| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9993739889922224, | |
| "eval_steps": 500, | |
| "global_step": 1559, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0006410352719642222, | |
| "grad_norm": 0.10498046875, | |
| "learning_rate": 1.282051282051282e-06, | |
| "loss": 1.8493, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.003205176359821111, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 6.41025641025641e-06, | |
| "loss": 1.8865, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.006410352719642222, | |
| "grad_norm": 0.1005859375, | |
| "learning_rate": 1.282051282051282e-05, | |
| "loss": 1.8383, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.009615529079463333, | |
| "grad_norm": 0.10693359375, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 1.8385, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.012820705439284444, | |
| "grad_norm": 0.1103515625, | |
| "learning_rate": 2.564102564102564e-05, | |
| "loss": 1.8346, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.016025881799105555, | |
| "grad_norm": 0.1298828125, | |
| "learning_rate": 3.205128205128206e-05, | |
| "loss": 1.8127, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.019231058158926666, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 1.7981, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.022436234518747777, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 4.4871794871794874e-05, | |
| "loss": 1.7907, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.025641410878568888, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 5.128205128205128e-05, | |
| "loss": 1.7468, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.02884658723839, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 5.769230769230769e-05, | |
| "loss": 1.7105, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03205176359821111, | |
| "grad_norm": 0.126953125, | |
| "learning_rate": 6.410256410256412e-05, | |
| "loss": 1.6887, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.035256939958032224, | |
| "grad_norm": 0.107421875, | |
| "learning_rate": 7.051282051282052e-05, | |
| "loss": 1.6757, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.03846211631785333, | |
| "grad_norm": 0.10009765625, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 1.6736, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.041667292677674446, | |
| "grad_norm": 0.078125, | |
| "learning_rate": 8.333333333333334e-05, | |
| "loss": 1.6252, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.04487246903749555, | |
| "grad_norm": 0.06201171875, | |
| "learning_rate": 8.974358974358975e-05, | |
| "loss": 1.5655, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04807764539731667, | |
| "grad_norm": 0.049072265625, | |
| "learning_rate": 9.615384615384617e-05, | |
| "loss": 1.5646, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.051282821757137775, | |
| "grad_norm": 0.04345703125, | |
| "learning_rate": 0.00010256410256410256, | |
| "loss": 1.5861, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05448799811695889, | |
| "grad_norm": 0.039794921875, | |
| "learning_rate": 0.00010897435897435896, | |
| "loss": 1.5379, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.05769317447678, | |
| "grad_norm": 0.0400390625, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 1.5557, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.06089835083660111, | |
| "grad_norm": 0.037841796875, | |
| "learning_rate": 0.00012179487179487179, | |
| "loss": 1.5102, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.06410352719642222, | |
| "grad_norm": 0.038330078125, | |
| "learning_rate": 0.00012820512820512823, | |
| "loss": 1.5048, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06730870355624333, | |
| "grad_norm": 0.033935546875, | |
| "learning_rate": 0.00013461538461538464, | |
| "loss": 1.5127, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.07051387991606445, | |
| "grad_norm": 0.03173828125, | |
| "learning_rate": 0.00014102564102564104, | |
| "loss": 1.5161, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.07371905627588556, | |
| "grad_norm": 0.0301513671875, | |
| "learning_rate": 0.00014743589743589745, | |
| "loss": 1.4948, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.07692423263570666, | |
| "grad_norm": 0.03125, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 1.4584, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.08012940899552777, | |
| "grad_norm": 0.029052734375, | |
| "learning_rate": 0.00016025641025641028, | |
| "loss": 1.4704, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.08333458535534889, | |
| "grad_norm": 0.0279541015625, | |
| "learning_rate": 0.0001666666666666667, | |
| "loss": 1.4411, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.08653976171517, | |
| "grad_norm": 0.0263671875, | |
| "learning_rate": 0.0001730769230769231, | |
| "loss": 1.4723, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.0897449380749911, | |
| "grad_norm": 0.02685546875, | |
| "learning_rate": 0.0001794871794871795, | |
| "loss": 1.4505, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.09295011443481223, | |
| "grad_norm": 0.0291748046875, | |
| "learning_rate": 0.0001858974358974359, | |
| "loss": 1.4367, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.09615529079463334, | |
| "grad_norm": 0.0262451171875, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 1.4291, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09936046715445444, | |
| "grad_norm": 0.0390625, | |
| "learning_rate": 0.00019871794871794874, | |
| "loss": 1.4075, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.10256564351427555, | |
| "grad_norm": 0.03857421875, | |
| "learning_rate": 0.00019999598882613538, | |
| "loss": 1.4203, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.10577081987409667, | |
| "grad_norm": 0.029541015625, | |
| "learning_rate": 0.00019997969398381457, | |
| "loss": 1.4188, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.10897599623391778, | |
| "grad_norm": 0.025146484375, | |
| "learning_rate": 0.00019995086681563726, | |
| "loss": 1.4512, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.11218117259373889, | |
| "grad_norm": 0.025146484375, | |
| "learning_rate": 0.0001999095109350519, | |
| "loss": 1.417, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.11538634895356, | |
| "grad_norm": 0.02734375, | |
| "learning_rate": 0.0001998556315259648, | |
| "loss": 1.4309, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.11859152531338112, | |
| "grad_norm": 0.0255126953125, | |
| "learning_rate": 0.00019978923534209054, | |
| "loss": 1.4201, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.12179670167320222, | |
| "grad_norm": 0.0286865234375, | |
| "learning_rate": 0.00019971033070610518, | |
| "loss": 1.4187, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.12500187803302334, | |
| "grad_norm": 0.030517578125, | |
| "learning_rate": 0.0001996189275086033, | |
| "loss": 1.4153, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.12820705439284444, | |
| "grad_norm": 0.0272216796875, | |
| "learning_rate": 0.00019951503720685784, | |
| "loss": 1.4279, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.13141223075266556, | |
| "grad_norm": 0.0267333984375, | |
| "learning_rate": 0.0001993986728233844, | |
| "loss": 1.4052, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.13461740711248665, | |
| "grad_norm": 0.0264892578125, | |
| "learning_rate": 0.0001992698489443085, | |
| "loss": 1.3943, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.13782258347230777, | |
| "grad_norm": 0.032470703125, | |
| "learning_rate": 0.0001991285817175375, | |
| "loss": 1.3931, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.1410277598321289, | |
| "grad_norm": 0.0291748046875, | |
| "learning_rate": 0.0001989748888507363, | |
| "loss": 1.3931, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.14423293619195, | |
| "grad_norm": 0.03125, | |
| "learning_rate": 0.00019880878960910772, | |
| "loss": 1.3899, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.1474381125517711, | |
| "grad_norm": 0.0322265625, | |
| "learning_rate": 0.0001986303048129778, | |
| "loss": 1.4305, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.15064328891159223, | |
| "grad_norm": 0.033203125, | |
| "learning_rate": 0.0001984394568351858, | |
| "loss": 1.4028, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.15384846527141333, | |
| "grad_norm": 0.03369140625, | |
| "learning_rate": 0.00019823626959827997, | |
| "loss": 1.3758, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.15705364163123445, | |
| "grad_norm": 0.041015625, | |
| "learning_rate": 0.0001980207685715186, | |
| "loss": 1.407, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.16025881799105554, | |
| "grad_norm": 0.034912109375, | |
| "learning_rate": 0.00019779298076767795, | |
| "loss": 1.3923, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.16346399435087666, | |
| "grad_norm": 0.047119140625, | |
| "learning_rate": 0.00019755293473966572, | |
| "loss": 1.3967, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.16666917071069778, | |
| "grad_norm": 0.043701171875, | |
| "learning_rate": 0.00019730066057694235, | |
| "loss": 1.4007, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.16987434707051888, | |
| "grad_norm": 0.050048828125, | |
| "learning_rate": 0.00019703618990174918, | |
| "loss": 1.3978, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.17307952343034, | |
| "grad_norm": 0.048095703125, | |
| "learning_rate": 0.00019675955586514468, | |
| "loss": 1.3744, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.17628469979016112, | |
| "grad_norm": 0.033935546875, | |
| "learning_rate": 0.00019647079314284897, | |
| "loss": 1.3929, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.1794898761499822, | |
| "grad_norm": 0.033935546875, | |
| "learning_rate": 0.0001961699379308974, | |
| "loss": 1.4031, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.18269505250980334, | |
| "grad_norm": 0.04052734375, | |
| "learning_rate": 0.0001958570279411032, | |
| "loss": 1.3813, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.18590022886962446, | |
| "grad_norm": 0.052734375, | |
| "learning_rate": 0.00019553210239633056, | |
| "loss": 1.3956, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.18910540522944555, | |
| "grad_norm": 0.048095703125, | |
| "learning_rate": 0.00019519520202557797, | |
| "loss": 1.3988, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.19231058158926667, | |
| "grad_norm": 0.037109375, | |
| "learning_rate": 0.00019484636905887296, | |
| "loss": 1.3925, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.19551575794908777, | |
| "grad_norm": 0.036865234375, | |
| "learning_rate": 0.00019448564722197853, | |
| "loss": 1.376, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.1987209343089089, | |
| "grad_norm": 0.04052734375, | |
| "learning_rate": 0.00019411308173091228, | |
| "loss": 1.3974, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.20192611066873, | |
| "grad_norm": 0.052490234375, | |
| "learning_rate": 0.0001937287192862787, | |
| "loss": 1.3765, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.2051312870285511, | |
| "grad_norm": 0.059326171875, | |
| "learning_rate": 0.00019333260806741502, | |
| "loss": 1.3769, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.20833646338837222, | |
| "grad_norm": 0.052490234375, | |
| "learning_rate": 0.00019292479772635237, | |
| "loss": 1.3792, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.21154163974819334, | |
| "grad_norm": 0.048583984375, | |
| "learning_rate": 0.00019250533938159166, | |
| "loss": 1.3968, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.21474681610801444, | |
| "grad_norm": 0.040283203125, | |
| "learning_rate": 0.00019207428561169608, | |
| "loss": 1.38, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.21795199246783556, | |
| "grad_norm": 0.043701171875, | |
| "learning_rate": 0.0001916316904487005, | |
| "loss": 1.3737, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.22115716882765665, | |
| "grad_norm": 0.03759765625, | |
| "learning_rate": 0.00019117760937133844, | |
| "loss": 1.4065, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.22436234518747777, | |
| "grad_norm": 0.038330078125, | |
| "learning_rate": 0.00019071209929808806, | |
| "loss": 1.4012, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.2275675215472989, | |
| "grad_norm": 0.041748046875, | |
| "learning_rate": 0.00019023521858003742, | |
| "loss": 1.3941, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.23077269790712, | |
| "grad_norm": 0.037841796875, | |
| "learning_rate": 0.00018974702699357029, | |
| "loss": 1.4072, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2339778742669411, | |
| "grad_norm": 0.03759765625, | |
| "learning_rate": 0.00018924758573287315, | |
| "loss": 1.3531, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.23718305062676223, | |
| "grad_norm": 0.03662109375, | |
| "learning_rate": 0.00018873695740226468, | |
| "loss": 1.3682, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.24038822698658333, | |
| "grad_norm": 0.047607421875, | |
| "learning_rate": 0.0001882152060083484, | |
| "loss": 1.3796, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.24359340334640445, | |
| "grad_norm": 0.041015625, | |
| "learning_rate": 0.00018768239695198945, | |
| "loss": 1.3835, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.24679857970622554, | |
| "grad_norm": 0.04541015625, | |
| "learning_rate": 0.0001871385970201168, | |
| "loss": 1.3678, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.2500037560660467, | |
| "grad_norm": 0.04345703125, | |
| "learning_rate": 0.00018658387437735135, | |
| "loss": 1.3778, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.2532089324258678, | |
| "grad_norm": 0.06396484375, | |
| "learning_rate": 0.00018601829855746185, | |
| "loss": 1.3811, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.2564141087856889, | |
| "grad_norm": 0.057373046875, | |
| "learning_rate": 0.00018544194045464886, | |
| "loss": 1.3851, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.25961928514551, | |
| "grad_norm": 0.0458984375, | |
| "learning_rate": 0.0001848548723146581, | |
| "loss": 1.3865, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.2628244615053311, | |
| "grad_norm": 0.047119140625, | |
| "learning_rate": 0.00018425716772572473, | |
| "loss": 1.3638, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.2660296378651522, | |
| "grad_norm": 0.04443359375, | |
| "learning_rate": 0.00018364890160934904, | |
| "loss": 1.3918, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.2692348142249733, | |
| "grad_norm": 0.042236328125, | |
| "learning_rate": 0.00018303015021090525, | |
| "loss": 1.3794, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.27243999058479446, | |
| "grad_norm": 0.06005859375, | |
| "learning_rate": 0.00018240099109008412, | |
| "loss": 1.3836, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.27564516694461555, | |
| "grad_norm": 0.05419921875, | |
| "learning_rate": 0.000181761503111171, | |
| "loss": 1.3676, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.27885034330443664, | |
| "grad_norm": 0.04443359375, | |
| "learning_rate": 0.0001811117664331604, | |
| "loss": 1.3513, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.2820555196642578, | |
| "grad_norm": 0.047607421875, | |
| "learning_rate": 0.00018045186249970784, | |
| "loss": 1.3602, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.2852606960240789, | |
| "grad_norm": 0.043212890625, | |
| "learning_rate": 0.00017978187402892148, | |
| "loss": 1.3468, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.2884658723839, | |
| "grad_norm": 0.05078125, | |
| "learning_rate": 0.00017910188500299304, | |
| "loss": 1.3651, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.29167104874372113, | |
| "grad_norm": 0.04296875, | |
| "learning_rate": 0.00017841198065767107, | |
| "loss": 1.3763, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.2948762251035422, | |
| "grad_norm": 0.044921875, | |
| "learning_rate": 0.00017771224747157652, | |
| "loss": 1.3597, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.2980814014633633, | |
| "grad_norm": 0.0654296875, | |
| "learning_rate": 0.00017700277315536305, | |
| "loss": 1.3558, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.30128657782318446, | |
| "grad_norm": 0.052978515625, | |
| "learning_rate": 0.00017628364664072218, | |
| "loss": 1.3534, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.30449175418300556, | |
| "grad_norm": 0.04248046875, | |
| "learning_rate": 0.00017555495806923635, | |
| "loss": 1.3525, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.30769693054282665, | |
| "grad_norm": 0.044189453125, | |
| "learning_rate": 0.00017481679878107926, | |
| "loss": 1.3715, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.3109021069026478, | |
| "grad_norm": 0.058837890625, | |
| "learning_rate": 0.00017406926130356692, | |
| "loss": 1.3689, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.3141072832624689, | |
| "grad_norm": 0.095703125, | |
| "learning_rate": 0.00017331243933955918, | |
| "loss": 1.3686, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.31731245962229, | |
| "grad_norm": 0.059326171875, | |
| "learning_rate": 0.00017254642775571438, | |
| "loss": 1.3784, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.3205176359821111, | |
| "grad_norm": 0.07373046875, | |
| "learning_rate": 0.00017177132257059787, | |
| "loss": 1.3488, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.32372281234193223, | |
| "grad_norm": 0.0439453125, | |
| "learning_rate": 0.00017098722094264617, | |
| "loss": 1.3789, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.3269279887017533, | |
| "grad_norm": 0.052490234375, | |
| "learning_rate": 0.00017019422115798833, | |
| "loss": 1.3414, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.3301331650615744, | |
| "grad_norm": 0.0458984375, | |
| "learning_rate": 0.0001693924226181259, | |
| "loss": 1.3667, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.33333834142139557, | |
| "grad_norm": 0.05322265625, | |
| "learning_rate": 0.00016858192582747304, | |
| "loss": 1.3749, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.33654351778121666, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 0.00016776283238075851, | |
| "loss": 1.3929, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.33974869414103775, | |
| "grad_norm": 0.050537109375, | |
| "learning_rate": 0.00016693524495029068, | |
| "loss": 1.3527, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.3429538705008589, | |
| "grad_norm": 0.059814453125, | |
| "learning_rate": 0.00016609926727308806, | |
| "loss": 1.3577, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.34615904686068, | |
| "grad_norm": 0.07861328125, | |
| "learning_rate": 0.00016525500413787554, | |
| "loss": 1.3639, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.3493642232205011, | |
| "grad_norm": 0.0595703125, | |
| "learning_rate": 0.00016440256137194965, | |
| "loss": 1.3608, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.35256939958032224, | |
| "grad_norm": 0.052978515625, | |
| "learning_rate": 0.0001635420458279131, | |
| "loss": 1.3324, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.35577457594014333, | |
| "grad_norm": 0.062255859375, | |
| "learning_rate": 0.0001626735653702809, | |
| "loss": 1.3283, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.3589797522999644, | |
| "grad_norm": 0.04931640625, | |
| "learning_rate": 0.00016179722886195967, | |
| "loss": 1.3287, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.3621849286597856, | |
| "grad_norm": 0.0703125, | |
| "learning_rate": 0.00016091314615060195, | |
| "loss": 1.3799, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.36539010501960667, | |
| "grad_norm": 0.051025390625, | |
| "learning_rate": 0.00016002142805483685, | |
| "loss": 1.3399, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.36859528137942776, | |
| "grad_norm": 0.05908203125, | |
| "learning_rate": 0.00015912218635037896, | |
| "loss": 1.3698, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.3718004577392489, | |
| "grad_norm": 0.05078125, | |
| "learning_rate": 0.0001582155337560177, | |
| "loss": 1.3378, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.37500563409907, | |
| "grad_norm": 0.051025390625, | |
| "learning_rate": 0.00015730158391948784, | |
| "loss": 1.337, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.3782108104588911, | |
| "grad_norm": 0.0498046875, | |
| "learning_rate": 0.0001563804514032242, | |
| "loss": 1.3527, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.3814159868187122, | |
| "grad_norm": 0.052734375, | |
| "learning_rate": 0.0001554522516700011, | |
| "loss": 1.3583, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.38462116317853334, | |
| "grad_norm": 0.06201171875, | |
| "learning_rate": 0.00015451710106845955, | |
| "loss": 1.3421, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.38782633953835444, | |
| "grad_norm": 0.050537109375, | |
| "learning_rate": 0.0001535751168185228, | |
| "loss": 1.3577, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.39103151589817553, | |
| "grad_norm": 0.05517578125, | |
| "learning_rate": 0.00015262641699670328, | |
| "loss": 1.3706, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.3942366922579967, | |
| "grad_norm": 0.054931640625, | |
| "learning_rate": 0.0001516711205213016, | |
| "loss": 1.3439, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.3974418686178178, | |
| "grad_norm": 0.0478515625, | |
| "learning_rate": 0.00015070934713750042, | |
| "loss": 1.3353, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.40064704497763887, | |
| "grad_norm": 0.048583984375, | |
| "learning_rate": 0.00014974121740235456, | |
| "loss": 1.3489, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.40385222133746, | |
| "grad_norm": 0.057373046875, | |
| "learning_rate": 0.00014876685266967924, | |
| "loss": 1.3481, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.4070573976972811, | |
| "grad_norm": 0.053466796875, | |
| "learning_rate": 0.00014778637507483866, | |
| "loss": 1.3533, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.4102625740571022, | |
| "grad_norm": 0.06494140625, | |
| "learning_rate": 0.0001467999075194363, | |
| "loss": 1.3522, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.41346775041692335, | |
| "grad_norm": 0.06689453125, | |
| "learning_rate": 0.00014580757365590963, | |
| "loss": 1.3712, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.41667292677674445, | |
| "grad_norm": 0.053955078125, | |
| "learning_rate": 0.00014480949787203014, | |
| "loss": 1.3606, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.41987810313656554, | |
| "grad_norm": 0.046875, | |
| "learning_rate": 0.0001438058052753118, | |
| "loss": 1.3488, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.4230832794963867, | |
| "grad_norm": 0.058837890625, | |
| "learning_rate": 0.00014279662167732867, | |
| "loss": 1.342, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.4262884558562078, | |
| "grad_norm": 0.07080078125, | |
| "learning_rate": 0.00014178207357794486, | |
| "loss": 1.3712, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.4294936322160289, | |
| "grad_norm": 0.05029296875, | |
| "learning_rate": 0.00014076228814945778, | |
| "loss": 1.3227, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.43269880857585, | |
| "grad_norm": 0.06982421875, | |
| "learning_rate": 0.00013973739322065728, | |
| "loss": 1.3201, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.4359039849356711, | |
| "grad_norm": 0.05029296875, | |
| "learning_rate": 0.00013870751726080256, | |
| "loss": 1.3406, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.4391091612954922, | |
| "grad_norm": 0.06201171875, | |
| "learning_rate": 0.00013767278936351854, | |
| "loss": 1.3636, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.4423143376553133, | |
| "grad_norm": 0.0458984375, | |
| "learning_rate": 0.0001366333392306143, | |
| "loss": 1.3576, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.44551951401513445, | |
| "grad_norm": 0.06005859375, | |
| "learning_rate": 0.00013558929715582515, | |
| "loss": 1.3517, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.44872469037495555, | |
| "grad_norm": 0.05126953125, | |
| "learning_rate": 0.00013454079400848027, | |
| "loss": 1.3376, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.45192986673477664, | |
| "grad_norm": 0.059326171875, | |
| "learning_rate": 0.00013348796121709862, | |
| "loss": 1.3633, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.4551350430945978, | |
| "grad_norm": 0.05078125, | |
| "learning_rate": 0.00013243093075291444, | |
| "loss": 1.3217, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.4583402194544189, | |
| "grad_norm": 0.056884765625, | |
| "learning_rate": 0.00013136983511333482, | |
| "loss": 1.3265, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.46154539581424, | |
| "grad_norm": 0.05859375, | |
| "learning_rate": 0.00013030480730533145, | |
| "loss": 1.3451, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.4647505721740611, | |
| "grad_norm": 0.054443359375, | |
| "learning_rate": 0.00012923598082876812, | |
| "loss": 1.376, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.4679557485338822, | |
| "grad_norm": 0.058349609375, | |
| "learning_rate": 0.0001281634896596669, | |
| "loss": 1.3524, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.4711609248937033, | |
| "grad_norm": 0.0634765625, | |
| "learning_rate": 0.00012708746823341446, | |
| "loss": 1.3599, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.47436610125352446, | |
| "grad_norm": 0.053466796875, | |
| "learning_rate": 0.00012600805142791042, | |
| "loss": 1.3416, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.47757127761334556, | |
| "grad_norm": 0.055419921875, | |
| "learning_rate": 0.000124925374546661, | |
| "loss": 1.3574, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.48077645397316665, | |
| "grad_norm": 0.052978515625, | |
| "learning_rate": 0.0001238395733018187, | |
| "loss": 1.3574, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.4839816303329878, | |
| "grad_norm": 0.053466796875, | |
| "learning_rate": 0.00012275078379717089, | |
| "loss": 1.3341, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.4871868066928089, | |
| "grad_norm": 0.0556640625, | |
| "learning_rate": 0.00012165914251107952, | |
| "loss": 1.3241, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.49039198305263, | |
| "grad_norm": 0.054443359375, | |
| "learning_rate": 0.00012056478627937365, | |
| "loss": 1.3788, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.4935971594124511, | |
| "grad_norm": 0.049560546875, | |
| "learning_rate": 0.00011946785227819726, | |
| "loss": 1.3581, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.49680233577227223, | |
| "grad_norm": 0.05615234375, | |
| "learning_rate": 0.00011836847800681443, | |
| "loss": 1.3328, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.5000075121320934, | |
| "grad_norm": 0.0556640625, | |
| "learning_rate": 0.00011726680127037401, | |
| "loss": 1.3533, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.5032126884919145, | |
| "grad_norm": 0.05419921875, | |
| "learning_rate": 0.00011616296016263582, | |
| "loss": 1.3622, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.5064178648517356, | |
| "grad_norm": 0.049072265625, | |
| "learning_rate": 0.00011505709304866084, | |
| "loss": 1.3446, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.5096230412115567, | |
| "grad_norm": 0.0712890625, | |
| "learning_rate": 0.00011394933854746733, | |
| "loss": 1.3384, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.5128282175713778, | |
| "grad_norm": 0.055908203125, | |
| "learning_rate": 0.00011283983551465511, | |
| "loss": 1.3378, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5160333939311988, | |
| "grad_norm": 0.060791015625, | |
| "learning_rate": 0.00011172872302500017, | |
| "loss": 1.3656, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.51923857029102, | |
| "grad_norm": 0.0791015625, | |
| "learning_rate": 0.00011061614035502193, | |
| "loss": 1.3521, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.5224437466508411, | |
| "grad_norm": 0.05859375, | |
| "learning_rate": 0.00010950222696552486, | |
| "loss": 1.3614, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.5256489230106622, | |
| "grad_norm": 0.08203125, | |
| "learning_rate": 0.00010838712248411753, | |
| "loss": 1.3314, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.5288540993704833, | |
| "grad_norm": 0.05322265625, | |
| "learning_rate": 0.00010727096668771036, | |
| "loss": 1.338, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.5320592757303044, | |
| "grad_norm": 0.0556640625, | |
| "learning_rate": 0.0001061538994849946, | |
| "loss": 1.3611, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.5352644520901255, | |
| "grad_norm": 0.06201171875, | |
| "learning_rate": 0.00010503606089890529, | |
| "loss": 1.3175, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.5384696284499466, | |
| "grad_norm": 0.05712890625, | |
| "learning_rate": 0.00010391759104906928, | |
| "loss": 1.3525, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.5416748048097678, | |
| "grad_norm": 0.0498046875, | |
| "learning_rate": 0.00010279863013424154, | |
| "loss": 1.3313, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.5448799811695889, | |
| "grad_norm": 0.051025390625, | |
| "learning_rate": 0.00010167931841473142, | |
| "loss": 1.3349, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.54808515752941, | |
| "grad_norm": 0.06298828125, | |
| "learning_rate": 0.00010055979619482112, | |
| "loss": 1.3408, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.5512903338892311, | |
| "grad_norm": 0.058837890625, | |
| "learning_rate": 9.944020380517889e-05, | |
| "loss": 1.3175, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.5544955102490522, | |
| "grad_norm": 0.050048828125, | |
| "learning_rate": 9.832068158526862e-05, | |
| "loss": 1.3375, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.5577006866088733, | |
| "grad_norm": 0.0498046875, | |
| "learning_rate": 9.720136986575848e-05, | |
| "loss": 1.3475, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.5609058629686945, | |
| "grad_norm": 0.051513671875, | |
| "learning_rate": 9.608240895093076e-05, | |
| "loss": 1.3295, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.5641110393285156, | |
| "grad_norm": 0.046142578125, | |
| "learning_rate": 9.496393910109472e-05, | |
| "loss": 1.3429, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.5673162156883367, | |
| "grad_norm": 0.04443359375, | |
| "learning_rate": 9.384610051500545e-05, | |
| "loss": 1.3293, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.5705213920481578, | |
| "grad_norm": 0.052734375, | |
| "learning_rate": 9.272903331228968e-05, | |
| "loss": 1.3498, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.5737265684079789, | |
| "grad_norm": 0.062255859375, | |
| "learning_rate": 9.161287751588248e-05, | |
| "loss": 1.3351, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.5769317447678, | |
| "grad_norm": 0.064453125, | |
| "learning_rate": 9.049777303447516e-05, | |
| "loss": 1.353, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.5801369211276212, | |
| "grad_norm": 0.0556640625, | |
| "learning_rate": 8.938385964497808e-05, | |
| "loss": 1.3363, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.5833420974874423, | |
| "grad_norm": 0.06201171875, | |
| "learning_rate": 8.827127697499984e-05, | |
| "loss": 1.3696, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.5865472738472634, | |
| "grad_norm": 0.080078125, | |
| "learning_rate": 8.71601644853449e-05, | |
| "loss": 1.3481, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.5897524502070844, | |
| "grad_norm": 0.06884765625, | |
| "learning_rate": 8.605066145253268e-05, | |
| "loss": 1.3256, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.5929576265669055, | |
| "grad_norm": 0.051513671875, | |
| "learning_rate": 8.494290695133917e-05, | |
| "loss": 1.3544, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.5961628029267266, | |
| "grad_norm": 0.05810546875, | |
| "learning_rate": 8.383703983736419e-05, | |
| "loss": 1.3443, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.5993679792865477, | |
| "grad_norm": 0.06103515625, | |
| "learning_rate": 8.2733198729626e-05, | |
| "loss": 1.3816, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.6025731556463689, | |
| "grad_norm": 0.046142578125, | |
| "learning_rate": 8.163152199318558e-05, | |
| "loss": 1.3247, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.60577833200619, | |
| "grad_norm": 0.053466796875, | |
| "learning_rate": 8.053214772180277e-05, | |
| "loss": 1.3532, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.6089835083660111, | |
| "grad_norm": 0.05419921875, | |
| "learning_rate": 7.94352137206264e-05, | |
| "loss": 1.3443, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.6121886847258322, | |
| "grad_norm": 0.047119140625, | |
| "learning_rate": 7.83408574889205e-05, | |
| "loss": 1.3327, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.6153938610856533, | |
| "grad_norm": 0.0537109375, | |
| "learning_rate": 7.724921620282916e-05, | |
| "loss": 1.334, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.6185990374454744, | |
| "grad_norm": 0.0703125, | |
| "learning_rate": 7.616042669818133e-05, | |
| "loss": 1.3572, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.6218042138052956, | |
| "grad_norm": 0.055419921875, | |
| "learning_rate": 7.507462545333903e-05, | |
| "loss": 1.3322, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.6250093901651167, | |
| "grad_norm": 0.07958984375, | |
| "learning_rate": 7.399194857208961e-05, | |
| "loss": 1.3222, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.6282145665249378, | |
| "grad_norm": 0.05078125, | |
| "learning_rate": 7.291253176658561e-05, | |
| "loss": 1.3375, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.6314197428847589, | |
| "grad_norm": 0.08251953125, | |
| "learning_rate": 7.183651034033313e-05, | |
| "loss": 1.3397, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.63462491924458, | |
| "grad_norm": 0.04931640625, | |
| "learning_rate": 7.07640191712319e-05, | |
| "loss": 1.34, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.6378300956044011, | |
| "grad_norm": 0.049072265625, | |
| "learning_rate": 6.969519269466857e-05, | |
| "loss": 1.3344, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.6410352719642222, | |
| "grad_norm": 0.052490234375, | |
| "learning_rate": 6.863016488666517e-05, | |
| "loss": 1.3475, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6442404483240434, | |
| "grad_norm": 0.04736328125, | |
| "learning_rate": 6.756906924708558e-05, | |
| "loss": 1.3317, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.6474456246838645, | |
| "grad_norm": 0.050537109375, | |
| "learning_rate": 6.651203878290139e-05, | |
| "loss": 1.3243, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.6506508010436856, | |
| "grad_norm": 0.053955078125, | |
| "learning_rate": 6.545920599151975e-05, | |
| "loss": 1.3351, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.6538559774035066, | |
| "grad_norm": 0.058837890625, | |
| "learning_rate": 6.441070284417487e-05, | |
| "loss": 1.3536, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.6570611537633277, | |
| "grad_norm": 0.060791015625, | |
| "learning_rate": 6.336666076938572e-05, | |
| "loss": 1.3064, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.6602663301231488, | |
| "grad_norm": 0.056396484375, | |
| "learning_rate": 6.232721063648148e-05, | |
| "loss": 1.3496, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.66347150648297, | |
| "grad_norm": 0.0478515625, | |
| "learning_rate": 6.12924827391975e-05, | |
| "loss": 1.3487, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 0.6666766828427911, | |
| "grad_norm": 0.05126953125, | |
| "learning_rate": 6.026260677934272e-05, | |
| "loss": 1.3241, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.6698818592026122, | |
| "grad_norm": 0.0478515625, | |
| "learning_rate": 5.9237711850542246e-05, | |
| "loss": 1.3454, | |
| "step": 1045 | |
| }, | |
| { | |
| "epoch": 0.6730870355624333, | |
| "grad_norm": 0.046142578125, | |
| "learning_rate": 5.8217926422055126e-05, | |
| "loss": 1.3364, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.6762922119222544, | |
| "grad_norm": 0.054443359375, | |
| "learning_rate": 5.7203378322671355e-05, | |
| "loss": 1.3152, | |
| "step": 1055 | |
| }, | |
| { | |
| "epoch": 0.6794973882820755, | |
| "grad_norm": 0.0546875, | |
| "learning_rate": 5.619419472468823e-05, | |
| "loss": 1.3486, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.6827025646418967, | |
| "grad_norm": 0.05029296875, | |
| "learning_rate": 5.519050212796986e-05, | |
| "loss": 1.3301, | |
| "step": 1065 | |
| }, | |
| { | |
| "epoch": 0.6859077410017178, | |
| "grad_norm": 0.051513671875, | |
| "learning_rate": 5.419242634409039e-05, | |
| "loss": 1.3279, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.6891129173615389, | |
| "grad_norm": 0.0478515625, | |
| "learning_rate": 5.32000924805637e-05, | |
| "loss": 1.3415, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 0.69231809372136, | |
| "grad_norm": 0.04638671875, | |
| "learning_rate": 5.2213624925161386e-05, | |
| "loss": 1.3449, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.6955232700811811, | |
| "grad_norm": 0.04541015625, | |
| "learning_rate": 5.123314733032074e-05, | |
| "loss": 1.3442, | |
| "step": 1085 | |
| }, | |
| { | |
| "epoch": 0.6987284464410022, | |
| "grad_norm": 0.04736328125, | |
| "learning_rate": 5.0258782597645446e-05, | |
| "loss": 1.3309, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.7019336228008233, | |
| "grad_norm": 0.0478515625, | |
| "learning_rate": 4.929065286249959e-05, | |
| "loss": 1.3564, | |
| "step": 1095 | |
| }, | |
| { | |
| "epoch": 0.7051387991606445, | |
| "grad_norm": 0.048095703125, | |
| "learning_rate": 4.832887947869841e-05, | |
| "loss": 1.3578, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.7083439755204656, | |
| "grad_norm": 0.047119140625, | |
| "learning_rate": 4.737358300329673e-05, | |
| "loss": 1.3417, | |
| "step": 1105 | |
| }, | |
| { | |
| "epoch": 0.7115491518802867, | |
| "grad_norm": 0.05029296875, | |
| "learning_rate": 4.642488318147723e-05, | |
| "loss": 1.3259, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.7147543282401078, | |
| "grad_norm": 0.052001953125, | |
| "learning_rate": 4.548289893154051e-05, | |
| "loss": 1.3568, | |
| "step": 1115 | |
| }, | |
| { | |
| "epoch": 0.7179595045999289, | |
| "grad_norm": 0.047607421875, | |
| "learning_rate": 4.4547748329998925e-05, | |
| "loss": 1.3211, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.72116468095975, | |
| "grad_norm": 0.05126953125, | |
| "learning_rate": 4.361954859677584e-05, | |
| "loss": 1.3398, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 0.7243698573195712, | |
| "grad_norm": 0.048095703125, | |
| "learning_rate": 4.2698416080512204e-05, | |
| "loss": 1.3266, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.7275750336793922, | |
| "grad_norm": 0.050048828125, | |
| "learning_rate": 4.1784466243982324e-05, | |
| "loss": 1.3447, | |
| "step": 1135 | |
| }, | |
| { | |
| "epoch": 0.7307802100392133, | |
| "grad_norm": 0.052001953125, | |
| "learning_rate": 4.0877813649621076e-05, | |
| "loss": 1.3385, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.7339853863990344, | |
| "grad_norm": 0.04638671875, | |
| "learning_rate": 3.997857194516319e-05, | |
| "loss": 1.3403, | |
| "step": 1145 | |
| }, | |
| { | |
| "epoch": 0.7371905627588555, | |
| "grad_norm": 0.05078125, | |
| "learning_rate": 3.9086853849398065e-05, | |
| "loss": 1.3503, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.7403957391186766, | |
| "grad_norm": 0.06396484375, | |
| "learning_rate": 3.8202771138040336e-05, | |
| "loss": 1.3354, | |
| "step": 1155 | |
| }, | |
| { | |
| "epoch": 0.7436009154784978, | |
| "grad_norm": 0.05078125, | |
| "learning_rate": 3.732643462971912e-05, | |
| "loss": 1.3258, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.7468060918383189, | |
| "grad_norm": 0.049560546875, | |
| "learning_rate": 3.6457954172086896e-05, | |
| "loss": 1.3493, | |
| "step": 1165 | |
| }, | |
| { | |
| "epoch": 0.75001126819814, | |
| "grad_norm": 0.046875, | |
| "learning_rate": 3.559743862805034e-05, | |
| "loss": 1.3275, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.7532164445579611, | |
| "grad_norm": 0.045654296875, | |
| "learning_rate": 3.47449958621245e-05, | |
| "loss": 1.3148, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 0.7564216209177822, | |
| "grad_norm": 0.051513671875, | |
| "learning_rate": 3.390073272691198e-05, | |
| "loss": 1.3338, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.7596267972776033, | |
| "grad_norm": 0.049072265625, | |
| "learning_rate": 3.306475504970931e-05, | |
| "loss": 1.2935, | |
| "step": 1185 | |
| }, | |
| { | |
| "epoch": 0.7628319736374244, | |
| "grad_norm": 0.04833984375, | |
| "learning_rate": 3.2237167619241495e-05, | |
| "loss": 1.3275, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.7660371499972456, | |
| "grad_norm": 0.056396484375, | |
| "learning_rate": 3.141807417252697e-05, | |
| "loss": 1.3461, | |
| "step": 1195 | |
| }, | |
| { | |
| "epoch": 0.7692423263570667, | |
| "grad_norm": 0.04345703125, | |
| "learning_rate": 3.060757738187409e-05, | |
| "loss": 1.3394, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.7724475027168878, | |
| "grad_norm": 0.053955078125, | |
| "learning_rate": 2.980577884201169e-05, | |
| "loss": 1.3511, | |
| "step": 1205 | |
| }, | |
| { | |
| "epoch": 0.7756526790767089, | |
| "grad_norm": 0.04736328125, | |
| "learning_rate": 2.9012779057353855e-05, | |
| "loss": 1.3213, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.77885785543653, | |
| "grad_norm": 0.0576171875, | |
| "learning_rate": 2.822867742940214e-05, | |
| "loss": 1.3384, | |
| "step": 1215 | |
| }, | |
| { | |
| "epoch": 0.7820630317963511, | |
| "grad_norm": 0.04833984375, | |
| "learning_rate": 2.745357224428563e-05, | |
| "loss": 1.343, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.7852682081561723, | |
| "grad_norm": 0.049560546875, | |
| "learning_rate": 2.6687560660440858e-05, | |
| "loss": 1.3541, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 0.7884733845159934, | |
| "grad_norm": 0.047607421875, | |
| "learning_rate": 2.593073869643312e-05, | |
| "loss": 1.3491, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.7916785608758145, | |
| "grad_norm": 0.04248046875, | |
| "learning_rate": 2.518320121892076e-05, | |
| "loss": 1.3439, | |
| "step": 1235 | |
| }, | |
| { | |
| "epoch": 0.7948837372356355, | |
| "grad_norm": 0.04736328125, | |
| "learning_rate": 2.4445041930763678e-05, | |
| "loss": 1.3236, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.7980889135954566, | |
| "grad_norm": 0.0478515625, | |
| "learning_rate": 2.371635335927781e-05, | |
| "loss": 1.3505, | |
| "step": 1245 | |
| }, | |
| { | |
| "epoch": 0.8012940899552777, | |
| "grad_norm": 0.0517578125, | |
| "learning_rate": 2.2997226844636977e-05, | |
| "loss": 1.3223, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.8044992663150989, | |
| "grad_norm": 0.046630859375, | |
| "learning_rate": 2.2287752528423468e-05, | |
| "loss": 1.3282, | |
| "step": 1255 | |
| }, | |
| { | |
| "epoch": 0.80770444267492, | |
| "grad_norm": 0.046875, | |
| "learning_rate": 2.1588019342328968e-05, | |
| "loss": 1.3294, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.8109096190347411, | |
| "grad_norm": 0.0439453125, | |
| "learning_rate": 2.089811499700699e-05, | |
| "loss": 1.3356, | |
| "step": 1265 | |
| }, | |
| { | |
| "epoch": 0.8141147953945622, | |
| "grad_norm": 0.045654296875, | |
| "learning_rate": 2.021812597107855e-05, | |
| "loss": 1.3486, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.8173199717543833, | |
| "grad_norm": 0.04931640625, | |
| "learning_rate": 1.954813750029216e-05, | |
| "loss": 1.3492, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 0.8205251481142044, | |
| "grad_norm": 0.05126953125, | |
| "learning_rate": 1.8888233566839653e-05, | |
| "loss": 1.329, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.8237303244740255, | |
| "grad_norm": 0.048095703125, | |
| "learning_rate": 1.8238496888828982e-05, | |
| "loss": 1.317, | |
| "step": 1285 | |
| }, | |
| { | |
| "epoch": 0.8269355008338467, | |
| "grad_norm": 0.051513671875, | |
| "learning_rate": 1.759900890991589e-05, | |
| "loss": 1.3177, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.8301406771936678, | |
| "grad_norm": 0.0458984375, | |
| "learning_rate": 1.696984978909476e-05, | |
| "loss": 1.323, | |
| "step": 1295 | |
| }, | |
| { | |
| "epoch": 0.8333458535534889, | |
| "grad_norm": 0.0439453125, | |
| "learning_rate": 1.6351098390650966e-05, | |
| "loss": 1.3517, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.83655102991331, | |
| "grad_norm": 0.052978515625, | |
| "learning_rate": 1.5742832274275288e-05, | |
| "loss": 1.35, | |
| "step": 1305 | |
| }, | |
| { | |
| "epoch": 0.8397562062731311, | |
| "grad_norm": 0.049072265625, | |
| "learning_rate": 1.514512768534193e-05, | |
| "loss": 1.3614, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.8429613826329522, | |
| "grad_norm": 0.0439453125, | |
| "learning_rate": 1.4558059545351143e-05, | |
| "loss": 1.3389, | |
| "step": 1315 | |
| }, | |
| { | |
| "epoch": 0.8461665589927734, | |
| "grad_norm": 0.04541015625, | |
| "learning_rate": 1.3981701442538153e-05, | |
| "loss": 1.3272, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.8493717353525945, | |
| "grad_norm": 0.048583984375, | |
| "learning_rate": 1.3416125622648668e-05, | |
| "loss": 1.3324, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 0.8525769117124156, | |
| "grad_norm": 0.04541015625, | |
| "learning_rate": 1.286140297988323e-05, | |
| "loss": 1.3352, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.8557820880722367, | |
| "grad_norm": 0.04443359375, | |
| "learning_rate": 1.231760304801054e-05, | |
| "loss": 1.3361, | |
| "step": 1335 | |
| }, | |
| { | |
| "epoch": 0.8589872644320578, | |
| "grad_norm": 0.047119140625, | |
| "learning_rate": 1.1784793991651621e-05, | |
| "loss": 1.3252, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.8621924407918788, | |
| "grad_norm": 0.044189453125, | |
| "learning_rate": 1.1263042597735362e-05, | |
| "loss": 1.3468, | |
| "step": 1345 | |
| }, | |
| { | |
| "epoch": 0.8653976171517, | |
| "grad_norm": 0.046630859375, | |
| "learning_rate": 1.0752414267126875e-05, | |
| "loss": 1.3301, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.8686027935115211, | |
| "grad_norm": 0.05029296875, | |
| "learning_rate": 1.0252973006429733e-05, | |
| "loss": 1.36, | |
| "step": 1355 | |
| }, | |
| { | |
| "epoch": 0.8718079698713422, | |
| "grad_norm": 0.047119140625, | |
| "learning_rate": 9.764781419962577e-06, | |
| "loss": 1.3482, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.8750131462311633, | |
| "grad_norm": 0.04638671875, | |
| "learning_rate": 9.287900701911944e-06, | |
| "loss": 1.3232, | |
| "step": 1365 | |
| }, | |
| { | |
| "epoch": 0.8782183225909844, | |
| "grad_norm": 0.04931640625, | |
| "learning_rate": 8.822390628661582e-06, | |
| "loss": 1.3571, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.8814234989508055, | |
| "grad_norm": 0.044921875, | |
| "learning_rate": 8.368309551299536e-06, | |
| "loss": 1.3274, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 0.8846286753106266, | |
| "grad_norm": 0.04541015625, | |
| "learning_rate": 7.92571438830394e-06, | |
| "loss": 1.3656, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.8878338516704478, | |
| "grad_norm": 0.046142578125, | |
| "learning_rate": 7.494660618408378e-06, | |
| "loss": 1.3659, | |
| "step": 1385 | |
| }, | |
| { | |
| "epoch": 0.8910390280302689, | |
| "grad_norm": 0.04541015625, | |
| "learning_rate": 7.075202273647652e-06, | |
| "loss": 1.3305, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.89424420439009, | |
| "grad_norm": 0.046875, | |
| "learning_rate": 6.667391932584999e-06, | |
| "loss": 1.36, | |
| "step": 1395 | |
| }, | |
| { | |
| "epoch": 0.8974493807499111, | |
| "grad_norm": 0.0458984375, | |
| "learning_rate": 6.271280713721317e-06, | |
| "loss": 1.3382, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.9006545571097322, | |
| "grad_norm": 0.04638671875, | |
| "learning_rate": 5.886918269087716e-06, | |
| "loss": 1.326, | |
| "step": 1405 | |
| }, | |
| { | |
| "epoch": 0.9038597334695533, | |
| "grad_norm": 0.046875, | |
| "learning_rate": 5.514352778021492e-06, | |
| "loss": 1.3602, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.9070649098293745, | |
| "grad_norm": 0.046142578125, | |
| "learning_rate": 5.153630941127063e-06, | |
| "loss": 1.3407, | |
| "step": 1415 | |
| }, | |
| { | |
| "epoch": 0.9102700861891956, | |
| "grad_norm": 0.046875, | |
| "learning_rate": 4.804797974422026e-06, | |
| "loss": 1.3241, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.9134752625490167, | |
| "grad_norm": 0.050537109375, | |
| "learning_rate": 4.4678976036694355e-06, | |
| "loss": 1.3324, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 0.9166804389088378, | |
| "grad_norm": 0.04443359375, | |
| "learning_rate": 4.142972058896811e-06, | |
| "loss": 1.3267, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.9198856152686589, | |
| "grad_norm": 0.044921875, | |
| "learning_rate": 3.830062069102602e-06, | |
| "loss": 1.3496, | |
| "step": 1435 | |
| }, | |
| { | |
| "epoch": 0.92309079162848, | |
| "grad_norm": 0.046630859375, | |
| "learning_rate": 3.529206857151035e-06, | |
| "loss": 1.3481, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.9262959679883012, | |
| "grad_norm": 0.04345703125, | |
| "learning_rate": 3.240444134855347e-06, | |
| "loss": 1.3433, | |
| "step": 1445 | |
| }, | |
| { | |
| "epoch": 0.9295011443481223, | |
| "grad_norm": 0.045654296875, | |
| "learning_rate": 2.963810098250841e-06, | |
| "loss": 1.3555, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.9327063207079433, | |
| "grad_norm": 0.044921875, | |
| "learning_rate": 2.6993394230576674e-06, | |
| "loss": 1.3218, | |
| "step": 1455 | |
| }, | |
| { | |
| "epoch": 0.9359114970677644, | |
| "grad_norm": 0.04638671875, | |
| "learning_rate": 2.4470652603343023e-06, | |
| "loss": 1.346, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.9391166734275855, | |
| "grad_norm": 0.044677734375, | |
| "learning_rate": 2.2070192323220607e-06, | |
| "loss": 1.3551, | |
| "step": 1465 | |
| }, | |
| { | |
| "epoch": 0.9423218497874066, | |
| "grad_norm": 0.0439453125, | |
| "learning_rate": 1.9792314284813986e-06, | |
| "loss": 1.3262, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.9455270261472277, | |
| "grad_norm": 0.04736328125, | |
| "learning_rate": 1.763730401720065e-06, | |
| "loss": 1.3257, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 0.9487322025070489, | |
| "grad_norm": 0.046142578125, | |
| "learning_rate": 1.5605431648141878e-06, | |
| "loss": 1.3158, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.95193737886687, | |
| "grad_norm": 0.044677734375, | |
| "learning_rate": 1.3696951870222018e-06, | |
| "loss": 1.3637, | |
| "step": 1485 | |
| }, | |
| { | |
| "epoch": 0.9551425552266911, | |
| "grad_norm": 0.053466796875, | |
| "learning_rate": 1.1912103908922945e-06, | |
| "loss": 1.3337, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.9583477315865122, | |
| "grad_norm": 0.050048828125, | |
| "learning_rate": 1.0251111492637244e-06, | |
| "loss": 1.3557, | |
| "step": 1495 | |
| }, | |
| { | |
| "epoch": 0.9615529079463333, | |
| "grad_norm": 0.05126953125, | |
| "learning_rate": 8.714182824624883e-07, | |
| "loss": 1.3373, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9647580843061544, | |
| "grad_norm": 0.0458984375, | |
| "learning_rate": 7.301510556914859e-07, | |
| "loss": 1.3274, | |
| "step": 1505 | |
| }, | |
| { | |
| "epoch": 0.9679632606659756, | |
| "grad_norm": 0.05224609375, | |
| "learning_rate": 6.01327176615607e-07, | |
| "loss": 1.3894, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.9711684370257967, | |
| "grad_norm": 0.045166015625, | |
| "learning_rate": 4.84962793142163e-07, | |
| "loss": 1.3419, | |
| "step": 1515 | |
| }, | |
| { | |
| "epoch": 0.9743736133856178, | |
| "grad_norm": 0.044677734375, | |
| "learning_rate": 3.8107249139672783e-07, | |
| "loss": 1.3321, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.9775787897454389, | |
| "grad_norm": 0.0478515625, | |
| "learning_rate": 2.89669293894812e-07, | |
| "loss": 1.3497, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 0.98078396610526, | |
| "grad_norm": 0.049072265625, | |
| "learning_rate": 2.1076465790946798e-07, | |
| "loss": 1.3518, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.9839891424650811, | |
| "grad_norm": 0.04638671875, | |
| "learning_rate": 1.443684740351947e-07, | |
| "loss": 1.3224, | |
| "step": 1535 | |
| }, | |
| { | |
| "epoch": 0.9871943188249022, | |
| "grad_norm": 0.049072265625, | |
| "learning_rate": 9.048906494811826e-08, | |
| "loss": 1.3513, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.9903994951847234, | |
| "grad_norm": 0.050048828125, | |
| "learning_rate": 4.9133184362748497e-08, | |
| "loss": 1.3494, | |
| "step": 1545 | |
| }, | |
| { | |
| "epoch": 0.9936046715445445, | |
| "grad_norm": 0.04443359375, | |
| "learning_rate": 2.0306016185456244e-08, | |
| "loss": 1.3344, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.9968098479043656, | |
| "grad_norm": 0.047119140625, | |
| "learning_rate": 4.011173864637563e-09, | |
| "loss": 1.3662, | |
| "step": 1555 | |
| }, | |
| { | |
| "epoch": 0.9993739889922224, | |
| "eval_loss": 1.4191993474960327, | |
| "eval_runtime": 1938.5869, | |
| "eval_samples_per_second": 7.3, | |
| "eval_steps_per_second": 7.3, | |
| "step": 1559 | |
| }, | |
| { | |
| "epoch": 0.9993739889922224, | |
| "step": 1559, | |
| "total_flos": 3.232184148701479e+18, | |
| "train_loss": 0.016414370117774753, | |
| "train_runtime": 2971.8566, | |
| "train_samples_per_second": 67.189, | |
| "train_steps_per_second": 0.525 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1559, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 20, | |
| "total_flos": 3.232184148701479e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |