hajeong67's picture
Upload folder using huggingface_hub
08f7e22 verified
raw
history blame
73 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 4200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.007142857142857143,
"grad_norm": 2.519733190536499,
"learning_rate": 9.976190476190477e-05,
"loss": 1.2999,
"step": 10
},
{
"epoch": 0.014285714285714285,
"grad_norm": 2.4232161045074463,
"learning_rate": 9.952380952380953e-05,
"loss": 1.1228,
"step": 20
},
{
"epoch": 0.02142857142857143,
"grad_norm": 2.7126998901367188,
"learning_rate": 9.92857142857143e-05,
"loss": 0.9913,
"step": 30
},
{
"epoch": 0.02857142857142857,
"grad_norm": 7.3811235427856445,
"learning_rate": 9.904761904761905e-05,
"loss": 0.9306,
"step": 40
},
{
"epoch": 0.03571428571428571,
"grad_norm": 6.5697197914123535,
"learning_rate": 9.880952380952381e-05,
"loss": 0.7457,
"step": 50
},
{
"epoch": 0.04285714285714286,
"grad_norm": 1.2972687482833862,
"learning_rate": 9.857142857142858e-05,
"loss": 0.6408,
"step": 60
},
{
"epoch": 0.05,
"grad_norm": 3.600625991821289,
"learning_rate": 9.833333333333333e-05,
"loss": 0.6541,
"step": 70
},
{
"epoch": 0.05714285714285714,
"grad_norm": 1.5594621896743774,
"learning_rate": 9.80952380952381e-05,
"loss": 0.538,
"step": 80
},
{
"epoch": 0.06428571428571428,
"grad_norm": 4.595429420471191,
"learning_rate": 9.785714285714286e-05,
"loss": 0.556,
"step": 90
},
{
"epoch": 0.07142857142857142,
"grad_norm": 2.620643138885498,
"learning_rate": 9.761904761904762e-05,
"loss": 0.5521,
"step": 100
},
{
"epoch": 0.07857142857142857,
"grad_norm": 3.503641366958618,
"learning_rate": 9.738095238095239e-05,
"loss": 0.5912,
"step": 110
},
{
"epoch": 0.08571428571428572,
"grad_norm": 1.5058848857879639,
"learning_rate": 9.714285714285715e-05,
"loss": 0.5421,
"step": 120
},
{
"epoch": 0.09285714285714286,
"grad_norm": 3.7034990787506104,
"learning_rate": 9.69047619047619e-05,
"loss": 0.5568,
"step": 130
},
{
"epoch": 0.1,
"grad_norm": 3.714170455932617,
"learning_rate": 9.666666666666667e-05,
"loss": 0.5286,
"step": 140
},
{
"epoch": 0.10714285714285714,
"grad_norm": 2.069328546524048,
"learning_rate": 9.642857142857143e-05,
"loss": 0.4755,
"step": 150
},
{
"epoch": 0.11428571428571428,
"grad_norm": 1.2878583669662476,
"learning_rate": 9.61904761904762e-05,
"loss": 0.4817,
"step": 160
},
{
"epoch": 0.12142857142857143,
"grad_norm": 3.1945483684539795,
"learning_rate": 9.595238095238096e-05,
"loss": 0.5035,
"step": 170
},
{
"epoch": 0.12857142857142856,
"grad_norm": 1.9643166065216064,
"learning_rate": 9.571428571428573e-05,
"loss": 0.4932,
"step": 180
},
{
"epoch": 0.1357142857142857,
"grad_norm": 3.5002968311309814,
"learning_rate": 9.547619047619049e-05,
"loss": 0.5531,
"step": 190
},
{
"epoch": 0.14285714285714285,
"grad_norm": 4.7095794677734375,
"learning_rate": 9.523809523809524e-05,
"loss": 0.4601,
"step": 200
},
{
"epoch": 0.15,
"grad_norm": 1.4850050210952759,
"learning_rate": 9.5e-05,
"loss": 0.432,
"step": 210
},
{
"epoch": 0.15714285714285714,
"grad_norm": 2.860565423965454,
"learning_rate": 9.476190476190476e-05,
"loss": 0.4699,
"step": 220
},
{
"epoch": 0.16428571428571428,
"grad_norm": 1.8467603921890259,
"learning_rate": 9.452380952380952e-05,
"loss": 0.42,
"step": 230
},
{
"epoch": 0.17142857142857143,
"grad_norm": 1.6448352336883545,
"learning_rate": 9.428571428571429e-05,
"loss": 0.431,
"step": 240
},
{
"epoch": 0.17857142857142858,
"grad_norm": 1.8279732465744019,
"learning_rate": 9.404761904761905e-05,
"loss": 0.4413,
"step": 250
},
{
"epoch": 0.18571428571428572,
"grad_norm": 3.0112857818603516,
"learning_rate": 9.380952380952381e-05,
"loss": 0.4263,
"step": 260
},
{
"epoch": 0.19285714285714287,
"grad_norm": 1.7717293500900269,
"learning_rate": 9.357142857142858e-05,
"loss": 0.4829,
"step": 270
},
{
"epoch": 0.2,
"grad_norm": 4.515851974487305,
"learning_rate": 9.333333333333334e-05,
"loss": 0.4194,
"step": 280
},
{
"epoch": 0.20714285714285716,
"grad_norm": 3.998619556427002,
"learning_rate": 9.309523809523811e-05,
"loss": 0.4422,
"step": 290
},
{
"epoch": 0.21428571428571427,
"grad_norm": 3.5681912899017334,
"learning_rate": 9.285714285714286e-05,
"loss": 0.4677,
"step": 300
},
{
"epoch": 0.22142857142857142,
"grad_norm": 3.286348342895508,
"learning_rate": 9.261904761904762e-05,
"loss": 0.4152,
"step": 310
},
{
"epoch": 0.22857142857142856,
"grad_norm": 2.018721580505371,
"learning_rate": 9.238095238095239e-05,
"loss": 0.5065,
"step": 320
},
{
"epoch": 0.2357142857142857,
"grad_norm": 1.8489034175872803,
"learning_rate": 9.214285714285714e-05,
"loss": 0.392,
"step": 330
},
{
"epoch": 0.24285714285714285,
"grad_norm": 1.3958369493484497,
"learning_rate": 9.19047619047619e-05,
"loss": 0.3988,
"step": 340
},
{
"epoch": 0.25,
"grad_norm": 2.9772961139678955,
"learning_rate": 9.166666666666667e-05,
"loss": 0.3955,
"step": 350
},
{
"epoch": 0.2571428571428571,
"grad_norm": 3.2752914428710938,
"learning_rate": 9.142857142857143e-05,
"loss": 0.3998,
"step": 360
},
{
"epoch": 0.2642857142857143,
"grad_norm": 1.8811057806015015,
"learning_rate": 9.11904761904762e-05,
"loss": 0.4775,
"step": 370
},
{
"epoch": 0.2714285714285714,
"grad_norm": 2.3310513496398926,
"learning_rate": 9.095238095238096e-05,
"loss": 0.3703,
"step": 380
},
{
"epoch": 0.2785714285714286,
"grad_norm": 2.967597484588623,
"learning_rate": 9.071428571428571e-05,
"loss": 0.4177,
"step": 390
},
{
"epoch": 0.2857142857142857,
"grad_norm": 2.671673536300659,
"learning_rate": 9.047619047619048e-05,
"loss": 0.3823,
"step": 400
},
{
"epoch": 0.29285714285714287,
"grad_norm": 4.647867679595947,
"learning_rate": 9.023809523809524e-05,
"loss": 0.3888,
"step": 410
},
{
"epoch": 0.3,
"grad_norm": 3.4788432121276855,
"learning_rate": 9e-05,
"loss": 0.3856,
"step": 420
},
{
"epoch": 0.30714285714285716,
"grad_norm": 3.308420181274414,
"learning_rate": 8.976190476190477e-05,
"loss": 0.4143,
"step": 430
},
{
"epoch": 0.3142857142857143,
"grad_norm": 3.499880075454712,
"learning_rate": 8.952380952380953e-05,
"loss": 0.4857,
"step": 440
},
{
"epoch": 0.32142857142857145,
"grad_norm": 1.7907614707946777,
"learning_rate": 8.92857142857143e-05,
"loss": 0.356,
"step": 450
},
{
"epoch": 0.32857142857142857,
"grad_norm": 3.0111501216888428,
"learning_rate": 8.904761904761905e-05,
"loss": 0.358,
"step": 460
},
{
"epoch": 0.3357142857142857,
"grad_norm": 2.1389689445495605,
"learning_rate": 8.880952380952381e-05,
"loss": 0.3331,
"step": 470
},
{
"epoch": 0.34285714285714286,
"grad_norm": 1.8007326126098633,
"learning_rate": 8.857142857142857e-05,
"loss": 0.3148,
"step": 480
},
{
"epoch": 0.35,
"grad_norm": 3.603548288345337,
"learning_rate": 8.833333333333333e-05,
"loss": 0.4513,
"step": 490
},
{
"epoch": 0.35714285714285715,
"grad_norm": 2.270984411239624,
"learning_rate": 8.80952380952381e-05,
"loss": 0.3392,
"step": 500
},
{
"epoch": 0.36428571428571427,
"grad_norm": 1.5367858409881592,
"learning_rate": 8.785714285714286e-05,
"loss": 0.3665,
"step": 510
},
{
"epoch": 0.37142857142857144,
"grad_norm": 1.5397262573242188,
"learning_rate": 8.761904761904762e-05,
"loss": 0.3875,
"step": 520
},
{
"epoch": 0.37857142857142856,
"grad_norm": 3.6661577224731445,
"learning_rate": 8.738095238095239e-05,
"loss": 0.3617,
"step": 530
},
{
"epoch": 0.38571428571428573,
"grad_norm": 2.5946245193481445,
"learning_rate": 8.714285714285715e-05,
"loss": 0.3318,
"step": 540
},
{
"epoch": 0.39285714285714285,
"grad_norm": 2.508056640625,
"learning_rate": 8.690476190476192e-05,
"loss": 0.3681,
"step": 550
},
{
"epoch": 0.4,
"grad_norm": 5.406072616577148,
"learning_rate": 8.666666666666667e-05,
"loss": 0.3589,
"step": 560
},
{
"epoch": 0.40714285714285714,
"grad_norm": 2.0363543033599854,
"learning_rate": 8.642857142857143e-05,
"loss": 0.4328,
"step": 570
},
{
"epoch": 0.4142857142857143,
"grad_norm": 1.796236515045166,
"learning_rate": 8.61904761904762e-05,
"loss": 0.4005,
"step": 580
},
{
"epoch": 0.42142857142857143,
"grad_norm": 5.6172943115234375,
"learning_rate": 8.595238095238096e-05,
"loss": 0.3414,
"step": 590
},
{
"epoch": 0.42857142857142855,
"grad_norm": 2.953678846359253,
"learning_rate": 8.571428571428571e-05,
"loss": 0.3917,
"step": 600
},
{
"epoch": 0.4357142857142857,
"grad_norm": 1.8276687860488892,
"learning_rate": 8.547619047619048e-05,
"loss": 0.4057,
"step": 610
},
{
"epoch": 0.44285714285714284,
"grad_norm": 1.8668811321258545,
"learning_rate": 8.523809523809524e-05,
"loss": 0.4153,
"step": 620
},
{
"epoch": 0.45,
"grad_norm": 2.87809419631958,
"learning_rate": 8.5e-05,
"loss": 0.3911,
"step": 630
},
{
"epoch": 0.45714285714285713,
"grad_norm": 4.863674640655518,
"learning_rate": 8.476190476190477e-05,
"loss": 0.4458,
"step": 640
},
{
"epoch": 0.4642857142857143,
"grad_norm": 1.7150845527648926,
"learning_rate": 8.452380952380952e-05,
"loss": 0.3951,
"step": 650
},
{
"epoch": 0.4714285714285714,
"grad_norm": 2.0149145126342773,
"learning_rate": 8.428571428571429e-05,
"loss": 0.3238,
"step": 660
},
{
"epoch": 0.4785714285714286,
"grad_norm": 2.3146402835845947,
"learning_rate": 8.404761904761905e-05,
"loss": 0.3129,
"step": 670
},
{
"epoch": 0.4857142857142857,
"grad_norm": 1.8723537921905518,
"learning_rate": 8.380952380952382e-05,
"loss": 0.2969,
"step": 680
},
{
"epoch": 0.4928571428571429,
"grad_norm": 1.6206679344177246,
"learning_rate": 8.357142857142858e-05,
"loss": 0.3673,
"step": 690
},
{
"epoch": 0.5,
"grad_norm": 2.490722179412842,
"learning_rate": 8.333333333333334e-05,
"loss": 0.3493,
"step": 700
},
{
"epoch": 0.5071428571428571,
"grad_norm": 1.7868791818618774,
"learning_rate": 8.309523809523811e-05,
"loss": 0.423,
"step": 710
},
{
"epoch": 0.5142857142857142,
"grad_norm": 2.3152015209198,
"learning_rate": 8.285714285714287e-05,
"loss": 0.3448,
"step": 720
},
{
"epoch": 0.5214285714285715,
"grad_norm": 1.934523105621338,
"learning_rate": 8.261904761904762e-05,
"loss": 0.372,
"step": 730
},
{
"epoch": 0.5285714285714286,
"grad_norm": 2.421624183654785,
"learning_rate": 8.238095238095238e-05,
"loss": 0.3602,
"step": 740
},
{
"epoch": 0.5357142857142857,
"grad_norm": 2.0156285762786865,
"learning_rate": 8.214285714285714e-05,
"loss": 0.334,
"step": 750
},
{
"epoch": 0.5428571428571428,
"grad_norm": 1.6513855457305908,
"learning_rate": 8.19047619047619e-05,
"loss": 0.3553,
"step": 760
},
{
"epoch": 0.55,
"grad_norm": 1.950451374053955,
"learning_rate": 8.166666666666667e-05,
"loss": 0.3696,
"step": 770
},
{
"epoch": 0.5571428571428572,
"grad_norm": 2.4662423133850098,
"learning_rate": 8.142857142857143e-05,
"loss": 0.3232,
"step": 780
},
{
"epoch": 0.5642857142857143,
"grad_norm": 2.060654401779175,
"learning_rate": 8.11904761904762e-05,
"loss": 0.3294,
"step": 790
},
{
"epoch": 0.5714285714285714,
"grad_norm": 3.3581550121307373,
"learning_rate": 8.095238095238096e-05,
"loss": 0.3022,
"step": 800
},
{
"epoch": 0.5785714285714286,
"grad_norm": 1.6778148412704468,
"learning_rate": 8.071428571428573e-05,
"loss": 0.2808,
"step": 810
},
{
"epoch": 0.5857142857142857,
"grad_norm": 5.509958267211914,
"learning_rate": 8.047619047619048e-05,
"loss": 0.3584,
"step": 820
},
{
"epoch": 0.5928571428571429,
"grad_norm": 2.3136510848999023,
"learning_rate": 8.023809523809524e-05,
"loss": 0.2967,
"step": 830
},
{
"epoch": 0.6,
"grad_norm": 2.757749557495117,
"learning_rate": 8e-05,
"loss": 0.3091,
"step": 840
},
{
"epoch": 0.6071428571428571,
"grad_norm": 1.6453288793563843,
"learning_rate": 7.976190476190477e-05,
"loss": 0.2882,
"step": 850
},
{
"epoch": 0.6142857142857143,
"grad_norm": 3.599712371826172,
"learning_rate": 7.952380952380952e-05,
"loss": 0.35,
"step": 860
},
{
"epoch": 0.6214285714285714,
"grad_norm": 1.203050136566162,
"learning_rate": 7.928571428571429e-05,
"loss": 0.3147,
"step": 870
},
{
"epoch": 0.6285714285714286,
"grad_norm": 3.0582826137542725,
"learning_rate": 7.904761904761905e-05,
"loss": 0.369,
"step": 880
},
{
"epoch": 0.6357142857142857,
"grad_norm": 1.7133578062057495,
"learning_rate": 7.880952380952382e-05,
"loss": 0.4404,
"step": 890
},
{
"epoch": 0.6428571428571429,
"grad_norm": 2.1532819271087646,
"learning_rate": 7.857142857142858e-05,
"loss": 0.3249,
"step": 900
},
{
"epoch": 0.65,
"grad_norm": 3.529282808303833,
"learning_rate": 7.833333333333333e-05,
"loss": 0.3449,
"step": 910
},
{
"epoch": 0.6571428571428571,
"grad_norm": 3.412729501724243,
"learning_rate": 7.80952380952381e-05,
"loss": 0.3678,
"step": 920
},
{
"epoch": 0.6642857142857143,
"grad_norm": 1.8676691055297852,
"learning_rate": 7.785714285714286e-05,
"loss": 0.3522,
"step": 930
},
{
"epoch": 0.6714285714285714,
"grad_norm": 1.8748834133148193,
"learning_rate": 7.761904761904762e-05,
"loss": 0.2561,
"step": 940
},
{
"epoch": 0.6785714285714286,
"grad_norm": 3.1622543334960938,
"learning_rate": 7.738095238095239e-05,
"loss": 0.3245,
"step": 950
},
{
"epoch": 0.6857142857142857,
"grad_norm": 2.17108416557312,
"learning_rate": 7.714285714285715e-05,
"loss": 0.335,
"step": 960
},
{
"epoch": 0.6928571428571428,
"grad_norm": 3.686673402786255,
"learning_rate": 7.690476190476192e-05,
"loss": 0.319,
"step": 970
},
{
"epoch": 0.7,
"grad_norm": 7.108774662017822,
"learning_rate": 7.666666666666667e-05,
"loss": 0.3337,
"step": 980
},
{
"epoch": 0.7071428571428572,
"grad_norm": 2.532336711883545,
"learning_rate": 7.642857142857143e-05,
"loss": 0.377,
"step": 990
},
{
"epoch": 0.7142857142857143,
"grad_norm": 1.8768038749694824,
"learning_rate": 7.619047619047618e-05,
"loss": 0.3745,
"step": 1000
},
{
"epoch": 0.7214285714285714,
"grad_norm": 1.8926101922988892,
"learning_rate": 7.595238095238095e-05,
"loss": 0.3503,
"step": 1010
},
{
"epoch": 0.7285714285714285,
"grad_norm": 2.2763147354125977,
"learning_rate": 7.571428571428571e-05,
"loss": 0.307,
"step": 1020
},
{
"epoch": 0.7357142857142858,
"grad_norm": 3.663257360458374,
"learning_rate": 7.547619047619048e-05,
"loss": 0.2876,
"step": 1030
},
{
"epoch": 0.7428571428571429,
"grad_norm": 1.844648003578186,
"learning_rate": 7.523809523809524e-05,
"loss": 0.3512,
"step": 1040
},
{
"epoch": 0.75,
"grad_norm": 2.5800540447235107,
"learning_rate": 7.500000000000001e-05,
"loss": 0.301,
"step": 1050
},
{
"epoch": 0.7571428571428571,
"grad_norm": 2.4674232006073,
"learning_rate": 7.476190476190477e-05,
"loss": 0.3635,
"step": 1060
},
{
"epoch": 0.7642857142857142,
"grad_norm": 3.5119080543518066,
"learning_rate": 7.452380952380952e-05,
"loss": 0.3688,
"step": 1070
},
{
"epoch": 0.7714285714285715,
"grad_norm": 2.5902345180511475,
"learning_rate": 7.428571428571429e-05,
"loss": 0.3136,
"step": 1080
},
{
"epoch": 0.7785714285714286,
"grad_norm": 1.734082818031311,
"learning_rate": 7.404761904761905e-05,
"loss": 0.3366,
"step": 1090
},
{
"epoch": 0.7857142857142857,
"grad_norm": 2.0863406658172607,
"learning_rate": 7.380952380952382e-05,
"loss": 0.331,
"step": 1100
},
{
"epoch": 0.7928571428571428,
"grad_norm": 2.9227652549743652,
"learning_rate": 7.357142857142858e-05,
"loss": 0.3131,
"step": 1110
},
{
"epoch": 0.8,
"grad_norm": 5.268605709075928,
"learning_rate": 7.333333333333333e-05,
"loss": 0.3212,
"step": 1120
},
{
"epoch": 0.8071428571428572,
"grad_norm": 2.633431911468506,
"learning_rate": 7.30952380952381e-05,
"loss": 0.2707,
"step": 1130
},
{
"epoch": 0.8142857142857143,
"grad_norm": 1.652354121208191,
"learning_rate": 7.285714285714286e-05,
"loss": 0.3037,
"step": 1140
},
{
"epoch": 0.8214285714285714,
"grad_norm": 3.1987786293029785,
"learning_rate": 7.261904761904762e-05,
"loss": 0.3603,
"step": 1150
},
{
"epoch": 0.8285714285714286,
"grad_norm": 1.7624555826187134,
"learning_rate": 7.238095238095238e-05,
"loss": 0.308,
"step": 1160
},
{
"epoch": 0.8357142857142857,
"grad_norm": 2.6379318237304688,
"learning_rate": 7.214285714285714e-05,
"loss": 0.3068,
"step": 1170
},
{
"epoch": 0.8428571428571429,
"grad_norm": 2.6688101291656494,
"learning_rate": 7.19047619047619e-05,
"loss": 0.3134,
"step": 1180
},
{
"epoch": 0.85,
"grad_norm": 3.698643207550049,
"learning_rate": 7.166666666666667e-05,
"loss": 0.2973,
"step": 1190
},
{
"epoch": 0.8571428571428571,
"grad_norm": 2.9004178047180176,
"learning_rate": 7.142857142857143e-05,
"loss": 0.3116,
"step": 1200
},
{
"epoch": 0.8642857142857143,
"grad_norm": 2.3414883613586426,
"learning_rate": 7.11904761904762e-05,
"loss": 0.334,
"step": 1210
},
{
"epoch": 0.8714285714285714,
"grad_norm": 3.6203651428222656,
"learning_rate": 7.095238095238096e-05,
"loss": 0.3241,
"step": 1220
},
{
"epoch": 0.8785714285714286,
"grad_norm": 2.7472829818725586,
"learning_rate": 7.071428571428573e-05,
"loss": 0.2727,
"step": 1230
},
{
"epoch": 0.8857142857142857,
"grad_norm": 3.2332115173339844,
"learning_rate": 7.047619047619048e-05,
"loss": 0.3329,
"step": 1240
},
{
"epoch": 0.8928571428571429,
"grad_norm": 1.968595027923584,
"learning_rate": 7.023809523809524e-05,
"loss": 0.3478,
"step": 1250
},
{
"epoch": 0.9,
"grad_norm": 2.3712966442108154,
"learning_rate": 7e-05,
"loss": 0.3964,
"step": 1260
},
{
"epoch": 0.9071428571428571,
"grad_norm": 1.7704048156738281,
"learning_rate": 6.976190476190476e-05,
"loss": 0.3983,
"step": 1270
},
{
"epoch": 0.9142857142857143,
"grad_norm": 1.8879201412200928,
"learning_rate": 6.952380952380952e-05,
"loss": 0.3296,
"step": 1280
},
{
"epoch": 0.9214285714285714,
"grad_norm": 4.0522780418396,
"learning_rate": 6.928571428571429e-05,
"loss": 0.3722,
"step": 1290
},
{
"epoch": 0.9285714285714286,
"grad_norm": 1.7849159240722656,
"learning_rate": 6.904761904761905e-05,
"loss": 0.3022,
"step": 1300
},
{
"epoch": 0.9357142857142857,
"grad_norm": 3.376678705215454,
"learning_rate": 6.880952380952382e-05,
"loss": 0.3429,
"step": 1310
},
{
"epoch": 0.9428571428571428,
"grad_norm": 4.100617408752441,
"learning_rate": 6.857142857142858e-05,
"loss": 0.2793,
"step": 1320
},
{
"epoch": 0.95,
"grad_norm": 3.8000288009643555,
"learning_rate": 6.833333333333333e-05,
"loss": 0.3114,
"step": 1330
},
{
"epoch": 0.9571428571428572,
"grad_norm": 4.846517562866211,
"learning_rate": 6.80952380952381e-05,
"loss": 0.3377,
"step": 1340
},
{
"epoch": 0.9642857142857143,
"grad_norm": 2.0567409992218018,
"learning_rate": 6.785714285714286e-05,
"loss": 0.3198,
"step": 1350
},
{
"epoch": 0.9714285714285714,
"grad_norm": 2.3197948932647705,
"learning_rate": 6.761904761904763e-05,
"loss": 0.2817,
"step": 1360
},
{
"epoch": 0.9785714285714285,
"grad_norm": 2.7312827110290527,
"learning_rate": 6.738095238095239e-05,
"loss": 0.3378,
"step": 1370
},
{
"epoch": 0.9857142857142858,
"grad_norm": 3.115516185760498,
"learning_rate": 6.714285714285714e-05,
"loss": 0.3638,
"step": 1380
},
{
"epoch": 0.9928571428571429,
"grad_norm": 1.0742294788360596,
"learning_rate": 6.69047619047619e-05,
"loss": 0.3176,
"step": 1390
},
{
"epoch": 1.0,
"grad_norm": 2.3872761726379395,
"learning_rate": 6.666666666666667e-05,
"loss": 0.2921,
"step": 1400
},
{
"epoch": 1.0,
"eval_loss": 0.2929375469684601,
"eval_runtime": 68.5646,
"eval_samples_per_second": 2.917,
"eval_steps_per_second": 0.365,
"step": 1400
},
{
"epoch": 1.0071428571428571,
"grad_norm": 2.1662676334381104,
"learning_rate": 6.642857142857143e-05,
"loss": 0.2885,
"step": 1410
},
{
"epoch": 1.0142857142857142,
"grad_norm": 4.116220474243164,
"learning_rate": 6.619047619047619e-05,
"loss": 0.2951,
"step": 1420
},
{
"epoch": 1.0214285714285714,
"grad_norm": 4.550313472747803,
"learning_rate": 6.595238095238095e-05,
"loss": 0.3078,
"step": 1430
},
{
"epoch": 1.0285714285714285,
"grad_norm": 1.7829937934875488,
"learning_rate": 6.571428571428571e-05,
"loss": 0.3309,
"step": 1440
},
{
"epoch": 1.0357142857142858,
"grad_norm": 2.684500217437744,
"learning_rate": 6.547619047619048e-05,
"loss": 0.2994,
"step": 1450
},
{
"epoch": 1.042857142857143,
"grad_norm": 1.3234150409698486,
"learning_rate": 6.523809523809524e-05,
"loss": 0.3782,
"step": 1460
},
{
"epoch": 1.05,
"grad_norm": 2.6253223419189453,
"learning_rate": 6.500000000000001e-05,
"loss": 0.3094,
"step": 1470
},
{
"epoch": 1.0571428571428572,
"grad_norm": 3.272584915161133,
"learning_rate": 6.476190476190477e-05,
"loss": 0.3406,
"step": 1480
},
{
"epoch": 1.0642857142857143,
"grad_norm": 3.765272378921509,
"learning_rate": 6.452380952380954e-05,
"loss": 0.3525,
"step": 1490
},
{
"epoch": 1.0714285714285714,
"grad_norm": 4.5847578048706055,
"learning_rate": 6.428571428571429e-05,
"loss": 0.3296,
"step": 1500
},
{
"epoch": 1.0785714285714285,
"grad_norm": 2.448350191116333,
"learning_rate": 6.404761904761904e-05,
"loss": 0.2704,
"step": 1510
},
{
"epoch": 1.0857142857142856,
"grad_norm": 2.5641868114471436,
"learning_rate": 6.38095238095238e-05,
"loss": 0.277,
"step": 1520
},
{
"epoch": 1.092857142857143,
"grad_norm": 1.8631259202957153,
"learning_rate": 6.357142857142857e-05,
"loss": 0.2915,
"step": 1530
},
{
"epoch": 1.1,
"grad_norm": 1.7823829650878906,
"learning_rate": 6.333333333333333e-05,
"loss": 0.279,
"step": 1540
},
{
"epoch": 1.1071428571428572,
"grad_norm": 3.2730162143707275,
"learning_rate": 6.30952380952381e-05,
"loss": 0.3179,
"step": 1550
},
{
"epoch": 1.1142857142857143,
"grad_norm": 2.495410203933716,
"learning_rate": 6.285714285714286e-05,
"loss": 0.326,
"step": 1560
},
{
"epoch": 1.1214285714285714,
"grad_norm": 1.844226598739624,
"learning_rate": 6.261904761904763e-05,
"loss": 0.2931,
"step": 1570
},
{
"epoch": 1.1285714285714286,
"grad_norm": 3.5361154079437256,
"learning_rate": 6.238095238095239e-05,
"loss": 0.2932,
"step": 1580
},
{
"epoch": 1.1357142857142857,
"grad_norm": 1.9047455787658691,
"learning_rate": 6.214285714285714e-05,
"loss": 0.2779,
"step": 1590
},
{
"epoch": 1.1428571428571428,
"grad_norm": 6.630354881286621,
"learning_rate": 6.19047619047619e-05,
"loss": 0.3563,
"step": 1600
},
{
"epoch": 1.15,
"grad_norm": 4.267199993133545,
"learning_rate": 6.166666666666667e-05,
"loss": 0.3146,
"step": 1610
},
{
"epoch": 1.157142857142857,
"grad_norm": 4.094973087310791,
"learning_rate": 6.142857142857143e-05,
"loss": 0.3096,
"step": 1620
},
{
"epoch": 1.1642857142857144,
"grad_norm": 2.1625826358795166,
"learning_rate": 6.11904761904762e-05,
"loss": 0.2893,
"step": 1630
},
{
"epoch": 1.1714285714285715,
"grad_norm": 4.5134758949279785,
"learning_rate": 6.0952380952380964e-05,
"loss": 0.3037,
"step": 1640
},
{
"epoch": 1.1785714285714286,
"grad_norm": 1.987043023109436,
"learning_rate": 6.0714285714285715e-05,
"loss": 0.3427,
"step": 1650
},
{
"epoch": 1.1857142857142857,
"grad_norm": 1.7972902059555054,
"learning_rate": 6.047619047619047e-05,
"loss": 0.3012,
"step": 1660
},
{
"epoch": 1.1928571428571428,
"grad_norm": 2.2862155437469482,
"learning_rate": 6.023809523809524e-05,
"loss": 0.2877,
"step": 1670
},
{
"epoch": 1.2,
"grad_norm": 2.5027403831481934,
"learning_rate": 6e-05,
"loss": 0.3172,
"step": 1680
},
{
"epoch": 1.207142857142857,
"grad_norm": 2.698021173477173,
"learning_rate": 5.9761904761904766e-05,
"loss": 0.2689,
"step": 1690
},
{
"epoch": 1.2142857142857142,
"grad_norm": 1.9757822751998901,
"learning_rate": 5.9523809523809524e-05,
"loss": 0.2845,
"step": 1700
},
{
"epoch": 1.2214285714285715,
"grad_norm": 1.7587430477142334,
"learning_rate": 5.928571428571429e-05,
"loss": 0.3264,
"step": 1710
},
{
"epoch": 1.2285714285714286,
"grad_norm": 2.466726541519165,
"learning_rate": 5.904761904761905e-05,
"loss": 0.2311,
"step": 1720
},
{
"epoch": 1.2357142857142858,
"grad_norm": 2.041752815246582,
"learning_rate": 5.880952380952382e-05,
"loss": 0.2829,
"step": 1730
},
{
"epoch": 1.2428571428571429,
"grad_norm": 4.565363883972168,
"learning_rate": 5.8571428571428575e-05,
"loss": 0.2356,
"step": 1740
},
{
"epoch": 1.25,
"grad_norm": 3.1162221431732178,
"learning_rate": 5.833333333333334e-05,
"loss": 0.2719,
"step": 1750
},
{
"epoch": 1.2571428571428571,
"grad_norm": 4.6578779220581055,
"learning_rate": 5.8095238095238104e-05,
"loss": 0.2477,
"step": 1760
},
{
"epoch": 1.2642857142857142,
"grad_norm": 2.782367706298828,
"learning_rate": 5.785714285714287e-05,
"loss": 0.3718,
"step": 1770
},
{
"epoch": 1.2714285714285714,
"grad_norm": 3.4065630435943604,
"learning_rate": 5.761904761904762e-05,
"loss": 0.3211,
"step": 1780
},
{
"epoch": 1.2785714285714285,
"grad_norm": 2.6382477283477783,
"learning_rate": 5.738095238095238e-05,
"loss": 0.3476,
"step": 1790
},
{
"epoch": 1.2857142857142856,
"grad_norm": 1.6404658555984497,
"learning_rate": 5.714285714285714e-05,
"loss": 0.3465,
"step": 1800
},
{
"epoch": 1.292857142857143,
"grad_norm": 2.675689935684204,
"learning_rate": 5.6904761904761906e-05,
"loss": 0.3027,
"step": 1810
},
{
"epoch": 1.3,
"grad_norm": 3.062359571456909,
"learning_rate": 5.666666666666667e-05,
"loss": 0.3043,
"step": 1820
},
{
"epoch": 1.3071428571428572,
"grad_norm": 3.5478060245513916,
"learning_rate": 5.642857142857143e-05,
"loss": 0.3022,
"step": 1830
},
{
"epoch": 1.3142857142857143,
"grad_norm": 3.2227487564086914,
"learning_rate": 5.619047619047619e-05,
"loss": 0.2776,
"step": 1840
},
{
"epoch": 1.3214285714285714,
"grad_norm": 1.7066378593444824,
"learning_rate": 5.595238095238096e-05,
"loss": 0.3345,
"step": 1850
},
{
"epoch": 1.3285714285714285,
"grad_norm": 3.7412548065185547,
"learning_rate": 5.571428571428572e-05,
"loss": 0.2886,
"step": 1860
},
{
"epoch": 1.3357142857142856,
"grad_norm": 6.295001983642578,
"learning_rate": 5.547619047619048e-05,
"loss": 0.2926,
"step": 1870
},
{
"epoch": 1.342857142857143,
"grad_norm": 2.9826364517211914,
"learning_rate": 5.5238095238095244e-05,
"loss": 0.2908,
"step": 1880
},
{
"epoch": 1.35,
"grad_norm": 2.300792932510376,
"learning_rate": 5.500000000000001e-05,
"loss": 0.3346,
"step": 1890
},
{
"epoch": 1.3571428571428572,
"grad_norm": 3.02064847946167,
"learning_rate": 5.4761904761904766e-05,
"loss": 0.298,
"step": 1900
},
{
"epoch": 1.3642857142857143,
"grad_norm": 2.3579671382904053,
"learning_rate": 5.4523809523809524e-05,
"loss": 0.3148,
"step": 1910
},
{
"epoch": 1.3714285714285714,
"grad_norm": 2.4837729930877686,
"learning_rate": 5.428571428571428e-05,
"loss": 0.2942,
"step": 1920
},
{
"epoch": 1.3785714285714286,
"grad_norm": 1.5465370416641235,
"learning_rate": 5.4047619047619046e-05,
"loss": 0.3141,
"step": 1930
},
{
"epoch": 1.3857142857142857,
"grad_norm": 3.10286808013916,
"learning_rate": 5.380952380952381e-05,
"loss": 0.2519,
"step": 1940
},
{
"epoch": 1.3928571428571428,
"grad_norm": 2.024423837661743,
"learning_rate": 5.3571428571428575e-05,
"loss": 0.2399,
"step": 1950
},
{
"epoch": 1.4,
"grad_norm": 3.297781467437744,
"learning_rate": 5.333333333333333e-05,
"loss": 0.3115,
"step": 1960
},
{
"epoch": 1.407142857142857,
"grad_norm": 12.920825004577637,
"learning_rate": 5.30952380952381e-05,
"loss": 0.3199,
"step": 1970
},
{
"epoch": 1.4142857142857144,
"grad_norm": 1.6162301301956177,
"learning_rate": 5.285714285714286e-05,
"loss": 0.228,
"step": 1980
},
{
"epoch": 1.4214285714285715,
"grad_norm": 3.950810194015503,
"learning_rate": 5.261904761904763e-05,
"loss": 0.3363,
"step": 1990
},
{
"epoch": 1.4285714285714286,
"grad_norm": 1.5367783308029175,
"learning_rate": 5.2380952380952384e-05,
"loss": 0.2328,
"step": 2000
},
{
"epoch": 1.4357142857142857,
"grad_norm": 1.4729127883911133,
"learning_rate": 5.214285714285715e-05,
"loss": 0.2892,
"step": 2010
},
{
"epoch": 1.4428571428571428,
"grad_norm": 3.7027993202209473,
"learning_rate": 5.1904761904761913e-05,
"loss": 0.2598,
"step": 2020
},
{
"epoch": 1.45,
"grad_norm": 4.046818256378174,
"learning_rate": 5.166666666666667e-05,
"loss": 0.3184,
"step": 2030
},
{
"epoch": 1.457142857142857,
"grad_norm": 5.93725061416626,
"learning_rate": 5.142857142857143e-05,
"loss": 0.2777,
"step": 2040
},
{
"epoch": 1.4642857142857144,
"grad_norm": 2.02740740776062,
"learning_rate": 5.119047619047619e-05,
"loss": 0.2632,
"step": 2050
},
{
"epoch": 1.4714285714285715,
"grad_norm": 1.166384220123291,
"learning_rate": 5.095238095238095e-05,
"loss": 0.2457,
"step": 2060
},
{
"epoch": 1.4785714285714286,
"grad_norm": 2.251250982284546,
"learning_rate": 5.0714285714285716e-05,
"loss": 0.2996,
"step": 2070
},
{
"epoch": 1.4857142857142858,
"grad_norm": 1.698370337486267,
"learning_rate": 5.047619047619048e-05,
"loss": 0.2418,
"step": 2080
},
{
"epoch": 1.4928571428571429,
"grad_norm": 1.665677785873413,
"learning_rate": 5.023809523809524e-05,
"loss": 0.2718,
"step": 2090
},
{
"epoch": 1.5,
"grad_norm": 2.417782783508301,
"learning_rate": 5e-05,
"loss": 0.263,
"step": 2100
},
{
"epoch": 1.5071428571428571,
"grad_norm": 3.639488935470581,
"learning_rate": 4.976190476190477e-05,
"loss": 0.2855,
"step": 2110
},
{
"epoch": 1.5142857142857142,
"grad_norm": 3.0365843772888184,
"learning_rate": 4.9523809523809525e-05,
"loss": 0.3073,
"step": 2120
},
{
"epoch": 1.5214285714285714,
"grad_norm": 4.151659965515137,
"learning_rate": 4.928571428571429e-05,
"loss": 0.2918,
"step": 2130
},
{
"epoch": 1.5285714285714285,
"grad_norm": 11.354039192199707,
"learning_rate": 4.904761904761905e-05,
"loss": 0.3428,
"step": 2140
},
{
"epoch": 1.5357142857142856,
"grad_norm": 3.684797525405884,
"learning_rate": 4.880952380952381e-05,
"loss": 0.2791,
"step": 2150
},
{
"epoch": 1.5428571428571427,
"grad_norm": 3.1418874263763428,
"learning_rate": 4.8571428571428576e-05,
"loss": 0.2386,
"step": 2160
},
{
"epoch": 1.55,
"grad_norm": 2.303208827972412,
"learning_rate": 4.8333333333333334e-05,
"loss": 0.3338,
"step": 2170
},
{
"epoch": 1.5571428571428572,
"grad_norm": 7.349155902862549,
"learning_rate": 4.80952380952381e-05,
"loss": 0.2595,
"step": 2180
},
{
"epoch": 1.5642857142857143,
"grad_norm": 2.4980363845825195,
"learning_rate": 4.785714285714286e-05,
"loss": 0.2499,
"step": 2190
},
{
"epoch": 1.5714285714285714,
"grad_norm": 3.0202884674072266,
"learning_rate": 4.761904761904762e-05,
"loss": 0.2608,
"step": 2200
},
{
"epoch": 1.5785714285714287,
"grad_norm": 7.224872589111328,
"learning_rate": 4.738095238095238e-05,
"loss": 0.3299,
"step": 2210
},
{
"epoch": 1.5857142857142859,
"grad_norm": 3.9374120235443115,
"learning_rate": 4.714285714285714e-05,
"loss": 0.3279,
"step": 2220
},
{
"epoch": 1.592857142857143,
"grad_norm": 5.184910774230957,
"learning_rate": 4.690476190476191e-05,
"loss": 0.2691,
"step": 2230
},
{
"epoch": 1.6,
"grad_norm": 5.877132415771484,
"learning_rate": 4.666666666666667e-05,
"loss": 0.2903,
"step": 2240
},
{
"epoch": 1.6071428571428572,
"grad_norm": 3.351426601409912,
"learning_rate": 4.642857142857143e-05,
"loss": 0.2771,
"step": 2250
},
{
"epoch": 1.6142857142857143,
"grad_norm": 1.7859790325164795,
"learning_rate": 4.6190476190476194e-05,
"loss": 0.349,
"step": 2260
},
{
"epoch": 1.6214285714285714,
"grad_norm": 1.9868186712265015,
"learning_rate": 4.595238095238095e-05,
"loss": 0.3505,
"step": 2270
},
{
"epoch": 1.6285714285714286,
"grad_norm": 3.075303316116333,
"learning_rate": 4.5714285714285716e-05,
"loss": 0.2753,
"step": 2280
},
{
"epoch": 1.6357142857142857,
"grad_norm": 2.1696481704711914,
"learning_rate": 4.547619047619048e-05,
"loss": 0.2501,
"step": 2290
},
{
"epoch": 1.6428571428571428,
"grad_norm": 2.9309706687927246,
"learning_rate": 4.523809523809524e-05,
"loss": 0.2904,
"step": 2300
},
{
"epoch": 1.65,
"grad_norm": 3.5121254920959473,
"learning_rate": 4.5e-05,
"loss": 0.2786,
"step": 2310
},
{
"epoch": 1.657142857142857,
"grad_norm": 1.3729180097579956,
"learning_rate": 4.476190476190477e-05,
"loss": 0.2589,
"step": 2320
},
{
"epoch": 1.6642857142857141,
"grad_norm": 4.851173400878906,
"learning_rate": 4.4523809523809525e-05,
"loss": 0.3489,
"step": 2330
},
{
"epoch": 1.6714285714285713,
"grad_norm": 4.376727104187012,
"learning_rate": 4.428571428571428e-05,
"loss": 0.3188,
"step": 2340
},
{
"epoch": 1.6785714285714286,
"grad_norm": 2.563443183898926,
"learning_rate": 4.404761904761905e-05,
"loss": 0.3499,
"step": 2350
},
{
"epoch": 1.6857142857142857,
"grad_norm": 2.768203020095825,
"learning_rate": 4.380952380952381e-05,
"loss": 0.2616,
"step": 2360
},
{
"epoch": 1.6928571428571428,
"grad_norm": 4.015028476715088,
"learning_rate": 4.3571428571428576e-05,
"loss": 0.2851,
"step": 2370
},
{
"epoch": 1.7,
"grad_norm": 1.4554102420806885,
"learning_rate": 4.3333333333333334e-05,
"loss": 0.274,
"step": 2380
},
{
"epoch": 1.7071428571428573,
"grad_norm": 2.553905963897705,
"learning_rate": 4.30952380952381e-05,
"loss": 0.2692,
"step": 2390
},
{
"epoch": 1.7142857142857144,
"grad_norm": 3.319626569747925,
"learning_rate": 4.2857142857142856e-05,
"loss": 0.3143,
"step": 2400
},
{
"epoch": 1.7214285714285715,
"grad_norm": 2.207327127456665,
"learning_rate": 4.261904761904762e-05,
"loss": 0.2548,
"step": 2410
},
{
"epoch": 1.7285714285714286,
"grad_norm": 3.1698238849639893,
"learning_rate": 4.2380952380952385e-05,
"loss": 0.3354,
"step": 2420
},
{
"epoch": 1.7357142857142858,
"grad_norm": 5.368778228759766,
"learning_rate": 4.214285714285714e-05,
"loss": 0.2665,
"step": 2430
},
{
"epoch": 1.7428571428571429,
"grad_norm": 1.1383056640625,
"learning_rate": 4.190476190476191e-05,
"loss": 0.2947,
"step": 2440
},
{
"epoch": 1.75,
"grad_norm": 3.5954413414001465,
"learning_rate": 4.166666666666667e-05,
"loss": 0.3362,
"step": 2450
},
{
"epoch": 1.7571428571428571,
"grad_norm": 3.8164236545562744,
"learning_rate": 4.1428571428571437e-05,
"loss": 0.2375,
"step": 2460
},
{
"epoch": 1.7642857142857142,
"grad_norm": 4.327922344207764,
"learning_rate": 4.119047619047619e-05,
"loss": 0.2776,
"step": 2470
},
{
"epoch": 1.7714285714285714,
"grad_norm": 2.2782583236694336,
"learning_rate": 4.095238095238095e-05,
"loss": 0.2407,
"step": 2480
},
{
"epoch": 1.7785714285714285,
"grad_norm": 1.7074918746948242,
"learning_rate": 4.0714285714285717e-05,
"loss": 0.2514,
"step": 2490
},
{
"epoch": 1.7857142857142856,
"grad_norm": 4.881974220275879,
"learning_rate": 4.047619047619048e-05,
"loss": 0.3229,
"step": 2500
},
{
"epoch": 1.7928571428571427,
"grad_norm": 4.011612892150879,
"learning_rate": 4.023809523809524e-05,
"loss": 0.3093,
"step": 2510
},
{
"epoch": 1.8,
"grad_norm": 2.8738279342651367,
"learning_rate": 4e-05,
"loss": 0.2844,
"step": 2520
},
{
"epoch": 1.8071428571428572,
"grad_norm": 2.006197214126587,
"learning_rate": 3.976190476190476e-05,
"loss": 0.2299,
"step": 2530
},
{
"epoch": 1.8142857142857143,
"grad_norm": 2.935206651687622,
"learning_rate": 3.9523809523809526e-05,
"loss": 0.2737,
"step": 2540
},
{
"epoch": 1.8214285714285714,
"grad_norm": 2.227921962738037,
"learning_rate": 3.928571428571429e-05,
"loss": 0.2654,
"step": 2550
},
{
"epoch": 1.8285714285714287,
"grad_norm": 5.1243367195129395,
"learning_rate": 3.904761904761905e-05,
"loss": 0.3842,
"step": 2560
},
{
"epoch": 1.8357142857142859,
"grad_norm": 2.0849359035491943,
"learning_rate": 3.880952380952381e-05,
"loss": 0.2521,
"step": 2570
},
{
"epoch": 1.842857142857143,
"grad_norm": 2.884455442428589,
"learning_rate": 3.857142857142858e-05,
"loss": 0.2784,
"step": 2580
},
{
"epoch": 1.85,
"grad_norm": 2.0989837646484375,
"learning_rate": 3.8333333333333334e-05,
"loss": 0.2359,
"step": 2590
},
{
"epoch": 1.8571428571428572,
"grad_norm": 3.074004650115967,
"learning_rate": 3.809523809523809e-05,
"loss": 0.2629,
"step": 2600
},
{
"epoch": 1.8642857142857143,
"grad_norm": 3.4337799549102783,
"learning_rate": 3.785714285714286e-05,
"loss": 0.2449,
"step": 2610
},
{
"epoch": 1.8714285714285714,
"grad_norm": 4.196152687072754,
"learning_rate": 3.761904761904762e-05,
"loss": 0.3082,
"step": 2620
},
{
"epoch": 1.8785714285714286,
"grad_norm": 2.139364719390869,
"learning_rate": 3.7380952380952386e-05,
"loss": 0.2816,
"step": 2630
},
{
"epoch": 1.8857142857142857,
"grad_norm": 6.65344762802124,
"learning_rate": 3.7142857142857143e-05,
"loss": 0.2729,
"step": 2640
},
{
"epoch": 1.8928571428571428,
"grad_norm": 1.8867508172988892,
"learning_rate": 3.690476190476191e-05,
"loss": 0.2668,
"step": 2650
},
{
"epoch": 1.9,
"grad_norm": 4.062267303466797,
"learning_rate": 3.6666666666666666e-05,
"loss": 0.2908,
"step": 2660
},
{
"epoch": 1.907142857142857,
"grad_norm": 4.145746231079102,
"learning_rate": 3.642857142857143e-05,
"loss": 0.3168,
"step": 2670
},
{
"epoch": 1.9142857142857141,
"grad_norm": 6.509150505065918,
"learning_rate": 3.619047619047619e-05,
"loss": 0.2744,
"step": 2680
},
{
"epoch": 1.9214285714285713,
"grad_norm": 2.091609239578247,
"learning_rate": 3.595238095238095e-05,
"loss": 0.2945,
"step": 2690
},
{
"epoch": 1.9285714285714286,
"grad_norm": 2.6082661151885986,
"learning_rate": 3.571428571428572e-05,
"loss": 0.2538,
"step": 2700
},
{
"epoch": 1.9357142857142857,
"grad_norm": 3.4130637645721436,
"learning_rate": 3.547619047619048e-05,
"loss": 0.2726,
"step": 2710
},
{
"epoch": 1.9428571428571428,
"grad_norm": 5.003239631652832,
"learning_rate": 3.523809523809524e-05,
"loss": 0.2697,
"step": 2720
},
{
"epoch": 1.95,
"grad_norm": 2.7061688899993896,
"learning_rate": 3.5e-05,
"loss": 0.2756,
"step": 2730
},
{
"epoch": 1.9571428571428573,
"grad_norm": 2.0089480876922607,
"learning_rate": 3.476190476190476e-05,
"loss": 0.2869,
"step": 2740
},
{
"epoch": 1.9642857142857144,
"grad_norm": 2.2568588256835938,
"learning_rate": 3.4523809523809526e-05,
"loss": 0.2687,
"step": 2750
},
{
"epoch": 1.9714285714285715,
"grad_norm": 2.0823915004730225,
"learning_rate": 3.428571428571429e-05,
"loss": 0.263,
"step": 2760
},
{
"epoch": 1.9785714285714286,
"grad_norm": 1.738771915435791,
"learning_rate": 3.404761904761905e-05,
"loss": 0.2283,
"step": 2770
},
{
"epoch": 1.9857142857142858,
"grad_norm": 4.0020623207092285,
"learning_rate": 3.380952380952381e-05,
"loss": 0.3286,
"step": 2780
},
{
"epoch": 1.9928571428571429,
"grad_norm": 2.4708666801452637,
"learning_rate": 3.357142857142857e-05,
"loss": 0.279,
"step": 2790
},
{
"epoch": 2.0,
"grad_norm": 1.950387716293335,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.3078,
"step": 2800
},
{
"epoch": 2.0,
"eval_loss": 0.25949418544769287,
"eval_runtime": 68.2677,
"eval_samples_per_second": 2.93,
"eval_steps_per_second": 0.366,
"step": 2800
},
{
"epoch": 2.007142857142857,
"grad_norm": 2.9532036781311035,
"learning_rate": 3.309523809523809e-05,
"loss": 0.2379,
"step": 2810
},
{
"epoch": 2.0142857142857142,
"grad_norm": 2.3207550048828125,
"learning_rate": 3.285714285714286e-05,
"loss": 0.3187,
"step": 2820
},
{
"epoch": 2.0214285714285714,
"grad_norm": 1.4604542255401611,
"learning_rate": 3.261904761904762e-05,
"loss": 0.2192,
"step": 2830
},
{
"epoch": 2.0285714285714285,
"grad_norm": 2.958681106567383,
"learning_rate": 3.2380952380952386e-05,
"loss": 0.2237,
"step": 2840
},
{
"epoch": 2.0357142857142856,
"grad_norm": 4.185633182525635,
"learning_rate": 3.2142857142857144e-05,
"loss": 0.3259,
"step": 2850
},
{
"epoch": 2.0428571428571427,
"grad_norm": 3.2991695404052734,
"learning_rate": 3.19047619047619e-05,
"loss": 0.1868,
"step": 2860
},
{
"epoch": 2.05,
"grad_norm": 2.3345556259155273,
"learning_rate": 3.1666666666666666e-05,
"loss": 0.2627,
"step": 2870
},
{
"epoch": 2.057142857142857,
"grad_norm": 1.6416515111923218,
"learning_rate": 3.142857142857143e-05,
"loss": 0.2417,
"step": 2880
},
{
"epoch": 2.064285714285714,
"grad_norm": 2.0688915252685547,
"learning_rate": 3.1190476190476195e-05,
"loss": 0.2359,
"step": 2890
},
{
"epoch": 2.0714285714285716,
"grad_norm": 2.668055534362793,
"learning_rate": 3.095238095238095e-05,
"loss": 0.2879,
"step": 2900
},
{
"epoch": 2.0785714285714287,
"grad_norm": 2.9602174758911133,
"learning_rate": 3.071428571428572e-05,
"loss": 0.2061,
"step": 2910
},
{
"epoch": 2.085714285714286,
"grad_norm": 3.5890145301818848,
"learning_rate": 3.0476190476190482e-05,
"loss": 0.266,
"step": 2920
},
{
"epoch": 2.092857142857143,
"grad_norm": 2.9162445068359375,
"learning_rate": 3.0238095238095236e-05,
"loss": 0.234,
"step": 2930
},
{
"epoch": 2.1,
"grad_norm": 5.290048599243164,
"learning_rate": 3e-05,
"loss": 0.2562,
"step": 2940
},
{
"epoch": 2.107142857142857,
"grad_norm": 4.895926475524902,
"learning_rate": 2.9761904761904762e-05,
"loss": 0.2958,
"step": 2950
},
{
"epoch": 2.1142857142857143,
"grad_norm": 5.531802177429199,
"learning_rate": 2.9523809523809526e-05,
"loss": 0.2388,
"step": 2960
},
{
"epoch": 2.1214285714285714,
"grad_norm": 2.0870862007141113,
"learning_rate": 2.9285714285714288e-05,
"loss": 0.3,
"step": 2970
},
{
"epoch": 2.1285714285714286,
"grad_norm": 1.2288978099822998,
"learning_rate": 2.9047619047619052e-05,
"loss": 0.2691,
"step": 2980
},
{
"epoch": 2.1357142857142857,
"grad_norm": 3.008167028427124,
"learning_rate": 2.880952380952381e-05,
"loss": 0.3721,
"step": 2990
},
{
"epoch": 2.142857142857143,
"grad_norm": 4.923089027404785,
"learning_rate": 2.857142857142857e-05,
"loss": 0.2516,
"step": 3000
},
{
"epoch": 2.15,
"grad_norm": 3.881120443344116,
"learning_rate": 2.8333333333333335e-05,
"loss": 0.275,
"step": 3010
},
{
"epoch": 2.157142857142857,
"grad_norm": 4.30492639541626,
"learning_rate": 2.8095238095238096e-05,
"loss": 0.2353,
"step": 3020
},
{
"epoch": 2.164285714285714,
"grad_norm": 3.907482862472534,
"learning_rate": 2.785714285714286e-05,
"loss": 0.2856,
"step": 3030
},
{
"epoch": 2.1714285714285713,
"grad_norm": 4.306923866271973,
"learning_rate": 2.7619047619047622e-05,
"loss": 0.2768,
"step": 3040
},
{
"epoch": 2.1785714285714284,
"grad_norm": 4.364368915557861,
"learning_rate": 2.7380952380952383e-05,
"loss": 0.2854,
"step": 3050
},
{
"epoch": 2.185714285714286,
"grad_norm": 3.7898383140563965,
"learning_rate": 2.714285714285714e-05,
"loss": 0.2168,
"step": 3060
},
{
"epoch": 2.192857142857143,
"grad_norm": 5.487813472747803,
"learning_rate": 2.6904761904761905e-05,
"loss": 0.2849,
"step": 3070
},
{
"epoch": 2.2,
"grad_norm": 3.3784303665161133,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.2431,
"step": 3080
},
{
"epoch": 2.2071428571428573,
"grad_norm": 4.644072532653809,
"learning_rate": 2.642857142857143e-05,
"loss": 0.3144,
"step": 3090
},
{
"epoch": 2.2142857142857144,
"grad_norm": 9.652532577514648,
"learning_rate": 2.6190476190476192e-05,
"loss": 0.3005,
"step": 3100
},
{
"epoch": 2.2214285714285715,
"grad_norm": 2.1053450107574463,
"learning_rate": 2.5952380952380957e-05,
"loss": 0.2487,
"step": 3110
},
{
"epoch": 2.2285714285714286,
"grad_norm": 3.275733709335327,
"learning_rate": 2.5714285714285714e-05,
"loss": 0.2863,
"step": 3120
},
{
"epoch": 2.2357142857142858,
"grad_norm": 2.46889066696167,
"learning_rate": 2.5476190476190476e-05,
"loss": 0.2497,
"step": 3130
},
{
"epoch": 2.242857142857143,
"grad_norm": 4.135228633880615,
"learning_rate": 2.523809523809524e-05,
"loss": 0.2368,
"step": 3140
},
{
"epoch": 2.25,
"grad_norm": 5.32862663269043,
"learning_rate": 2.5e-05,
"loss": 0.2336,
"step": 3150
},
{
"epoch": 2.257142857142857,
"grad_norm": 5.055444717407227,
"learning_rate": 2.4761904761904762e-05,
"loss": 0.2474,
"step": 3160
},
{
"epoch": 2.2642857142857142,
"grad_norm": 9.018756866455078,
"learning_rate": 2.4523809523809523e-05,
"loss": 0.2637,
"step": 3170
},
{
"epoch": 2.2714285714285714,
"grad_norm": 2.1222236156463623,
"learning_rate": 2.4285714285714288e-05,
"loss": 0.2242,
"step": 3180
},
{
"epoch": 2.2785714285714285,
"grad_norm": 5.100944519042969,
"learning_rate": 2.404761904761905e-05,
"loss": 0.2634,
"step": 3190
},
{
"epoch": 2.2857142857142856,
"grad_norm": 5.248310565948486,
"learning_rate": 2.380952380952381e-05,
"loss": 0.2557,
"step": 3200
},
{
"epoch": 2.2928571428571427,
"grad_norm": 4.782702922821045,
"learning_rate": 2.357142857142857e-05,
"loss": 0.2313,
"step": 3210
},
{
"epoch": 2.3,
"grad_norm": 5.026177883148193,
"learning_rate": 2.3333333333333336e-05,
"loss": 0.3054,
"step": 3220
},
{
"epoch": 2.307142857142857,
"grad_norm": 2.8892881870269775,
"learning_rate": 2.3095238095238097e-05,
"loss": 0.2477,
"step": 3230
},
{
"epoch": 2.314285714285714,
"grad_norm": 3.408980131149292,
"learning_rate": 2.2857142857142858e-05,
"loss": 0.2381,
"step": 3240
},
{
"epoch": 2.3214285714285716,
"grad_norm": 2.3508706092834473,
"learning_rate": 2.261904761904762e-05,
"loss": 0.1975,
"step": 3250
},
{
"epoch": 2.3285714285714287,
"grad_norm": 1.3427135944366455,
"learning_rate": 2.2380952380952384e-05,
"loss": 0.2334,
"step": 3260
},
{
"epoch": 2.335714285714286,
"grad_norm": 3.8124921321868896,
"learning_rate": 2.214285714285714e-05,
"loss": 0.2983,
"step": 3270
},
{
"epoch": 2.342857142857143,
"grad_norm": 3.923583507537842,
"learning_rate": 2.1904761904761906e-05,
"loss": 0.278,
"step": 3280
},
{
"epoch": 2.35,
"grad_norm": 3.14457106590271,
"learning_rate": 2.1666666666666667e-05,
"loss": 0.2409,
"step": 3290
},
{
"epoch": 2.357142857142857,
"grad_norm": 4.7329230308532715,
"learning_rate": 2.1428571428571428e-05,
"loss": 0.2861,
"step": 3300
},
{
"epoch": 2.3642857142857143,
"grad_norm": 5.68926477432251,
"learning_rate": 2.1190476190476193e-05,
"loss": 0.2904,
"step": 3310
},
{
"epoch": 2.3714285714285714,
"grad_norm": 8.324594497680664,
"learning_rate": 2.0952380952380954e-05,
"loss": 0.3381,
"step": 3320
},
{
"epoch": 2.3785714285714286,
"grad_norm": 5.043862819671631,
"learning_rate": 2.0714285714285718e-05,
"loss": 0.2893,
"step": 3330
},
{
"epoch": 2.3857142857142857,
"grad_norm": 1.9703952074050903,
"learning_rate": 2.0476190476190476e-05,
"loss": 0.1997,
"step": 3340
},
{
"epoch": 2.392857142857143,
"grad_norm": 3.2141919136047363,
"learning_rate": 2.023809523809524e-05,
"loss": 0.2547,
"step": 3350
},
{
"epoch": 2.4,
"grad_norm": 3.0841588973999023,
"learning_rate": 2e-05,
"loss": 0.2527,
"step": 3360
},
{
"epoch": 2.407142857142857,
"grad_norm": 2.329394817352295,
"learning_rate": 1.9761904761904763e-05,
"loss": 0.2228,
"step": 3370
},
{
"epoch": 2.414285714285714,
"grad_norm": 2.290874719619751,
"learning_rate": 1.9523809523809524e-05,
"loss": 0.2613,
"step": 3380
},
{
"epoch": 2.4214285714285713,
"grad_norm": 1.3011747598648071,
"learning_rate": 1.928571428571429e-05,
"loss": 0.22,
"step": 3390
},
{
"epoch": 2.4285714285714284,
"grad_norm": 3.5783472061157227,
"learning_rate": 1.9047619047619046e-05,
"loss": 0.2729,
"step": 3400
},
{
"epoch": 2.435714285714286,
"grad_norm": 3.1245083808898926,
"learning_rate": 1.880952380952381e-05,
"loss": 0.2551,
"step": 3410
},
{
"epoch": 2.442857142857143,
"grad_norm": 4.986431121826172,
"learning_rate": 1.8571428571428572e-05,
"loss": 0.3029,
"step": 3420
},
{
"epoch": 2.45,
"grad_norm": 1.8483643531799316,
"learning_rate": 1.8333333333333333e-05,
"loss": 0.2458,
"step": 3430
},
{
"epoch": 2.4571428571428573,
"grad_norm": 5.947587490081787,
"learning_rate": 1.8095238095238094e-05,
"loss": 0.2848,
"step": 3440
},
{
"epoch": 2.4642857142857144,
"grad_norm": 5.204590320587158,
"learning_rate": 1.785714285714286e-05,
"loss": 0.3195,
"step": 3450
},
{
"epoch": 2.4714285714285715,
"grad_norm": 2.213702440261841,
"learning_rate": 1.761904761904762e-05,
"loss": 0.2624,
"step": 3460
},
{
"epoch": 2.4785714285714286,
"grad_norm": 2.896759033203125,
"learning_rate": 1.738095238095238e-05,
"loss": 0.3234,
"step": 3470
},
{
"epoch": 2.4857142857142858,
"grad_norm": 2.1793391704559326,
"learning_rate": 1.7142857142857145e-05,
"loss": 0.2136,
"step": 3480
},
{
"epoch": 2.492857142857143,
"grad_norm": 5.157558917999268,
"learning_rate": 1.6904761904761906e-05,
"loss": 0.2655,
"step": 3490
},
{
"epoch": 2.5,
"grad_norm": 2.9449174404144287,
"learning_rate": 1.6666666666666667e-05,
"loss": 0.2846,
"step": 3500
},
{
"epoch": 2.507142857142857,
"grad_norm": 3.97674298286438,
"learning_rate": 1.642857142857143e-05,
"loss": 0.2423,
"step": 3510
},
{
"epoch": 2.5142857142857142,
"grad_norm": 5.1696553230285645,
"learning_rate": 1.6190476190476193e-05,
"loss": 0.2468,
"step": 3520
},
{
"epoch": 2.5214285714285714,
"grad_norm": 2.9929051399230957,
"learning_rate": 1.595238095238095e-05,
"loss": 0.2912,
"step": 3530
},
{
"epoch": 2.5285714285714285,
"grad_norm": 3.287910223007202,
"learning_rate": 1.5714285714285715e-05,
"loss": 0.2436,
"step": 3540
},
{
"epoch": 2.5357142857142856,
"grad_norm": 4.533077239990234,
"learning_rate": 1.5476190476190476e-05,
"loss": 0.3301,
"step": 3550
},
{
"epoch": 2.5428571428571427,
"grad_norm": 4.241918087005615,
"learning_rate": 1.5238095238095241e-05,
"loss": 0.3111,
"step": 3560
},
{
"epoch": 2.55,
"grad_norm": 2.0946638584136963,
"learning_rate": 1.5e-05,
"loss": 0.2098,
"step": 3570
},
{
"epoch": 2.557142857142857,
"grad_norm": 4.646461486816406,
"learning_rate": 1.4761904761904763e-05,
"loss": 0.1745,
"step": 3580
},
{
"epoch": 2.564285714285714,
"grad_norm": 4.6408233642578125,
"learning_rate": 1.4523809523809526e-05,
"loss": 0.212,
"step": 3590
},
{
"epoch": 2.571428571428571,
"grad_norm": 6.365333080291748,
"learning_rate": 1.4285714285714285e-05,
"loss": 0.2445,
"step": 3600
},
{
"epoch": 2.5785714285714287,
"grad_norm": 5.597777366638184,
"learning_rate": 1.4047619047619048e-05,
"loss": 0.2973,
"step": 3610
},
{
"epoch": 2.585714285714286,
"grad_norm": 4.016028881072998,
"learning_rate": 1.3809523809523811e-05,
"loss": 0.2514,
"step": 3620
},
{
"epoch": 2.592857142857143,
"grad_norm": 3.058223009109497,
"learning_rate": 1.357142857142857e-05,
"loss": 0.2766,
"step": 3630
},
{
"epoch": 2.6,
"grad_norm": 2.698791027069092,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.2042,
"step": 3640
},
{
"epoch": 2.607142857142857,
"grad_norm": 5.222252368927002,
"learning_rate": 1.3095238095238096e-05,
"loss": 0.2953,
"step": 3650
},
{
"epoch": 2.6142857142857143,
"grad_norm": 1.2256240844726562,
"learning_rate": 1.2857142857142857e-05,
"loss": 0.2938,
"step": 3660
},
{
"epoch": 2.6214285714285714,
"grad_norm": 2.3339574337005615,
"learning_rate": 1.261904761904762e-05,
"loss": 0.2417,
"step": 3670
},
{
"epoch": 2.6285714285714286,
"grad_norm": 6.824151039123535,
"learning_rate": 1.2380952380952381e-05,
"loss": 0.259,
"step": 3680
},
{
"epoch": 2.6357142857142857,
"grad_norm": 5.956028461456299,
"learning_rate": 1.2142857142857144e-05,
"loss": 0.2088,
"step": 3690
},
{
"epoch": 2.642857142857143,
"grad_norm": 1.825620174407959,
"learning_rate": 1.1904761904761905e-05,
"loss": 0.2426,
"step": 3700
},
{
"epoch": 2.65,
"grad_norm": 1.8482383489608765,
"learning_rate": 1.1666666666666668e-05,
"loss": 0.2352,
"step": 3710
},
{
"epoch": 2.657142857142857,
"grad_norm": 5.8184919357299805,
"learning_rate": 1.1428571428571429e-05,
"loss": 0.2556,
"step": 3720
},
{
"epoch": 2.664285714285714,
"grad_norm": 3.534730911254883,
"learning_rate": 1.1190476190476192e-05,
"loss": 0.3155,
"step": 3730
},
{
"epoch": 2.6714285714285713,
"grad_norm": 1.4020763635635376,
"learning_rate": 1.0952380952380953e-05,
"loss": 0.21,
"step": 3740
},
{
"epoch": 2.678571428571429,
"grad_norm": 3.11566162109375,
"learning_rate": 1.0714285714285714e-05,
"loss": 0.2764,
"step": 3750
},
{
"epoch": 2.685714285714286,
"grad_norm": 4.930715084075928,
"learning_rate": 1.0476190476190477e-05,
"loss": 0.3339,
"step": 3760
},
{
"epoch": 2.692857142857143,
"grad_norm": 4.198614597320557,
"learning_rate": 1.0238095238095238e-05,
"loss": 0.2917,
"step": 3770
},
{
"epoch": 2.7,
"grad_norm": 1.6880650520324707,
"learning_rate": 1e-05,
"loss": 0.2212,
"step": 3780
},
{
"epoch": 2.7071428571428573,
"grad_norm": 3.580629587173462,
"learning_rate": 9.761904761904762e-06,
"loss": 0.2547,
"step": 3790
},
{
"epoch": 2.7142857142857144,
"grad_norm": 2.764923095703125,
"learning_rate": 9.523809523809523e-06,
"loss": 0.2286,
"step": 3800
},
{
"epoch": 2.7214285714285715,
"grad_norm": 2.3037569522857666,
"learning_rate": 9.285714285714286e-06,
"loss": 0.1979,
"step": 3810
},
{
"epoch": 2.7285714285714286,
"grad_norm": 3.4528393745422363,
"learning_rate": 9.047619047619047e-06,
"loss": 0.2377,
"step": 3820
},
{
"epoch": 2.7357142857142858,
"grad_norm": 2.326822519302368,
"learning_rate": 8.80952380952381e-06,
"loss": 0.2457,
"step": 3830
},
{
"epoch": 2.742857142857143,
"grad_norm": 3.2105257511138916,
"learning_rate": 8.571428571428573e-06,
"loss": 0.2553,
"step": 3840
},
{
"epoch": 2.75,
"grad_norm": 7.470506191253662,
"learning_rate": 8.333333333333334e-06,
"loss": 0.2285,
"step": 3850
},
{
"epoch": 2.757142857142857,
"grad_norm": 7.341104984283447,
"learning_rate": 8.095238095238097e-06,
"loss": 0.2454,
"step": 3860
},
{
"epoch": 2.7642857142857142,
"grad_norm": 5.3889055252075195,
"learning_rate": 7.857142857142858e-06,
"loss": 0.2754,
"step": 3870
},
{
"epoch": 2.7714285714285714,
"grad_norm": 4.78093957901001,
"learning_rate": 7.6190476190476205e-06,
"loss": 0.3796,
"step": 3880
},
{
"epoch": 2.7785714285714285,
"grad_norm": 2.088247060775757,
"learning_rate": 7.380952380952382e-06,
"loss": 0.2137,
"step": 3890
},
{
"epoch": 2.7857142857142856,
"grad_norm": 2.5746066570281982,
"learning_rate": 7.142857142857143e-06,
"loss": 0.3069,
"step": 3900
},
{
"epoch": 2.7928571428571427,
"grad_norm": 1.7371402978897095,
"learning_rate": 6.9047619047619055e-06,
"loss": 0.2522,
"step": 3910
},
{
"epoch": 2.8,
"grad_norm": 1.5857880115509033,
"learning_rate": 6.666666666666667e-06,
"loss": 0.3117,
"step": 3920
},
{
"epoch": 2.807142857142857,
"grad_norm": 3.37313175201416,
"learning_rate": 6.428571428571429e-06,
"loss": 0.3156,
"step": 3930
},
{
"epoch": 2.814285714285714,
"grad_norm": 4.180731773376465,
"learning_rate": 6.190476190476191e-06,
"loss": 0.279,
"step": 3940
},
{
"epoch": 2.821428571428571,
"grad_norm": 5.761970520019531,
"learning_rate": 5.9523809523809525e-06,
"loss": 0.2478,
"step": 3950
},
{
"epoch": 2.8285714285714287,
"grad_norm": 3.018902063369751,
"learning_rate": 5.7142857142857145e-06,
"loss": 0.2224,
"step": 3960
},
{
"epoch": 2.835714285714286,
"grad_norm": 2.3872735500335693,
"learning_rate": 5.4761904761904765e-06,
"loss": 0.1913,
"step": 3970
},
{
"epoch": 2.842857142857143,
"grad_norm": 4.139348983764648,
"learning_rate": 5.2380952380952384e-06,
"loss": 0.2815,
"step": 3980
},
{
"epoch": 2.85,
"grad_norm": 7.203916072845459,
"learning_rate": 5e-06,
"loss": 0.3189,
"step": 3990
},
{
"epoch": 2.857142857142857,
"grad_norm": 5.5615715980529785,
"learning_rate": 4.7619047619047615e-06,
"loss": 0.2584,
"step": 4000
},
{
"epoch": 2.8642857142857143,
"grad_norm": 1.6023494005203247,
"learning_rate": 4.5238095238095235e-06,
"loss": 0.2672,
"step": 4010
},
{
"epoch": 2.8714285714285714,
"grad_norm": 5.98526668548584,
"learning_rate": 4.285714285714286e-06,
"loss": 0.2492,
"step": 4020
},
{
"epoch": 2.8785714285714286,
"grad_norm": 2.8703975677490234,
"learning_rate": 4.047619047619048e-06,
"loss": 0.2164,
"step": 4030
},
{
"epoch": 2.8857142857142857,
"grad_norm": 4.569535732269287,
"learning_rate": 3.8095238095238102e-06,
"loss": 0.2991,
"step": 4040
},
{
"epoch": 2.892857142857143,
"grad_norm": 3.0604588985443115,
"learning_rate": 3.5714285714285714e-06,
"loss": 0.2416,
"step": 4050
},
{
"epoch": 2.9,
"grad_norm": 3.326860189437866,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.2035,
"step": 4060
},
{
"epoch": 2.907142857142857,
"grad_norm": 4.7924675941467285,
"learning_rate": 3.0952380952380953e-06,
"loss": 0.3382,
"step": 4070
},
{
"epoch": 2.914285714285714,
"grad_norm": 1.4087116718292236,
"learning_rate": 2.8571428571428573e-06,
"loss": 0.1943,
"step": 4080
},
{
"epoch": 2.9214285714285713,
"grad_norm": 2.154977798461914,
"learning_rate": 2.6190476190476192e-06,
"loss": 0.2088,
"step": 4090
},
{
"epoch": 2.928571428571429,
"grad_norm": 2.5638303756713867,
"learning_rate": 2.3809523809523808e-06,
"loss": 0.2024,
"step": 4100
},
{
"epoch": 2.935714285714286,
"grad_norm": 5.622275352478027,
"learning_rate": 2.142857142857143e-06,
"loss": 0.2493,
"step": 4110
},
{
"epoch": 2.942857142857143,
"grad_norm": 4.387233734130859,
"learning_rate": 1.9047619047619051e-06,
"loss": 0.2102,
"step": 4120
},
{
"epoch": 2.95,
"grad_norm": 3.105154514312744,
"learning_rate": 1.6666666666666667e-06,
"loss": 0.2266,
"step": 4130
},
{
"epoch": 2.9571428571428573,
"grad_norm": 2.8728506565093994,
"learning_rate": 1.4285714285714286e-06,
"loss": 0.1978,
"step": 4140
},
{
"epoch": 2.9642857142857144,
"grad_norm": 4.60084867477417,
"learning_rate": 1.1904761904761904e-06,
"loss": 0.2214,
"step": 4150
},
{
"epoch": 2.9714285714285715,
"grad_norm": 3.7164621353149414,
"learning_rate": 9.523809523809526e-07,
"loss": 0.2272,
"step": 4160
},
{
"epoch": 2.9785714285714286,
"grad_norm": 2.304844856262207,
"learning_rate": 7.142857142857143e-07,
"loss": 0.2652,
"step": 4170
},
{
"epoch": 2.9857142857142858,
"grad_norm": 2.9933526515960693,
"learning_rate": 4.761904761904763e-07,
"loss": 0.2606,
"step": 4180
},
{
"epoch": 2.992857142857143,
"grad_norm": 2.8444643020629883,
"learning_rate": 2.3809523809523814e-07,
"loss": 0.2501,
"step": 4190
},
{
"epoch": 3.0,
"grad_norm": 3.5711305141448975,
"learning_rate": 0.0,
"loss": 0.2887,
"step": 4200
},
{
"epoch": 3.0,
"eval_loss": 0.23848384618759155,
"eval_runtime": 68.3028,
"eval_samples_per_second": 2.928,
"eval_steps_per_second": 0.366,
"step": 4200
}
],
"logging_steps": 10,
"max_steps": 4200,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 2.5133825654784e+16,
"train_batch_size": 2,
"trial_name": null,
"trial_params": null
}