoch": 7.232,
"grad_norm": 813415.3125,
"learning_rate": 0.0008428300094966762,
"loss": 0.2185,
"step": 2260
},
{
"epoch": 7.264,
"grad_norm": 471922.9375,
"learning_rate": 0.0008416429249762583,
"loss": 0.2074,
"step": 2270
},
{
"epoch": 7.296,
"grad_norm": 451017.125,
"learning_rate": 0.0008404558404558405,
"loss": 0.1841,
"step": 2280
},
{
"epoch": 7.328,
"grad_norm": 506415.46875,
"learning_rate": 0.0008392687559354227,
"loss": 0.1943,
"step": 2290
},
{
"epoch": 7.36,
"grad_norm": 444371.15625,
"learning_rate": 0.0008380816714150048,
"loss": 0.217,
"step": 2300
},
{
"epoch": 7.392,
"grad_norm": 377225.34375,
"learning_rate": 0.0008368945868945869,
"loss": 0.1975,
"step": 2310
},
{
"epoch": 7.424,
"grad_norm": 594714.4375,
"learning_rate": 0.000835707502374169,
"loss": 0.1969,
"step": 2320
},
{
"epoch": 7.456,
"grad_norm": 500928.84375,
"learning_rate": 0.0008345204178537512,
"loss": 0.2038,
"step": 2330
},
{
"epoch": 7.4879999999999995,
"grad_norm": 460287.4375,
"learning_rate": 0.0008333333333333334,
"loss": 0.2156,
"step": 2340
},
{
"epoch": 7.52,
"grad_norm": 448026.53125,
"learning_rate": 0.0008321462488129154,
"loss": 0.207,
"step": 2350
},
{
"epoch": 7.552,
"grad_norm": 556125.625,
"learning_rate": 0.0008309591642924976,
"loss": 0.232,
"step": 2360
},
{
"epoch": 7.584,
"grad_norm": 332308.3125,
"learning_rate": 0.0008297720797720798,
"loss": 0.2147,
"step": 2370
},
{
"epoch": 7.616,
"grad_norm": 400010.5625,
"learning_rate": 0.000828584995251662,
"loss": 0.2073,
"step": 2380
},
{
"epoch": 7.648,
"grad_norm": 467695.78125,
"learning_rate": 0.000827397910731244,
"loss": 0.2018,
"step": 2390
},
{
"epoch": 7.68,
"grad_norm": 446946.625,
"learning_rate": 0.0008262108262108262,
"loss": 0.2226,
"step": 2400
},
{
"epoch": 7.712,
"grad_norm": 557084.5,
"learning_rate": 0.0008250237416904083,
"loss": 0.2278,
"step": 2410
},
{
"epoch": 7.744,
"grad_norm": 429602.84375,
"learning_rate": 0.0008238366571699906,
"loss": 0.1832,
"step": 2420
},
{
"epoch": 7.776,
"grad_norm": 388821.59375,
"learning_rate": 0.0008226495726495727,
"loss": 0.2202,
"step": 2430
},
{
"epoch": 7.808,
"grad_norm": 681185.4375,
"learning_rate": 0.0008214624881291548,
"loss": 0.2078,
"step": 2440
},
{
"epoch": 7.84,
"grad_norm": 763735.875,
"learning_rate": 0.000820275403608737,
"loss": 0.2242,
"step": 2450
},
{
"epoch": 7.872,
"grad_norm": 565001.625,
"learning_rate": 0.0008190883190883192,
"loss": 0.2186,
"step": 2460
},
{
"epoch": 7.904,
"grad_norm": 458239.75,
"learning_rate": 0.0008179012345679012,
"loss": 0.1843,
"step": 2470
},
{
"epoch": 7.936,
"grad_norm": 609510.125,
"learning_rate": 0.0008167141500474834,
"loss": 0.1953,
"step": 2480
},
{
"epoch": 7.968,
"grad_norm": 578130.5625,
"learning_rate": 0.0008155270655270656,
"loss": 0.2354,
"step": 2490
},
{
"epoch": 8.0,
"grad_norm": 743331.375,
"learning_rate": 0.0008143399810066477,
"loss": 0.207,
"step": 2500
},
{
"epoch": 8.0,
"eval_accuracy": 0.7695,
"eval_loss": 0.540494441986084,
"eval_runtime": 58.3615,
"eval_samples_per_second": 68.538,
"eval_steps_per_second": 2.142,
"step": 2500
},
{
"epoch": 8.032,
"grad_norm": 439252.84375,
"learning_rate": 0.0008131528964862298,
"loss": 0.1953,
"step": 2510
},
{
"epoch": 8.064,
"grad_norm": 574105.1875,
"learning_rate": 0.000811965811965812,
"loss": 0.2288,
"step": 2520
},
{
"epoch": 8.096,
"grad_norm": 392970.59375,
"learning_rate": 0.0008107787274453941,
"loss": 0.1998,
"step": 2530
},
{
"epoch": 8.128,
"grad_norm": 415349.78125,
"learning_rate": 0.0008095916429249762,
"loss": 0.1898,
"step": 2540
},
{
"epoch": 8.16,
"grad_norm": 402935.21875,
"learning_rate": 0.0008084045584045584,
"loss": 0.209,
"step": 2550
},
{
"epoch": 8.192,
"grad_norm": 710684.375,
"learning_rate": 0.0008072174738841405,
"loss": 0.2232,
"step": 2560
},
{
"epoch": 8.224,
"grad_norm": 522874.5625,
"learning_rate": 0.0008060303893637228,
"loss": 0.2241,
"step": 2570
},
{
"epoch": 8.256,
"grad_norm": 469906.8125,
"learning_rate": 0.0008048433048433049,
"loss": 0.2161,
"step": 2580
},
{
"epoch": 8.288,
"grad_norm": 496419.34375,
"learning_rate": 0.000803656220322887,
"loss": 0.2131,
"step": 2590
},
{
"epoch": 8.32,
"grad_norm": 641309.9375,
"learning_rate": 0.0008024691358024692,
"loss": 0.2189,
"step": 2600
},
{
"epoch": 8.352,
"grad_norm": 543925.3125,
"learning_rate": 0.0008012820512820514,
"loss": 0.2152,
"step": 2610
},
{
"epoch": 8.384,
"grad_norm": 412617.3125,
"learning_rate": 0.0008000949667616334,
"loss": 0.1995,
"step": 2620
},
{
"epoch": 8.416,
"grad_norm": 595042.1875,
"learning_rate": 0.0007989078822412156,
"loss": 0.2212,
"step": 2630
},
{
"epoch": 8.448,
"grad_norm": 439380.8125,
"learning_rate": 0.0007977207977207978,
"loss": 0.2038,
"step": 2640
},
{
"epoch": 8.48,
"grad_norm": 476969.78125,
"learning_rate": 0.0007965337132003799,
"loss": 0.2097,
"step": 2650
},
{
"epoch": 8.512,
"grad_norm": 371208.65625,
"learning_rate": 0.000795346628679962,
"loss": 0.2195,
"step": 2660
},
{
"epoch": 8.544,
"grad_norm": 363928.875,
"learning_rate": 0.0007941595441595441,
"loss": 0.2062,
"step": 2670
},
{
"epoch": 8.576,
"grad_norm": 482233.5625,
"learning_rate": 0.0007929724596391263,
"loss": 0.2212,
"step": 2680
},
{
"epoch": 8.608,
"grad_norm": 675145.5625,
"learning_rate": 0.0007917853751187085,
"loss": 0.2247,
"step": 2690
},
{
"epoch": 8.64,
"grad_norm": 385214.0625,
"learning_rate": 0.0007905982905982905,
"loss": 0.2006,
"step": 2700
},
{
"epoch": 8.672,
"grad_norm": 552452.0625,
"learning_rate": 0.0007894112060778727,
"loss": 0.2046,
"step": 2710
},
{
"epoch": 8.704,
"grad_norm": 638284.4375,
"learning_rate": 0.0007882241215574549,
"loss": 0.2222,
"step": 2720
},
{
"epoch": 8.736,
"grad_norm": 314355.84375,
"learning_rate": 0.0007870370370370372,
"loss": 0.2059,
"step": 2730
},
{
"epoch": 8.768,
"grad_norm": 499934.5625,
"learning_rate": 0.0007858499525166192,
"loss": 0.2003,
"step": 2740
},
{
"epoch": 8.8,
"grad_norm": 514626.78125,
"learning_rate": 0.0007846628679962014,
"loss": 0.2202,
"step": 2750
},
{
"epoch": 8.832,
"grad_norm": 743981.875,
"learning_rate": 0.0007834757834757835,
"loss": 0.223,
"step": 2760
},
{
"epoch": 8.864,
"grad_norm": 372136.40625,
"learning_rate": 0.0007822886989553656,
"loss": 0.205,
"step": 2770
},
{
"epoch": 8.896,
"grad_norm": 524279.8125,
"learning_rate": 0.0007811016144349478,
"loss": 0.2241,
"step": 2780
},
{
"epoch": 8.928,
"grad_norm": 526485.4375,
"learning_rate": 0.0007799145299145299,
"loss": 0.2046,
"step": 2790
},
{
"epoch": 8.96,
"grad_norm": 388172.5625,
"learning_rate": 0.0007787274453941121,
"loss": 0.2174,
"step": 2800
},
{
"epoch": 8.992,
"grad_norm": 657450.8125,
"learning_rate": 0.0007775403608736942,
"loss": 0.2273,
"step": 2810
},
{
"epoch": 8.9984,
"eval_accuracy": 0.76725,
"eval_loss": 0.5568080544471741,
"eval_runtime": 58.1506,
"eval_samples_per_second": 68.787,
"eval_steps_per_second": 2.15,
"step": 2812
},
{
"epoch": 9.024,
"grad_norm": 381848.6875,
"learning_rate": 0.0007763532763532763,
"loss": 0.1682,
"step": 2820
},
{
"epoch": 9.056,
"grad_norm": 488400.375,
"learning_rate": 0.0007751661918328585,
"loss": 0.2072,
"step": 2830
},
{
"epoch": 9.088,
"grad_norm": 557510.875,
"learning_rate": 0.0007739791073124407,
"loss": 0.2127,
"step": 2840
},
{
"epoch": 9.12,
"grad_norm": 689299.9375,
"learning_rate": 0.0007727920227920227,
"loss": 0.2191,
"step": 2850
},
{
"epoch": 9.152,
"grad_norm": 543505.4375,
"learning_rate": 0.0007716049382716049,
"loss": 0.2011,
"step": 2860
},
{
"epoch": 9.184,
"grad_norm": 514595.4375,
"learning_rate": 0.0007704178537511871,
"loss": 0.1982,
"step": 2870
},
{
"epoch": 9.216,
"grad_norm": 489803.34375,
"learning_rate": 0.0007692307692307693,
"loss": 0.2097,
"step": 2880
},
{
"epoch": 9.248,
"grad_norm": 415434.8125,
"learning_rate": 0.0007680436847103514,
"loss": 0.1999,
"step": 2890
},
{
"epoch": 9.28,
"grad_norm": 405549.75,
"learning_rate": 0.0007668566001899336,
"loss": 0.227,
"step": 2900
},
{
"epoch": 9.312,
"grad_norm": 474433.0625,
"learning_rate": 0.0007656695156695157,
"loss": 0.2027,
"step": 2910
},
{
"epoch": 9.344,
"grad_norm": 332759.625,
"learning_rate": 0.0007644824311490979,
"loss": 0.208,
"step": 2920
},
{
"epoch": 9.376,
"grad_norm": 523231.1875,
"learning_rate": 0.00076329534662868,
"loss": 0.2236,
"step": 2930
},
{
"epoch": 9.408,
"grad_norm": 756157.0625,
"learning_rate": 0.0007621082621082621,
"loss": 0.2244,
"step": 2940
},
{
"epoch": 9.44,
"grad_norm": 666299.875,
"learning_rate": 0.0007609211775878443,
"loss": 0.2434,
"step": 2950
},
{
"epoch": 9.472,
"grad_norm": 387654.9375,
"learning_rate": 0.0007597340930674265,
"loss": 0.1967,
"step": 2960
},
{
"epoch": 9.504,
"grad_norm": 641448.8125,
"learning_rate": 0.0007585470085470085,
"loss": 0.2211,
"step": 2970
},
{
"epoch": 9.536,
"grad_norm": 626212.25,
"learning_rate": 0.0007573599240265907,
"loss": 0.2041,
"step": 2980
},
{
"epoch": 9.568,
"grad_norm": 473671.46875,
"learning_rate": 0.0007561728395061729,
"loss": 0.2098,
"step": 2990
},
{
"epoch": 9.6,
"grad_norm": 595403.9375,
"learning_rate": 0.0007549857549857549,
"loss": 0.2197,
"step": 3000
},
{
"epoch": 9.632,
"grad_norm": 495355.15625,
"learning_rate": 0.0007537986704653371,
"loss": 0.2275,
"step": 3010
},
{
"epoch": 9.664,
"grad_norm": 433815.03125,
"learning_rate": 0.0007526115859449192,
"loss": 0.2411,
"step": 3020
},
{
"epoch": 9.696,
"grad_norm": 386762.625,
"learning_rate": 0.0007514245014245015,
"loss": 0.1899,
"step": 3030
},
{
"epoch": 9.728,
"grad_norm": 407145.6875,
"learning_rate": 0.0007502374169040836,
"loss": 0.2029,
"step": 3040
},
{
"epoch": 9.76,
"grad_norm": 573868.8125,
"learning_rate": 0.0007490503323836657,
"loss": 0.2023,
"step": 3050
},
{
"epoch": 9.792,
"grad_norm": 457500.875,
"learning_rate": 0.0007478632478632479,
"loss": 0.2094,
"step": 3060
},
{
"epoch": 9.824,
"grad_norm": 477359.5,
"learning_rate": 0.0007466761633428301,
"loss": 0.2024,
"step": 3070
},
{
"epoch": 9.856,
"grad_norm": 406626.96875,
"learning_rate": 0.0007454890788224121,
"loss": 0.2074,
"step": 3080
},
{
"epoch": 9.888,
"grad_norm": 440002.4375,
"learning_rate": 0.0007443019943019943,
"loss": 0.1979,
"step": 3090
},
{
"epoch": 9.92,
"grad_norm": 465033.5625,
"learning_rate": 0.0007431149097815765,
"loss": 0.2019,
"step": 3100
},
{
"epoch": 9.952,
"grad_norm": 627469.5625,
"learning_rate": 0.0007419278252611586,
"loss": 0.2066,
"step": 3110
},
{
"epoch": 9.984,
"grad_norm": 495403.3125,
"learning_rate": 0.0007407407407407407,
"loss": 0.2136,
"step": 3120
},
{
"epoch": 10.0,
"eval_accuracy": 0.77275,
"eval_loss": 0.5483222007751465,
"eval_runtime": 58.3061,
"eval_samples_per_second": 68.603,
"eval_steps_per_second": 2.144,
"step": 3125
},
{
"epoch": 10.016,
"grad_norm": 450484.65625,
"learning_rate": 0.0007395536562203229,
"loss": 0.2118,
"step": 3130
},
{
"epoch": 10.048,
"grad_norm": 329785.25,
"learning_rate": 0.000738366571699905,
"loss": 0.2155,
"step": 3140
},
{
"epoch": 10.08,
"grad_norm": 587364.0,
"learning_rate": 0.0007371794871794872,
"loss": 0.2,
"step": 3150
},
{
"epoch": 10.112,
"grad_norm": 518025.4375,
"learning_rate": 0.0007359924026590693,
"loss": 0.2157,
"step": 3160
},
{
"epoch": 10.144,
"grad_norm": 416655.78125,
"learning_rate": 0.0007348053181386514,
"loss": 0.1865,
"step": 3170
},
{
"epoch": 10.176,
"grad_norm": 570281.9375,
"learning_rate": 0.0007336182336182337,
"loss": 0.2008,
"step": 3180
},
{
"epoch": 10.208,
"grad_norm": 401606.65625,
"learning_rate": 0.0007324311490978159,
"loss": 0.1793,
"step": 3190
},
{
"epoch": 10.24,
"grad_norm": 859104.125,
"learning_rate": 0.0007312440645773979,
"loss": 0.2221,
"step": 3200
},
{
"epoch": 10.272,
"grad_norm": 652583.1875,
"learning_rate": 0.0007300569800569801,
"loss": 0.2338,
"step": 3210
},
{
"epoch": 10.304,
"grad_norm": 495821.96875,
"learning_rate": 0.0007288698955365623,
"loss": 0.1917,
"step": 3220
},
{
"epoch": 10.336,
"grad_norm": 357223.5,
"learning_rate": 0.0007276828110161444,
"loss": 0.1903,
"step": 3230
},
{
"epoch": 10.368,
"grad_norm": 603392.0,
"learning_rate": 0.0007264957264957265,
"loss": 0.2161,
"step": 3240
},
{
"epoch": 10.4,
"grad_norm": 329822.15625,
"learning_rate": 0.0007253086419753087,
"loss": 0.1862,
"step": 3250
},
{
"epoch": 10.432,
"grad_norm": 600281.1875,
"learning_rate": 0.0007241215574548908,
"loss": 0.2185,
"step": 3260
},
{
"epoch": 10.464,
"grad_norm": 553877.1875,
"learning_rate": 0.0007229344729344729,
"loss": 0.2306,
"step": 3270
},
{
"epoch": 10.496,
"grad_norm": 571845.375,
"learning_rate": 0.0007217473884140551,
"loss": 0.2141,
"step": 3280
},
{
"epoch": 10.528,
"grad_norm": 544081.0,
"learning_rate": 0.0007205603038936372,
"loss": 0.1899,
"step": 3290
},
{
"epoch": 10.56,
"grad_norm": 331369.15625,
"learning_rate": 0.0007193732193732194,
"loss": 0.2077,
"step": 3300
},
{
"epoch": 10.592,
"grad_norm": 647373.4375,
"learning_rate": 0.0007181861348528014,
"loss": 0.2368,
"step": 3310
},
{
"epoch": 10.624,
"grad_norm": 522459.0625,
"learning_rate": 0.0007169990503323836,
"loss": 0.2045,
"step": 3320
},
{
"epoch": 10.656,
"grad_norm": 426327.875,
"learning_rate": 0.0007158119658119658,
"loss": 0.2038,
"step": 3330
},
{
"epoch": 10.688,
"grad_norm": 511659.5,
"learning_rate": 0.0007146248812915481,
"loss": 0.2097,
"step": 3340
},
{
"epoch": 10.72,
"grad_norm": 475094.4375,
"learning_rate": 0.0007134377967711301,
"loss": 0.2063,
"step": 3350
},
{
"epoch": 10.752,
"grad_norm": 652052.625,
"learning_rate": 0.0007122507122507123,
"loss": 0.2022,
"step": 3360
},
{
"epoch": 10.784,
"grad_norm": 462668.75,
"learning_rate": 0.0007110636277302945,
"loss": 0.2168,
"step": 3370
},
{
"epoch": 10.816,
"grad_norm": 405784.875,
"learning_rate": 0.0007098765432098766,
"loss": 0.2003,
"step": 3380
},
{
"epoch": 10.848,
"grad_norm": 721382.25,
"learning_rate": 0.0007086894586894587,
"loss": 0.2271,
"step": 3390
},
{
"epoch": 10.88,
"grad_norm": 494971.40625,
"learning_rate": 0.0007075023741690408,
"loss": 0.2471,
"step": 3400
},
{
"epoch": 10.912,
"grad_norm": 414839.5625,
"learning_rate": 0.000706315289648623,
"loss": 0.2309,
"step": 3410
},
{
"epoch": 10.943999999999999,
"grad_norm": 496315.90625,
"learning_rate": 0.0007051282051282052,
"loss": 0.2112,
"step": 3420
},
{
"epoch": 10.975999999999999,
"grad_norm": 473891.03125,
"learning_rate": 0.0007039411206077872,
"loss": 0.2184,
"step": 3430
},
{
"epoch": 10.9984,
"eval_accuracy": 0.7665,
"eval_loss": 0.5606116652488708,
"eval_runtime": 58.0847,
"eval_samples_per_second": 68.865,
"eval_steps_per_second": 2.152,
"step": 3437
},
{
"epoch": 11.008,
"grad_norm": 612277.8125,
"learning_rate": 0.0007027540360873694,
"loss": 0.2077,
"step": 3440
},
{
"epoch": 11.04,
"grad_norm": 470802.4375,
"learning_rate": 0.0007015669515669516,
"loss": 0.212,
"step": 3450
},
{
"epoch": 11.072,
"grad_norm": 446870.0,
"learning_rate": 0.0007003798670465337,
"loss": 0.2173,
"step": 3460
},
{
"epoch": 11.104,
"grad_norm": 309706.625,
"learning_rate": 0.0006991927825261158,
"loss": 0.2054,
"step": 3470
},
{
"epoch": 11.136,
"grad_norm": 427292.09375,
"learning_rate": 0.000698005698005698,
"loss": 0.1961,
"step": 3480
},
{
"epoch": 11.168,
"grad_norm": 650260.3125,
"learning_rate": 0.0006968186134852802,
"loss": 0.2146,
"step": 3490
},
{
"epoch": 11.2,
"grad_norm": 416737.78125,
"learning_rate": 0.0006956315289648623,
"loss": 0.1953,
"step": 3500
},
{
"epoch": 11.232,
"grad_norm": 568974.875,
"learning_rate": 0.0006944444444444445,
"loss": 0.1921,
"step": 3510
},
{
"epoch": 11.264,
"grad_norm": 630401.9375,
"learning_rate": 0.0006932573599240266,
"loss": 0.2157,
"step": 3520
},
{
"epoch": 11.296,
"grad_norm": 449739.40625,
"learning_rate": 0.0006920702754036088,
"loss": 0.2191,
"step": 3530
},
{
"epoch": 11.328,
"grad_norm": 504943.21875,
"learning_rate": 0.0006908831908831909,
"loss": 0.2161,
"step": 3540
},
{
"epoch": 11.36,
"grad_norm": 621631.25,
"learning_rate": 0.000689696106362773,
"loss": 0.2137,
"step": 3550
},
{
"epoch": 11.392,
"grad_norm": 474948.28125,
"learning_rate": 0.0006885090218423552,
"loss": 0.2206,
"step": 3560
},
{
"epoch": 11.424,
"grad_norm": 534952.875,
"learning_rate": 0.0006873219373219374,
"loss": 0.2098,
"step": 3570
},
{
"epoch": 11.456,
"grad_norm": 449058.09375,
"learning_rate": 0.0006861348528015194,
"loss": 0.2134,
"step": 3580
},
{
"epoch": 11.488,
"grad_norm": 601122.25,
"learning_rate": 0.0006849477682811016,
"loss": 0.2208,
"step": 3590
},
{
"epoch": 11.52,
"grad_norm": 363794.46875,
"learning_rate": 0.0006837606837606838,
"loss": 0.2073,
"step": 3600
},
{
"epoch": 11.552,
"grad_norm": 483605.5,
"learning_rate": 0.0006825735992402659,
"loss": 0.2195,
"step": 3610
},
{
"epoch": 11.584,
"grad_norm": 533998.8125,
"learning_rate": 0.000681386514719848,
"loss": 0.2024,
"step": 3620
},
{
"epoch": 11.616,
"grad_norm": 435305.09375,
"learning_rate": 0.0006801994301994302,
"loss": 0.2183,
"step": 3630
},
{
"epoch": 11.648,
"grad_norm": 567111.9375,
"learning_rate": 0.0006790123456790124,
"loss": 0.1902,
"step": 3640
},
{
"epoch": 11.68,
"grad_norm": 405747.65625,
"learning_rate": 0.0006778252611585946,
"loss": 0.2238,
"step": 3650
},
{
"epoch": 11.712,
"grad_norm": 575645.875,
"learning_rate": 0.0006766381766381767,
"loss": 0.2022,
"step": 3660
},
{
"epoch": 11.744,
"grad_norm": 559379.5,
"learning_rate": 0.0006754510921177588,
"loss": 0.23,
"step": 3670
},
{
"epoch": 11.776,
"grad_norm": 496193.0,
"learning_rate": 0.000674264007597341,
"loss": 0.2015,
"step": 3680
},
{
"epoch": 11.808,
"grad_norm": 470825.0,
"learning_rate": 0.0006730769230769232,
"loss": 0.2249,
"step": 3690
},
{
"epoch": 11.84,
"grad_norm": 507620.0,
"learning_rate": 0.0006718898385565052,
"loss": 0.2111,
"step": 3700
},
{
"epoch": 11.872,
"grad_norm": 468785.75,
"learning_rate": 0.0006707027540360874,
"loss": 0.2051,
"step": 3710
},
{
"epoch": 11.904,
"grad_norm": 522254.5,
"learning_rate": 0.0006695156695156696,
"loss": 0.2035,
"step": 3720
},
{
"epoch": 11.936,
"grad_norm": 453727.625,
"learning_rate": 0.0006683285849952516,
"loss": 0.2061,
"step": 3730
},
{
"epoch": 11.968,
"grad_norm": 518330.09375,
"learning_rate": 0.0006671415004748338,
"loss": 0.215,
"step": 3740
},
{
"epoch": 12.0,
"grad_norm": 498891.5,
"learning_rate": 0.0006659544159544159,
"loss": 0.212,
"step": 3750
},
{
"epoch": 12.0,
"eval_accuracy": 0.761,
"eval_loss": 0.5577554106712341,
"eval_runtime": 58.301,
"eval_samples_per_second": 68.609,
"eval_steps_per_second": 2.144,
"step": 3750
},
{
"epoch": 12.032,
"grad_norm": 663838.375,
"learning_rate": 0.0006647673314339981,
"loss": 0.2126,
"step": 3760
},
{
"epoch": 12.064,
"grad_norm": 532076.6875,
"learning_rate": 0.0006635802469135802,
"loss": 0.2208,
"step": 3770
},
{
"epoch": 12.096,
"grad_norm": 609268.625,
"learning_rate": 0.0006623931623931623,
"loss": 0.1939,
"step": 3780
},
{
"epoch": 12.128,
"grad_norm": 383783.15625,
"learning_rate": 0.0006612060778727446,
"loss": 0.2174,
"step": 3790
},
{
"epoch": 12.16,
"grad_norm": 483893.375,
"learning_rate": 0.0006600189933523268,
"loss": 0.2083,
"step": 3800
},
{
"epoch": 12.192,
"grad_norm": 527741.625,
"learning_rate": 0.0006588319088319088,
"loss": 0.1918,
"step": 3810
},
{
"epoch": 12.224,
"grad_norm": 472716.28125,
"learning_rate": 0.000657644824311491,
"loss": 0.2262,
"step": 3820
},
{
"epoch": 12.256,
"grad_norm": 521136.75,
"learning_rate": 0.0006564577397910732,
"loss": 0.2014,
"step": 3830
},
{
"epoch": 12.288,
"grad_norm": 350996.34375,
"learning_rate": 0.0006552706552706553,
"loss": 0.2096,
"step": 3840
},
{
"epoch": 12.32,
"grad_norm": 372622.5625,
"learning_rate": 0.0006540835707502374,
"loss": 0.2027,
"step": 3850
},
{
"epoch": 12.352,
"grad_norm": 639510.3125,
"learning_rate": 0.0006528964862298196,
"loss": 0.2272,
"step": 3860
},
{
"epoch": 12.384,
"grad_norm": 417869.4375,
"learning_rate": 0.0006517094017094017,
"loss": 0.1993,
"step": 3870
},
{
"epoch": 12.416,
"grad_norm": 666939.6875,
"learning_rate": 0.0006505223171889839,
"loss": 0.2285,
"step": 3880
},
{
"epoch": 12.448,
"grad_norm": 392265.34375,
"learning_rate": 0.000649335232668566,
"loss": 0.2296,
"step": 3890
},
{
"epoch": 12.48,
"grad_norm": 490617.21875,
"learning_rate": 0.0006481481481481481,
"loss": 0.2158,
"step": 3900
},
{
"epoch": 12.512,
"grad_norm": 432442.40625,
"learning_rate": 0.0006469610636277303,
"loss": 0.206,
"step": 3910
},
{
"epoch": 12.544,
"grad_norm": 562610.0,
"learning_rate": 0.0006457739791073125,
"loss": 0.2248,
"step": 3920
},
{
"epoch": 12.576,
"grad_norm": 646817.1875,
"learning_rate": 0.0006445868945868945,
"loss": 0.2283,
"step": 3930
},
{
"epoch": 12.608,
"grad_norm": 499516.1875,
"learning_rate": 0.0006433998100664767,
"loss": 0.2219,
"step": 3940
},
{
"epoch": 12.64,
"grad_norm": 319734.6875,
"learning_rate": 0.000642212725546059,
"loss": 0.1956,
"step": 3950
},
{
"epoch": 12.672,
"grad_norm": 598636.3125,
"learning_rate": 0.0006410256410256411,
"loss": 0.2137,
"step": 3960
},
{
"epoch": 12.704,
"grad_norm": 490632.96875,
"learning_rate": 0.0006398385565052232,
"loss": 0.2297,
"step": 3970
},
{
"epoch": 12.736,
"grad_norm": 1125840.875,
"learning_rate": 0.0006386514719848054,
"loss": 0.2152,
"step": 3980
},
{
"epoch": 12.768,
"grad_norm": 550740.3125,
"learning_rate": 0.0006374643874643875,
"loss": 0.2294,
"step": 3990
},
{
"epoch": 12.8,
"grad_norm": 529733.125,
"learning_rate": 0.0006362773029439696,
"loss": 0.2093,
"step": 4000
},
{
"epoch": 12.832,
"grad_norm": 1042909.75,
"learning_rate": 0.0006350902184235518,
"loss": 0.1975,
"step": 4010
},
{
"epoch": 12.864,
"grad_norm": 885337.75,
"learning_rate": 0.0006339031339031339,
"loss": 0.2048,
"step": 4020
},
{
"epoch": 12.896,
"grad_norm": 1166082.375,
"learning_rate": 0.0006327160493827161,
"loss": 0.2052,
"step": 4030
},
{
"epoch": 12.928,
"grad_norm": 719802.6875,
"learning_rate": 0.0006315289648622981,
"loss": 0.2104,
"step": 4040
},
{
"epoch": 12.96,
"grad_norm": 1076805.5,
"learning_rate": 0.0006303418803418803,
"loss": 0.2183,
"step": 4050
},
{
"epoch": 12.992,
"grad_norm": 852746.0,
"learning_rate": 0.0006291547958214625,
"loss": 0.1903,
"step": 4060
},
{
"epoch": 12.9984,
"eval_accuracy": 0.769,
"eval_loss": 0.5370960235595703,
"eval_runtime": 58.6162,
"eval_samples_per_second": 68.241,
"eval_steps_per_second": 2.133,
"step": 4062
},
{
"epoch": 13.024,
"grad_norm": 654611.75,
"learning_rate": 0.0006279677113010446,
"loss": 0.2046,
"step": 4070
},
{
"epoch": 13.056,
"grad_norm": 891781.0625,
"learning_rate": 0.0006267806267806267,
"loss": 0.2293,
"step": 4080
},
{
"epoch": 13.088,
"grad_norm": 1099780.75,
"learning_rate": 0.0006255935422602089,
"loss": 0.2096,
"step": 4090
},
{
"epoch": 13.12,
"grad_norm": 1592056.875,
"learning_rate": 0.0006244064577397912,
"loss": 0.2054,
"step": 4100
},
{
"epoch": 13.152,
"grad_norm": 1195057.75,
"learning_rate": 0.0006232193732193733,
"loss": 0.215,
"step": 4110
},
{
"epoch": 13.184,
"grad_norm": 864565.3125,
"learning_rate": 0.0006220322886989554,
"loss": 0.1961,
"step": 4120
},
{
"epoch": 13.216,
"grad_norm": 1274778.875,
"learning_rate": 0.0006208452041785375,
"loss": 0.21,
"step": 4130
},
{
"epoch": 13.248,
"grad_norm": 1391209.125,
"learning_rate": 0.0006196581196581197,
"loss": 0.2291,
"step": 4140
},
{
"epoch": 13.28,
"grad_norm": 1173761.25,
"learning_rate": 0.0006184710351377019,
"loss": 0.2186,
"step": 4150
},
{
"epoch": 13.312,
"grad_norm": 1076280.625,
"learning_rate": 0.0006172839506172839,
"loss": 0.2193,
"step": 4160
},
{
"epoch": 13.344,
"grad_norm": 809341.9375,
"learning_rate": 0.0006160968660968661,
"loss": 0.1914,
"step": 4170
},
{
"epoch": 13.376,
"grad_norm": 1084379.0,
"learning_rate": 0.0006149097815764483,
"loss": 0.2064,
"step": 4180
},
{
"epoch": 13.408,
"grad_norm": 833949.4375,
"learning_rate": 0.0006137226970560304,
"loss": 0.2139,
"step": 4190
},
{
"epoch": 13.44,
"grad_norm": 1337724.25,
"learning_rate": 0.0006125356125356125,
"loss": 0.1961,
"step": 4200
},
{
"epoch": 13.472,
"grad_norm": 964052.1875,
"learning_rate": 0.0006113485280151947,
"loss": 0.2057,
"step": 4210
},
{
"epoch": 13.504,
"grad_norm": 1420036.875,
"learning_rate": 0.0006101614434947768,
"loss": 0.2339,
"step": 4220
},
{
"epoch": 13.536,
"grad_norm": 1049277.375,
"learning_rate": 0.0006089743589743589,
"loss": 0.1966,
"step": 4230
},
{
"epoch": 13.568,
"grad_norm": 1342290.875,
"learning_rate": 0.0006077872744539411,
"loss": 0.2193,
"step": 4240
},
{
"epoch": 13.6,
"grad_norm": 1058454.25,
"learning_rate": 0.0006066001899335233,
"loss": 0.2211,
"step": 4250
},
{
"epoch": 13.632,
"grad_norm": 802834.5,
"learning_rate": 0.0006054131054131055,
"loss": 0.206,
"step": 4260
},
{
"epoch": 13.664,
"grad_norm": 939160.625,
"learning_rate": 0.0006042260208926876,
"loss": 0.2077,
"step": 4270
},
{
"epoch": 13.696,
"grad_norm": 1321520.75,
"learning_rate": 0.0006030389363722697,
"loss": 0.2347,
"step": 4280
},
{
"epoch": 13.728,
"grad_norm": Infinity,
"learning_rate": 0.0006018518518518519,
"loss": 0.2057,
"step": 4290
},
{
"epoch": 13.76,
"grad_norm": 1250876.625,
"learning_rate": 0.0006006647673314341,
"loss": 0.2216,
"step": 4300
},
{
"epoch": 13.792,
"grad_norm": 1065398.25,
"learning_rate": 0.0005994776828110161,
"loss": 0.2212,
"step": 4310
},
{
"epoch": 13.824,
"grad_norm": 923487.375,
"learning_rate": 0.0005982905982905983,
"loss": 0.2153,
"step": 4320
},
{
"epoch": 13.856,
"grad_norm": 951445.8125,
"learning_rate": 0.0005971035137701805,
"loss": 0.2128,
"step": 4330
},
{
"epoch": 13.888,
"grad_norm": 1154133.5,
"learning_rate": 0.0005959164292497626,
"loss": 0.2115,
"step": 4340
},
{
"epoch": 13.92,
"grad_norm": 991618.625,
"learning_rate": 0.0005947293447293447,
"loss": 0.2101,
"step": 4350
},
{
"epoch": 13.952,
"grad_norm": 974332.625,
"learning_rate": 0.0005935422602089269,
"loss": 0.2084,
"step": 4360
},
{
"epoch": 13.984,
"grad_norm": 1015135.1875,
"learning_rate": 0.000592355175688509,
"loss": 0.2487,
"step": 4370
},
{
"epoch": 14.0,
"eval_accuracy": 0.7645,
"eval_loss": 0.5582425594329834,
"eval_runtime": 61.8491,
"eval_samples_per_second": 64.674,
"eval_steps_per_second": 2.021,
"step": 4375
},
{
"epoch": 14.016,
"grad_norm": NaN,
"learning_rate": 0.0005911680911680912,
"loss": 0.2202,
"step": 4380
},
{
"epoch": 14.048,
"grad_norm": 725532.25,
"learning_rate": 0.0005899810066476732,
"loss": 0.1936,
"step": 4390
},
{
"epoch": 14.08,
"grad_norm": Infinity,
"learning_rate": 0.0005887939221272554,
"loss": 0.2276,
"step": 4400
},
{
"epoch": 14.112,
"grad_norm": Infinity,
"learning_rate": 0.0005876068376068377,
"loss": 0.215,
"step": 4410
},
{
"epoch": 14.144,
"grad_norm": 687965.625,
"learning_rate": 0.0005864197530864199,
"loss": 0.2159,
"step": 4420
},
{
"epoch": 14.176,
"grad_norm": 1061211.25,
"learning_rate": 0.0005852326685660019,
"loss": 0.193,
"step": 4430
},
{
"epoch": 14.208,
"grad_norm": 872460.75,
"learning_rate": 0.0005840455840455841,
"loss": 0.2167,
"step": 4440
},
{
"epoch": 14.24,
"grad_norm": 1020573.6875,
"learning_rate": 0.0005828584995251662,
"loss": 0.1868,
"step": 4450
},
{
"epoch": 14.272,
"grad_norm": Infinity,
"learning_rate": 0.0005816714150047484,
"loss": 0.1987,
"step": 4460
},
{
"epoch": 14.304,
"grad_norm": 1052701.25,
"learning_rate": 0.0005804843304843305,
"loss": 0.2031,
"step": 4470
},
{
"epoch": 14.336,
"grad_norm": Infinity,
"learning_rate": 0.0005792972459639126,
"loss": 0.2252,
"step": 4480
},
{
"epoch": 14.368,
"grad_norm": 679741.25,
"learning_rate": 0.0005781101614434948,
"loss": 0.1823,
"step": 4490
},
{
"epoch": 14.4,
"grad_norm": 1170840.375,
"learning_rate": 0.0005769230769230769,
"loss": 0.2216,
"step": 4500
},
{
"epoch": 14.432,
"grad_norm": Infinity,
"learning_rate": 0.000575735992402659,
"loss": 0.235,
"step": 4510
},
{
"epoch": 14.464,
"grad_norm": Infinity,
"learning_rate": 0.0005745489078822412,
"loss": 0.1994,
"step": 4520
},
{
"epoch": 14.496,
"grad_norm": 1133208.125,
"learning_rate": 0.0005733618233618234,
"loss": 0.2242,
"step": 4530
},
{
"epoch": 14.528,
"grad_norm": 1266796.75,
"learning_rate": 0.0005721747388414054,
"loss": 0.2125,
"step": 4540
},
{
"epoch": 14.56,
"grad_norm": 1040882.375,
"learning_rate": 0.0005709876543209876,
"loss": 0.1994,
"step": 4550
},
{
"epoch": 14.592,
"grad_norm": 896784.9375,
"learning_rate": 0.0005698005698005699,
"loss": 0.2121,
"step": 4560
},
{
"epoch": 14.624,
"grad_norm": Infinity,
"learning_rate": 0.000568613485280152,
"loss": 0.2284,
"step": 4570
},
{
"epoch": 14.656,
"grad_norm": 1164985.0,
"learning_rate": 0.0005674264007597341,
"loss": 0.2249,
"step": 4580
},
{
"epoch": 14.688,
"grad_norm": Infinity,
"learning_rate": 0.0005662393162393163,
"loss": 0.2214,
"step": 4590
},
{
"epoch": 14.72,
"grad_norm": 939342.125,
"learning_rate": 0.0005650522317188984,
"loss": 0.2015,
"step": 4600
},
{
"epoch": 14.752,
"grad_norm": 636754.0625,
"learning_rate": 0.0005638651471984806,
"loss": 0.2103,
"step": 4610
},
{
"epoch": 14.784,
"grad_norm": 1226811.0,
"learning_rate": 0.0005626780626780627,
"loss": 0.2122,
"step": 4620
},
{
"epoch": 14.816,
"grad_norm": 1043980.875,
"learning_rate": 0.0005614909781576448,
"loss": 0.1971,
"step": 4630
},
{
"epoch": 14.848,
"grad_norm": 1471149.375,
"learning_rate": 0.000560303893637227,
"loss": 0.2176,
"step": 4640
},
{
"epoch": 14.88,
"grad_norm": 935474.1875,
"learning_rate": 0.0005591168091168092,
"loss": 0.2057,
"step": 4650
},
{
"epoch": 14.912,
"grad_norm": 938039.5625,
"learning_rate": 0.0005579297245963912,
"loss": 0.2064,
"step": 4660
},
{
"epoch": 14.943999999999999,
"grad_norm": Infinity,
"learning_rate": 0.0005567426400759734,
"loss": 0.2427,
"step": 4670
},
{
"epoch": 14.975999999999999,
"grad_norm": 1067872.875,
"learning_rate": 0.0005555555555555556,
"loss": 0.2025,
"step": 4680
},
{
"epoch": 14.9984,
"eval_accuracy": 0.77775,
"eval_loss": 0.5413510203361511,
"eval_runtime": 58.0126,
"eval_samples_per_second": 68.951,
"eval_steps_per_second": 2.155,
"step": 4687
},
{
"epoch": 15.008,
"grad_norm": 1081480.5,
"learning_rate": 0.0005543684710351377,
"loss": 0.2015,
"step": 4690
},
{
"epoch": 15.04,
"grad_norm": 1213699.125,
"learning_rate": 0.0005531813865147198,
"loss": 0.2256,
"step": 4700
},
{
"epoch": 15.072,
"grad_norm": 1344064.75,
"learning_rate": 0.0005519943019943021,
"loss": 0.2214,
"step": 4710
},
{
"epoch": 15.104,
"grad_norm": 849799.5,
"learning_rate": 0.0005508072174738842,
"loss": 0.2023,
"step": 4720
},
{
"epoch": 15.136,
"grad_norm": 982086.5,
"learning_rate": 0.0005496201329534663,
"loss": 0.1956,
"step": 4730
},
{
"epoch": 15.168,
"grad_norm": 744843.3125,
"learning_rate": 0.0005484330484330485,
"loss": 0.203,
"step": 4740
},
{
"epoch": 15.2,
"grad_norm": Infinity,
"learning_rate": 0.0005472459639126306,
"loss": 0.1863,
"step": 4750
},
{
"epoch": 15.232,
"grad_norm": 1000041.625,
"learning_rate": 0.0005460588793922128,
"loss": 0.2262,
"step": 4760
},
{
"epoch": 15.264,
"grad_norm": Infinity,
"learning_rate": 0.0005448717948717948,
"loss": 0.1778,
"step": 4770
},
{
"epoch": 15.296,
"grad_norm": 761818.25,
"learning_rate": 0.000543684710351377,
"loss": 0.2134,
"step": 4780
},
{
"epoch": 15.328,
"grad_norm": Infinity,
"learning_rate": 0.0005424976258309592,
"loss": 0.2246,
"step": 4790
},
{
"epoch": 15.36,
"grad_norm": 1044292.875,
"learning_rate": 0.0005413105413105413,
"loss": 0.2246,
"step": 4800
},
{
"epoch": 15.392,
"grad_norm": 740541.6875,
"learning_rate": 0.0005401234567901234,
"loss": 0.1943,
"step": 4810
},
{
"epoch": 15.424,
"grad_norm": 980434.4375,
"learning_rate": 0.0005389363722697056,
"loss": 0.1904,
"step": 4820
},
{
"epoch": 15.456,
"grad_norm": 1246758.625,
"learning_rate": 0.0005377492877492877,
"loss": 0.2293,
"step": 4830
},
{
"epoch": 15.488,
"grad_norm": 1481804.875,
"learning_rate": 0.0005365622032288699,
"loss": 0.2251,
"step": 4840
},
{
"epoch": 15.52,
"grad_norm": 956031.5,
"learning_rate": 0.000535375118708452,
"loss": 0.2239,
"step": 4850
},
{
"epoch": 15.552,
"grad_norm": 1267528.25,
"learning_rate": 0.0005341880341880342,
"loss": 0.2263,
"step": 4860
},
{
"epoch": 15.584,
"grad_norm": Infinity,
"learning_rate": 0.0005330009496676164,
"loss": 0.2041,
"step": 4870
},
{
"epoch": 15.616,
"grad_norm": 925829.875,
"learning_rate": 0.0005318138651471986,
"loss": 0.1838,
"step": 4880
},
{
"epoch": 15.648,
"grad_norm": Infinity,
"learning_rate": 0.0005306267806267806,
"loss": 0.2163,
"step": 4890
},
{
"epoch": 15.68,
"grad_norm": 1228792.375,
"learning_rate": 0.0005294396961063628,
"loss": 0.2032,
"step": 4900
},
{
"epoch": 15.712,
"grad_norm": 838715.125,
"learning_rate": 0.000528252611585945,
"loss": 0.1991,
"step": 4910
},
{
"epoch": 15.744,
"grad_norm": 964459.0,
"learning_rate": 0.0005270655270655271,
"loss": 0.2263,
"step": 4920
},
{
"epoch": 15.776,
"grad_norm": 798003.25,
"learning_rate": 0.0005258784425451092,
"loss": 0.199,
"step": 4930
},
{
"epoch": 15.808,
"grad_norm": Infinity,
"learning_rate": 0.0005246913580246914,
"loss": 0.2218,
"step": 4940
},
{
"epoch": 15.84,
"grad_norm": 1483930.875,
"learning_rate": 0.0005235042735042735,
"loss": 0.2356,
"step": 4950
},
{
"epoch": 15.872,
"grad_norm": Infinity,
"learning_rate": 0.0005223171889838556,
"loss": 0.2247,
"step": 4960
},
{
"epoch": 15.904,
"grad_norm": Infinity,
"learning_rate": 0.0005211301044634378,
"loss": 0.1986,
"step": 4970
},
{
"epoch": 15.936,
"grad_norm": 1194618.5,
"learning_rate": 0.0005199430199430199,
"loss": 0.2393,
"step": 4980
},
{
"epoch": 15.968,
"grad_norm": 837843.75,
"learning_rate": 0.0005187559354226021,
"loss": 0.2107,
"step": 4990
},
{
"epoch": 16.0,
"grad_norm": 912574.75,
"learning_rate": 0.0005175688509021842,
"loss": 0.2207,
"step": 5000
},
{
"epoch": 16.0,
"eval_accuracy": 0.7685,
"eval_loss": 0.5375534892082214,
"eval_runtime": 58.3981,
"eval_samples_per_second": 68.495,
"eval_steps_per_second": 2.14,
"step": 5000
},
{
"epoch": 16.032,
"grad_norm": Infinity,
"learning_rate": 0.0005163817663817663,
"loss": 0.2148,
"step": 5010
},
{
"epoch": 16.064,
"grad_norm": Infinity,
"learning_rate": 0.0005151946818613486,
"loss": 0.2201,
"step": 5020
},
{
"epoch": 16.096,
"grad_norm": 924105.5,
"learning_rate": 0.0005140075973409308,
"loss": 0.2183,
"step": 5030
},
{
"epoch": 16.128,
"grad_norm": 1160855.125,
"learning_rate": 0.0005128205128205128,
"loss": 0.2145,
"step": 5040
},
{
"epoch": 16.16,
"grad_norm": 1011611.5,
"learning_rate": 0.000511633428300095,
"loss": 0.2183,
"step": 5050
},
{
"epoch": 16.192,
"grad_norm": 1249247.375,
"learning_rate": 0.0005104463437796772,
"loss": 0.2077,
"step": 5060
},
{
"epoch": 16.224,
"grad_norm": Infinity,
"learning_rate": 0.0005092592592592593,
"loss": 0.204,
"step": 5070
},
{
"epoch": 16.256,
"grad_norm": 1121213.75,
"learning_rate": 0.0005080721747388414,
"loss": 0.2116,
"step": 5080
},
{
"epoch": 16.288,
"grad_norm": 938717.25,
"learning_rate": 0.0005068850902184235,
"loss": 0.2033,
"step": 5090
},
{
"epoch": 16.32,
"grad_norm": Infinity,
"learning_rate": 0.0005056980056980057,
"loss": 0.2219,
"step": 5100
},
{
"epoch": 16.352,
"grad_norm": 952882.625,
"learning_rate": 0.0005045109211775879,
"loss": 0.226,
"step": 5110
},
{
"epoch": 16.384,
"grad_norm": 1363143.5,
"learning_rate": 0.0005033238366571699,
"loss": 0.2174,
"step": 5120
},
{
"epoch": 16.416,
"grad_norm": 917260.875,
"learning_rate": 0.0005021367521367521,
"loss": 0.1846,
"step": 5130
},
{
"epoch": 16.448,
"grad_norm": Infinity,
"learning_rate": 0.0005009496676163343,
"loss": 0.1981,
"step": 5140
},
{
"epoch": 16.48,
"grad_norm": 1652643.0,
"learning_rate": 0.0004997625830959164,
"loss": 0.2138,
"step": 5150
},
{
"epoch": 16.512,
"grad_norm": 942965.1875,
"learning_rate": 0.0004985754985754986,
"loss": 0.207,
"step": 5160
},
{
"epoch": 16.544,
"grad_norm": 1058725.375,
"learning_rate": 0.0004973884140550808,
"loss": 0.1956,
"step": 5170
},
{
"epoch": 16.576,
"grad_norm": 1011255.875,
"learning_rate": 0.000496201329534663,
"loss": 0.2182,
"step": 5180
},
{
"epoch": 16.608,
"grad_norm": 949710.75,
"learning_rate": 0.000495014245014245,
"loss": 0.205,
"step": 5190
},
{
"epoch": 16.64,
"grad_norm": 939276.4375,
"learning_rate": 0.0004938271604938272,
"loss": 0.2186,
"step": 5200
},
{
"epoch": 16.672,
"grad_norm": 1041804.3125,
"learning_rate": 0.0004926400759734093,
"loss": 0.2115,
"step": 5210
},
{
"epoch": 16.704,
"grad_norm": 976855.5625,
"learning_rate": 0.0004914529914529914,
"loss": 0.2076,
"step": 5220
},
{
"epoch": 16.736,
"grad_norm": 881195.3125,
"learning_rate": 0.0004902659069325736,
"loss": 0.2018,
"step": 5230
},
{
"epoch": 16.768,
"grad_norm": 1045999.0,
"learning_rate": 0.0004890788224121557,
"loss": 0.2095,
"step": 5240
},
{
"epoch": 16.8,
"grad_norm": 1055538.875,
"learning_rate": 0.0004878917378917379,
"loss": 0.2254,
"step": 5250
},
{
"epoch": 16.832,
"grad_norm": 1142565.625,
"learning_rate": 0.00048670465337132,
"loss": 0.2343,
"step": 5260
},
{
"epoch": 16.864,
"grad_norm": 1099992.25,
"learning_rate": 0.0004855175688509022,
"loss": 0.219,
"step": 5270
},
{
"epoch": 16.896,
"grad_norm": Infinity,
"learning_rate": 0.00048433048433048435,
"loss": 0.2458,
"step": 5280
},
{
"epoch": 16.928,
"grad_norm": 946468.9375,
"learning_rate": 0.0004831433998100665,
"loss": 0.1974,
"step": 5290
},
{
"epoch": 16.96,
"grad_norm": 1041877.25,
"learning_rate": 0.00048195631528964863,
"loss": 0.217,
"step": 5300
},
{
"epoch": 16.992,
"grad_norm": 1171594.625,
"learning_rate": 0.0004807692307692308,
"loss": 0.2012,
"step": 5310
},
{
"epoch": 16.9984,
"eval_accuracy": 0.77025,
"eval_loss": 0.548855721950531,
"eval_runtime": 58.086,
"eval_samples_per_second": 68.863,
"eval_steps_per_second": 2.152,
"step": 5312
},
{
"epoch": 17.024,
"grad_norm": 980497.25,
"learning_rate": 0.0004795821462488129,
"loss": 0.2157,
"step": 5320
},
{
"epoch": 17.056,
"grad_norm": 748758.875,
"learning_rate": 0.0004783950617283951,
"loss": 0.2023,
"step": 5330
},
{
"epoch": 17.088,
"grad_norm": 1243409.375,
"learning_rate": 0.0004772079772079772,
"loss": 0.2254,
"step": 5340
},
{
"epoch": 17.12,
"grad_norm": 738619.375,
"learning_rate": 0.00047602089268755936,
"loss": 0.1961,
"step": 5350
},
{
"epoch": 17.152,
"grad_norm": 1088919.75,
"learning_rate": 0.0004748338081671415,
"loss": 0.2067,
"step": 5360
},
{
"epoch": 17.184,
"grad_norm": 966854.0,
"learning_rate": 0.0004736467236467237,
"loss": 0.2142,
"step": 5370
},
{
"epoch": 17.216,
"grad_norm": 713046.25,
"learning_rate": 0.0004724596391263058,
"loss": 0.2114,
"step": 5380
},
{
"epoch": 17.248,
"grad_norm": 962303.5625,
"learning_rate": 0.00047127255460588797,
"loss": 0.2236,
"step": 5390
},
{
"epoch": 17.28,
"grad_norm": 1023706.8125,
"learning_rate": 0.0004700854700854701,
"loss": 0.1875,
"step": 5400
},
{
"epoch": 17.312,
"grad_norm": 950098.0625,
"learning_rate": 0.00046889838556505225,
"loss": 0.197,
"step": 5410
},
{
"epoch": 17.344,
"grad_norm": 904872.25,
"learning_rate": 0.00046771130104463437,
"loss": 0.2195,
"step": 5420
},
{
"epoch": 17.376,
"grad_norm": Infinity,
"learning_rate": 0.0004665242165242166,
"loss": 0.1911,
"step": 5430
},
{
"epoch": 17.408,
"grad_norm": Infinity,
"learning_rate": 0.0004653371320037987,
"loss": 0.2201,
"step": 5440
},
{
"epoch": 17.44,
"grad_norm": 922537.875,
"learning_rate": 0.0004641500474833808,
"loss": 0.2058,
"step": 5450
},
{
"epoch": 17.472,
"grad_norm": 1170017.125,
"learning_rate": 0.000462962962962963,
"loss": 0.2242,
"step": 5460
},
{
"epoch": 17.504,
"grad_norm": 967685.5,
"learning_rate": 0.0004617758784425451,
"loss": 0.2135,
"step": 5470
},
{
"epoch": 17.536,
"grad_norm": 866546.0,
"learning_rate": 0.00046058879392212726,
"loss": 0.206,
"step": 5480
},
{
"epoch": 17.568,
"grad_norm": 1577800.875,
"learning_rate": 0.0004594017094017094,
"loss": 0.2175,
"step": 5490
},
{
"epoch": 17.6,
"grad_norm": NaN,
"learning_rate": 0.00045821462488129154,
"loss": 0.2106,
"step": 5500
},
{
"epoch": 17.632,
"grad_norm": 1285103.625,
"learning_rate": 0.0004570275403608737,
"loss": 0.229,
"step": 5510
},
{
"epoch": 17.664,
"grad_norm": Infinity,
"learning_rate": 0.0004558404558404559,
"loss": 0.203,
"step": 5520
},
{
"epoch": 17.696,
"grad_norm": Infinity,
"learning_rate": 0.000454653371320038,
"loss": 0.2219,
"step": 5530
},
{
"epoch": 17.728,
"grad_norm": 1200600.875,
"learning_rate": 0.00045346628679962016,
"loss": 0.2182,
"step": 5540
},
{
"epoch": 17.76,
"grad_norm": 1468528.75,
"learning_rate": 0.00045227920227920227,
"loss": 0.1966,
"step": 5550
},
{
"epoch": 17.792,
"grad_norm": 955159.25,
"learning_rate": 0.00045109211775878444,
"loss": 0.241,
"step": 5560
},
{
"epoch": 17.824,
"grad_norm": 996693.75,
"learning_rate": 0.00044990503323836655,
"loss": 0.1966,
"step": 5570
},
{
"epoch": 17.856,
"grad_norm": 1332255.875,
"learning_rate": 0.0004487179487179487,
"loss": 0.2169,
"step": 5580
},
{
"epoch": 17.888,
"grad_norm": 1062334.375,
"learning_rate": 0.0004475308641975309,
"loss": 0.2109,
"step": 5590
},
{
"epoch": 17.92,
"grad_norm": Infinity,
"learning_rate": 0.00044634377967711305,
"loss": 0.2107,
"step": 5600
},
{
"epoch": 17.951999999999998,
"grad_norm": 916163.4375,
"learning_rate": 0.00044515669515669517,
"loss": 0.2343,
"step": 5610
},
{
"epoch": 17.984,
"grad_norm": 1330298.25,
"learning_rate": 0.00044396961063627733,
"loss": 0.2198,
"step": 5620
},
{
"epoch": 18.0,
"eval_accuracy": 0.77525,
"eval_loss": 0.5560150146484375,
"eval_runtime": 58.4205,
"eval_samples_per_second": 68.469,
"eval_steps_per_second": 2.14,
"step": 5625
},
{
"epoch": 18.016,
"grad_norm": 1105357.875,
"learning_rate": 0.00044278252611585945,
"loss": 0.1996,
"step": 5630
},
{
"epoch": 18.048,
"grad_norm": 1458348.125,
"learning_rate": 0.0004415954415954416,
"loss": 0.2199,
"step": 5640
},
{
"epoch": 18.08,
"grad_norm": 970556.25,
"learning_rate": 0.0004404083570750237,
"loss": 0.2191,
"step": 5650
},
{
"epoch": 18.112,
"grad_norm": Infinity,
"learning_rate": 0.00043922127255460595,
"loss": 0.2307,
"step": 5660
},
{
"epoch": 18.144,
"grad_norm": 976057.1875,
"learning_rate": 0.00043803418803418806,
"loss": 0.1966,
"step": 5670
},
{
"epoch": 18.176,
"grad_norm": 833928.25,
"learning_rate": 0.00043684710351377023,
"loss": 0.2268,
"step": 5680
},
{
"epoch": 18.208,
"grad_norm": 931736.9375,
"learning_rate": 0.00043566001899335234,
"loss": 0.2261,
"step": 5690
},
{
"epoch": 18.24,
"grad_norm": 1063323.25,
"learning_rate": 0.00043447293447293445,
"loss": 0.1793,
"step": 5700
},
{
"epoch": 18.272,
"grad_norm": Infinity,
"learning_rate": 0.0004332858499525166,
"loss": 0.2064,
"step": 5710
},
{
"epoch": 18.304,
"grad_norm": Infinity,
"learning_rate": 0.00043209876543209873,
"loss": 0.2174,
"step": 5720
},
{
"epoch": 18.336,
"grad_norm": 745713.25,
"learning_rate": 0.0004309116809116809,
"loss": 0.1981,
"step": 5730
},
{
"epoch": 18.368,
"grad_norm": 960917.375,
"learning_rate": 0.00042972459639126307,
"loss": 0.2114,
"step": 5740
},
{
"epoch": 18.4,
"grad_norm": Infinity,
"learning_rate": 0.00042853751187084524,
"loss": 0.2171,
"step": 5750
},
{
"epoch": 18.432,
"grad_norm": 825171.125,
"learning_rate": 0.00042735042735042735,
"loss": 0.1824,
"step": 5760
},
{
"epoch": 18.464,
"grad_norm": 769558.375,
"learning_rate": 0.0004261633428300095,
"loss": 0.2074,
"step": 5770
},
{
"epoch": 18.496,
"grad_norm": 772380.4375,
"learning_rate": 0.00042497625830959163,
"loss": 0.2226,
"step": 5780
},
{
"epoch": 18.528,
"grad_norm": 1635508.0,
"learning_rate": 0.0004237891737891738,
"loss": 0.2241,
"step": 5790
},
{
"epoch": 18.56,
"grad_norm": 977012.4375,
"learning_rate": 0.0004226020892687559,
"loss": 0.1885,
"step": 5800
},
{
"epoch": 18.592,
"grad_norm": 1104309.375,
"learning_rate": 0.0004214150047483381,
"loss": 0.2285,
"step": 5810
},
{
"epoch": 18.624,
"grad_norm": 1027584.875,
"learning_rate": 0.00042022792022792025,
"loss": 0.2192,
"step": 5820
},
{
"epoch": 18.656,
"grad_norm": 1208884.875,
"learning_rate": 0.0004190408357075024,
"loss": 0.2229,
"step": 5830
},
{
"epoch": 18.688,
"grad_norm": 1337715.5,
"learning_rate": 0.0004178537511870845,
"loss": 0.2235,
"step": 5840
},
{
"epoch": 18.72,
"grad_norm": 1082770.625,
"learning_rate": 0.0004166666666666667,
"loss": 0.1985,
"step": 5850
},
{
"epoch": 18.752,
"grad_norm": 1084099.75,
"learning_rate": 0.0004154795821462488,
"loss": 0.2001,
"step": 5860
},
{
"epoch": 18.784,
"grad_norm": 970450.75,
"learning_rate": 0.000414292497625831,
"loss": 0.2216,
"step": 5870
},
{
"epoch": 18.816,
"grad_norm": 811773.875,
"learning_rate": 0.0004131054131054131,
"loss": 0.2318,
"step": 5880
},
{
"epoch": 18.848,
"grad_norm": 839901.625,
"learning_rate": 0.0004119183285849953,
"loss": 0.2338,
"step": 5890
},
{
"epoch": 18.88,
"grad_norm": 1070109.25,
"learning_rate": 0.0004107312440645774,
"loss": 0.2,
"step": 5900
},
{
"epoch": 18.912,
"grad_norm": 795279.5,
"learning_rate": 0.0004095441595441596,
"loss": 0.1919,
"step": 5910
},
{
"epoch": 18.944,
"grad_norm": 1289021.25,
"learning_rate": 0.0004083570750237417,
"loss": 0.1973,
"step": 5920
},
{
"epoch": 18.976,
"grad_norm": 1276458.875,
"learning_rate": 0.00040716999050332387,
"loss": 0.2171,
"step": 5930
},
{
"epoch": 18.9984,
"eval_accuracy": 0.7725,
"eval_loss": 0.557004988193512,
"eval_runtime": 57.9489,
"eval_samples_per_second": 69.026,
"eval_steps_per_second": 2.157,
"step": 5937
},
{
"epoch": 19.008,
"grad_norm": Infinity,
"learning_rate": 0.000405982905982906,
"loss": 0.2159,
"step": 5940
},
{
"epoch": 19.04,
"grad_norm": 974318.875,
"learning_rate": 0.0004047958214624881,
"loss": 0.2077,
"step": 5950
},
{
"epoch": 19.072,
"grad_norm": 1039002.875,
"learning_rate": 0.00040360873694207026,
"loss": 0.2189,
"step": 5960
},
{
"epoch": 19.104,
"grad_norm": 852144.6875,
"learning_rate": 0.00040242165242165243,
"loss": 0.1993,
"step": 5970
},
{
"epoch": 19.136,
"grad_norm": 1005178.375,
"learning_rate": 0.0004012345679012346,
"loss": 0.2116,
"step": 5980
},
{
"epoch": 19.168,
"grad_norm": 1007026.75,
"learning_rate": 0.0004000474833808167,
"loss": 0.1993,
"step": 5990
},
{
"epoch": 19.2,
"grad_norm": 1615725.875,
"learning_rate": 0.0003988603988603989,
"loss": 0.2174,
"step": 6000
},
{
"epoch": 19.232,
"grad_norm": Infinity,
"learning_rate": 0.000397673314339981,
"loss": 0.2001,
"step": 6010
},
{
"epoch": 19.264,
"grad_norm": Infinity,
"learning_rate": 0.00039648622981956316,
"loss": 0.1909,
"step": 6020
},
{
"epoch": 19.296,
"grad_norm": Infinity,
"learning_rate": 0.00039529914529914527,
"loss": 0.1939,
"step": 6030
},
{
"epoch": 19.328,
"grad_norm": Infinity,
"learning_rate": 0.00039411206077872744,
"loss": 0.216,
"step": 6040
},
{
"epoch": 19.36,
"grad_norm": NaN,
"learning_rate": 0.0003929249762583096,
"loss": 0.2347,
"step": 6050
},
{
"epoch": 19.392,
"grad_norm": Infinity,
"learning_rate": 0.0003917378917378918,
"loss": 0.2205,
"step": 6060
},
{
"epoch": 19.424,
"grad_norm": Infinity,
"learning_rate": 0.0003905508072174739,
"loss": 0.2024,
"step": 6070
},
{
"epoch": 19.456,
"grad_norm": NaN,
"learning_rate": 0.00038936372269705605,
"loss": 0.2393,
"step": 6080
},
{
"epoch": 19.488,
"grad_norm": Infinity,
"learning_rate": 0.00038817663817663817,
"loss": 0.2286,
"step": 6090
},
{
"epoch": 19.52,
"grad_norm": Infinity,
"learning_rate": 0.00038698955365622033,
"loss": 0.223,
"step": 6100
},
{
"epoch": 19.552,
"grad_norm": Infinity,
"learning_rate": 0.00038580246913580245,
"loss": 0.2015,
"step": 6110
},
{
"epoch": 19.584,
"grad_norm": Infinity,
"learning_rate": 0.00038461538461538467,
"loss": 0.2163,
"step": 6120
},
{
"epoch": 19.616,
"grad_norm": 1314457.375,
"learning_rate": 0.0003834283000949668,
"loss": 0.2002,
"step": 6130
},
{
"epoch": 19.648,
"grad_norm": Infinity,
"learning_rate": 0.00038224121557454895,
"loss": 0.2215,
"step": 6140
},
{
"epoch": 19.68,
"grad_norm": NaN,
"learning_rate": 0.00038105413105413106,
"loss": 0.2117,
"step": 6150
},
{
"epoch": 19.712,
"grad_norm": NaN,
"learning_rate": 0.00037986704653371323,
"loss": 0.2196,
"step": 6160
},
{
"epoch": 19.744,
"grad_norm": Infinity,
"learning_rate": 0.00037867996201329534,
"loss": 0.2303,
"step": 6170
},
{
"epoch": 19.776,
"grad_norm": Infinity,
"learning_rate": 0.00037749287749287746,
"loss": 0.2374,
"step": 6180
},
{
"epoch": 19.808,
"grad_norm": NaN,
"learning_rate": 0.0003763057929724596,
"loss": 0.2177,
"step": 6190
},
{
"epoch": 19.84,
"grad_norm": NaN,
"learning_rate": 0.0003751187084520418,
"loss": 0.2005,
"step": 6200
},
{
"epoch": 19.872,
"grad_norm": Infinity,
"learning_rate": 0.00037393162393162396,
"loss": 0.2228,
"step": 6210
},
{
"epoch": 19.904,
"grad_norm": Infinity,
"learning_rate": 0.00037274453941120607,
"loss": 0.201,
"step": 6220
},
{
"epoch": 19.936,
"grad_norm": NaN,
"learning_rate": 0.00037155745489078824,
"loss": 0.205,
"step": 6230
},
{
"epoch": 19.968,
"grad_norm": 1964826.75,
"learning_rate": 0.00037037037037037035,
"loss": 0.2213,
"step": 6240
},
{
"epoch": 20.0,
"grad_norm": Infinity,
"learning_rate": 0.0003691832858499525,
"loss": 0.2116,
"step": 6250
},
{
"epoch": 20.0,
"eval_accuracy": 0.7625,
"eval_loss": 0.5622299909591675,
"eval_runtime": 58.6144,
"eval_samples_per_second": 68.243,
"eval_steps_per_second": 2.133,
"step": 6250
},
{
"epoch": 20.032,
"grad_norm": NaN,
"learning_rate": 0.00036799620132953463,
"loss": 0.2141,
"step": 6260
},
{
"epoch": 20.064,
"grad_norm": Infinity,
"learning_rate": 0.00036680911680911685,
"loss": 0.2003,
"step": 6270
},
{
"epoch": 20.096,
"grad_norm": Infinity,
"learning_rate": 0.00036562203228869897,
"loss": 0.2081,
"step": 6280
},
{
"epoch": 20.128,
"grad_norm": 1561355.625,
"learning_rate": 0.00036443494776828113,
"loss": 0.2078,
"step": 6290
},
{
"epoch": 20.16,
"grad_norm": Infinity,
"learning_rate": 0.00036324786324786325,
"loss": 0.2105,
"step": 6300
},
{
"epoch": 20.192,
"grad_norm": Infinity,
"learning_rate": 0.0003620607787274454,
"loss": 0.2395,
"step": 6310
},
{
"epoch": 20.224,
"grad_norm": Infinity,
"learning_rate": 0.00036087369420702753,
"loss": 0.2105,
"step": 6320
},
{
"epoch": 20.256,
"grad_norm": Infinity,
"learning_rate": 0.0003596866096866097,
"loss": 0.216,
"step": 6330
},
{
"epoch": 20.288,
"grad_norm": Infinity,
"learning_rate": 0.0003584995251661918,
"loss": 0.224,
"step": 6340
},
{
"epoch": 20.32,
"grad_norm": Infinity,
"learning_rate": 0.00035731244064577403,
"loss": 0.2054,
"step": 6350
},
{
"epoch": 20.352,
"grad_norm": 1508851.0,
"learning_rate": 0.00035612535612535614,
"loss": 0.2151,
"step": 6360
},
{
"epoch": 20.384,
"grad_norm": 1946181.375,
"learning_rate": 0.0003549382716049383,
"loss": 0.2173,
"step": 6370
},
{
"epoch": 20.416,
"grad_norm": Infinity,
"learning_rate": 0.0003537511870845204,
"loss": 0.2035,
"step": 6380
},
{
"epoch": 20.448,
"grad_norm": Infinity,
"learning_rate": 0.0003525641025641026,
"loss": 0.2215,
"step": 6390
},
{
"epoch": 20.48,
"grad_norm": Infinity,
"learning_rate": 0.0003513770180436847,
"loss": 0.205,
"step": 6400
},
{
"epoch": 20.512,
"grad_norm": Infinity,
"learning_rate": 0.00035018993352326687,
"loss": 0.1974,
"step": 6410
},
{
"epoch": 20.544,
"grad_norm": 1950701.25,
"learning_rate": 0.000349002849002849,
"loss": 0.2087,
"step": 6420
},
{
"epoch": 20.576,
"grad_norm": Infinity,
"learning_rate": 0.00034781576448243115,
"loss": 0.2136,
"step": 6430
},
{
"epoch": 20.608,
"grad_norm": NaN,
"learning_rate": 0.0003466286799620133,
"loss": 0.2051,
"step": 6440
},
{
"epoch": 20.64,
"grad_norm": Infinity,
"learning_rate": 0.00034544159544159543,
"loss": 0.2181,
"step": 6450
},
{
"epoch": 20.672,
"grad_norm": Infinity,
"learning_rate": 0.0003442545109211776,
"loss": 0.2367,
"step": 6460
},
{
"epoch": 20.704,
"grad_norm": Infinity,
"learning_rate": 0.0003430674264007597,
"loss": 0.1962,
"step": 6470
},
{
"epoch": 20.736,
"grad_norm": 1704853.25,
"learning_rate": 0.0003418803418803419,
"loss": 0.201,
"step": 6480
},
{
"epoch": 20.768,
"grad_norm": Infinity,
"learning_rate": 0.000340693257359924,
"loss": 0.2186,
"step": 6490
},
{
"epoch": 20.8,
"grad_norm": Infinity,
"learning_rate": 0.0003395061728395062,
"loss": 0.2131,
"step": 6500
},
{
"epoch": 20.832,
"grad_norm": Infinity,
"learning_rate": 0.00033831908831908833,
"loss": 0.223,
"step": 6510
},
{
"epoch": 20.864,
"grad_norm": 1689417.75,
"learning_rate": 0.0003371320037986705,
"loss": 0.2086,
"step": 6520
},
{
"epoch": 20.896,
"grad_norm": Infinity,
"learning_rate": 0.0003359449192782526,
"loss": 0.2109,
"step": 6530
},
{
"epoch": 20.928,
"grad_norm": Infinity,
"learning_rate": 0.0003347578347578348,
"loss": 0.2014,
"step": 6540
},
{
"epoch": 20.96,
"grad_norm": Infinity,
"learning_rate": 0.0003335707502374169,
"loss": 0.2174,
"step": 6550
},
{
"epoch": 20.992,
"grad_norm": Infinity,
"learning_rate": 0.00033238366571699906,
"loss": 0.2162,
"step": 6560
},
{
"epoch": 20.9984,
"eval_accuracy": 0.76675,
"eval_loss": 0.5587143898010254,
"eval_runtime": 59.4419,
"eval_samples_per_second": 67.293,
"eval_steps_per_second": 2.103,
"step": 6562
},
{
"epoch": 21.024,
"grad_norm": 1291432.5,
"learning_rate": 0.00033119658119658117,
"loss": 0.2018,
"step": 6570
},
{
"epoch": 21.056,
"grad_norm": NaN,
"learning_rate": 0.0003300094966761634,
"loss": 0.2017,
"step": 6580
},
{
"epoch": 21.088,
"grad_norm": Infinity,
"learning_rate": 0.0003288224121557455,
"loss": 0.2061,
"step": 6590
},
{
"epoch": 21.12,
"grad_norm": NaN,
"learning_rate": 0.00032763532763532767,
"loss": 0.2068,
"step": 6600
},
{
"epoch": 21.152,
"grad_norm": Infinity,
"learning_rate": 0.0003264482431149098,
"loss": 0.2055,
"step": 6610
},
{
"epoch": 21.184,
"grad_norm": Infinity,
"learning_rate": 0.00032526115859449195,
"loss": 0.2069,
"step": 6620
},
{
"epoch": 21.216,
"grad_norm": 1809235.625,
"learning_rate": 0.00032407407407407406,
"loss": 0.1952,
"step": 6630
},
{
"epoch": 21.248,
"grad_norm": NaN,
"learning_rate": 0.00032288698955365623,
"loss": 0.2042,
"step": 6640
},
{
"epoch": 21.28,
"grad_norm": Infinity,
"learning_rate": 0.00032169990503323835,
"loss": 0.2071,
"step": 6650
},
{
"epoch": 21.312,
"grad_norm": Infinity,
"learning_rate": 0.00032051282051282057,
"loss": 0.2131,
"step": 6660
},
{
"epoch": 21.344,
"grad_norm": Infinity,
"learning_rate": 0.0003193257359924027,
"loss": 0.2297,
"step": 6670
},
{
"epoch": 21.376,
"grad_norm": NaN,
"learning_rate": 0.0003181386514719848,
"loss": 0.2083,
"step": 6680
},
{
"epoch": 21.408,
"grad_norm": 1783178.125,
"learning_rate": 0.00031695156695156696,
"loss": 0.1937,
"step": 6690
},
{
"epoch": 21.44,
"grad_norm": NaN,
"learning_rate": 0.0003157644824311491,
"loss": 0.2248,
"step": 6700
},
{
"epoch": 21.472,
"grad_norm": Infinity,
"learning_rate": 0.00031457739791073124,
"loss": 0.2114,
"step": 6710
},
{
"epoch": 21.504,
"grad_norm": Infinity,
"learning_rate": 0.00031339031339031335,
"loss": 0.2521,
"step": 6720
},
{
"epoch": 21.536,
"grad_norm": NaN,
"learning_rate": 0.0003122032288698956,
"loss": 0.2058,
"step": 6730
},
{
"epoch": 21.568,
"grad_norm": NaN,
"learning_rate": 0.0003110161443494777,
"loss": 0.1923,
"step": 6740
},
{
"epoch": 21.6,
"grad_norm": Infinity,
"learning_rate": 0.00030982905982905986,
"loss": 0.2405,
"step": 6750
},
{
"epoch": 21.632,
"grad_norm": 1427268.125,
"learning_rate": 0.00030864197530864197,
"loss": 0.187,
"step": 6760
},
{
"epoch": 21.664,
"grad_norm": 1644058.125,
"learning_rate": 0.00030745489078822414,
"loss": 0.2343,
"step": 6770
},
{
"epoch": 21.696,
"grad_norm": 2057941.25,
"learning_rate": 0.00030626780626780625,
"loss": 0.2114,
"step": 6780
},
{
"epoch": 21.728,
"grad_norm": Infinity,
"learning_rate": 0.0003050807217473884,
"loss": 0.2144,
"step": 6790
},
{
"epoch": 21.76,
"grad_norm": Infinity,
"learning_rate": 0.00030389363722697053,
"loss": 0.2318,
"step": 6800
},
{
"epoch": 21.792,
"grad_norm": Infinity,
"learning_rate": 0.00030270655270655275,
"loss": 0.2327,
"step": 6810
},
{
"epoch": 21.824,
"grad_norm": Infinity,
"learning_rate": 0.00030151946818613486,
"loss": 0.2207,
"step": 6820
},
{
"epoch": 21.856,
"grad_norm": Infinity,
"learning_rate": 0.00030033238366571703,
"loss": 0.2115,
"step": 6830
},
{
"epoch": 21.888,
"grad_norm": Infinity,
"learning_rate": 0.00029914529914529915,
"loss": 0.1802,
"step": 6840
},
{
"epoch": 21.92,
"grad_norm": Infinity,
"learning_rate": 0.0002979582146248813,
"loss": 0.211,
"step": 6850
},
{
"epoch": 21.951999999999998,
"grad_norm": Infinity,
"learning_rate": 0.0002967711301044634,
"loss": 0.2275,
"step": 6860
},
{
"epoch": 21.984,
"grad_norm": Infinity,
"learning_rate": 0.0002955840455840456,
"loss": 0.224,
"step": 6870
},
{
"epoch": 22.0,
"eval_accuracy": 0.77125,
"eval_loss": 0.545555591583252,
"eval_runtime": 58.3787,
"eval_samples_per_second": 68.518,
"eval_steps_per_second": 2.141,
"step": 6875
},
{
"epoch": 22.016,
"grad_norm": 2120620.25,
"learning_rate": 0.0002943969610636277,
"loss": 0.2007,
"step": 6880
},
{
"epoch": 22.048,
"grad_norm": Infinity,
"learning_rate": 0.00029320987654320993,
"loss": 0.2052,
"step": 6890
},
{
"epoch": 22.08,
"grad_norm": NaN,
"learning_rate": 0.00029202279202279204,
"loss": 0.2109,
"step": 6900
},
{
"epoch": 22.112,
"grad_norm": Infinity,
"learning_rate": 0.0002908357075023742,
"loss": 0.1884,
"step": 6910
},
{
"epoch": 22.144,
"grad_norm": Infinity,
"learning_rate": 0.0002896486229819563,
"loss": 0.2099,
"step": 6920
},
{
"epoch": 22.176,
"grad_norm": Infinity,
"learning_rate": 0.00028846153846153843,
"loss": 0.2147,
"step": 6930
},
{
"epoch": 22.208,
"grad_norm": Infinity,
"learning_rate": 0.0002872744539411206,
"loss": 0.2116,
"step": 6940
},
{
"epoch": 22.24,
"grad_norm": Infinity,
"learning_rate": 0.0002860873694207027,
"loss": 0.2044,
"step": 6950
},
{
"epoch": 22.272,
"grad_norm": 1670116.875,
"learning_rate": 0.00028490028490028494,
"loss": 0.2107,
"step": 6960
},
{
"epoch": 22.304,
"grad_norm": Infinity,
"learning_rate": 0.00028371320037986705,
"loss": 0.1955,
"step": 6970
},
{
"epoch": 22.336,
"grad_norm": Infinity,
"learning_rate": 0.0002825261158594492,
"loss": 0.214,
"step": 6980
},
{
"epoch": 22.368,
"grad_norm": Infinity,
"learning_rate": 0.00028133903133903133,
"loss": 0.2164,
"step": 6990
},
{
"epoch": 22.4,
"grad_norm": Infinity,
"learning_rate": 0.0002801519468186135,
"loss": 0.2199,
"step": 7000
},
{
"epoch": 22.432,
"grad_norm": Infinity,
"learning_rate": 0.0002789648622981956,
"loss": 0.2066,
"step": 7010
},
{
"epoch": 22.464,
"grad_norm": 2138083.0,
"learning_rate": 0.0002777777777777778,
"loss": 0.2208,
"step": 7020
},
{
"epoch": 22.496,
"grad_norm": Infinity,
"learning_rate": 0.0002765906932573599,
"loss": 0.2207,
"step": 7030
},
{
"epoch": 22.528,
"grad_norm": Infinity,
"learning_rate": 0.0002754036087369421,
"loss": 0.2121,
"step": 7040
},
{
"epoch": 22.56,
"grad_norm": Infinity,
"learning_rate": 0.0002742165242165242,
"loss": 0.2049,
"step": 7050
},
{
"epoch": 22.592,
"grad_norm": Infinity,
"learning_rate": 0.0002730294396961064,
"loss": 0.2268,
"step": 7060
},
{
"epoch": 22.624,
"grad_norm": Infinity,
"learning_rate": 0.0002718423551756885,
"loss": 0.2251,
"step": 7070
},
{
"epoch": 22.656,
"grad_norm": 1488781.25,
"learning_rate": 0.0002706552706552707,
"loss": 0.2041,
"step": 7080
},
{
"epoch": 22.688,
"grad_norm": Infinity,
"learning_rate": 0.0002694681861348528,
"loss": 0.2164,
"step": 7090
},
{
"epoch": 22.72,
"grad_norm": Infinity,
"learning_rate": 0.00026828110161443495,
"loss": 0.2033,
"step": 7100
},
{
"epoch": 22.752,
"grad_norm": Infinity,
"learning_rate": 0.0002670940170940171,
"loss": 0.1891,
"step": 7110
},
{
"epoch": 22.784,
"grad_norm": Infinity,
"learning_rate": 0.0002659069325735993,
"loss": 0.2017,
"step": 7120
},
{
"epoch": 22.816,
"grad_norm": Infinity,
"learning_rate": 0.0002647198480531814,
"loss": 0.234,
"step": 7130
},
{
"epoch": 22.848,
"grad_norm": Infinity,
"learning_rate": 0.00026353276353276357,
"loss": 0.2086,
"step": 7140
},
{
"epoch": 22.88,
"grad_norm": Infinity,
"learning_rate": 0.0002623456790123457,
"loss": 0.2041,
"step": 7150
},
{
"epoch": 22.912,
"grad_norm": Infinity,
"learning_rate": 0.0002611585944919278,
"loss": 0.2287,
"step": 7160
},
{
"epoch": 22.944,
"grad_norm": NaN,
"learning_rate": 0.00025997150997150996,
"loss": 0.2055,
"step": 7170
},
{
"epoch": 22.976,
"grad_norm": Infinity,
"learning_rate": 0.0002587844254510921,
"loss": 0.212,
"step": 7180
},
{
"epoch": 22.9984,
"eval_accuracy": 0.76525,
"eval_loss": 0.5647286772727966,
"eval_runtime": 58.2038,
"eval_samples_per_second": 68.724,
"eval_steps_per_second": 2.148,
"step": 7187
},
{
"epoch": 23.008,
"grad_norm": Infinity,
"learning_rate": 0.0002575973409306743,
"loss": 0.2204,
"step": 7190
},
{
"epoch": 23.04,
"grad_norm": Infinity,
"learning_rate": 0.0002564102564102564,
"loss": 0.2154,
"step": 7200
},
{
"epoch": 23.072,
"grad_norm": Infinity,
"learning_rate": 0.0002552231718898386,
"loss": 0.2123,
"step": 7210
},
{
"epoch": 23.104,
"grad_norm": Infinity,
"learning_rate": 0.0002540360873694207,
"loss": 0.2144,
"step": 7220
},
{
"epoch": 23.136,
"grad_norm": Infinity,
"learning_rate": 0.00025284900284900286,
"loss": 0.193,
"step": 7230
},
{
"epoch": 23.168,
"grad_norm": 1738463.0,
"learning_rate": 0.00025166191832858497,
"loss": 0.1986,
"step": 7240
},
{
"epoch": 23.2,
"grad_norm": Infinity,
"learning_rate": 0.00025047483380816714,
"loss": 0.2222,
"step": 7250
},
{
"epoch": 23.232,
"grad_norm": Infinity,
"learning_rate": 0.0002492877492877493,
"loss": 0.2246,
"step": 7260
},
{
"epoch": 23.264,
"grad_norm": NaN,
"learning_rate": 0.0002481006647673315,
"loss": 0.2103,
"step": 7270
},
{
"epoch": 23.296,
"grad_norm": Infinity,
"learning_rate": 0.0002469135802469136,
"loss": 0.208,
"step": 7280
},
{
"epoch": 23.328,
"grad_norm": Infinity,
"learning_rate": 0.0002457264957264957,
"loss": 0.2187,
"step": 7290
},
{
"epoch": 23.36,
"grad_norm": Infinity,
"learning_rate": 0.00024453941120607787,
"loss": 0.211,
"step": 7300
},
{
"epoch": 23.392,
"grad_norm": 1851251.625,
"learning_rate": 0.00024335232668566,
"loss": 0.1992,
"step": 7310
},
{
"epoch": 23.424,
"grad_norm": Infinity,
"learning_rate": 0.00024216524216524217,
"loss": 0.1985,
"step": 7320
},
{
"epoch": 23.456,
"grad_norm": Infinity,
"learning_rate": 0.00024097815764482431,
"loss": 0.2002,
"step": 7330
},
{
"epoch": 23.488,
"grad_norm": Infinity,
"learning_rate": 0.00023979107312440645,
"loss": 0.2025,
"step": 7340
},
{
"epoch": 23.52,
"grad_norm": Infinity,
"learning_rate": 0.0002386039886039886,
"loss": 0.1979,
"step": 7350
},
{
"epoch": 23.552,
"grad_norm": Infinity,
"learning_rate": 0.00023741690408357076,
"loss": 0.1975,
"step": 7360
},
{
"epoch": 23.584,
"grad_norm": Infinity,
"learning_rate": 0.0002362298195631529,
"loss": 0.2347,
"step": 7370
},
{
"epoch": 23.616,
"grad_norm": Infinity,
"learning_rate": 0.00023504273504273504,
"loss": 0.2143,
"step": 7380
},
{
"epoch": 23.648,
"grad_norm": Infinity,
"learning_rate": 0.00023385565052231718,
"loss": 0.2299,
"step": 7390
},
{
"epoch": 23.68,
"grad_norm": Infinity,
"learning_rate": 0.00023266856600189935,
"loss": 0.2,
"step": 7400
},
{
"epoch": 23.712,
"grad_norm": Infinity,
"learning_rate": 0.0002314814814814815,
"loss": 0.2225,
"step": 7410
},
{
"epoch": 23.744,
"grad_norm": Infinity,
"learning_rate": 0.00023029439696106363,
"loss": 0.2049,
"step": 7420
},
{
"epoch": 23.776,
"grad_norm": Infinity,
"learning_rate": 0.00022910731244064577,
"loss": 0.2258,
"step": 7430
},
{
"epoch": 23.808,
"grad_norm": Infinity,
"learning_rate": 0.00022792022792022794,
"loss": 0.2008,
"step": 7440
},
{
"epoch": 23.84,
"grad_norm": Infinity,
"learning_rate": 0.00022673314339981008,
"loss": 0.2079,
"step": 7450
},
{
"epoch": 23.872,
"grad_norm": Infinity,
"learning_rate": 0.00022554605887939222,
"loss": 0.1949,
"step": 7460
},
{
"epoch": 23.904,
"grad_norm": Infinity,
"learning_rate": 0.00022435897435897436,
"loss": 0.1853,
"step": 7470
},
{
"epoch": 23.936,
"grad_norm": NaN,
"learning_rate": 0.00022317188983855653,
"loss": 0.2076,
"step": 7480
},
{
"epoch": 23.968,
"grad_norm": Infinity,
"learning_rate": 0.00022198480531813867,
"loss": 0.2082,
"step": 7490
},
{
"epoch": 24.0,
"grad_norm": Infinity,
"learning_rate": 0.0002207977207977208,
"loss": 0.2084,
"step": 7500
},
{
"epoch": 24.0,
"eval_accuracy": 0.76725,
"eval_loss": 0.5532636046409607,
"eval_runtime": 58.6171,
"eval_samples_per_second": 68.239,
"eval_steps_per_second": 2.132,
"step": 7500
},
{
"epoch": 24.032,
"grad_norm": Infinity,
"learning_rate": 0.00021961063627730297,
"loss": 0.2117,
"step": 7510
},
{
"epoch": 24.064,
"grad_norm": NaN,
"learning_rate": 0.00021842355175688511,
"loss": 0.2147,
"step": 7520
},
{
"epoch": 24.096,
"grad_norm": Infinity,
"learning_rate": 0.00021723646723646723,
"loss": 0.2179,
"step": 7530
},
{
"epoch": 24.128,
"grad_norm": Infinity,
"learning_rate": 0.00021604938271604937,
"loss": 0.1879,
"step": 7540
},
{
"epoch": 24.16,
"grad_norm": Infinity,
"learning_rate": 0.00021486229819563153,
"loss": 0.1989,
"step": 7550
},
{
"epoch": 24.192,
"grad_norm": Infinity,
"learning_rate": 0.00021367521367521368,
"loss": 0.2281,
"step": 7560
},
{
"epoch": 24.224,
"grad_norm": Infinity,
"learning_rate": 0.00021248812915479582,
"loss": 0.2159,
"step": 7570
},
{
"epoch": 24.256,
"grad_norm": Infinity,
"learning_rate": 0.00021130104463437796,
"loss": 0.2186,
"step": 7580
},
{
"epoch": 24.288,
"grad_norm": Infinity,
"learning_rate": 0.00021011396011396012,
"loss": 0.2257,
"step": 7590
},
{
"epoch": 24.32,
"grad_norm": Infinity,
"learning_rate": 0.00020892687559354226,
"loss": 0.2098,
"step": 7600
},
{
"epoch": 24.352,
"grad_norm": 1342396.875,
"learning_rate": 0.0002077397910731244,
"loss": 0.2101,
"step": 7610
},
{
"epoch": 24.384,
"grad_norm": Infinity,
"learning_rate": 0.00020655270655270654,
"loss": 0.213,
"step": 7620
},
{
"epoch": 24.416,
"grad_norm": Infinity,
"learning_rate": 0.0002053656220322887,
"loss": 0.211,
"step": 7630
},
{
"epoch": 24.448,
"grad_norm": NaN,
"learning_rate": 0.00020417853751187085,
"loss": 0.187,
"step": 7640
},
{
"epoch": 24.48,
"grad_norm": Infinity,
"learning_rate": 0.000202991452991453,
"loss": 0.2285,
"step": 7650
},
{
"epoch": 24.512,
"grad_norm": 1773714.0,
"learning_rate": 0.00020180436847103513,
"loss": 0.2177,
"step": 7660
},
{
"epoch": 24.544,
"grad_norm": Infinity,
"learning_rate": 0.0002006172839506173,
"loss": 0.2064,
"step": 7670
},
{
"epoch": 24.576,
"grad_norm": Infinity,
"learning_rate": 0.00019943019943019944,
"loss": 0.2344,
"step": 7680
},
{
"epoch": 24.608,
"grad_norm": Infinity,
"learning_rate": 0.00019824311490978158,
"loss": 0.2185,
"step": 7690
},
{
"epoch": 24.64,
"grad_norm": Infinity,
"learning_rate": 0.00019705603038936372,
"loss": 0.2115,
"step": 7700
},
{
"epoch": 24.672,
"grad_norm": Infinity,
"learning_rate": 0.0001958689458689459,
"loss": 0.2214,
"step": 7710
},
{
"epoch": 24.704,
"grad_norm": Infinity,
"learning_rate": 0.00019468186134852803,
"loss": 0.2071,
"step": 7720
},
{
"epoch": 24.736,
"grad_norm": Infinity,
"learning_rate": 0.00019349477682811017,
"loss": 0.2109,
"step": 7730
},
{
"epoch": 24.768,
"grad_norm": Infinity,
"learning_rate": 0.00019230769230769233,
"loss": 0.195,
"step": 7740
},
{
"epoch": 24.8,
"grad_norm": 1795345.25,
"learning_rate": 0.00019112060778727447,
"loss": 0.2347,
"step": 7750
},
{
"epoch": 24.832,
"grad_norm": Infinity,
"learning_rate": 0.00018993352326685662,
"loss": 0.2028,
"step": 7760
},
{
"epoch": 24.864,
"grad_norm": Infinity,
"learning_rate": 0.00018874643874643873,
"loss": 0.231,
"step": 7770
},
{
"epoch": 24.896,
"grad_norm": Infinity,
"learning_rate": 0.0001875593542260209,
"loss": 0.2069,
"step": 7780
},
{
"epoch": 24.928,
"grad_norm": Infinity,
"learning_rate": 0.00018637226970560304,
"loss": 0.1928,
"step": 7790
},
{
"epoch": 24.96,
"grad_norm": 1809381.25,
"learning_rate": 0.00018518518518518518,
"loss": 0.1927,
"step": 7800
},
{
"epoch": 24.992,
"grad_norm": Infinity,
"learning_rate": 0.00018399810066476732,
"loss": 0.2226,
"step": 7810
},
{
"epoch": 24.9984,
"eval_accuracy": 0.7705,
"eval_loss": 0.5434042811393738,
"eval_runtime": 59.0686,
"eval_samples_per_second": 67.718,
"eval_steps_per_second": 2.116,
"step": 7812
},
{
"epoch": 25.024,
"grad_norm": Infinity,
"learning_rate": 0.00018281101614434948,
"loss": 0.2133,
"step": 7820
},
{
"epoch": 25.056,
"grad_norm": Infinity,
"learning_rate": 0.00018162393162393162,
"loss": 0.2168,
"step": 7830
},
{
"epoch": 25.088,
"grad_norm": Infinity,
"learning_rate": 0.00018043684710351376,
"loss": 0.2239,
"step": 7840
},
{
"epoch": 25.12,
"grad_norm": Infinity,
"learning_rate": 0.0001792497625830959,
"loss": 0.2393,
"step": 7850
},
{
"epoch": 25.152,
"grad_norm": Infinity,
"learning_rate": 0.00017806267806267807,
"loss": 0.192,
"step": 7860
},
{
"epoch": 25.184,
"grad_norm": Infinity,
"learning_rate": 0.0001768755935422602,
"loss": 0.2059,
"step": 7870
},
{
"epoch": 25.216,
"grad_norm": Infinity,
"learning_rate": 0.00017568850902184235,
"loss": 0.2137,
"step": 7880
},
{
"epoch": 25.248,
"grad_norm": Infinity,
"learning_rate": 0.0001745014245014245,
"loss": 0.2123,
"step": 7890
},
{
"epoch": 25.28,
"grad_norm": Infinity,
"learning_rate": 0.00017331433998100666,
"loss": 0.2252,
"step": 7900
},
{
"epoch": 25.312,
"grad_norm": 1587581.25,
"learning_rate": 0.0001721272554605888,
"loss": 0.2132,
"step": 7910
},
{
"epoch": 25.344,
"grad_norm": Infinity,
"learning_rate": 0.00017094017094017094,
"loss": 0.2104,
"step": 7920
},
{
"epoch": 25.376,
"grad_norm": NaN,
"learning_rate": 0.0001697530864197531,
"loss": 0.1865,
"step": 7930
},
{
"epoch": 25.408,
"grad_norm": Infinity,
"learning_rate": 0.00016856600189933525,
"loss": 0.2034,
"step": 7940
},
{
"epoch": 25.44,
"grad_norm": 1961501.125,
"learning_rate": 0.0001673789173789174,
"loss": 0.2157,
"step": 7950
},
{
"epoch": 25.472,
"grad_norm": Infinity,
"learning_rate": 0.00016619183285849953,
"loss": 0.2035,
"step": 7960
},
{
"epoch": 25.504,
"grad_norm": 1725835.5,
"learning_rate": 0.0001650047483380817,
"loss": 0.2063,
"step": 7970
},
{
"epoch": 25.536,
"grad_norm": Infinity,
"learning_rate": 0.00016381766381766384,
"loss": 0.2128,
"step": 7980
},
{
"epoch": 25.568,
"grad_norm": NaN,
"learning_rate": 0.00016263057929724598,
"loss": 0.2029,
"step": 7990
},
{
"epoch": 25.6,
"grad_norm": Infinity,
"learning_rate": 0.00016144349477682812,
"loss": 0.2183,
"step": 8000
},
{
"epoch": 25.632,
"grad_norm": NaN,
"learning_rate": 0.00016025641025641028,
"loss": 0.2269,
"step": 8010
},
{
"epoch": 25.664,
"grad_norm": Infinity,
"learning_rate": 0.0001590693257359924,
"loss": 0.2039,
"step": 8020
},
{
"epoch": 25.696,
"grad_norm": NaN,
"learning_rate": 0.00015788224121557454,
"loss": 0.2344,
"step": 8030
},
{
"epoch": 25.728,
"grad_norm": Infinity,
"learning_rate": 0.00015669515669515668,
"loss": 0.1994,
"step": 8040
},
{
"epoch": 25.76,
"grad_norm": NaN,
"learning_rate": 0.00015550807217473884,
"loss": 0.2006,
"step": 8050
},
{
"epoch": 25.792,
"grad_norm": NaN,
"learning_rate": 0.00015432098765432098,
"loss": 0.2082,
"step": 8060
},
{
"epoch": 25.824,
"grad_norm": NaN,
"learning_rate": 0.00015313390313390312,
"loss": 0.1903,
"step": 8070
},
{
"epoch": 25.856,
"grad_norm": NaN,
"learning_rate": 0.00015194681861348526,
"loss": 0.2106,
"step": 8080
},
{
"epoch": 25.888,
"grad_norm": NaN,
"learning_rate": 0.00015075973409306743,
"loss": 0.2059,
"step": 8090
},
{
"epoch": 25.92,
"grad_norm": NaN,
"learning_rate": 0.00014957264957264957,
"loss": 0.2104,
"step": 8100
},
{
"epoch": 25.951999999999998,
"grad_norm": NaN,
"learning_rate": 0.0001483855650522317,
"loss": 0.2069,
"step": 8110
},
{
"epoch": 25.984,
"grad_norm": Infinity,
"learning_rate": 0.00014719848053181385,
"loss": 0.2173,
"step": 8120
},
{
"epoch": 26.0,
"eval_accuracy": 0.7675,
"eval_loss": 0.5737802386283875,
"eval_runtime": 60.4948,
"eval_samples_per_second": 66.121,
"eval_steps_per_second": 2.066,
"step": 8125
},
{
"epoch": 26.016,
"grad_norm": NaN,
"learning_rate": 0.00014601139601139602,
"loss": 0.2187,
"step": 8130
},
{
"epoch": 26.048,
"grad_norm": NaN,
"learning_rate": 0.00014482431149097816,
"loss": 0.2001,
"step": 8140
},
{
"epoch": 26.08,
"grad_norm": NaN,
"learning_rate": 0.0001436372269705603,
"loss": 0.2392,
"step": 8150
},
{
"epoch": 26.112,
"grad_norm": NaN,
"learning_rate": 0.00014245014245014247,
"loss": 0.2125,
"step": 8160
},
{
"epoch": 26.144,
"grad_norm": NaN,
"learning_rate": 0.0001412630579297246,
"loss": 0.2118,
"step": 8170
},
{
"epoch": 26.176,
"grad_norm": NaN,
"learning_rate": 0.00014007597340930675,
"loss": 0.2093,
"step": 8180
},
{
"epoch": 26.208,
"grad_norm": NaN,
"learning_rate": 0.0001388888888888889,
"loss": 0.2059,
"step": 8190
},
{
"epoch": 26.24,
"grad_norm": NaN,
"learning_rate": 0.00013770180436847106,
"loss": 0.2307,
"step": 8200
},
{
"epoch": 26.272,
"grad_norm": NaN,
"learning_rate": 0.0001365147198480532,
"loss": 0.199,
"step": 8210
},
{
"epoch": 26.304,
"grad_norm": NaN,
"learning_rate": 0.00013532763532763534,
"loss": 0.2089,
"step": 8220
},
{
"epoch": 26.336,
"grad_norm": NaN,
"learning_rate": 0.00013414055080721748,
"loss": 0.2205,
"step": 8230
},
{
"epoch": 26.368,
"grad_norm": NaN,
"learning_rate": 0.00013295346628679964,
"loss": 0.2175,
"step": 8240
},
{
"epoch": 26.4,
"grad_norm": Infinity,
"learning_rate": 0.00013176638176638178,
"loss": 0.2002,
"step": 8250
},
{
"epoch": 26.432,
"grad_norm": NaN,
"learning_rate": 0.0001305792972459639,
"loss": 0.2078,
"step": 8260
},
{
"epoch": 26.464,
"grad_norm": Infinity,
"learning_rate": 0.00012939221272554604,
"loss": 0.2287,
"step": 8270
},
{
"epoch": 26.496,
"grad_norm": NaN,
"learning_rate": 0.0001282051282051282,
"loss": 0.2089,
"step": 8280
},
{
"epoch": 26.528,
"grad_norm": NaN,
"learning_rate": 0.00012701804368471035,
"loss": 0.2076,
"step": 8290
},
{
"epoch": 26.56,
"grad_norm": NaN,
"learning_rate": 0.00012583095916429249,
"loss": 0.2157,
"step": 8300
},
{
"epoch": 26.592,
"grad_norm": NaN,
"learning_rate": 0.00012464387464387465,
"loss": 0.2178,
"step": 8310
},
{
"epoch": 26.624,
"grad_norm": Infinity,
"learning_rate": 0.0001234567901234568,
"loss": 0.2168,
"step": 8320
},
{
"epoch": 26.656,
"grad_norm": NaN,
"learning_rate": 0.00012226970560303893,
"loss": 0.2004,
"step": 8330
},
{
"epoch": 26.688,
"grad_norm": NaN,
"learning_rate": 0.00012108262108262109,
"loss": 0.2131,
"step": 8340
},
{
"epoch": 26.72,
"grad_norm": NaN,
"learning_rate": 0.00011989553656220323,
"loss": 0.2059,
"step": 8350
},
{
"epoch": 26.752,
"grad_norm": NaN,
"learning_rate": 0.00011870845204178538,
"loss": 0.2218,
"step": 8360
},
{
"epoch": 26.784,
"grad_norm": NaN,
"learning_rate": 0.00011752136752136752,
"loss": 0.1975,
"step": 8370
},
{
"epoch": 26.816,
"grad_norm": NaN,
"learning_rate": 0.00011633428300094968,
"loss": 0.2113,
"step": 8380
},
{
"epoch": 26.848,
"grad_norm": NaN,
"learning_rate": 0.00011514719848053182,
"loss": 0.2346,
"step": 8390
},
{
"epoch": 26.88,
"grad_norm": NaN,
"learning_rate": 0.00011396011396011397,
"loss": 0.2067,
"step": 8400
},
{
"epoch": 26.912,
"grad_norm": NaN,
"learning_rate": 0.00011277302943969611,
"loss": 0.2158,
"step": 8410
},
{
"epoch": 26.944,
"grad_norm": NaN,
"learning_rate": 0.00011158594491927826,
"loss": 0.2046,
"step": 8420
},
{
"epoch": 26.976,
"grad_norm": NaN,
"learning_rate": 0.0001103988603988604,
"loss": 0.2216,
"step": 8430
},
{
"epoch": 26.9984,
"eval_accuracy": 0.76725,
"eval_loss": 0.5557342171669006,
"eval_runtime": 54.1898,
"eval_samples_per_second": 73.815,
"eval_steps_per_second": 2.307,
"step": 8437
},
{
"epoch": 27.008,
"grad_norm": NaN,
"learning_rate": 0.00010921177587844256,
"loss": 0.2184,
"step": 8440
},
{
"epoch": 27.04,
"grad_norm": NaN,
"learning_rate": 0.00010802469135802468,
"loss": 0.2286,
"step": 8450
},
{
"epoch": 27.072,
"grad_norm": Infinity,
"learning_rate": 0.00010683760683760684,
"loss": 0.2076,
"step": 8460
},
{
"epoch": 27.104,
"grad_norm": NaN,
"learning_rate": 0.00010565052231718898,
"loss": 0.2199,
"step": 8470
},
{
"epoch": 27.136,
"grad_norm": NaN,
"learning_rate": 0.00010446343779677113,
"loss": 0.2072,
"step": 8480
},
{
"epoch": 27.168,
"grad_norm": Infinity,
"learning_rate": 0.00010327635327635327,
"loss": 0.2274,
"step": 8490
},
{
"epoch": 27.2,
"grad_norm": NaN,
"learning_rate": 0.00010208926875593543,
"loss": 0.2152,
"step": 8500
},
{
"epoch": 27.232,
"grad_norm": NaN,
"learning_rate": 0.00010090218423551757,
"loss": 0.2265,
"step": 8510
},
{
"epoch": 27.264,
"grad_norm": NaN,
"learning_rate": 9.971509971509972e-05,
"loss": 0.2348,
"step": 8520
},
{
"epoch": 27.296,
"grad_norm": NaN,
"learning_rate": 9.852801519468186e-05,
"loss": 0.2254,
"step": 8530
},
{
"epoch": 27.328,
"grad_norm": NaN,
"learning_rate": 9.734093067426401e-05,
"loss": 0.2017,
"step": 8540
},
{
"epoch": 27.36,
"grad_norm": NaN,
"learning_rate": 9.615384615384617e-05,
"loss": 0.1982,
"step": 8550
},
{
"epoch": 27.392,
"grad_norm": Infinity,
"learning_rate": 9.496676163342831e-05,
"loss": 0.2028,
"step": 8560
},
{
"epoch": 27.424,
"grad_norm": NaN,
"learning_rate": 9.377967711301045e-05,
"loss": 0.2082,
"step": 8570
},
{
"epoch": 27.456,
"grad_norm": NaN,
"learning_rate": 9.259259259259259e-05,
"loss": 0.1861,
"step": 8580
},
{
"epoch": 27.488,
"grad_norm": NaN,
"learning_rate": 9.140550807217474e-05,
"loss": 0.2172,
"step": 8590
},
{
"epoch": 27.52,
"grad_norm": NaN,
"learning_rate": 9.021842355175688e-05,
"loss": 0.1983,
"step": 8600
},
{
"epoch": 27.552,
"grad_norm": NaN,
"learning_rate": 8.903133903133904e-05,
"loss": 0.2247,
"step": 8610
},
{
"epoch": 27.584,
"grad_norm": NaN,
"learning_rate": 8.784425451092118e-05,
"loss": 0.2131,
"step": 8620
},
{
"epoch": 27.616,
"grad_norm": NaN,
"learning_rate": 8.665716999050333e-05,
"loss": 0.2163,
"step": 8630
},
{
"epoch": 27.648,
"grad_norm": NaN,
"learning_rate": 8.547008547008547e-05,
"loss": 0.2069,
"step": 8640
},
{
"epoch": 27.68,
"grad_norm": NaN,
"learning_rate": 8.428300094966762e-05,
"loss": 0.1939,
"step": 8650
},
{
"epoch": 27.712,
"grad_norm": NaN,
"learning_rate": 8.309591642924976e-05,
"loss": 0.2181,
"step": 8660
},
{
"epoch": 27.744,
"grad_norm": NaN,
"learning_rate": 8.190883190883192e-05,
"loss": 0.1991,
"step": 8670
},
{
"epoch": 27.776,
"grad_norm": NaN,
"learning_rate": 8.072174738841406e-05,
"loss": 0.2245,
"step": 8680
},
{
"epoch": 27.808,
"grad_norm": Infinity,
"learning_rate": 7.95346628679962e-05,
"loss": 0.2061,
"step": 8690
},
{
"epoch": 27.84,
"grad_norm": NaN,
"learning_rate": 7.834757834757834e-05,
"loss": 0.2171,
"step": 8700
},
{
"epoch": 27.872,
"grad_norm": Infinity,
"learning_rate": 7.716049382716049e-05,
"loss": 0.2166,
"step": 8710
},
{
"epoch": 27.904,
"grad_norm": NaN,
"learning_rate": 7.597340930674263e-05,
"loss": 0.2272,
"step": 8720
},
{
"epoch": 27.936,
"grad_norm": Infinity,
"learning_rate": 7.478632478632479e-05,
"loss": 0.1896,
"step": 8730
},
{
"epoch": 27.968,
"grad_norm": NaN,
"learning_rate": 7.359924026590693e-05,
"loss": 0.1975,
"step": 8740
},
{
"epoch": 28.0,
"grad_norm": NaN,
"learning_rate": 7.241215574548908e-05,
"loss": 0.1918,
"step": 8750
},
{
"epoch": 28.0,
"eval_accuracy": 0.7705,
"eval_loss": 0.5501910448074341,
"eval_runtime": 51.9391,
"eval_samples_per_second": 77.013,
"eval_steps_per_second": 2.407,
"step": 8750
},
{
"epoch": 28.032,
"grad_norm": Infinity,
"learning_rate": 7.122507122507123e-05,
"loss": 0.2219,
"step": 8760
},
{
"epoch": 28.064,
"grad_norm": NaN,
"learning_rate": 7.003798670465337e-05,
"loss": 0.2083,
"step": 8770
},
{
"epoch": 28.096,
"grad_norm": NaN,
"learning_rate": 6.885090218423553e-05,
"loss": 0.2184,
"step": 8780
},
{
"epoch": 28.128,
"grad_norm": NaN,
"learning_rate": 6.766381766381767e-05,
"loss": 0.2101,
"step": 8790
},
{
"epoch": 28.16,
"grad_norm": NaN,
"learning_rate": 6.647673314339982e-05,
"loss": 0.2107,
"step": 8800
},
{
"epoch": 28.192,
"grad_norm": NaN,
"learning_rate": 6.528964862298195e-05,
"loss": 0.2187,
"step": 8810
},
{
"epoch": 28.224,
"grad_norm": NaN,
"learning_rate": 6.41025641025641e-05,
"loss": 0.2333,
"step": 8820
},
{
"epoch": 28.256,
"grad_norm": NaN,
"learning_rate": 6.291547958214624e-05,
"loss": 0.1879,
"step": 8830
},
{
"epoch": 28.288,
"grad_norm": NaN,
"learning_rate": 6.17283950617284e-05,
"loss": 0.223,
"step": 8840
},
{
"epoch": 28.32,
"grad_norm": NaN,
"learning_rate": 6.0541310541310544e-05,
"loss": 0.2429,
"step": 8850
},
{
"epoch": 28.352,
"grad_norm": NaN,
"learning_rate": 5.935422602089269e-05,
"loss": 0.2101,
"step": 8860
},
{
"epoch": 28.384,
"grad_norm": NaN,
"learning_rate": 5.816714150047484e-05,
"loss": 0.2148,
"step": 8870
},
{
"epoch": 28.416,
"grad_norm": Infinity,
"learning_rate": 5.6980056980056985e-05,
"loss": 0.2202,
"step": 8880
},
{
"epoch": 28.448,
"grad_norm": NaN,
"learning_rate": 5.579297245963913e-05,
"loss": 0.227,
"step": 8890
},
{
"epoch": 28.48,
"grad_norm": NaN,
"learning_rate": 5.460588793922128e-05,
"loss": 0.2122,
"step": 8900
},
{
"epoch": 28.512,
"grad_norm": NaN,
"learning_rate": 5.341880341880342e-05,
"loss": 0.2169,
"step": 8910
},
{
"epoch": 28.544,
"grad_norm": NaN,
"learning_rate": 5.2231718898385566e-05,
"loss": 0.2016,
"step": 8920
},
{
"epoch": 28.576,
"grad_norm": NaN,
"learning_rate": 5.104463437796771e-05,
"loss": 0.2147,
"step": 8930
},
{
"epoch": 28.608,
"grad_norm": NaN,
"learning_rate": 4.985754985754986e-05,
"loss": 0.2198,
"step": 8940
},
{
"epoch": 28.64,
"grad_norm": NaN,
"learning_rate": 4.867046533713201e-05,
"loss": 0.2045,
"step": 8950
},
{
"epoch": 28.672,
"grad_norm": NaN,
"learning_rate": 4.7483380816714154e-05,
"loss": 0.2219,
"step": 8960
},
{
"epoch": 28.704,
"grad_norm": NaN,
"learning_rate": 4.6296296296296294e-05,
"loss": 0.2139,
"step": 8970
},
{
"epoch": 28.736,
"grad_norm": NaN,
"learning_rate": 4.510921177587844e-05,
"loss": 0.1975,
"step": 8980
},
{
"epoch": 28.768,
"grad_norm": NaN,
"learning_rate": 4.392212725546059e-05,
"loss": 0.2197,
"step": 8990
},
{
"epoch": 28.8,
"grad_norm": NaN,
"learning_rate": 4.2735042735042735e-05,
"loss": 0.2083,
"step": 9000
},
{
"epoch": 28.832,
"grad_norm": NaN,
"learning_rate": 4.154795821462488e-05,
"loss": 0.2399,
"step": 9010
},
{
"epoch": 28.864,
"grad_norm": NaN,
"learning_rate": 4.036087369420703e-05,
"loss": 0.2088,
"step": 9020
},
{
"epoch": 28.896,
"grad_norm": NaN,
"learning_rate": 3.917378917378917e-05,
"loss": 0.1972,
"step": 9030
},
{
"epoch": 28.928,
"grad_norm": Infinity,
"learning_rate": 3.7986704653371316e-05,
"loss": 0.205,
"step": 9040
},
{
"epoch": 28.96,
"grad_norm": NaN,
"learning_rate": 3.679962013295346e-05,
"loss": 0.1864,
"step": 9050
},
{
"epoch": 28.992,
"grad_norm": NaN,
"learning_rate": 3.561253561253562e-05,
"loss": 0.199,
"step": 9060
},
{
"epoch": 28.9984,
"eval_accuracy": 0.7675,
"eval_loss": 0.5455857515335083,
"eval_runtime": 52.0571,
"eval_samples_per_second": 76.839,
"eval_steps_per_second": 2.401,
"step": 9062
},
{
"epoch": 29.024,
"grad_norm": NaN,
"learning_rate": 3.4425451092117764e-05,
"loss": 0.1993,
"step": 9070
},
{
"epoch": 29.056,
"grad_norm": NaN,
"learning_rate": 3.323836657169991e-05,
"loss": 0.1937,
"step": 9080
},
{
"epoch": 29.088,
"grad_norm": NaN,
"learning_rate": 3.205128205128205e-05,
"loss": 0.2244,
"step": 9090
},
{
"epoch": 29.12,
"grad_norm": NaN,
"learning_rate": 3.08641975308642e-05,
"loss": 0.2198,
"step": 9100
},
{
"epoch": 29.152,
"grad_norm": NaN,
"learning_rate": 2.9677113010446345e-05,
"loss": 0.1889,
"step": 9110
},
{
"epoch": 29.184,
"grad_norm": NaN,
"learning_rate": 2.8490028490028492e-05,
"loss": 0.2253,
"step": 9120
},
{
"epoch": 29.216,
"grad_norm": Infinity,
"learning_rate": 2.730294396961064e-05,
"loss": 0.2155,
"step": 9130
},
{
"epoch": 29.248,
"grad_norm": NaN,
"learning_rate": 2.6115859449192783e-05,
"loss": 0.2084,
"step": 9140
},
{
"epoch": 29.28,
"grad_norm": NaN,
"learning_rate": 2.492877492877493e-05,
"loss": 0.1883,
"step": 9150
},
{
"epoch": 29.312,
"grad_norm": NaN,
"learning_rate": 2.3741690408357077e-05,
"loss": 0.225,
"step": 9160
},
{
"epoch": 29.344,
"grad_norm": NaN,
"learning_rate": 2.255460588793922e-05,
"loss": 0.2065,
"step": 9170
},
{
"epoch": 29.376,
"grad_norm": NaN,
"learning_rate": 2.1367521367521368e-05,
"loss": 0.2185,
"step": 9180
},
{
"epoch": 29.408,
"grad_norm": NaN,
"learning_rate": 2.0180436847103515e-05,
"loss": 0.2292,
"step": 9190
},
{
"epoch": 29.44,
"grad_norm": NaN,
"learning_rate": 1.8993352326685658e-05,
"loss": 0.2212,
"step": 9200
},
{
"epoch": 29.472,
"grad_norm": NaN,
"learning_rate": 1.780626780626781e-05,
"loss": 0.2327,
"step": 9210
},
{
"epoch": 29.504,
"grad_norm": NaN,
"learning_rate": 1.6619183285849956e-05,
"loss": 0.1956,
"step": 9220
},
{
"epoch": 29.536,
"grad_norm": NaN,
"learning_rate": 1.54320987654321e-05,
"loss": 0.2135,
"step": 9230
},
{
"epoch": 29.568,
"grad_norm": NaN,
"learning_rate": 1.4245014245014246e-05,
"loss": 0.206,
"step": 9240
},
{
"epoch": 29.6,
"grad_norm": NaN,
"learning_rate": 1.3057929724596391e-05,
"loss": 0.2007,
"step": 9250
},
{
"epoch": 29.632,
"grad_norm": NaN,
"learning_rate": 1.1870845204178538e-05,
"loss": 0.2297,
"step": 9260
},
{
"epoch": 29.664,
"grad_norm": NaN,
"learning_rate": 1.0683760683760684e-05,
"loss": 0.2129,
"step": 9270
},
{
"epoch": 29.696,
"grad_norm": NaN,
"learning_rate": 9.496676163342829e-06,
"loss": 0.1998,
"step": 9280
},
{
"epoch": 29.728,
"grad_norm": NaN,
"learning_rate": 8.309591642924978e-06,
"loss": 0.2104,
"step": 9290
},
{
"epoch": 29.76,
"grad_norm": NaN,
"learning_rate": 7.122507122507123e-06,
"loss": 0.2138,
"step": 9300
},
{
"epoch": 29.792,
"grad_norm": NaN,
"learning_rate": 5.935422602089269e-06,
"loss": 0.2111,
"step": 9310
},
{
"epoch": 29.824,
"grad_norm": NaN,
"learning_rate": 4.7483380816714145e-06,
"loss": 0.2009,
"step": 9320
},
{
"epoch": 29.856,
"grad_norm": NaN,
"learning_rate": 3.5612535612535615e-06,
"loss": 0.2021,
"step": 9330
},
{
"epoch": 29.888,
"grad_norm": NaN,
"learning_rate": 2.3741690408357073e-06,
"loss": 0.1985,
"step": 9340
},
{
"epoch": 29.92,
"grad_norm": NaN,
"learning_rate": 1.1870845204178536e-06,
"loss": 0.2092,
"step": 9350
},
{
"epoch": 29.951999999999998,
"grad_norm": NaN,
"learning_rate": 0.0,
"loss": 0.21,
"step": 9360
},
{
"epoch": 29.951999999999998,
"eval_accuracy": 0.7715,
"eval_loss": 0.5483142733573914,
"eval_runtime": 52.0007,
"eval_samples_per_second": 76.922,
"eval_steps_per_second": 2.404,
"step": 9360
},
{
"epoch": 29.951999999999998,
"step": 9360,
"total_flos": 2.9779463284174356e+19,
"train_loss": 0.21163268028161464,
"train_runtime": 43715.1441,
"train_samples_per_second": 27.45,
"train_steps_per_second": 0.214
}
],
"logging_steps": 10,
"max_steps": 9360,
"num_input_tokens_seen": 0,
"num_train_epochs": 30,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.9779463284174356e+19,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}