| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.996510275300504, | |
| "eval_steps": 500, | |
| "global_step": 966, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.031019775106630478, | |
| "grad_norm": 4.631979942321777, | |
| "learning_rate": 5.517241379310345e-06, | |
| "loss": 3.6051, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.062039550213260956, | |
| "grad_norm": 1.5514965057373047, | |
| "learning_rate": 1.2413793103448277e-05, | |
| "loss": 2.545, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09305932531989143, | |
| "grad_norm": 1.3494954109191895, | |
| "learning_rate": 1.9310344827586207e-05, | |
| "loss": 2.1342, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.12407910042652191, | |
| "grad_norm": 1.3906075954437256, | |
| "learning_rate": 2.620689655172414e-05, | |
| "loss": 2.0661, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.15509887553315238, | |
| "grad_norm": 1.3953957557678223, | |
| "learning_rate": 3.310344827586207e-05, | |
| "loss": 2.0186, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.18611865063978286, | |
| "grad_norm": 1.6760238409042358, | |
| "learning_rate": 4e-05, | |
| "loss": 1.7399, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.21713842574641334, | |
| "grad_norm": 1.727597951889038, | |
| "learning_rate": 4.689655172413793e-05, | |
| "loss": 1.6954, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.24815820085304383, | |
| "grad_norm": 1.604373812675476, | |
| "learning_rate": 5.379310344827586e-05, | |
| "loss": 1.6684, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2791779759596743, | |
| "grad_norm": 1.6111727952957153, | |
| "learning_rate": 6.068965517241379e-05, | |
| "loss": 1.6644, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.31019775106630476, | |
| "grad_norm": 1.4779224395751953, | |
| "learning_rate": 6.758620689655173e-05, | |
| "loss": 1.562, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.34121752617293527, | |
| "grad_norm": 1.940397024154663, | |
| "learning_rate": 7.448275862068966e-05, | |
| "loss": 1.5314, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3722373012795657, | |
| "grad_norm": 1.7813657522201538, | |
| "learning_rate": 8.137931034482759e-05, | |
| "loss": 1.5245, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4032570763861962, | |
| "grad_norm": 2.0770983695983887, | |
| "learning_rate": 8.827586206896552e-05, | |
| "loss": 1.5824, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.4342768514928267, | |
| "grad_norm": 1.6462304592132568, | |
| "learning_rate": 9.517241379310345e-05, | |
| "loss": 1.4695, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.46529662659945714, | |
| "grad_norm": 2.1206908226013184, | |
| "learning_rate": 9.999670548734657e-05, | |
| "loss": 1.5062, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.49631640170608765, | |
| "grad_norm": 2.2103354930877686, | |
| "learning_rate": 9.993814845038307e-05, | |
| "loss": 1.3786, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5273361768127182, | |
| "grad_norm": 2.300985336303711, | |
| "learning_rate": 9.980647870476639e-05, | |
| "loss": 1.4175, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5583559519193486, | |
| "grad_norm": 2.7707035541534424, | |
| "learning_rate": 9.960188902359786e-05, | |
| "loss": 1.4232, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5893757270259791, | |
| "grad_norm": 2.5883567333221436, | |
| "learning_rate": 9.932467893952367e-05, | |
| "loss": 1.3551, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6203955021326095, | |
| "grad_norm": 2.2564728260040283, | |
| "learning_rate": 9.897525430619965e-05, | |
| "loss": 1.3123, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.65141527723924, | |
| "grad_norm": 2.3438875675201416, | |
| "learning_rate": 9.855412670409493e-05, | |
| "loss": 1.265, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.6824350523458705, | |
| "grad_norm": 2.057579755783081, | |
| "learning_rate": 9.806191269150479e-05, | |
| "loss": 1.2062, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.713454827452501, | |
| "grad_norm": 2.852278470993042, | |
| "learning_rate": 9.749933290186913e-05, | |
| "loss": 1.2265, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7444746025591314, | |
| "grad_norm": 2.790354013442993, | |
| "learning_rate": 9.686721098871789e-05, | |
| "loss": 1.2365, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7754943776657619, | |
| "grad_norm": 2.6249144077301025, | |
| "learning_rate": 9.61664724197888e-05, | |
| "loss": 1.1627, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8065141527723924, | |
| "grad_norm": 2.421135663986206, | |
| "learning_rate": 9.539814312208195e-05, | |
| "loss": 1.2203, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8375339278790229, | |
| "grad_norm": 2.2487196922302246, | |
| "learning_rate": 9.45633479798359e-05, | |
| "loss": 1.1503, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.8685537029856534, | |
| "grad_norm": 3.0988595485687256, | |
| "learning_rate": 9.366330918762361e-05, | |
| "loss": 1.1575, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.8995734780922838, | |
| "grad_norm": 2.629244327545166, | |
| "learning_rate": 9.269934446097986e-05, | |
| "loss": 1.1328, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9305932531989143, | |
| "grad_norm": 3.2107417583465576, | |
| "learning_rate": 9.16728651071797e-05, | |
| "loss": 1.1052, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9616130283055447, | |
| "grad_norm": 2.520418405532837, | |
| "learning_rate": 9.058537395899252e-05, | |
| "loss": 1.0805, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.9926328034121753, | |
| "grad_norm": 2.8015685081481934, | |
| "learning_rate": 8.943846317443673e-05, | |
| "loss": 1.0565, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.0236525785188058, | |
| "grad_norm": 3.1201586723327637, | |
| "learning_rate": 8.823381190575654e-05, | |
| "loss": 0.8105, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.0546723536254363, | |
| "grad_norm": 2.659348487854004, | |
| "learning_rate": 8.69731838410335e-05, | |
| "loss": 0.7056, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.0856921287320667, | |
| "grad_norm": 2.9386487007141113, | |
| "learning_rate": 8.565842462203197e-05, | |
| "loss": 0.7194, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.1167119038386972, | |
| "grad_norm": 3.119506359100342, | |
| "learning_rate": 8.42914591420592e-05, | |
| "loss": 0.6653, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.1477316789453276, | |
| "grad_norm": 3.151764154434204, | |
| "learning_rate": 8.287428872779583e-05, | |
| "loss": 0.6434, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.1787514540519581, | |
| "grad_norm": 2.7570505142211914, | |
| "learning_rate": 8.140898820922307e-05, | |
| "loss": 0.6401, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.2097712291585885, | |
| "grad_norm": 3.0509796142578125, | |
| "learning_rate": 7.989770288193614e-05, | |
| "loss": 0.6535, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.240791004265219, | |
| "grad_norm": 2.9341208934783936, | |
| "learning_rate": 7.834264536629148e-05, | |
| "loss": 0.6892, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.2718107793718496, | |
| "grad_norm": 4.456368446350098, | |
| "learning_rate": 7.674609236798621e-05, | |
| "loss": 0.6351, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.30283055447848, | |
| "grad_norm": 3.2823486328125, | |
| "learning_rate": 7.511038134481237e-05, | |
| "loss": 0.6424, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.3338503295851105, | |
| "grad_norm": 3.2044315338134766, | |
| "learning_rate": 7.343790708446609e-05, | |
| "loss": 0.6297, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.3648701046917409, | |
| "grad_norm": 2.9563002586364746, | |
| "learning_rate": 7.173111819842222e-05, | |
| "loss": 0.55, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.3958898797983714, | |
| "grad_norm": 2.863718271255493, | |
| "learning_rate": 6.999251353700718e-05, | |
| "loss": 0.564, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.426909654905002, | |
| "grad_norm": 2.617347240447998, | |
| "learning_rate": 6.822463853091911e-05, | |
| "loss": 0.5635, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.4579294300116323, | |
| "grad_norm": 2.8376879692077637, | |
| "learning_rate": 6.643008146455114e-05, | |
| "loss": 0.5589, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.488949205118263, | |
| "grad_norm": 2.8635995388031006, | |
| "learning_rate": 6.46114696865741e-05, | |
| "loss": 0.5668, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.5199689802248932, | |
| "grad_norm": 2.637256622314453, | |
| "learning_rate": 6.277146576332657e-05, | |
| "loss": 0.5273, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.5509887553315238, | |
| "grad_norm": 3.0905144214630127, | |
| "learning_rate": 6.091276358064408e-05, | |
| "loss": 0.5076, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.5820085304381544, | |
| "grad_norm": 3.2317402362823486, | |
| "learning_rate": 5.903808439983428e-05, | |
| "loss": 0.4969, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.613028305544785, | |
| "grad_norm": 3.3929502964019775, | |
| "learning_rate": 5.7150172873572906e-05, | |
| "loss": 0.4832, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.6440480806514153, | |
| "grad_norm": 3.057800531387329, | |
| "learning_rate": 5.525179302755303e-05, | |
| "loss": 0.4986, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.6750678557580456, | |
| "grad_norm": 3.7113149166107178, | |
| "learning_rate": 5.3345724213771145e-05, | |
| "loss": 0.492, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 1.7060876308646762, | |
| "grad_norm": 3.076077699661255, | |
| "learning_rate": 5.143475704137433e-05, | |
| "loss": 0.4645, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 1.7371074059713068, | |
| "grad_norm": 3.2166106700897217, | |
| "learning_rate": 4.9521689291026406e-05, | |
| "loss": 0.4516, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 1.7681271810779373, | |
| "grad_norm": 2.9462897777557373, | |
| "learning_rate": 4.760932181877439e-05, | |
| "loss": 0.4372, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 1.7991469561845677, | |
| "grad_norm": 2.8754382133483887, | |
| "learning_rate": 4.570045445541253e-05, | |
| "loss": 0.4232, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 1.830166731291198, | |
| "grad_norm": 3.5116498470306396, | |
| "learning_rate": 4.379788190734712e-05, | |
| "loss": 0.4268, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 1.8611865063978286, | |
| "grad_norm": 3.0945634841918945, | |
| "learning_rate": 4.190438966496407e-05, | |
| "loss": 0.4093, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.8922062815044591, | |
| "grad_norm": 2.920431613922119, | |
| "learning_rate": 4.002274992448911e-05, | |
| "loss": 0.3945, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 1.9232260566110897, | |
| "grad_norm": 2.7673826217651367, | |
| "learning_rate": 3.815571752931162e-05, | |
| "loss": 0.3948, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 1.95424583171772, | |
| "grad_norm": 2.909179925918579, | |
| "learning_rate": 3.630602593671405e-05, | |
| "loss": 0.3731, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 1.9852656068243504, | |
| "grad_norm": 2.851840019226074, | |
| "learning_rate": 3.4476383215912114e-05, | |
| "loss": 0.3476, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.016285381930981, | |
| "grad_norm": 2.012725353240967, | |
| "learning_rate": 3.266946808326466e-05, | |
| "loss": 0.2479, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.0473051570376115, | |
| "grad_norm": 2.0801613330841064, | |
| "learning_rate": 3.0887925980458154e-05, | |
| "loss": 0.1408, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.078324932144242, | |
| "grad_norm": 1.9638011455535889, | |
| "learning_rate": 2.913436520140731e-05, | |
| "loss": 0.1383, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.1093447072508726, | |
| "grad_norm": 2.354614019393921, | |
| "learning_rate": 2.74113530735426e-05, | |
| "loss": 0.1431, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.1403644823575028, | |
| "grad_norm": 1.8083250522613525, | |
| "learning_rate": 2.5721412199075372e-05, | |
| "loss": 0.1253, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.1713842574641333, | |
| "grad_norm": 2.4336416721343994, | |
| "learning_rate": 2.4067016761743515e-05, | |
| "loss": 0.1316, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.202404032570764, | |
| "grad_norm": 2.073843240737915, | |
| "learning_rate": 2.2450588904444968e-05, | |
| "loss": 0.1251, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.2334238076773945, | |
| "grad_norm": 1.7229952812194824, | |
| "learning_rate": 2.0874495183062503e-05, | |
| "loss": 0.1151, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.2644435827840246, | |
| "grad_norm": 1.7413322925567627, | |
| "learning_rate": 1.9341043101671412e-05, | |
| "loss": 0.1176, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.295463357890655, | |
| "grad_norm": 1.87288236618042, | |
| "learning_rate": 1.7852477734202954e-05, | |
| "loss": 0.1099, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.3264831329972857, | |
| "grad_norm": 1.6436331272125244, | |
| "learning_rate": 1.641097843750952e-05, | |
| "loss": 0.1061, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.3575029081039163, | |
| "grad_norm": 2.2512550354003906, | |
| "learning_rate": 1.5018655660644055e-05, | |
| "loss": 0.0849, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.388522683210547, | |
| "grad_norm": 1.8482258319854736, | |
| "learning_rate": 1.3677547855024907e-05, | |
| "loss": 0.1089, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.419542458317177, | |
| "grad_norm": 1.5719361305236816, | |
| "learning_rate": 1.2389618490009775e-05, | |
| "loss": 0.0837, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.4505622334238075, | |
| "grad_norm": 1.8144450187683105, | |
| "learning_rate": 1.1156753178248564e-05, | |
| "loss": 0.1133, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.481582008530438, | |
| "grad_norm": 2.3764026165008545, | |
| "learning_rate": 9.980756915023332e-06, | |
| "loss": 0.09, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.5126017836370687, | |
| "grad_norm": 1.60674250125885, | |
| "learning_rate": 8.863351435617395e-06, | |
| "loss": 0.0831, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.543621558743699, | |
| "grad_norm": 1.7593117952346802, | |
| "learning_rate": 7.806172694582487e-06, | |
| "loss": 0.0951, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.5746413338503293, | |
| "grad_norm": 2.17861008644104, | |
| "learning_rate": 6.810768470594553e-06, | |
| "loss": 0.0938, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 2.60566110895696, | |
| "grad_norm": 1.4292868375778198, | |
| "learning_rate": 5.878596100404743e-06, | |
| "loss": 0.0738, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 2.6366808840635905, | |
| "grad_norm": 1.06781005859375, | |
| "learning_rate": 5.01102034520326e-06, | |
| "loss": 0.0893, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 2.667700659170221, | |
| "grad_norm": 1.4993703365325928, | |
| "learning_rate": 4.209311392519955e-06, | |
| "loss": 0.0706, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 2.6987204342768516, | |
| "grad_norm": 1.5031639337539673, | |
| "learning_rate": 3.4746429965867967e-06, | |
| "loss": 0.0752, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 2.7297402093834817, | |
| "grad_norm": 1.7316648960113525, | |
| "learning_rate": 2.808090759885207e-06, | |
| "loss": 0.075, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 2.7607599844901123, | |
| "grad_norm": 1.4538629055023193, | |
| "learning_rate": 2.2106305583936617e-06, | |
| "loss": 0.0768, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 2.791779759596743, | |
| "grad_norm": 0.8957967162132263, | |
| "learning_rate": 1.6831371128416983e-06, | |
| "loss": 0.0726, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 2.8227995347033734, | |
| "grad_norm": 1.7930282354354858, | |
| "learning_rate": 1.2263827080616074e-06, | |
| "loss": 0.0655, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 2.853819309810004, | |
| "grad_norm": 1.678768277168274, | |
| "learning_rate": 8.410360623130554e-07, | |
| "loss": 0.0743, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 2.884839084916634, | |
| "grad_norm": 1.4549870491027832, | |
| "learning_rate": 5.276613482359138e-07, | |
| "loss": 0.0773, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 2.9158588600232647, | |
| "grad_norm": 1.54071044921875, | |
| "learning_rate": 2.867173668646583e-07, | |
| "loss": 0.0804, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 2.9468786351298952, | |
| "grad_norm": 1.3247941732406616, | |
| "learning_rate": 1.1855687591376675e-07, | |
| "loss": 0.0756, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 2.977898410236526, | |
| "grad_norm": 1.7488839626312256, | |
| "learning_rate": 2.342607331733415e-08, | |
| "loss": 0.0805, | |
| "step": 960 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 966, | |
| "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": 9.822238727303332e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |