| { | |
| "best_metric": 0.30429786443710327, | |
| "best_model_checkpoint": "/mnt/storage1/ziya/VQA/M3D/LaMed/output/Llama3.2_3B_Slake_BiomedCLIP_KG2/checkpoint-49175", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 49175, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.030503304524656837, | |
| "grad_norm": 219.84915161132812, | |
| "learning_rate": 5.083884087442806e-07, | |
| "loss": 6.7276, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.061006609049313675, | |
| "grad_norm": 198.07276916503906, | |
| "learning_rate": 1.0167768174885612e-06, | |
| "loss": 4.6638, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.09150991357397051, | |
| "grad_norm": 206.43557739257812, | |
| "learning_rate": 1.525165226232842e-06, | |
| "loss": 2.7355, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.12201321809862735, | |
| "grad_norm": 197.00241088867188, | |
| "learning_rate": 2.0335536349771224e-06, | |
| "loss": 1.7593, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.1525165226232842, | |
| "grad_norm": 130.2694091796875, | |
| "learning_rate": 2.541942043721403e-06, | |
| "loss": 1.3763, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.18301982714794102, | |
| "grad_norm": 125.28648376464844, | |
| "learning_rate": 3.050330452465684e-06, | |
| "loss": 1.2219, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.21352313167259787, | |
| "grad_norm": 34.316184997558594, | |
| "learning_rate": 3.558718861209965e-06, | |
| "loss": 1.0971, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.2440264361972547, | |
| "grad_norm": 98.0787582397461, | |
| "learning_rate": 4.067107269954245e-06, | |
| "loss": 1.1094, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.27452974072191155, | |
| "grad_norm": 84.89276123046875, | |
| "learning_rate": 4.575495678698526e-06, | |
| "loss": 0.9373, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.3050330452465684, | |
| "grad_norm": 59.95817565917969, | |
| "learning_rate": 5.083884087442806e-06, | |
| "loss": 1.0445, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.3355363497712252, | |
| "grad_norm": 6.108051776885986, | |
| "learning_rate": 5.5922724961870875e-06, | |
| "loss": 0.9383, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.36603965429588203, | |
| "grad_norm": 0.15768121182918549, | |
| "learning_rate": 6.100660904931368e-06, | |
| "loss": 0.8041, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.3965429588205389, | |
| "grad_norm": 61.489315032958984, | |
| "learning_rate": 6.609049313675649e-06, | |
| "loss": 0.8024, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.42704626334519574, | |
| "grad_norm": 59.69186782836914, | |
| "learning_rate": 7.11743772241993e-06, | |
| "loss": 0.8968, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.45754956786985257, | |
| "grad_norm": 33.70270538330078, | |
| "learning_rate": 7.62582613116421e-06, | |
| "loss": 0.8929, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.4880528723945094, | |
| "grad_norm": 0.0485120490193367, | |
| "learning_rate": 8.13421453990849e-06, | |
| "loss": 0.854, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.5185561769191662, | |
| "grad_norm": 22.23134994506836, | |
| "learning_rate": 8.642602948652772e-06, | |
| "loss": 0.7011, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.5490594814438231, | |
| "grad_norm": 31.907052993774414, | |
| "learning_rate": 9.150991357397053e-06, | |
| "loss": 0.7926, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 0.5795627859684799, | |
| "grad_norm": 0.41279709339141846, | |
| "learning_rate": 9.659379766141333e-06, | |
| "loss": 0.6784, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 0.6100660904931368, | |
| "grad_norm": 5.743080139160156, | |
| "learning_rate": 9.9999933570978e-06, | |
| "loss": 0.6052, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.6405693950177936, | |
| "grad_norm": 0.039274271577596664, | |
| "learning_rate": 9.999892097428444e-06, | |
| "loss": 0.6623, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 0.6710726995424504, | |
| "grad_norm": 7.821774959564209, | |
| "learning_rate": 9.999668840572521e-06, | |
| "loss": 0.7683, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 0.7015760040671073, | |
| "grad_norm": 45.965476989746094, | |
| "learning_rate": 9.99932359197749e-06, | |
| "loss": 0.6642, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 0.7320793085917641, | |
| "grad_norm": 11.506255149841309, | |
| "learning_rate": 9.998856360067404e-06, | |
| "loss": 0.5841, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 0.762582613116421, | |
| "grad_norm": 4.3167243003845215, | |
| "learning_rate": 9.998267156242705e-06, | |
| "loss": 0.5843, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.7930859176410778, | |
| "grad_norm": 0.1081758663058281, | |
| "learning_rate": 9.997555994879945e-06, | |
| "loss": 0.6163, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 0.8235892221657346, | |
| "grad_norm": 8.694897651672363, | |
| "learning_rate": 9.996722893331437e-06, | |
| "loss": 0.5964, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 0.8540925266903915, | |
| "grad_norm": 1.5994528532028198, | |
| "learning_rate": 9.995767871924826e-06, | |
| "loss": 0.5306, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 0.8845958312150483, | |
| "grad_norm": 32.59883499145508, | |
| "learning_rate": 9.994690953962602e-06, | |
| "loss": 0.5799, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 0.9150991357397051, | |
| "grad_norm": 18.219099044799805, | |
| "learning_rate": 9.993492165721523e-06, | |
| "loss": 0.537, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.945602440264362, | |
| "grad_norm": 38.14984893798828, | |
| "learning_rate": 9.99217153645198e-06, | |
| "loss": 0.5993, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 0.9761057447890188, | |
| "grad_norm": 26.25863265991211, | |
| "learning_rate": 9.990729098377278e-06, | |
| "loss": 0.5018, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.6256594061851501, | |
| "eval_runtime": 103.5099, | |
| "eval_samples_per_second": 20.278, | |
| "eval_steps_per_second": 1.353, | |
| "step": 9835 | |
| }, | |
| { | |
| "epoch": 1.0066090493136757, | |
| "grad_norm": 0.06917725503444672, | |
| "learning_rate": 9.989164886692855e-06, | |
| "loss": 0.5179, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 1.0371123538383324, | |
| "grad_norm": 0.5146661996841431, | |
| "learning_rate": 9.987478939565419e-06, | |
| "loss": 0.5132, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 1.0676156583629894, | |
| "grad_norm": 0.00010405463399365544, | |
| "learning_rate": 9.985671298132018e-06, | |
| "loss": 0.553, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.0981189628876462, | |
| "grad_norm": 3.304703712463379, | |
| "learning_rate": 9.98374200649904e-06, | |
| "loss": 0.4167, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 1.128622267412303, | |
| "grad_norm": 0.6084465384483337, | |
| "learning_rate": 9.981691111741132e-06, | |
| "loss": 0.4076, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 1.1591255719369598, | |
| "grad_norm": 0.05212326720356941, | |
| "learning_rate": 9.979518663900047e-06, | |
| "loss": 0.5073, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 1.1896288764616167, | |
| "grad_norm": 21.804567337036133, | |
| "learning_rate": 9.977224715983439e-06, | |
| "loss": 0.4453, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 1.2201321809862735, | |
| "grad_norm": 32.33522415161133, | |
| "learning_rate": 9.974809323963552e-06, | |
| "loss": 0.5258, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.2506354855109303, | |
| "grad_norm": 0.14623811841011047, | |
| "learning_rate": 9.972272546775862e-06, | |
| "loss": 0.446, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 1.281138790035587, | |
| "grad_norm": 0.28084316849708557, | |
| "learning_rate": 9.969614446317645e-06, | |
| "loss": 0.4267, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 1.311642094560244, | |
| "grad_norm": 2.6160593032836914, | |
| "learning_rate": 9.966835087446457e-06, | |
| "loss": 0.5175, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 1.3421453990849008, | |
| "grad_norm": 0.21234674751758575, | |
| "learning_rate": 9.963934537978547e-06, | |
| "loss": 0.4142, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 1.3726487036095576, | |
| "grad_norm": 0.09984096884727478, | |
| "learning_rate": 9.960912868687222e-06, | |
| "loss": 0.3752, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.4031520081342146, | |
| "grad_norm": 0.003108405042439699, | |
| "learning_rate": 9.957770153301104e-06, | |
| "loss": 0.4626, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 1.4336553126588714, | |
| "grad_norm": 7.253620424307883e-05, | |
| "learning_rate": 9.954506468502335e-06, | |
| "loss": 0.3995, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 1.4641586171835281, | |
| "grad_norm": 7.312223434448242, | |
| "learning_rate": 9.951121893924704e-06, | |
| "loss": 0.5808, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 1.4946619217081851, | |
| "grad_norm": 10.09123420715332, | |
| "learning_rate": 9.94761651215171e-06, | |
| "loss": 0.4186, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 1.525165226232842, | |
| "grad_norm": 6.39027214050293, | |
| "learning_rate": 9.943990408714542e-06, | |
| "loss": 0.4374, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.5556685307574987, | |
| "grad_norm": 0.5071086287498474, | |
| "learning_rate": 9.94024367208999e-06, | |
| "loss": 0.3927, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 1.5861718352821557, | |
| "grad_norm": 0.00023602554574608803, | |
| "learning_rate": 9.936376393698296e-06, | |
| "loss": 0.453, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 1.6166751398068124, | |
| "grad_norm": 4.159149646759033, | |
| "learning_rate": 9.93238866790091e-06, | |
| "loss": 0.4594, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 1.6471784443314692, | |
| "grad_norm": 0.02954951673746109, | |
| "learning_rate": 9.928280591998201e-06, | |
| "loss": 0.4708, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 1.6776817488561262, | |
| "grad_norm": 0.036519214510917664, | |
| "learning_rate": 9.924052266227069e-06, | |
| "loss": 0.348, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 1.708185053380783, | |
| "grad_norm": 14.24988079071045, | |
| "learning_rate": 9.919703793758511e-06, | |
| "loss": 0.4279, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 1.7386883579054397, | |
| "grad_norm": 17.464754104614258, | |
| "learning_rate": 9.9152352806951e-06, | |
| "loss": 0.4133, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 1.7691916624300967, | |
| "grad_norm": 0.008350013755261898, | |
| "learning_rate": 9.91064683606839e-06, | |
| "loss": 0.3868, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 1.7996949669547533, | |
| "grad_norm": 0.14084838330745697, | |
| "learning_rate": 9.90593857183627e-06, | |
| "loss": 0.41, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 1.8301982714794103, | |
| "grad_norm": 0.011409441009163857, | |
| "learning_rate": 9.901110602880211e-06, | |
| "loss": 0.3632, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 1.8607015760040673, | |
| "grad_norm": 0.2167450338602066, | |
| "learning_rate": 9.896163047002486e-06, | |
| "loss": 0.4203, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 1.8912048805287238, | |
| "grad_norm": 0.12340361624956131, | |
| "learning_rate": 9.891096024923274e-06, | |
| "loss": 0.3836, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 1.9217081850533808, | |
| "grad_norm": 0.011780009604990482, | |
| "learning_rate": 9.885909660277735e-06, | |
| "loss": 0.3638, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 1.9522114895780376, | |
| "grad_norm": 1.290787696838379, | |
| "learning_rate": 9.880604079612977e-06, | |
| "loss": 0.3716, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 1.9827147941026944, | |
| "grad_norm": 1.6361336747650057e-05, | |
| "learning_rate": 9.87517941238498e-06, | |
| "loss": 0.3499, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.41577816009521484, | |
| "eval_runtime": 103.2846, | |
| "eval_samples_per_second": 20.322, | |
| "eval_steps_per_second": 1.355, | |
| "step": 19670 | |
| }, | |
| { | |
| "epoch": 2.0132180986273513, | |
| "grad_norm": 69.40552520751953, | |
| "learning_rate": 9.869635790955423e-06, | |
| "loss": 0.3971, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 2.0437214031520083, | |
| "grad_norm": 0.0739259347319603, | |
| "learning_rate": 9.863973350588472e-06, | |
| "loss": 0.3106, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 2.074224707676665, | |
| "grad_norm": 6.068602085113525, | |
| "learning_rate": 9.85819222944747e-06, | |
| "loss": 0.3609, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 2.104728012201322, | |
| "grad_norm": 15.744518280029297, | |
| "learning_rate": 9.852292568591557e-06, | |
| "loss": 0.355, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 2.135231316725979, | |
| "grad_norm": 12.378608703613281, | |
| "learning_rate": 9.846274511972251e-06, | |
| "loss": 0.306, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 2.1657346212506354, | |
| "grad_norm": 10.521739959716797, | |
| "learning_rate": 9.840138206429911e-06, | |
| "loss": 0.3215, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 2.1962379257752924, | |
| "grad_norm": 8.210835456848145, | |
| "learning_rate": 9.833883801690179e-06, | |
| "loss": 0.2643, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 2.226741230299949, | |
| "grad_norm": 0.28642842173576355, | |
| "learning_rate": 9.827511450360295e-06, | |
| "loss": 0.3466, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 2.257244534824606, | |
| "grad_norm": 0.010792361572384834, | |
| "learning_rate": 9.821021307925406e-06, | |
| "loss": 0.3405, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 2.287747839349263, | |
| "grad_norm": 0.46060484647750854, | |
| "learning_rate": 9.814413532744753e-06, | |
| "loss": 0.3429, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.3182511438739195, | |
| "grad_norm": 0.04725227132439613, | |
| "learning_rate": 9.80768828604781e-06, | |
| "loss": 0.3277, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 2.3487544483985765, | |
| "grad_norm": 4.012476921081543, | |
| "learning_rate": 9.800845731930356e-06, | |
| "loss": 0.3004, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 2.3792577529232335, | |
| "grad_norm": 12.063570976257324, | |
| "learning_rate": 9.793886037350461e-06, | |
| "loss": 0.351, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 2.40976105744789, | |
| "grad_norm": 0.0161123126745224, | |
| "learning_rate": 9.786809372124425e-06, | |
| "loss": 0.3215, | |
| "step": 23700 | |
| }, | |
| { | |
| "epoch": 2.440264361972547, | |
| "grad_norm": 0.012391666881740093, | |
| "learning_rate": 9.779615908922622e-06, | |
| "loss": 0.3581, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 2.470767666497204, | |
| "grad_norm": 3.5457940101623535, | |
| "learning_rate": 9.772305823265294e-06, | |
| "loss": 0.339, | |
| "step": 24300 | |
| }, | |
| { | |
| "epoch": 2.5012709710218606, | |
| "grad_norm": 0.04232852905988693, | |
| "learning_rate": 9.764879293518266e-06, | |
| "loss": 0.3523, | |
| "step": 24600 | |
| }, | |
| { | |
| "epoch": 2.5317742755465176, | |
| "grad_norm": 0.2725585997104645, | |
| "learning_rate": 9.757336500888599e-06, | |
| "loss": 0.2807, | |
| "step": 24900 | |
| }, | |
| { | |
| "epoch": 2.562277580071174, | |
| "grad_norm": 15.362088203430176, | |
| "learning_rate": 9.749677629420157e-06, | |
| "loss": 0.3227, | |
| "step": 25200 | |
| }, | |
| { | |
| "epoch": 2.592780884595831, | |
| "grad_norm": 0.04092194885015488, | |
| "learning_rate": 9.74190286598913e-06, | |
| "loss": 0.2848, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 2.623284189120488, | |
| "grad_norm": 18.27284049987793, | |
| "learning_rate": 9.734012400299463e-06, | |
| "loss": 0.3182, | |
| "step": 25800 | |
| }, | |
| { | |
| "epoch": 2.6537874936451447, | |
| "grad_norm": 10.792008399963379, | |
| "learning_rate": 9.726006424878234e-06, | |
| "loss": 0.3344, | |
| "step": 26100 | |
| }, | |
| { | |
| "epoch": 2.6842907981698017, | |
| "grad_norm": 0.0012637190520763397, | |
| "learning_rate": 9.717885135070957e-06, | |
| "loss": 0.3022, | |
| "step": 26400 | |
| }, | |
| { | |
| "epoch": 2.7147941026944586, | |
| "grad_norm": 0.0459415428340435, | |
| "learning_rate": 9.709648729036805e-06, | |
| "loss": 0.3292, | |
| "step": 26700 | |
| }, | |
| { | |
| "epoch": 2.745297407219115, | |
| "grad_norm": 0.0021137718576937914, | |
| "learning_rate": 9.70129740774379e-06, | |
| "loss": 0.34, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 2.775800711743772, | |
| "grad_norm": 0.020368747413158417, | |
| "learning_rate": 9.69283137496385e-06, | |
| "loss": 0.2781, | |
| "step": 27300 | |
| }, | |
| { | |
| "epoch": 2.806304016268429, | |
| "grad_norm": 0.01668807491660118, | |
| "learning_rate": 9.68425083726788e-06, | |
| "loss": 0.3146, | |
| "step": 27600 | |
| }, | |
| { | |
| "epoch": 2.8368073207930857, | |
| "grad_norm": 18.376174926757812, | |
| "learning_rate": 9.67555600402069e-06, | |
| "loss": 0.3176, | |
| "step": 27900 | |
| }, | |
| { | |
| "epoch": 2.8673106253177427, | |
| "grad_norm": 0.012932281009852886, | |
| "learning_rate": 9.666747087375894e-06, | |
| "loss": 0.3183, | |
| "step": 28200 | |
| }, | |
| { | |
| "epoch": 2.8978139298423997, | |
| "grad_norm": 24.642900466918945, | |
| "learning_rate": 9.657824302270743e-06, | |
| "loss": 0.3418, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 2.9283172343670563, | |
| "grad_norm": 0.013992193154990673, | |
| "learning_rate": 9.64878786642087e-06, | |
| "loss": 0.3196, | |
| "step": 28800 | |
| }, | |
| { | |
| "epoch": 2.9588205388917133, | |
| "grad_norm": 0.001322018215432763, | |
| "learning_rate": 9.639638000314983e-06, | |
| "loss": 0.2828, | |
| "step": 29100 | |
| }, | |
| { | |
| "epoch": 2.9893238434163703, | |
| "grad_norm": 0.010299603454768658, | |
| "learning_rate": 9.630374927209485e-06, | |
| "loss": 0.3031, | |
| "step": 29400 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.384840726852417, | |
| "eval_runtime": 103.5367, | |
| "eval_samples_per_second": 20.273, | |
| "eval_steps_per_second": 1.352, | |
| "step": 29505 | |
| }, | |
| { | |
| "epoch": 3.019827147941027, | |
| "grad_norm": 0.03328376263380051, | |
| "learning_rate": 9.620998873123027e-06, | |
| "loss": 0.2759, | |
| "step": 29700 | |
| }, | |
| { | |
| "epoch": 3.050330452465684, | |
| "grad_norm": 13.360301971435547, | |
| "learning_rate": 9.611510066830984e-06, | |
| "loss": 0.1929, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 3.080833756990341, | |
| "grad_norm": 0.0005070280749350786, | |
| "learning_rate": 9.601908739859895e-06, | |
| "loss": 0.2731, | |
| "step": 30300 | |
| }, | |
| { | |
| "epoch": 3.1113370615149973, | |
| "grad_norm": 7.794924385962076e-06, | |
| "learning_rate": 9.592195126481786e-06, | |
| "loss": 0.2898, | |
| "step": 30600 | |
| }, | |
| { | |
| "epoch": 3.1418403660396543, | |
| "grad_norm": 0.000699843221809715, | |
| "learning_rate": 9.582369463708473e-06, | |
| "loss": 0.2485, | |
| "step": 30900 | |
| }, | |
| { | |
| "epoch": 3.1723436705643113, | |
| "grad_norm": 0.24882352352142334, | |
| "learning_rate": 9.572431991285775e-06, | |
| "loss": 0.2353, | |
| "step": 31200 | |
| }, | |
| { | |
| "epoch": 3.202846975088968, | |
| "grad_norm": 0.004962280858308077, | |
| "learning_rate": 9.562382951687658e-06, | |
| "loss": 0.2352, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 3.233350279613625, | |
| "grad_norm": 4.832521915435791, | |
| "learning_rate": 9.552222590110324e-06, | |
| "loss": 0.2363, | |
| "step": 31800 | |
| }, | |
| { | |
| "epoch": 3.263853584138282, | |
| "grad_norm": 0.019647739827632904, | |
| "learning_rate": 9.541951154466233e-06, | |
| "loss": 0.2931, | |
| "step": 32100 | |
| }, | |
| { | |
| "epoch": 3.2943568886629384, | |
| "grad_norm": 0.04419803246855736, | |
| "learning_rate": 9.53156889537804e-06, | |
| "loss": 0.2634, | |
| "step": 32400 | |
| }, | |
| { | |
| "epoch": 3.3248601931875954, | |
| "grad_norm": 0.00563009362667799, | |
| "learning_rate": 9.521076066172493e-06, | |
| "loss": 0.3196, | |
| "step": 32700 | |
| }, | |
| { | |
| "epoch": 3.3553634977122524, | |
| "grad_norm": 35.982086181640625, | |
| "learning_rate": 9.510472922874246e-06, | |
| "loss": 0.2312, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 3.385866802236909, | |
| "grad_norm": 0.023021390661597252, | |
| "learning_rate": 9.49975972419961e-06, | |
| "loss": 0.248, | |
| "step": 33300 | |
| }, | |
| { | |
| "epoch": 3.416370106761566, | |
| "grad_norm": 12.021758079528809, | |
| "learning_rate": 9.488936731550247e-06, | |
| "loss": 0.2552, | |
| "step": 33600 | |
| }, | |
| { | |
| "epoch": 3.4468734112862225, | |
| "grad_norm": 0.08628563582897186, | |
| "learning_rate": 9.47800420900679e-06, | |
| "loss": 0.2553, | |
| "step": 33900 | |
| }, | |
| { | |
| "epoch": 3.4773767158108795, | |
| "grad_norm": 1.5440913438796997, | |
| "learning_rate": 9.466962423322388e-06, | |
| "loss": 0.3116, | |
| "step": 34200 | |
| }, | |
| { | |
| "epoch": 3.5078800203355365, | |
| "grad_norm": 1.9582719687605277e-05, | |
| "learning_rate": 9.455811643916217e-06, | |
| "loss": 0.2573, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 3.538383324860193, | |
| "grad_norm": 0.0327552855014801, | |
| "learning_rate": 9.444552142866892e-06, | |
| "loss": 0.2714, | |
| "step": 34800 | |
| }, | |
| { | |
| "epoch": 3.56888662938485, | |
| "grad_norm": 23.4067440032959, | |
| "learning_rate": 9.433184194905831e-06, | |
| "loss": 0.2582, | |
| "step": 35100 | |
| }, | |
| { | |
| "epoch": 3.599389933909507, | |
| "grad_norm": 0.510364830493927, | |
| "learning_rate": 9.421708077410551e-06, | |
| "loss": 0.2953, | |
| "step": 35400 | |
| }, | |
| { | |
| "epoch": 3.6298932384341636, | |
| "grad_norm": 0.008372425101697445, | |
| "learning_rate": 9.410124070397908e-06, | |
| "loss": 0.3149, | |
| "step": 35700 | |
| }, | |
| { | |
| "epoch": 3.6603965429588206, | |
| "grad_norm": 0.005560519173741341, | |
| "learning_rate": 9.398432456517254e-06, | |
| "loss": 0.2407, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 3.690899847483477, | |
| "grad_norm": 4.856593608856201, | |
| "learning_rate": 9.386633521043545e-06, | |
| "loss": 0.2772, | |
| "step": 36300 | |
| }, | |
| { | |
| "epoch": 3.721403152008134, | |
| "grad_norm": 0.00459680799394846, | |
| "learning_rate": 9.374727551870377e-06, | |
| "loss": 0.2901, | |
| "step": 36600 | |
| }, | |
| { | |
| "epoch": 3.751906456532791, | |
| "grad_norm": 0.00823771022260189, | |
| "learning_rate": 9.362714839502973e-06, | |
| "loss": 0.3216, | |
| "step": 36900 | |
| }, | |
| { | |
| "epoch": 3.7824097610574476, | |
| "grad_norm": 0.011493176221847534, | |
| "learning_rate": 9.35059567705108e-06, | |
| "loss": 0.2682, | |
| "step": 37200 | |
| }, | |
| { | |
| "epoch": 3.8129130655821046, | |
| "grad_norm": 0.057664211839437485, | |
| "learning_rate": 9.33837036022182e-06, | |
| "loss": 0.2823, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 3.8434163701067616, | |
| "grad_norm": 11.62787914276123, | |
| "learning_rate": 9.326039187312485e-06, | |
| "loss": 0.1798, | |
| "step": 37800 | |
| }, | |
| { | |
| "epoch": 3.873919674631418, | |
| "grad_norm": 0.02458140067756176, | |
| "learning_rate": 9.313602459203248e-06, | |
| "loss": 0.2337, | |
| "step": 38100 | |
| }, | |
| { | |
| "epoch": 3.904422979156075, | |
| "grad_norm": 0.019079158082604408, | |
| "learning_rate": 9.301060479349826e-06, | |
| "loss": 0.2539, | |
| "step": 38400 | |
| }, | |
| { | |
| "epoch": 3.934926283680732, | |
| "grad_norm": 9.756794929504395, | |
| "learning_rate": 9.288413553776076e-06, | |
| "loss": 0.279, | |
| "step": 38700 | |
| }, | |
| { | |
| "epoch": 3.9654295882053887, | |
| "grad_norm": 3.1425323486328125, | |
| "learning_rate": 9.275661991066522e-06, | |
| "loss": 0.2767, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 3.9959328927300457, | |
| "grad_norm": 0.006602356676012278, | |
| "learning_rate": 9.262806102358834e-06, | |
| "loss": 0.2558, | |
| "step": 39300 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.3359658420085907, | |
| "eval_runtime": 110.2827, | |
| "eval_samples_per_second": 19.033, | |
| "eval_steps_per_second": 1.269, | |
| "step": 39340 | |
| }, | |
| { | |
| "epoch": 4.026436197254703, | |
| "grad_norm": 0.00417200056836009, | |
| "learning_rate": 9.249846201336235e-06, | |
| "loss": 0.2559, | |
| "step": 39600 | |
| }, | |
| { | |
| "epoch": 4.056939501779359, | |
| "grad_norm": 0.008380061946809292, | |
| "learning_rate": 9.236782604219838e-06, | |
| "loss": 0.2173, | |
| "step": 39900 | |
| }, | |
| { | |
| "epoch": 4.087442806304017, | |
| "grad_norm": 0.06265965104103088, | |
| "learning_rate": 9.223615629760943e-06, | |
| "loss": 0.2338, | |
| "step": 40200 | |
| }, | |
| { | |
| "epoch": 4.117946110828673, | |
| "grad_norm": 2.496182241884526e-05, | |
| "learning_rate": 9.21034559923325e-06, | |
| "loss": 0.2761, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 4.14844941535333, | |
| "grad_norm": 40.386531829833984, | |
| "learning_rate": 9.196972836425027e-06, | |
| "loss": 0.1951, | |
| "step": 40800 | |
| }, | |
| { | |
| "epoch": 4.178952719877987, | |
| "grad_norm": 0.024761863052845, | |
| "learning_rate": 9.183497667631199e-06, | |
| "loss": 0.1979, | |
| "step": 41100 | |
| }, | |
| { | |
| "epoch": 4.209456024402644, | |
| "grad_norm": 0.0945596694946289, | |
| "learning_rate": 9.169920421645401e-06, | |
| "loss": 0.1914, | |
| "step": 41400 | |
| }, | |
| { | |
| "epoch": 4.2399593289273, | |
| "grad_norm": 0.6701067686080933, | |
| "learning_rate": 9.156241429751947e-06, | |
| "loss": 0.2216, | |
| "step": 41700 | |
| }, | |
| { | |
| "epoch": 4.270462633451958, | |
| "grad_norm": 240.65187072753906, | |
| "learning_rate": 9.142461025717739e-06, | |
| "loss": 0.2277, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 4.300965937976614, | |
| "grad_norm": 0.0036695213057100773, | |
| "learning_rate": 9.128579545784142e-06, | |
| "loss": 0.2476, | |
| "step": 42300 | |
| }, | |
| { | |
| "epoch": 4.331469242501271, | |
| "grad_norm": 33.61179733276367, | |
| "learning_rate": 9.114597328658763e-06, | |
| "loss": 0.1958, | |
| "step": 42600 | |
| }, | |
| { | |
| "epoch": 4.361972547025927, | |
| "grad_norm": 52.01089859008789, | |
| "learning_rate": 9.100514715507196e-06, | |
| "loss": 0.217, | |
| "step": 42900 | |
| }, | |
| { | |
| "epoch": 4.392475851550585, | |
| "grad_norm": 0.0013590186135843396, | |
| "learning_rate": 9.086332049944692e-06, | |
| "loss": 0.2393, | |
| "step": 43200 | |
| }, | |
| { | |
| "epoch": 4.422979156075241, | |
| "grad_norm": 0.12093915045261383, | |
| "learning_rate": 9.072049678027778e-06, | |
| "loss": 0.2175, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 4.453482460599898, | |
| "grad_norm": 0.008124121464788914, | |
| "learning_rate": 9.057667948245816e-06, | |
| "loss": 0.262, | |
| "step": 43800 | |
| }, | |
| { | |
| "epoch": 4.483985765124555, | |
| "grad_norm": 0.7974226474761963, | |
| "learning_rate": 9.043187211512492e-06, | |
| "loss": 0.2742, | |
| "step": 44100 | |
| }, | |
| { | |
| "epoch": 4.514489069649212, | |
| "grad_norm": 13.317780494689941, | |
| "learning_rate": 9.028607821157256e-06, | |
| "loss": 0.2471, | |
| "step": 44400 | |
| }, | |
| { | |
| "epoch": 4.5449923741738685, | |
| "grad_norm": 0.0005867113941349089, | |
| "learning_rate": 9.013930132916709e-06, | |
| "loss": 0.252, | |
| "step": 44700 | |
| }, | |
| { | |
| "epoch": 4.575495678698526, | |
| "grad_norm": 0.053022801876068115, | |
| "learning_rate": 8.999154504925914e-06, | |
| "loss": 0.2319, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 4.6059989832231825, | |
| "grad_norm": 0.0034278552047908306, | |
| "learning_rate": 8.984281297709658e-06, | |
| "loss": 0.2185, | |
| "step": 45300 | |
| }, | |
| { | |
| "epoch": 4.636502287747839, | |
| "grad_norm": 0.008740383200347424, | |
| "learning_rate": 8.969310874173663e-06, | |
| "loss": 0.2231, | |
| "step": 45600 | |
| }, | |
| { | |
| "epoch": 4.6670055922724965, | |
| "grad_norm": 0.06879168748855591, | |
| "learning_rate": 8.95424359959572e-06, | |
| "loss": 0.1777, | |
| "step": 45900 | |
| }, | |
| { | |
| "epoch": 4.697508896797153, | |
| "grad_norm": 0.005040327087044716, | |
| "learning_rate": 8.939079841616785e-06, | |
| "loss": 0.2596, | |
| "step": 46200 | |
| }, | |
| { | |
| "epoch": 4.7280122013218095, | |
| "grad_norm": 0.0021866620518267155, | |
| "learning_rate": 8.923819970232003e-06, | |
| "loss": 0.2109, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 4.758515505846467, | |
| "grad_norm": 0.023091251030564308, | |
| "learning_rate": 8.90846435778169e-06, | |
| "loss": 0.1908, | |
| "step": 46800 | |
| }, | |
| { | |
| "epoch": 4.7890188103711235, | |
| "grad_norm": 0.0027208509854972363, | |
| "learning_rate": 8.893013378942232e-06, | |
| "loss": 0.2422, | |
| "step": 47100 | |
| }, | |
| { | |
| "epoch": 4.81952211489578, | |
| "grad_norm": 0.0003605252131819725, | |
| "learning_rate": 8.877467410716951e-06, | |
| "loss": 0.1941, | |
| "step": 47400 | |
| }, | |
| { | |
| "epoch": 4.8500254194204375, | |
| "grad_norm": 4.229369640350342, | |
| "learning_rate": 8.861826832426916e-06, | |
| "loss": 0.2363, | |
| "step": 47700 | |
| }, | |
| { | |
| "epoch": 4.880528723945094, | |
| "grad_norm": 0.01985393464565277, | |
| "learning_rate": 8.84609202570167e-06, | |
| "loss": 0.215, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 4.911032028469751, | |
| "grad_norm": 3.8484694957733154, | |
| "learning_rate": 8.830263374469927e-06, | |
| "loss": 0.2126, | |
| "step": 48300 | |
| }, | |
| { | |
| "epoch": 4.941535332994408, | |
| "grad_norm": 4.421438279678114e-05, | |
| "learning_rate": 8.814341264950207e-06, | |
| "loss": 0.2451, | |
| "step": 48600 | |
| }, | |
| { | |
| "epoch": 4.972038637519065, | |
| "grad_norm": 3.2253410816192627, | |
| "learning_rate": 8.798326085641407e-06, | |
| "loss": 0.261, | |
| "step": 48900 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.30429786443710327, | |
| "eval_runtime": 103.3201, | |
| "eval_samples_per_second": 20.316, | |
| "eval_steps_per_second": 1.355, | |
| "step": 49175 | |
| } | |
| ], | |
| "logging_steps": 300, | |
| "max_steps": 196700, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "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": 2.410768182018816e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |