| { | |
| "best_global_step": 2000, | |
| "best_metric": 0.8785109519958496, | |
| "best_model_checkpoint": "./results/checkpoint-2000", | |
| "epoch": 0.04941590401454804, | |
| "eval_steps": 500, | |
| "global_step": 2500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00019766361605819217, | |
| "grad_norm": 12.468854904174805, | |
| "learning_rate": 9e-07, | |
| "loss": 3.6886, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.00039532723211638433, | |
| "grad_norm": 52.02771759033203, | |
| "learning_rate": 1.9e-06, | |
| "loss": 3.597, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0005929908481745765, | |
| "grad_norm": 27.688196182250977, | |
| "learning_rate": 2.9e-06, | |
| "loss": 3.8032, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0007906544642327687, | |
| "grad_norm": 12.250176429748535, | |
| "learning_rate": 3.9e-06, | |
| "loss": 3.3572, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.000988318080290961, | |
| "grad_norm": 20.915935516357422, | |
| "learning_rate": 4.9000000000000005e-06, | |
| "loss": 3.3315, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.001185981696349153, | |
| "grad_norm": 11.775025367736816, | |
| "learning_rate": 5.9e-06, | |
| "loss": 2.7832, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0013836453124073452, | |
| "grad_norm": 12.706207275390625, | |
| "learning_rate": 6.900000000000001e-06, | |
| "loss": 2.3718, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.0015813089284655373, | |
| "grad_norm": 21.651674270629883, | |
| "learning_rate": 7.9e-06, | |
| "loss": 2.0937, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0017789725445237295, | |
| "grad_norm": 15.798470497131348, | |
| "learning_rate": 8.9e-06, | |
| "loss": 1.7039, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.001976636160581922, | |
| "grad_norm": 9.725250244140625, | |
| "learning_rate": 9.900000000000002e-06, | |
| "loss": 1.5623, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.002174299776640114, | |
| "grad_norm": 3.9574215412139893, | |
| "learning_rate": 1.09e-05, | |
| "loss": 1.6024, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.002371963392698306, | |
| "grad_norm": 8.425467491149902, | |
| "learning_rate": 1.19e-05, | |
| "loss": 1.0913, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.0025696270087564982, | |
| "grad_norm": 9.648313522338867, | |
| "learning_rate": 1.29e-05, | |
| "loss": 1.3395, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.0027672906248146904, | |
| "grad_norm": 0.9009546041488647, | |
| "learning_rate": 1.3900000000000002e-05, | |
| "loss": 0.999, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.0029649542408728825, | |
| "grad_norm": 10.469375610351562, | |
| "learning_rate": 1.49e-05, | |
| "loss": 0.8274, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.0031626178569310747, | |
| "grad_norm": 0.22269444167613983, | |
| "learning_rate": 1.59e-05, | |
| "loss": 0.8564, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.003360281472989267, | |
| "grad_norm": 7.279628276824951, | |
| "learning_rate": 1.69e-05, | |
| "loss": 1.0921, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.003557945089047459, | |
| "grad_norm": 0.6602649092674255, | |
| "learning_rate": 1.79e-05, | |
| "loss": 0.9095, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.003755608705105651, | |
| "grad_norm": 0.1962263137102127, | |
| "learning_rate": 1.8900000000000002e-05, | |
| "loss": 0.6642, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.003953272321163844, | |
| "grad_norm": 0.7642993927001953, | |
| "learning_rate": 1.9900000000000003e-05, | |
| "loss": 0.8014, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.004150935937222036, | |
| "grad_norm": 0.7219191193580627, | |
| "learning_rate": 2.09e-05, | |
| "loss": 1.0484, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.004348599553280228, | |
| "grad_norm": 7.698592185974121, | |
| "learning_rate": 2.19e-05, | |
| "loss": 1.2609, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.00454626316933842, | |
| "grad_norm": 0.9080265760421753, | |
| "learning_rate": 2.29e-05, | |
| "loss": 0.9157, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.004743926785396612, | |
| "grad_norm": 0.198422372341156, | |
| "learning_rate": 2.39e-05, | |
| "loss": 1.2826, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.004941590401454804, | |
| "grad_norm": 0.267986923456192, | |
| "learning_rate": 2.4900000000000002e-05, | |
| "loss": 1.0538, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.0051392540175129965, | |
| "grad_norm": 9.535926818847656, | |
| "learning_rate": 2.5900000000000003e-05, | |
| "loss": 0.8342, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.005336917633571189, | |
| "grad_norm": 0.20855465531349182, | |
| "learning_rate": 2.6900000000000003e-05, | |
| "loss": 0.4623, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.005534581249629381, | |
| "grad_norm": 0.4157664477825165, | |
| "learning_rate": 2.7900000000000004e-05, | |
| "loss": 0.5038, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.005732244865687573, | |
| "grad_norm": 8.263617515563965, | |
| "learning_rate": 2.8899999999999998e-05, | |
| "loss": 0.8435, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.005929908481745765, | |
| "grad_norm": 9.216032981872559, | |
| "learning_rate": 2.9900000000000002e-05, | |
| "loss": 0.7289, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.006127572097803957, | |
| "grad_norm": 8.40715503692627, | |
| "learning_rate": 3.09e-05, | |
| "loss": 1.1021, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.006325235713862149, | |
| "grad_norm": 0.4089910089969635, | |
| "learning_rate": 3.19e-05, | |
| "loss": 0.7686, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.0065228993299203415, | |
| "grad_norm": 8.227189064025879, | |
| "learning_rate": 3.29e-05, | |
| "loss": 0.9375, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.006720562945978534, | |
| "grad_norm": 8.776185989379883, | |
| "learning_rate": 3.3900000000000004e-05, | |
| "loss": 0.9814, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.006918226562036726, | |
| "grad_norm": 7.721053123474121, | |
| "learning_rate": 3.49e-05, | |
| "loss": 0.4118, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.007115890178094918, | |
| "grad_norm": 5.676611423492432, | |
| "learning_rate": 3.59e-05, | |
| "loss": 1.1953, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.00731355379415311, | |
| "grad_norm": 7.924537658691406, | |
| "learning_rate": 3.69e-05, | |
| "loss": 1.1895, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.007511217410211302, | |
| "grad_norm": 8.677846908569336, | |
| "learning_rate": 3.79e-05, | |
| "loss": 0.8925, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.007708881026269494, | |
| "grad_norm": 6.243840217590332, | |
| "learning_rate": 3.8900000000000004e-05, | |
| "loss": 0.6888, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.007906544642327687, | |
| "grad_norm": 7.967802047729492, | |
| "learning_rate": 3.99e-05, | |
| "loss": 1.4355, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.008104208258385879, | |
| "grad_norm": 8.38342571258545, | |
| "learning_rate": 4.09e-05, | |
| "loss": 1.0131, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.008301871874444072, | |
| "grad_norm": 0.07280826568603516, | |
| "learning_rate": 4.19e-05, | |
| "loss": 1.3299, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.008499535490502263, | |
| "grad_norm": 0.2540580630302429, | |
| "learning_rate": 4.29e-05, | |
| "loss": 0.6579, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.008697199106560456, | |
| "grad_norm": 7.481705188751221, | |
| "learning_rate": 4.39e-05, | |
| "loss": 0.9635, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.008894862722618647, | |
| "grad_norm": 0.20026662945747375, | |
| "learning_rate": 4.49e-05, | |
| "loss": 0.883, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.00909252633867684, | |
| "grad_norm": 6.689612865447998, | |
| "learning_rate": 4.5900000000000004e-05, | |
| "loss": 0.9595, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.009290189954735031, | |
| "grad_norm": 0.06714965403079987, | |
| "learning_rate": 4.69e-05, | |
| "loss": 0.7816, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.009487853570793224, | |
| "grad_norm": 7.251598834991455, | |
| "learning_rate": 4.79e-05, | |
| "loss": 0.5981, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.009685517186851416, | |
| "grad_norm": 6.720303058624268, | |
| "learning_rate": 4.89e-05, | |
| "loss": 0.9463, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.009883180802909609, | |
| "grad_norm": 6.996522903442383, | |
| "learning_rate": 4.99e-05, | |
| "loss": 0.7919, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.009883180802909609, | |
| "eval_loss": 0.9214622378349304, | |
| "eval_runtime": 245.7679, | |
| "eval_samples_per_second": 86.683, | |
| "eval_steps_per_second": 10.835, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.0100808444189678, | |
| "grad_norm": 0.3730940818786621, | |
| "learning_rate": 4.9997025245747756e-05, | |
| "loss": 0.8747, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.010278508035025993, | |
| "grad_norm": 6.384608268737793, | |
| "learning_rate": 4.999371996324526e-05, | |
| "loss": 1.1554, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.010476171651084184, | |
| "grad_norm": 6.133856773376465, | |
| "learning_rate": 4.999041468074276e-05, | |
| "loss": 1.1372, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.010673835267142377, | |
| "grad_norm": 0.2688162624835968, | |
| "learning_rate": 4.998710939824027e-05, | |
| "loss": 0.9546, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.010871498883200569, | |
| "grad_norm": 7.641685962677002, | |
| "learning_rate": 4.9983804115737774e-05, | |
| "loss": 0.9482, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.011069162499258762, | |
| "grad_norm": 6.708981037139893, | |
| "learning_rate": 4.998049883323528e-05, | |
| "loss": 0.6834, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.011266826115316953, | |
| "grad_norm": 7.363180637359619, | |
| "learning_rate": 4.997719355073278e-05, | |
| "loss": 0.7817, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.011464489731375146, | |
| "grad_norm": 6.942836284637451, | |
| "learning_rate": 4.997388826823029e-05, | |
| "loss": 1.0409, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.011662153347433339, | |
| "grad_norm": 5.417139530181885, | |
| "learning_rate": 4.997058298572779e-05, | |
| "loss": 1.1497, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.01185981696349153, | |
| "grad_norm": 0.11549051851034164, | |
| "learning_rate": 4.9967277703225295e-05, | |
| "loss": 0.671, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.012057480579549723, | |
| "grad_norm": 6.788494110107422, | |
| "learning_rate": 4.99639724207228e-05, | |
| "loss": 1.1224, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.012255144195607914, | |
| "grad_norm": 7.265660285949707, | |
| "learning_rate": 4.996066713822031e-05, | |
| "loss": 0.7574, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.012452807811666107, | |
| "grad_norm": 6.8422441482543945, | |
| "learning_rate": 4.995736185571781e-05, | |
| "loss": 0.668, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.012650471427724299, | |
| "grad_norm": 6.232938289642334, | |
| "learning_rate": 4.995405657321531e-05, | |
| "loss": 1.0247, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.012848135043782492, | |
| "grad_norm": 0.2170286625623703, | |
| "learning_rate": 4.995075129071282e-05, | |
| "loss": 0.8753, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.013045798659840683, | |
| "grad_norm": 5.531427383422852, | |
| "learning_rate": 4.9947446008210326e-05, | |
| "loss": 1.1531, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.013243462275898876, | |
| "grad_norm": 6.2044548988342285, | |
| "learning_rate": 4.994414072570783e-05, | |
| "loss": 0.8807, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.013441125891957067, | |
| "grad_norm": 6.0426836013793945, | |
| "learning_rate": 4.994083544320533e-05, | |
| "loss": 1.0564, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.01363878950801526, | |
| "grad_norm": 0.18100695312023163, | |
| "learning_rate": 4.993753016070284e-05, | |
| "loss": 0.845, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.013836453124073451, | |
| "grad_norm": 5.483642101287842, | |
| "learning_rate": 4.9934224878200344e-05, | |
| "loss": 1.1766, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.014034116740131644, | |
| "grad_norm": 4.922057628631592, | |
| "learning_rate": 4.993091959569785e-05, | |
| "loss": 1.0413, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.014231780356189836, | |
| "grad_norm": 5.475461959838867, | |
| "learning_rate": 4.9927614313195356e-05, | |
| "loss": 0.9466, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.014429443972248029, | |
| "grad_norm": 6.56538200378418, | |
| "learning_rate": 4.992430903069286e-05, | |
| "loss": 0.9341, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.01462710758830622, | |
| "grad_norm": 4.976065635681152, | |
| "learning_rate": 4.992100374819036e-05, | |
| "loss": 0.8329, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.014824771204364413, | |
| "grad_norm": 0.05167540907859802, | |
| "learning_rate": 4.9917698465687865e-05, | |
| "loss": 0.698, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.015022434820422604, | |
| "grad_norm": 0.05338999629020691, | |
| "learning_rate": 4.991439318318537e-05, | |
| "loss": 0.4403, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.015220098436480797, | |
| "grad_norm": 6.14265251159668, | |
| "learning_rate": 4.991108790068287e-05, | |
| "loss": 0.9864, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.015417762052538989, | |
| "grad_norm": 0.1925213485956192, | |
| "learning_rate": 4.9907782618180374e-05, | |
| "loss": 1.1037, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.015615425668597182, | |
| "grad_norm": 5.5837554931640625, | |
| "learning_rate": 4.990447733567788e-05, | |
| "loss": 0.8571, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.015813089284655375, | |
| "grad_norm": 3.6623053550720215, | |
| "learning_rate": 4.9901172053175386e-05, | |
| "loss": 0.9082, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.016010752900713566, | |
| "grad_norm": 0.05198395252227783, | |
| "learning_rate": 4.989786677067289e-05, | |
| "loss": 0.8335, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.016208416516771757, | |
| "grad_norm": 5.30173921585083, | |
| "learning_rate": 4.989456148817039e-05, | |
| "loss": 0.6003, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.01640608013282995, | |
| "grad_norm": 4.080887317657471, | |
| "learning_rate": 4.98912562056679e-05, | |
| "loss": 0.7937, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.016603743748888143, | |
| "grad_norm": 5.606449604034424, | |
| "learning_rate": 4.9887950923165404e-05, | |
| "loss": 1.2462, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.016801407364946334, | |
| "grad_norm": 6.573556900024414, | |
| "learning_rate": 4.988464564066291e-05, | |
| "loss": 0.819, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.016999070981004526, | |
| "grad_norm": 0.06031334027647972, | |
| "learning_rate": 4.988134035816042e-05, | |
| "loss": 0.6425, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.017196734597062717, | |
| "grad_norm": 4.252384662628174, | |
| "learning_rate": 4.987803507565792e-05, | |
| "loss": 0.8228, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.01739439821312091, | |
| "grad_norm": 0.12518936395645142, | |
| "learning_rate": 4.987472979315542e-05, | |
| "loss": 0.7576, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.017592061829179103, | |
| "grad_norm": 5.032276630401611, | |
| "learning_rate": 4.9871424510652925e-05, | |
| "loss": 0.8783, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.017789725445237294, | |
| "grad_norm": 0.059717316180467606, | |
| "learning_rate": 4.9868119228150435e-05, | |
| "loss": 0.8153, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.01798738906129549, | |
| "grad_norm": 3.093669891357422, | |
| "learning_rate": 4.986481394564794e-05, | |
| "loss": 0.6105, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.01818505267735368, | |
| "grad_norm": 3.593827962875366, | |
| "learning_rate": 4.986150866314544e-05, | |
| "loss": 0.6549, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.01838271629341187, | |
| "grad_norm": 0.11692095547914505, | |
| "learning_rate": 4.985820338064295e-05, | |
| "loss": 1.0544, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.018580379909470063, | |
| "grad_norm": 0.07805492728948593, | |
| "learning_rate": 4.985489809814045e-05, | |
| "loss": 0.8575, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.018778043525528258, | |
| "grad_norm": 3.621807098388672, | |
| "learning_rate": 4.9851592815637956e-05, | |
| "loss": 1.1574, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.01897570714158645, | |
| "grad_norm": 0.1467035412788391, | |
| "learning_rate": 4.984828753313546e-05, | |
| "loss": 0.9261, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.01917337075764464, | |
| "grad_norm": 5.1112895011901855, | |
| "learning_rate": 4.984498225063297e-05, | |
| "loss": 1.1392, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.01937103437370283, | |
| "grad_norm": 4.554998397827148, | |
| "learning_rate": 4.984167696813047e-05, | |
| "loss": 1.0693, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.019568697989761026, | |
| "grad_norm": 0.08779212087392807, | |
| "learning_rate": 4.9838371685627974e-05, | |
| "loss": 0.6191, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.019766361605819217, | |
| "grad_norm": 0.050688233226537704, | |
| "learning_rate": 4.983506640312548e-05, | |
| "loss": 0.9877, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.019766361605819217, | |
| "eval_loss": 0.9038341045379639, | |
| "eval_runtime": 236.3116, | |
| "eval_samples_per_second": 90.152, | |
| "eval_steps_per_second": 11.269, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.01996402522187741, | |
| "grad_norm": 0.0693553164601326, | |
| "learning_rate": 4.983176112062298e-05, | |
| "loss": 1.1317, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.0201616888379356, | |
| "grad_norm": 4.626270771026611, | |
| "learning_rate": 4.982845583812048e-05, | |
| "loss": 0.9338, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.020359352453993795, | |
| "grad_norm": 4.608981132507324, | |
| "learning_rate": 4.9825150555617986e-05, | |
| "loss": 1.0849, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.020557016070051986, | |
| "grad_norm": 4.660639762878418, | |
| "learning_rate": 4.9821845273115495e-05, | |
| "loss": 1.4633, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.020754679686110177, | |
| "grad_norm": 4.729643821716309, | |
| "learning_rate": 4.9818539990613e-05, | |
| "loss": 1.0748, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.02095234330216837, | |
| "grad_norm": 0.027232196182012558, | |
| "learning_rate": 4.98152347081105e-05, | |
| "loss": 0.7562, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.021150006918226563, | |
| "grad_norm": 0.018449297174811363, | |
| "learning_rate": 4.981192942560801e-05, | |
| "loss": 0.4423, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.021347670534284754, | |
| "grad_norm": 3.372770309448242, | |
| "learning_rate": 4.9808624143105514e-05, | |
| "loss": 0.7371, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.021545334150342946, | |
| "grad_norm": 0.24770286679267883, | |
| "learning_rate": 4.9805318860603016e-05, | |
| "loss": 1.4439, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.021742997766401137, | |
| "grad_norm": 0.11222311854362488, | |
| "learning_rate": 4.980201357810052e-05, | |
| "loss": 0.6073, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.021940661382459332, | |
| "grad_norm": 4.456133842468262, | |
| "learning_rate": 4.979870829559803e-05, | |
| "loss": 0.6697, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.022138324998517523, | |
| "grad_norm": 0.05747058987617493, | |
| "learning_rate": 4.979540301309553e-05, | |
| "loss": 0.949, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.022335988614575714, | |
| "grad_norm": 3.084164619445801, | |
| "learning_rate": 4.9792097730593035e-05, | |
| "loss": 0.8189, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.022533652230633906, | |
| "grad_norm": 3.456510305404663, | |
| "learning_rate": 4.9788792448090544e-05, | |
| "loss": 1.3608, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.0227313158466921, | |
| "grad_norm": 3.4595232009887695, | |
| "learning_rate": 4.978548716558805e-05, | |
| "loss": 0.8679, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.02292897946275029, | |
| "grad_norm": 4.478796005249023, | |
| "learning_rate": 4.978218188308555e-05, | |
| "loss": 0.8229, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.023126643078808483, | |
| "grad_norm": 0.07277049869298935, | |
| "learning_rate": 4.977887660058305e-05, | |
| "loss": 0.5847, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.023324306694866678, | |
| "grad_norm": 0.13703665137290955, | |
| "learning_rate": 4.977557131808056e-05, | |
| "loss": 1.0406, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.02352197031092487, | |
| "grad_norm": 3.7098865509033203, | |
| "learning_rate": 4.9772266035578065e-05, | |
| "loss": 0.4147, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.02371963392698306, | |
| "grad_norm": 0.03374309837818146, | |
| "learning_rate": 4.976896075307557e-05, | |
| "loss": 0.8773, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.02391729754304125, | |
| "grad_norm": 4.257831573486328, | |
| "learning_rate": 4.976565547057308e-05, | |
| "loss": 0.6989, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.024114961159099446, | |
| "grad_norm": 4.520907402038574, | |
| "learning_rate": 4.976235018807058e-05, | |
| "loss": 0.6721, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.024312624775157637, | |
| "grad_norm": 7.800019264221191, | |
| "learning_rate": 4.9759044905568084e-05, | |
| "loss": 0.9356, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.02451028839121583, | |
| "grad_norm": 3.6012537479400635, | |
| "learning_rate": 4.9755739623065586e-05, | |
| "loss": 0.9297, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.02470795200727402, | |
| "grad_norm": 4.273988246917725, | |
| "learning_rate": 4.975243434056309e-05, | |
| "loss": 1.1335, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.024905615623332215, | |
| "grad_norm": 0.15219643712043762, | |
| "learning_rate": 4.974912905806059e-05, | |
| "loss": 0.608, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.025103279239390406, | |
| "grad_norm": 4.302224636077881, | |
| "learning_rate": 4.9745823775558095e-05, | |
| "loss": 0.974, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.025300942855448597, | |
| "grad_norm": 0.09040108323097229, | |
| "learning_rate": 4.9742518493055605e-05, | |
| "loss": 0.871, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.02549860647150679, | |
| "grad_norm": 0.08080489933490753, | |
| "learning_rate": 4.973921321055311e-05, | |
| "loss": 1.0142, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.025696270087564983, | |
| "grad_norm": 4.567043781280518, | |
| "learning_rate": 4.973590792805061e-05, | |
| "loss": 0.9593, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.025893933703623175, | |
| "grad_norm": 3.6397976875305176, | |
| "learning_rate": 4.973260264554811e-05, | |
| "loss": 1.0869, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.026091597319681366, | |
| "grad_norm": 3.9633944034576416, | |
| "learning_rate": 4.972929736304562e-05, | |
| "loss": 0.8078, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.026289260935739557, | |
| "grad_norm": 2.7522361278533936, | |
| "learning_rate": 4.9725992080543126e-05, | |
| "loss": 1.0228, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.026486924551797752, | |
| "grad_norm": 3.5321545600891113, | |
| "learning_rate": 4.972268679804063e-05, | |
| "loss": 1.1676, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.026684588167855943, | |
| "grad_norm": 3.301154851913452, | |
| "learning_rate": 4.971938151553814e-05, | |
| "loss": 0.9431, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.026882251783914134, | |
| "grad_norm": 0.0989571139216423, | |
| "learning_rate": 4.971607623303564e-05, | |
| "loss": 1.0831, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.027079915399972326, | |
| "grad_norm": 3.8333899974823, | |
| "learning_rate": 4.9712770950533144e-05, | |
| "loss": 1.3761, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.02727757901603052, | |
| "grad_norm": 3.214996099472046, | |
| "learning_rate": 4.970946566803065e-05, | |
| "loss": 0.7668, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.02747524263208871, | |
| "grad_norm": 3.960110902786255, | |
| "learning_rate": 4.9706160385528156e-05, | |
| "loss": 1.002, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.027672906248146903, | |
| "grad_norm": 4.702463150024414, | |
| "learning_rate": 4.970285510302566e-05, | |
| "loss": 1.013, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.027870569864205094, | |
| "grad_norm": 4.348130226135254, | |
| "learning_rate": 4.969954982052316e-05, | |
| "loss": 0.828, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.02806823348026329, | |
| "grad_norm": 0.1650361716747284, | |
| "learning_rate": 4.9696244538020665e-05, | |
| "loss": 0.4713, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.02826589709632148, | |
| "grad_norm": 0.0727657675743103, | |
| "learning_rate": 4.9692939255518175e-05, | |
| "loss": 0.9447, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.02846356071237967, | |
| "grad_norm": 3.5271363258361816, | |
| "learning_rate": 4.968963397301568e-05, | |
| "loss": 1.0093, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.028661224328437863, | |
| "grad_norm": 0.10453560948371887, | |
| "learning_rate": 4.968632869051318e-05, | |
| "loss": 0.7617, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.028858887944496057, | |
| "grad_norm": 2.5674312114715576, | |
| "learning_rate": 4.968302340801069e-05, | |
| "loss": 1.0254, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.02905655156055425, | |
| "grad_norm": 4.187676429748535, | |
| "learning_rate": 4.967971812550819e-05, | |
| "loss": 1.119, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.02925421517661244, | |
| "grad_norm": 0.09666266292333603, | |
| "learning_rate": 4.9676412843005696e-05, | |
| "loss": 0.6603, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.029451878792670635, | |
| "grad_norm": 3.545473098754883, | |
| "learning_rate": 4.96731075605032e-05, | |
| "loss": 1.0676, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.029649542408728826, | |
| "grad_norm": 3.4705913066864014, | |
| "learning_rate": 4.96698022780007e-05, | |
| "loss": 0.6608, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.029649542408728826, | |
| "eval_loss": 0.8907366394996643, | |
| "eval_runtime": 236.1612, | |
| "eval_samples_per_second": 90.21, | |
| "eval_steps_per_second": 11.276, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.029847206024787017, | |
| "grad_norm": 0.06528951227664948, | |
| "learning_rate": 4.9666496995498204e-05, | |
| "loss": 0.8131, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.03004486964084521, | |
| "grad_norm": 0.11044718325138092, | |
| "learning_rate": 4.966319171299571e-05, | |
| "loss": 0.9144, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.030242533256903403, | |
| "grad_norm": 3.092930316925049, | |
| "learning_rate": 4.965988643049322e-05, | |
| "loss": 1.0863, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.030440196872961595, | |
| "grad_norm": 3.9734549522399902, | |
| "learning_rate": 4.965658114799072e-05, | |
| "loss": 0.5064, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.030637860489019786, | |
| "grad_norm": 0.035740889608860016, | |
| "learning_rate": 4.965327586548822e-05, | |
| "loss": 1.222, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.030835524105077977, | |
| "grad_norm": 0.06910470873117447, | |
| "learning_rate": 4.9649970582985725e-05, | |
| "loss": 0.8428, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.031033187721136172, | |
| "grad_norm": 3.261096715927124, | |
| "learning_rate": 4.9646665300483235e-05, | |
| "loss": 0.9363, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.031230851337194363, | |
| "grad_norm": 0.09766729176044464, | |
| "learning_rate": 4.964336001798074e-05, | |
| "loss": 0.803, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.03142851495325256, | |
| "grad_norm": 3.2366461753845215, | |
| "learning_rate": 4.964005473547824e-05, | |
| "loss": 1.2179, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.03162617856931075, | |
| "grad_norm": 0.09953774511814117, | |
| "learning_rate": 4.963674945297575e-05, | |
| "loss": 1.0656, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.03182384218536894, | |
| "grad_norm": 3.7796168327331543, | |
| "learning_rate": 4.963344417047325e-05, | |
| "loss": 1.0625, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.03202150580142713, | |
| "grad_norm": 0.16226442158222198, | |
| "learning_rate": 4.9630138887970756e-05, | |
| "loss": 0.8619, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.03221916941748532, | |
| "grad_norm": 3.7525415420532227, | |
| "learning_rate": 4.962683360546826e-05, | |
| "loss": 0.9852, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.032416833033543514, | |
| "grad_norm": 0.07482007145881653, | |
| "learning_rate": 4.962352832296577e-05, | |
| "loss": 0.8297, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.032614496649601706, | |
| "grad_norm": 0.1131066307425499, | |
| "learning_rate": 4.962022304046327e-05, | |
| "loss": 1.1424, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.0328121602656599, | |
| "grad_norm": 0.05459194257855415, | |
| "learning_rate": 4.9616917757960774e-05, | |
| "loss": 0.7202, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.033009823881718095, | |
| "grad_norm": 3.9901282787323, | |
| "learning_rate": 4.9613612475458284e-05, | |
| "loss": 0.8532, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.033207487497776286, | |
| "grad_norm": 4.42426061630249, | |
| "learning_rate": 4.961030719295579e-05, | |
| "loss": 0.977, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.03340515111383448, | |
| "grad_norm": 0.06116814911365509, | |
| "learning_rate": 4.960700191045329e-05, | |
| "loss": 0.5627, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.03360281472989267, | |
| "grad_norm": 3.5464470386505127, | |
| "learning_rate": 4.960369662795079e-05, | |
| "loss": 0.6769, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.03380047834595086, | |
| "grad_norm": 3.3844430446624756, | |
| "learning_rate": 4.96003913454483e-05, | |
| "loss": 1.448, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.03399814196200905, | |
| "grad_norm": 3.592660427093506, | |
| "learning_rate": 4.9597086062945805e-05, | |
| "loss": 0.9505, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.03419580557806724, | |
| "grad_norm": 3.355842113494873, | |
| "learning_rate": 4.959378078044331e-05, | |
| "loss": 1.0866, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.034393469194125434, | |
| "grad_norm": 3.0316247940063477, | |
| "learning_rate": 4.959047549794081e-05, | |
| "loss": 0.6762, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.03459113281018363, | |
| "grad_norm": 0.06703350692987442, | |
| "learning_rate": 4.9587170215438314e-05, | |
| "loss": 0.828, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.03478879642624182, | |
| "grad_norm": 3.7815728187561035, | |
| "learning_rate": 4.9583864932935816e-05, | |
| "loss": 1.3015, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.034986460042300015, | |
| "grad_norm": 3.3532164096832275, | |
| "learning_rate": 4.958055965043332e-05, | |
| "loss": 1.2062, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.035184123658358206, | |
| "grad_norm": 3.1324400901794434, | |
| "learning_rate": 4.957725436793083e-05, | |
| "loss": 0.9484, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.0353817872744164, | |
| "grad_norm": 0.03721948713064194, | |
| "learning_rate": 4.957394908542833e-05, | |
| "loss": 0.782, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.03557945089047459, | |
| "grad_norm": 2.5899410247802734, | |
| "learning_rate": 4.9570643802925835e-05, | |
| "loss": 0.9005, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.03577711450653278, | |
| "grad_norm": 0.14065319299697876, | |
| "learning_rate": 4.9567338520423344e-05, | |
| "loss": 1.2058, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.03597477812259098, | |
| "grad_norm": 3.255124568939209, | |
| "learning_rate": 4.956403323792085e-05, | |
| "loss": 0.6861, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.03617244173864917, | |
| "grad_norm": 2.9060869216918945, | |
| "learning_rate": 4.956072795541835e-05, | |
| "loss": 0.6472, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.03637010535470736, | |
| "grad_norm": 0.10534641146659851, | |
| "learning_rate": 4.955742267291585e-05, | |
| "loss": 1.1794, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.03656776897076555, | |
| "grad_norm": 3.3723835945129395, | |
| "learning_rate": 4.955411739041336e-05, | |
| "loss": 1.0936, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.03676543258682374, | |
| "grad_norm": 3.3948323726654053, | |
| "learning_rate": 4.9550812107910865e-05, | |
| "loss": 1.0723, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.036963096202881934, | |
| "grad_norm": 2.8475537300109863, | |
| "learning_rate": 4.954750682540837e-05, | |
| "loss": 1.1292, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.037160759818940126, | |
| "grad_norm": 3.0720722675323486, | |
| "learning_rate": 4.954420154290588e-05, | |
| "loss": 0.4767, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.03735842343499832, | |
| "grad_norm": 2.8738229274749756, | |
| "learning_rate": 4.954089626040338e-05, | |
| "loss": 0.8313, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.037556087051056515, | |
| "grad_norm": 0.04877200722694397, | |
| "learning_rate": 4.9537590977900884e-05, | |
| "loss": 0.8162, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.037753750667114706, | |
| "grad_norm": 0.15653006732463837, | |
| "learning_rate": 4.9534285695398386e-05, | |
| "loss": 1.117, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.0379514142831729, | |
| "grad_norm": 0.09898870438337326, | |
| "learning_rate": 4.9530980412895896e-05, | |
| "loss": 0.8124, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.03814907789923109, | |
| "grad_norm": 3.6042723655700684, | |
| "learning_rate": 4.95276751303934e-05, | |
| "loss": 0.978, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.03834674151528928, | |
| "grad_norm": 2.490565776824951, | |
| "learning_rate": 4.95243698478909e-05, | |
| "loss": 0.4846, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.03854440513134747, | |
| "grad_norm": 2.3419349193573, | |
| "learning_rate": 4.952106456538841e-05, | |
| "loss": 0.7827, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.03874206874740566, | |
| "grad_norm": 2.327219009399414, | |
| "learning_rate": 4.9517759282885914e-05, | |
| "loss": 0.9868, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.038939732363463854, | |
| "grad_norm": 0.18203556537628174, | |
| "learning_rate": 4.951445400038342e-05, | |
| "loss": 0.9514, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.03913739597952205, | |
| "grad_norm": 2.613198757171631, | |
| "learning_rate": 4.951114871788092e-05, | |
| "loss": 0.7549, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.03933505959558024, | |
| "grad_norm": 4.419915676116943, | |
| "learning_rate": 4.950784343537842e-05, | |
| "loss": 0.6902, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.039532723211638435, | |
| "grad_norm": 0.04059309512376785, | |
| "learning_rate": 4.9504538152875926e-05, | |
| "loss": 0.8878, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.039532723211638435, | |
| "eval_loss": 0.8785109519958496, | |
| "eval_runtime": 237.1008, | |
| "eval_samples_per_second": 89.852, | |
| "eval_steps_per_second": 11.232, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.039730386827696626, | |
| "grad_norm": 2.7908315658569336, | |
| "learning_rate": 4.950123287037343e-05, | |
| "loss": 1.0453, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.03992805044375482, | |
| "grad_norm": 0.19223953783512115, | |
| "learning_rate": 4.949792758787094e-05, | |
| "loss": 0.9602, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.04012571405981301, | |
| "grad_norm": 2.7216477394104004, | |
| "learning_rate": 4.949462230536844e-05, | |
| "loss": 1.1187, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.0403233776758712, | |
| "grad_norm": 3.306227922439575, | |
| "learning_rate": 4.9491317022865944e-05, | |
| "loss": 0.9302, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.0405210412919294, | |
| "grad_norm": 3.259765386581421, | |
| "learning_rate": 4.948801174036345e-05, | |
| "loss": 0.9832, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.04071870490798759, | |
| "grad_norm": 3.6598448753356934, | |
| "learning_rate": 4.9484706457860956e-05, | |
| "loss": 1.0918, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.04091636852404578, | |
| "grad_norm": 3.6276862621307373, | |
| "learning_rate": 4.948140117535846e-05, | |
| "loss": 0.8682, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.04111403214010397, | |
| "grad_norm": 3.547147035598755, | |
| "learning_rate": 4.947809589285596e-05, | |
| "loss": 0.5364, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.04131169575616216, | |
| "grad_norm": 3.0729477405548096, | |
| "learning_rate": 4.947479061035347e-05, | |
| "loss": 1.2961, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.041509359372220354, | |
| "grad_norm": 2.370206594467163, | |
| "learning_rate": 4.9471485327850975e-05, | |
| "loss": 0.6171, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.041707022988278546, | |
| "grad_norm": 3.900785446166992, | |
| "learning_rate": 4.946818004534848e-05, | |
| "loss": 1.3057, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.04190468660433674, | |
| "grad_norm": 3.226991653442383, | |
| "learning_rate": 4.946487476284598e-05, | |
| "loss": 1.0164, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.042102350220394935, | |
| "grad_norm": 0.5991417169570923, | |
| "learning_rate": 4.946156948034349e-05, | |
| "loss": 1.4067, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.042300013836453126, | |
| "grad_norm": 3.0779223442077637, | |
| "learning_rate": 4.945826419784099e-05, | |
| "loss": 0.7529, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.04249767745251132, | |
| "grad_norm": 0.05838579684495926, | |
| "learning_rate": 4.9454958915338496e-05, | |
| "loss": 0.7606, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.04269534106856951, | |
| "grad_norm": 3.432065486907959, | |
| "learning_rate": 4.9451653632836005e-05, | |
| "loss": 1.1331, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.0428930046846277, | |
| "grad_norm": 2.9518392086029053, | |
| "learning_rate": 4.944834835033351e-05, | |
| "loss": 0.6276, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.04309066830068589, | |
| "grad_norm": 0.06247523054480553, | |
| "learning_rate": 4.944504306783101e-05, | |
| "loss": 0.7537, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.04328833191674408, | |
| "grad_norm": 2.4885005950927734, | |
| "learning_rate": 4.9441737785328514e-05, | |
| "loss": 1.1748, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.043485995532802274, | |
| "grad_norm": 0.15998195111751556, | |
| "learning_rate": 4.9438432502826024e-05, | |
| "loss": 0.9554, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.04368365914886047, | |
| "grad_norm": 3.214895248413086, | |
| "learning_rate": 4.9435127220323526e-05, | |
| "loss": 0.8503, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.043881322764918664, | |
| "grad_norm": 0.03642280027270317, | |
| "learning_rate": 4.943182193782103e-05, | |
| "loss": 0.4855, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.044078986380976855, | |
| "grad_norm": 2.707956075668335, | |
| "learning_rate": 4.942851665531853e-05, | |
| "loss": 1.0405, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.044276649997035046, | |
| "grad_norm": 0.08170021325349808, | |
| "learning_rate": 4.9425211372816035e-05, | |
| "loss": 0.9093, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.04447431361309324, | |
| "grad_norm": 2.322641611099243, | |
| "learning_rate": 4.942190609031354e-05, | |
| "loss": 1.357, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.04467197722915143, | |
| "grad_norm": 2.5577518939971924, | |
| "learning_rate": 4.941860080781104e-05, | |
| "loss": 1.2809, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.04486964084520962, | |
| "grad_norm": 2.471879243850708, | |
| "learning_rate": 4.941529552530855e-05, | |
| "loss": 1.1575, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.04506730446126781, | |
| "grad_norm": 3.389317274093628, | |
| "learning_rate": 4.941199024280605e-05, | |
| "loss": 1.1552, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.04526496807732601, | |
| "grad_norm": 0.11907307803630829, | |
| "learning_rate": 4.9408684960303556e-05, | |
| "loss": 0.685, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.0454626316933842, | |
| "grad_norm": 3.5071091651916504, | |
| "learning_rate": 4.940537967780106e-05, | |
| "loss": 0.5446, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.04566029530944239, | |
| "grad_norm": 2.466038227081299, | |
| "learning_rate": 4.940207439529857e-05, | |
| "loss": 0.9161, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.04585795892550058, | |
| "grad_norm": 2.913646936416626, | |
| "learning_rate": 4.939876911279607e-05, | |
| "loss": 0.8943, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.046055622541558774, | |
| "grad_norm": 2.976774215698242, | |
| "learning_rate": 4.9395463830293574e-05, | |
| "loss": 1.4517, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.046253286157616966, | |
| "grad_norm": 0.20807339251041412, | |
| "learning_rate": 4.9392158547791084e-05, | |
| "loss": 0.535, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.04645094977367516, | |
| "grad_norm": 3.2078752517700195, | |
| "learning_rate": 4.938885326528859e-05, | |
| "loss": 0.561, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.046648613389733355, | |
| "grad_norm": 2.7821130752563477, | |
| "learning_rate": 4.938554798278609e-05, | |
| "loss": 1.0918, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.046846277005791546, | |
| "grad_norm": 0.18904615938663483, | |
| "learning_rate": 4.938224270028359e-05, | |
| "loss": 1.1504, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.04704394062184974, | |
| "grad_norm": 2.9167234897613525, | |
| "learning_rate": 4.93789374177811e-05, | |
| "loss": 0.8307, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.04724160423790793, | |
| "grad_norm": 2.7260210514068604, | |
| "learning_rate": 4.9375632135278605e-05, | |
| "loss": 0.7618, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.04743926785396612, | |
| "grad_norm": 2.3594138622283936, | |
| "learning_rate": 4.937232685277611e-05, | |
| "loss": 0.4669, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.04763693147002431, | |
| "grad_norm": 2.587303400039673, | |
| "learning_rate": 4.936902157027362e-05, | |
| "loss": 0.6653, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.0478345950860825, | |
| "grad_norm": 0.07142865657806396, | |
| "learning_rate": 4.936571628777112e-05, | |
| "loss": 0.8185, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.048032258702140694, | |
| "grad_norm": 3.059152603149414, | |
| "learning_rate": 4.936241100526862e-05, | |
| "loss": 0.9088, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.04822992231819889, | |
| "grad_norm": 0.1070665493607521, | |
| "learning_rate": 4.9359105722766126e-05, | |
| "loss": 1.1149, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.048427585934257084, | |
| "grad_norm": 2.317112684249878, | |
| "learning_rate": 4.9355800440263636e-05, | |
| "loss": 0.5108, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.048625249550315275, | |
| "grad_norm": 2.8447506427764893, | |
| "learning_rate": 4.935249515776114e-05, | |
| "loss": 1.0359, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.048822913166373466, | |
| "grad_norm": 3.0613043308258057, | |
| "learning_rate": 4.934918987525864e-05, | |
| "loss": 0.7497, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.04902057678243166, | |
| "grad_norm": 2.9260942935943604, | |
| "learning_rate": 4.9345884592756144e-05, | |
| "loss": 0.9751, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.04921824039848985, | |
| "grad_norm": 0.17592434585094452, | |
| "learning_rate": 4.934257931025365e-05, | |
| "loss": 0.9219, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.04941590401454804, | |
| "grad_norm": 0.07734190672636032, | |
| "learning_rate": 4.933927402775116e-05, | |
| "loss": 0.2931, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.04941590401454804, | |
| "eval_loss": 0.8793515563011169, | |
| "eval_runtime": 237.1956, | |
| "eval_samples_per_second": 89.816, | |
| "eval_steps_per_second": 11.227, | |
| "step": 2500 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 151773, | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.045168128e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |