| [ |
| { |
| "loss": 0.34830059051513673, |
| "grad_norm": 0.2803570032119751, |
| "learning_rate": 0.00013714285714285716, |
| "num_tokens": 640109.0, |
| "mean_token_accuracy": 0.9155976337194442, |
| "epoch": 0.10746910263299302, |
| "step": 25 |
| }, |
| { |
| "loss": 0.14681048393249513, |
| "grad_norm": 0.23165035247802734, |
| "learning_rate": 0.000199780703920947, |
| "num_tokens": 1262877.0, |
| "mean_token_accuracy": 0.9537364545464516, |
| "epoch": 0.21493820526598603, |
| "step": 50 |
| }, |
| { |
| "loss": 0.1139146327972412, |
| "grad_norm": 0.16550014913082123, |
| "learning_rate": 0.00019830242014201796, |
| "num_tokens": 1905162.0, |
| "mean_token_accuracy": 0.9621101367473602, |
| "epoch": 0.32240730789897903, |
| "step": 75 |
| }, |
| { |
| "loss": 0.11038614273071289, |
| "grad_norm": 0.18707048892974854, |
| "learning_rate": 0.0001954504062771555, |
| "num_tokens": 2537797.0, |
| "mean_token_accuracy": 0.9628356519341469, |
| "epoch": 0.42987641053197206, |
| "step": 100 |
| }, |
| { |
| "eval_loss": 0.1351652890443802, |
| "eval_runtime": 69.5984, |
| "eval_samples_per_second": 2.845, |
| "eval_steps_per_second": 1.422, |
| "eval_num_tokens": 2954175.0, |
| "eval_mean_token_accuracy": 0.9607259185627254, |
| "epoch": 0.4986566362170876, |
| "step": 116 |
| }, |
| { |
| "loss": 0.11258039474487305, |
| "grad_norm": 0.20463427901268005, |
| "learning_rate": 0.00019126451787870527, |
| "num_tokens": 3175008.0, |
| "mean_token_accuracy": 0.9623757800459862, |
| "epoch": 0.537345513164965, |
| "step": 125 |
| }, |
| { |
| "loss": 0.11596426963806153, |
| "grad_norm": 0.1986123025417328, |
| "learning_rate": 0.00018580325076824513, |
| "num_tokens": 3799800.0, |
| "mean_token_accuracy": 0.961990795135498, |
| "epoch": 0.6448146157979581, |
| "step": 150 |
| }, |
| { |
| "loss": 0.0997089672088623, |
| "grad_norm": 0.1511112004518509, |
| "learning_rate": 0.0001791429235849919, |
| "num_tokens": 4438381.0, |
| "mean_token_accuracy": 0.9641576319932937, |
| "epoch": 0.7522837184309511, |
| "step": 175 |
| }, |
| { |
| "loss": 0.09359555244445801, |
| "grad_norm": 0.13941654562950134, |
| "learning_rate": 0.0001713766112687139, |
| "num_tokens": 5075973.0, |
| "mean_token_accuracy": 0.9686441496014595, |
| "epoch": 0.8597528210639441, |
| "step": 200 |
| }, |
| { |
| "loss": 0.11163744926452637, |
| "grad_norm": 0.1599922776222229, |
| "learning_rate": 0.0001626128443812245, |
| "num_tokens": 5723214.0, |
| "mean_token_accuracy": 0.9624734339118004, |
| "epoch": 0.9672219236969372, |
| "step": 225 |
| }, |
| { |
| "eval_loss": 0.12350355833768845, |
| "eval_runtime": 69.5143, |
| "eval_samples_per_second": 2.848, |
| "eval_steps_per_second": 1.424, |
| "eval_num_tokens": 5899964.0, |
| "eval_mean_token_accuracy": 0.9637513997578862, |
| "epoch": 0.9973132724341752, |
| "step": 232 |
| }, |
| { |
| "loss": 0.07306031227111816, |
| "grad_norm": 0.1523396372795105, |
| "learning_rate": 0.00015297409244282694, |
| "num_tokens": 6364544.0, |
| "mean_token_accuracy": 0.9745264253035415, |
| "epoch": 1.0730789897904351, |
| "step": 250 |
| }, |
| { |
| "loss": 0.06787878513336182, |
| "grad_norm": 0.13754014670848846, |
| "learning_rate": 0.00014259505247837074, |
| "num_tokens": 6996524.0, |
| "mean_token_accuracy": 0.9755120638012886, |
| "epoch": 1.1805480924234282, |
| "step": 275 |
| }, |
| { |
| "loss": 0.08596912384033203, |
| "grad_norm": 0.17458459734916687, |
| "learning_rate": 0.0001316207666896824, |
| "num_tokens": 7621299.0, |
| "mean_token_accuracy": 0.9696658563613891, |
| "epoch": 1.2880171950564212, |
| "step": 300 |
| }, |
| { |
| "loss": 0.06656608581542969, |
| "grad_norm": 0.1313866376876831, |
| "learning_rate": 0.00012020459555901427, |
| "num_tokens": 8267290.0, |
| "mean_token_accuracy": 0.9760522067546844, |
| "epoch": 1.3954862976894142, |
| "step": 325 |
| }, |
| { |
| "eval_loss": 0.1249435767531395, |
| "eval_runtime": 69.7547, |
| "eval_samples_per_second": 2.839, |
| "eval_steps_per_second": 1.419, |
| "eval_num_tokens": 8834417.0, |
| "eval_mean_token_accuracy": 0.963239210422593, |
| "epoch": 1.4943578721117678, |
| "step": 348 |
| }, |
| { |
| "loss": 0.07149289608001709, |
| "grad_norm": 0.15927733480930328, |
| "learning_rate": 0.00010850607470843656, |
| "num_tokens": 8883001.0, |
| "mean_token_accuracy": 0.9746167114377022, |
| "epoch": 1.5029554003224073, |
| "step": 350 |
| }, |
| { |
| "loss": 0.0666344976425171, |
| "grad_norm": 0.14084048569202423, |
| "learning_rate": 9.668868546455486e-05, |
| "num_tokens": 9525128.0, |
| "mean_token_accuracy": 0.975580106973648, |
| "epoch": 1.6104245029554003, |
| "step": 375 |
| }, |
| { |
| "loss": 0.07050958156585693, |
| "grad_norm": 0.1653551161289215, |
| "learning_rate": 8.491757028386263e-05, |
| "num_tokens": 10162043.0, |
| "mean_token_accuracy": 0.9742929524183274, |
| "epoch": 1.7178936055883933, |
| "step": 400 |
| }, |
| { |
| "loss": 0.054537668228149414, |
| "grad_norm": 0.17465578019618988, |
| "learning_rate": 7.33572249645848e-05, |
| "num_tokens": 10802637.0, |
| "mean_token_accuracy": 0.9797186449170112, |
| "epoch": 1.8253627082213864, |
| "step": 425 |
| }, |
| { |
| "loss": 0.06608867645263672, |
| "grad_norm": 0.15113526582717896, |
| "learning_rate": 6.216919989526651e-05, |
| "num_tokens": 11430608.0, |
| "mean_token_accuracy": 0.9761568233370781, |
| "epoch": 1.9328318108543794, |
| "step": 450 |
| }, |
| { |
| "eval_loss": 0.11843688040971756, |
| "eval_runtime": 69.9946, |
| "eval_samples_per_second": 2.829, |
| "eval_steps_per_second": 1.414, |
| "eval_num_tokens": 11786762.0, |
| "eval_mean_token_accuracy": 0.965543883015411, |
| "epoch": 1.9930145083288555, |
| "step": 464 |
| }, |
| { |
| "loss": 0.05509011745452881, |
| "grad_norm": 0.09923101216554642, |
| "learning_rate": 5.1509842464076776e-05, |
| "num_tokens": 12054154.0, |
| "mean_token_accuracy": 0.9804502256630641, |
| "epoch": 2.0386888769478775, |
| "step": 475 |
| }, |
| { |
| "loss": 0.031655769348144534, |
| "grad_norm": 0.13326086103916168, |
| "learning_rate": 4.152811217759529e-05, |
| "num_tokens": 12697456.0, |
| "mean_token_accuracy": 0.9891643562912941, |
| "epoch": 2.1461579795808703, |
| "step": 500 |
| }, |
| { |
| "loss": 0.032743215560913086, |
| "grad_norm": 0.12852540612220764, |
| "learning_rate": 3.2363499021769526e-05, |
| "num_tokens": 13329332.0, |
| "mean_token_accuracy": 0.9882625249028206, |
| "epoch": 2.2536270822138635, |
| "step": 525 |
| }, |
| { |
| "loss": 0.029418470859527587, |
| "grad_norm": 0.12474379688501358, |
| "learning_rate": 2.4144074154968832e-05, |
| "num_tokens": 13972889.0, |
| "mean_token_accuracy": 0.9891040176153183, |
| "epoch": 2.3610961848468563, |
| "step": 550 |
| }, |
| { |
| "loss": 0.030236964225769044, |
| "grad_norm": 0.12814833223819733, |
| "learning_rate": 1.6984700173783175e-05, |
| "num_tokens": 14606119.0, |
| "mean_token_accuracy": 0.9886843663454056, |
| "epoch": 2.4685652874798496, |
| "step": 575 |
| }, |
| { |
| "eval_loss": 0.1397247463464737, |
| "eval_runtime": 69.3145, |
| "eval_samples_per_second": 2.857, |
| "eval_steps_per_second": 1.428, |
| "eval_num_tokens": 14728454.0, |
| "eval_mean_token_accuracy": 0.9646775379325404, |
| "epoch": 2.4900591080064483, |
| "step": 580 |
| }, |
| { |
| "loss": 0.030723834037780763, |
| "grad_norm": 0.12963370978832245, |
| "learning_rate": 1.0985425962260343e-05, |
| "num_tokens": 15237584.0, |
| "mean_token_accuracy": 0.9888986241817475, |
| "epoch": 2.5760343901128424, |
| "step": 600 |
| }, |
| { |
| "loss": 0.030379328727722168, |
| "grad_norm": 0.16792573034763336, |
| "learning_rate": 6.230088555808278e-06, |
| "num_tokens": 15876203.0, |
| "mean_token_accuracy": 0.9890140387415886, |
| "epoch": 2.6835034927458357, |
| "step": 625 |
| }, |
| { |
| "loss": 0.02967998743057251, |
| "grad_norm": 0.19673478603363037, |
| "learning_rate": 2.7851415580571692e-06, |
| "num_tokens": 16513501.0, |
| "mean_token_accuracy": 0.9892704981565476, |
| "epoch": 2.7909725953788285, |
| "step": 650 |
| }, |
| { |
| "loss": 0.030462250709533692, |
| "grad_norm": 0.11195345222949982, |
| "learning_rate": 6.987264830045526e-07, |
| "num_tokens": 17145749.0, |
| "mean_token_accuracy": 0.9888393118977546, |
| "epoch": 2.8984416980118217, |
| "step": 675 |
| }, |
| { |
| "eval_loss": 0.13951744139194489, |
| "eval_runtime": 69.3391, |
| "eval_samples_per_second": 2.856, |
| "eval_steps_per_second": 1.428, |
| "eval_num_tokens": 17672878.0, |
| "eval_mean_token_accuracy": 0.9649288859030213, |
| "epoch": 2.9887157442235357, |
| "step": 696 |
| }, |
| { |
| "eval_loss": 0.13946650922298431, |
| "eval_runtime": 69.3693, |
| "eval_samples_per_second": 2.854, |
| "eval_steps_per_second": 1.427, |
| "eval_num_tokens": 17736411.0, |
| "eval_mean_token_accuracy": 0.9649666162452313, |
| "epoch": 3.0, |
| "step": 699 |
| }, |
| { |
| "train_runtime": 14009.2214, |
| "train_samples_per_second": 0.797, |
| "train_steps_per_second": 0.05, |
| "total_flos": 8.799123393273508e+17, |
| "train_loss": 0.07888307489550676, |
| "epoch": 3.0, |
| "step": 699 |
| } |
| ] |
