| { | |
| "best_metric": 0.38091349601745605, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-150", | |
| "epoch": 0.12317799219872716, | |
| "eval_steps": 50, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.000821186614658181, | |
| "grad_norm": 2.0305542945861816, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7261, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.000821186614658181, | |
| "eval_loss": 1.465772271156311, | |
| "eval_runtime": 288.2262, | |
| "eval_samples_per_second": 7.116, | |
| "eval_steps_per_second": 1.78, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.001642373229316362, | |
| "grad_norm": 2.114654541015625, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7726, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.002463559843974543, | |
| "grad_norm": 2.2381458282470703, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8371, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.003284746458632724, | |
| "grad_norm": 1.9415267705917358, | |
| "learning_rate": 4e-05, | |
| "loss": 0.7885, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.0041059330732909054, | |
| "grad_norm": 1.484900712966919, | |
| "learning_rate": 5e-05, | |
| "loss": 0.7522, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.004927119687949086, | |
| "grad_norm": 1.259929895401001, | |
| "learning_rate": 6e-05, | |
| "loss": 0.7248, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.005748306302607267, | |
| "grad_norm": 0.9762012958526611, | |
| "learning_rate": 7e-05, | |
| "loss": 0.5922, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.006569492917265448, | |
| "grad_norm": 1.0204635858535767, | |
| "learning_rate": 8e-05, | |
| "loss": 0.5452, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.00739067953192363, | |
| "grad_norm": 0.8520223498344421, | |
| "learning_rate": 9e-05, | |
| "loss": 0.4675, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.008211866146581811, | |
| "grad_norm": 1.470017433166504, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4164, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.009033052761239991, | |
| "grad_norm": 0.8858131766319275, | |
| "learning_rate": 9.99983777858264e-05, | |
| "loss": 0.3847, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.009854239375898173, | |
| "grad_norm": 1.0039604902267456, | |
| "learning_rate": 9.999351124856874e-05, | |
| "loss": 0.3836, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.010675425990556354, | |
| "grad_norm": 0.6317289471626282, | |
| "learning_rate": 9.998540070400966e-05, | |
| "loss": 0.3325, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.011496612605214535, | |
| "grad_norm": 0.680807888507843, | |
| "learning_rate": 9.997404667843075e-05, | |
| "loss": 0.3493, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.012317799219872716, | |
| "grad_norm": 0.6124119758605957, | |
| "learning_rate": 9.995944990857849e-05, | |
| "loss": 0.3204, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.013138985834530896, | |
| "grad_norm": 0.6300580501556396, | |
| "learning_rate": 9.994161134161634e-05, | |
| "loss": 0.3341, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.013960172449189078, | |
| "grad_norm": 0.608344554901123, | |
| "learning_rate": 9.992053213506334e-05, | |
| "loss": 0.3386, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.01478135906384726, | |
| "grad_norm": 0.5785073041915894, | |
| "learning_rate": 9.989621365671902e-05, | |
| "loss": 0.3035, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.01560254567850544, | |
| "grad_norm": 0.6188136339187622, | |
| "learning_rate": 9.986865748457457e-05, | |
| "loss": 0.3162, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.016423732293163622, | |
| "grad_norm": 0.5587109923362732, | |
| "learning_rate": 9.983786540671051e-05, | |
| "loss": 0.273, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.017244918907821802, | |
| "grad_norm": 0.5472216606140137, | |
| "learning_rate": 9.980383942118066e-05, | |
| "loss": 0.3215, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.018066105522479982, | |
| "grad_norm": 0.4810258150100708, | |
| "learning_rate": 9.976658173588244e-05, | |
| "loss": 0.307, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.018887292137138165, | |
| "grad_norm": 0.5424895882606506, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 0.3065, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.019708478751796345, | |
| "grad_norm": 0.5074129700660706, | |
| "learning_rate": 9.968238114591566e-05, | |
| "loss": 0.2774, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.020529665366454525, | |
| "grad_norm": 0.5599948167800903, | |
| "learning_rate": 9.96354437049027e-05, | |
| "loss": 0.2691, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02135085198111271, | |
| "grad_norm": 0.6742061972618103, | |
| "learning_rate": 9.95852854910781e-05, | |
| "loss": 0.2898, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.02217203859577089, | |
| "grad_norm": 0.5526089072227478, | |
| "learning_rate": 9.953190975913647e-05, | |
| "loss": 0.3071, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.02299322521042907, | |
| "grad_norm": 0.5390534996986389, | |
| "learning_rate": 9.947531997255256e-05, | |
| "loss": 0.2906, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.023814411825087253, | |
| "grad_norm": 0.504539430141449, | |
| "learning_rate": 9.941551980335652e-05, | |
| "loss": 0.2688, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.024635598439745433, | |
| "grad_norm": 0.4898300766944885, | |
| "learning_rate": 9.935251313189564e-05, | |
| "loss": 0.2845, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.025456785054403613, | |
| "grad_norm": 0.49974775314331055, | |
| "learning_rate": 9.928630404658255e-05, | |
| "loss": 0.2702, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.026277971669061793, | |
| "grad_norm": 0.5819835662841797, | |
| "learning_rate": 9.921689684362989e-05, | |
| "loss": 0.2699, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.027099158283719976, | |
| "grad_norm": 0.6171815395355225, | |
| "learning_rate": 9.914429602677162e-05, | |
| "loss": 0.3477, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.027920344898378156, | |
| "grad_norm": 0.6020647883415222, | |
| "learning_rate": 9.906850630697068e-05, | |
| "loss": 0.2901, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.028741531513036336, | |
| "grad_norm": 0.7458943724632263, | |
| "learning_rate": 9.898953260211338e-05, | |
| "loss": 0.2498, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.02956271812769452, | |
| "grad_norm": 0.6588975787162781, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.2599, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.0303839047423527, | |
| "grad_norm": 0.6839740872383118, | |
| "learning_rate": 9.882205394146361e-05, | |
| "loss": 0.337, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.03120509135701088, | |
| "grad_norm": 0.6753020882606506, | |
| "learning_rate": 9.87335598531214e-05, | |
| "loss": 0.3069, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.032026277971669063, | |
| "grad_norm": 0.8709951043128967, | |
| "learning_rate": 9.864190351391822e-05, | |
| "loss": 0.3396, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.032847464586327244, | |
| "grad_norm": 0.6819374561309814, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 0.2963, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.033668651200985424, | |
| "grad_norm": 0.856820821762085, | |
| "learning_rate": 9.844912807753104e-05, | |
| "loss": 0.289, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.034489837815643604, | |
| "grad_norm": 0.7196516990661621, | |
| "learning_rate": 9.834802148926882e-05, | |
| "loss": 0.2858, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.035311024430301784, | |
| "grad_norm": 0.7757695913314819, | |
| "learning_rate": 9.824377766717759e-05, | |
| "loss": 0.2882, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.036132211044959964, | |
| "grad_norm": 0.7751405835151672, | |
| "learning_rate": 9.813640337548954e-05, | |
| "loss": 0.3174, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.03695339765961815, | |
| "grad_norm": 0.8659068942070007, | |
| "learning_rate": 9.802590558156862e-05, | |
| "loss": 0.2818, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.03777458427427633, | |
| "grad_norm": 1.336848258972168, | |
| "learning_rate": 9.791229145545831e-05, | |
| "loss": 0.3209, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.03859577088893451, | |
| "grad_norm": 0.9144354462623596, | |
| "learning_rate": 9.779556836941645e-05, | |
| "loss": 0.28, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.03941695750359269, | |
| "grad_norm": 0.8315229415893555, | |
| "learning_rate": 9.767574389743682e-05, | |
| "loss": 0.276, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.04023814411825087, | |
| "grad_norm": 0.967755913734436, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.3298, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.04105933073290905, | |
| "grad_norm": 1.0207139253616333, | |
| "learning_rate": 9.742682209735727e-05, | |
| "loss": 0.3058, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04105933073290905, | |
| "eval_loss": 0.44632911682128906, | |
| "eval_runtime": 291.2748, | |
| "eval_samples_per_second": 7.041, | |
| "eval_steps_per_second": 1.761, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04188051734756724, | |
| "grad_norm": 1.5864020586013794, | |
| "learning_rate": 9.729774092143627e-05, | |
| "loss": 0.4838, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.04270170396222542, | |
| "grad_norm": 1.2497475147247314, | |
| "learning_rate": 9.716559066288715e-05, | |
| "loss": 0.4411, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0435228905768836, | |
| "grad_norm": 0.9487363696098328, | |
| "learning_rate": 9.703037989675087e-05, | |
| "loss": 0.3657, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.04434407719154178, | |
| "grad_norm": 0.33357444405555725, | |
| "learning_rate": 9.689211739666023e-05, | |
| "loss": 0.2727, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.04516526380619996, | |
| "grad_norm": 0.4681488573551178, | |
| "learning_rate": 9.675081213427076e-05, | |
| "loss": 0.315, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.04598645042085814, | |
| "grad_norm": 0.4412896931171417, | |
| "learning_rate": 9.66064732786784e-05, | |
| "loss": 0.3051, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.04680763703551632, | |
| "grad_norm": 0.42806002497673035, | |
| "learning_rate": 9.645911019582467e-05, | |
| "loss": 0.296, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.047628823650174505, | |
| "grad_norm": 0.3795586824417114, | |
| "learning_rate": 9.630873244788883e-05, | |
| "loss": 0.3184, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.048450010264832685, | |
| "grad_norm": 0.34351328015327454, | |
| "learning_rate": 9.615534979266745e-05, | |
| "loss": 0.2832, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.049271196879490865, | |
| "grad_norm": 0.35215750336647034, | |
| "learning_rate": 9.599897218294122e-05, | |
| "loss": 0.3083, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.050092383494149045, | |
| "grad_norm": 0.3156924843788147, | |
| "learning_rate": 9.583960976582913e-05, | |
| "loss": 0.2577, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.050913570108807225, | |
| "grad_norm": 0.4058922231197357, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.3156, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.051734756723465405, | |
| "grad_norm": 0.3673064410686493, | |
| "learning_rate": 9.551197206565173e-05, | |
| "loss": 0.2932, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.052555943338123585, | |
| "grad_norm": 0.3226417601108551, | |
| "learning_rate": 9.534371804252728e-05, | |
| "loss": 0.2404, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.05337712995278177, | |
| "grad_norm": 0.3365015685558319, | |
| "learning_rate": 9.517252173051911e-05, | |
| "loss": 0.263, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.05419831656743995, | |
| "grad_norm": 0.38001781702041626, | |
| "learning_rate": 9.49983942383106e-05, | |
| "loss": 0.2973, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.05501950318209813, | |
| "grad_norm": 0.5203686356544495, | |
| "learning_rate": 9.482134686478519e-05, | |
| "loss": 0.2693, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.05584068979675631, | |
| "grad_norm": 0.32921281456947327, | |
| "learning_rate": 9.464139109829321e-05, | |
| "loss": 0.2742, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.05666187641141449, | |
| "grad_norm": 0.43800121545791626, | |
| "learning_rate": 9.445853861590647e-05, | |
| "loss": 0.2915, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.05748306302607267, | |
| "grad_norm": 0.32359007000923157, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 0.218, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05830424964073085, | |
| "grad_norm": 0.39324089884757996, | |
| "learning_rate": 9.408419115078471e-05, | |
| "loss": 0.2635, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.05912543625538904, | |
| "grad_norm": 0.45099887251853943, | |
| "learning_rate": 9.389272045892024e-05, | |
| "loss": 0.2402, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.05994662287004722, | |
| "grad_norm": 0.4052051603794098, | |
| "learning_rate": 9.36984016313259e-05, | |
| "loss": 0.2775, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.0607678094847054, | |
| "grad_norm": 0.41129499673843384, | |
| "learning_rate": 9.350124727707197e-05, | |
| "loss": 0.2564, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.06158899609936358, | |
| "grad_norm": 0.3723108470439911, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.2614, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.06241018271402176, | |
| "grad_norm": 0.39635589718818665, | |
| "learning_rate": 9.309848334400246e-05, | |
| "loss": 0.2825, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.06323136932867994, | |
| "grad_norm": 0.4818005859851837, | |
| "learning_rate": 9.289289989996133e-05, | |
| "loss": 0.2539, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.06405255594333813, | |
| "grad_norm": 0.4484040439128876, | |
| "learning_rate": 9.268453319711363e-05, | |
| "loss": 0.2857, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.0648737425579963, | |
| "grad_norm": 0.46060118079185486, | |
| "learning_rate": 9.247339675607605e-05, | |
| "loss": 0.2745, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.06569492917265449, | |
| "grad_norm": 0.3952708840370178, | |
| "learning_rate": 9.225950427718975e-05, | |
| "loss": 0.2954, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06651611578731266, | |
| "grad_norm": 0.4594082534313202, | |
| "learning_rate": 9.204286963963111e-05, | |
| "loss": 0.2647, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.06733730240197085, | |
| "grad_norm": 0.498677134513855, | |
| "learning_rate": 9.182350690051133e-05, | |
| "loss": 0.2741, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.06815848901662903, | |
| "grad_norm": 0.48790332674980164, | |
| "learning_rate": 9.160143029396422e-05, | |
| "loss": 0.217, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.06897967563128721, | |
| "grad_norm": 0.4500594437122345, | |
| "learning_rate": 9.13766542302225e-05, | |
| "loss": 0.2478, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.0698008622459454, | |
| "grad_norm": 0.477200984954834, | |
| "learning_rate": 9.114919329468282e-05, | |
| "loss": 0.276, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.07062204886060357, | |
| "grad_norm": 0.5421808362007141, | |
| "learning_rate": 9.091906224695935e-05, | |
| "loss": 0.3018, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.07144323547526175, | |
| "grad_norm": 0.473712295293808, | |
| "learning_rate": 9.068627601992598e-05, | |
| "loss": 0.2687, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.07226442208991993, | |
| "grad_norm": 0.5594393610954285, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.2515, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.07308560870457811, | |
| "grad_norm": 0.6386433839797974, | |
| "learning_rate": 9.021279861989885e-05, | |
| "loss": 0.2529, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.0739067953192363, | |
| "grad_norm": 0.5319457650184631, | |
| "learning_rate": 8.997213817017507e-05, | |
| "loss": 0.2646, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07472798193389447, | |
| "grad_norm": 0.595516562461853, | |
| "learning_rate": 8.972888398568772e-05, | |
| "loss": 0.2756, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.07554916854855266, | |
| "grad_norm": 0.6601677536964417, | |
| "learning_rate": 8.948305185085225e-05, | |
| "loss": 0.2805, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.07637035516321083, | |
| "grad_norm": 0.6162546873092651, | |
| "learning_rate": 8.92346577173636e-05, | |
| "loss": 0.2435, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.07719154177786902, | |
| "grad_norm": 0.6918389797210693, | |
| "learning_rate": 8.898371770316111e-05, | |
| "loss": 0.3159, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.07801272839252721, | |
| "grad_norm": 0.7181240916252136, | |
| "learning_rate": 8.873024809138272e-05, | |
| "loss": 0.2413, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.07883391500718538, | |
| "grad_norm": 1.1472549438476562, | |
| "learning_rate": 8.847426532930831e-05, | |
| "loss": 0.2949, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.07965510162184357, | |
| "grad_norm": 0.9359822273254395, | |
| "learning_rate": 8.821578602729242e-05, | |
| "loss": 0.316, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.08047628823650174, | |
| "grad_norm": 0.7444966435432434, | |
| "learning_rate": 8.795482695768658e-05, | |
| "loss": 0.3119, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.08129747485115993, | |
| "grad_norm": 1.2565944194793701, | |
| "learning_rate": 8.769140505375085e-05, | |
| "loss": 0.2877, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.0821186614658181, | |
| "grad_norm": 0.9129545092582703, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 0.3209, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0821186614658181, | |
| "eval_loss": 0.4635893702507019, | |
| "eval_runtime": 290.5387, | |
| "eval_samples_per_second": 7.059, | |
| "eval_steps_per_second": 1.766, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.08293984808047629, | |
| "grad_norm": 1.5546329021453857, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 0.5163, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.08376103469513448, | |
| "grad_norm": 1.350873589515686, | |
| "learning_rate": 8.688653405904652e-05, | |
| "loss": 0.4791, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.08458222130979265, | |
| "grad_norm": 0.984677255153656, | |
| "learning_rate": 8.661343332988869e-05, | |
| "loss": 0.3377, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.08540340792445084, | |
| "grad_norm": 0.3391437530517578, | |
| "learning_rate": 8.633795680751116e-05, | |
| "loss": 0.2633, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.08622459453910901, | |
| "grad_norm": 0.2502461373806, | |
| "learning_rate": 8.606012236719073e-05, | |
| "loss": 0.2567, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.0870457811537672, | |
| "grad_norm": 0.28781577944755554, | |
| "learning_rate": 8.577994803720606e-05, | |
| "loss": 0.2769, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.08786696776842537, | |
| "grad_norm": 0.30542901158332825, | |
| "learning_rate": 8.549745199766792e-05, | |
| "loss": 0.2965, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.08868815438308356, | |
| "grad_norm": 0.3351757526397705, | |
| "learning_rate": 8.521265257933948e-05, | |
| "loss": 0.2774, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.08950934099774174, | |
| "grad_norm": 0.2978907525539398, | |
| "learning_rate": 8.492556826244687e-05, | |
| "loss": 0.2744, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.09033052761239992, | |
| "grad_norm": 0.2558261752128601, | |
| "learning_rate": 8.463621767547998e-05, | |
| "loss": 0.2312, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.0911517142270581, | |
| "grad_norm": 0.30879148840904236, | |
| "learning_rate": 8.434461959398376e-05, | |
| "loss": 0.2845, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.09197290084171628, | |
| "grad_norm": 0.309334397315979, | |
| "learning_rate": 8.405079293933986e-05, | |
| "loss": 0.266, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.09279408745637446, | |
| "grad_norm": 0.3099028170108795, | |
| "learning_rate": 8.375475677753881e-05, | |
| "loss": 0.2692, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.09361527407103264, | |
| "grad_norm": 0.30877748131752014, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 0.284, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.09443646068569082, | |
| "grad_norm": 0.2932473123073578, | |
| "learning_rate": 8.315613291203976e-05, | |
| "loss": 0.268, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.09525764730034901, | |
| "grad_norm": 0.337277889251709, | |
| "learning_rate": 8.285358405218655e-05, | |
| "loss": 0.2448, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.09607883391500718, | |
| "grad_norm": 0.3811333477497101, | |
| "learning_rate": 8.25489033703452e-05, | |
| "loss": 0.293, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.09690002052966537, | |
| "grad_norm": 0.34305429458618164, | |
| "learning_rate": 8.224211063680853e-05, | |
| "loss": 0.2596, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.09772120714432354, | |
| "grad_norm": 0.3743656277656555, | |
| "learning_rate": 8.19332257589174e-05, | |
| "loss": 0.277, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.09854239375898173, | |
| "grad_norm": 0.2987310290336609, | |
| "learning_rate": 8.162226877976887e-05, | |
| "loss": 0.2484, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.0993635803736399, | |
| "grad_norm": 0.2911909222602844, | |
| "learning_rate": 8.130925987691569e-05, | |
| "loss": 0.2297, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.10018476698829809, | |
| "grad_norm": 0.29610538482666016, | |
| "learning_rate": 8.099421936105702e-05, | |
| "loss": 0.2224, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.10100595360295628, | |
| "grad_norm": 0.31938186287879944, | |
| "learning_rate": 8.067716767472045e-05, | |
| "loss": 0.2158, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.10182714021761445, | |
| "grad_norm": 0.3307250440120697, | |
| "learning_rate": 8.035812539093557e-05, | |
| "loss": 0.2443, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.10264832683227264, | |
| "grad_norm": 0.33900371193885803, | |
| "learning_rate": 8.003711321189895e-05, | |
| "loss": 0.2455, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.10346951344693081, | |
| "grad_norm": 0.33954647183418274, | |
| "learning_rate": 7.971415196763088e-05, | |
| "loss": 0.2413, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.104290700061589, | |
| "grad_norm": 0.3721504211425781, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 0.2724, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.10511188667624717, | |
| "grad_norm": 0.3788248300552368, | |
| "learning_rate": 7.906246623448183e-05, | |
| "loss": 0.2727, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.10593307329090536, | |
| "grad_norm": 0.3605785667896271, | |
| "learning_rate": 7.873378403255419e-05, | |
| "loss": 0.2781, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.10675425990556354, | |
| "grad_norm": 0.33415114879608154, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 0.2226, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.10757544652022172, | |
| "grad_norm": 0.43883877992630005, | |
| "learning_rate": 7.807084759519405e-05, | |
| "loss": 0.2474, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.1083966331348799, | |
| "grad_norm": 0.3597790598869324, | |
| "learning_rate": 7.773663637675694e-05, | |
| "loss": 0.2375, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.10921781974953808, | |
| "grad_norm": 0.41835817694664, | |
| "learning_rate": 7.740062536773352e-05, | |
| "loss": 0.2869, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.11003900636419627, | |
| "grad_norm": 0.40062186121940613, | |
| "learning_rate": 7.706283637139658e-05, | |
| "loss": 0.2451, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.11086019297885444, | |
| "grad_norm": 0.46534448862075806, | |
| "learning_rate": 7.672329130639005e-05, | |
| "loss": 0.2834, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.11168137959351263, | |
| "grad_norm": 0.4524616599082947, | |
| "learning_rate": 7.638201220530665e-05, | |
| "loss": 0.2579, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.11250256620817081, | |
| "grad_norm": 0.42796769738197327, | |
| "learning_rate": 7.603902121325813e-05, | |
| "loss": 0.2507, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.11332375282282899, | |
| "grad_norm": 0.6147335767745972, | |
| "learning_rate": 7.569434058643844e-05, | |
| "loss": 0.2547, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.11414493943748717, | |
| "grad_norm": 0.5195381045341492, | |
| "learning_rate": 7.534799269067953e-05, | |
| "loss": 0.2429, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.11496612605214535, | |
| "grad_norm": 0.549555242061615, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.2826, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.11578731266680353, | |
| "grad_norm": 0.5704975128173828, | |
| "learning_rate": 7.465038509514688e-05, | |
| "loss": 0.2789, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.1166084992814617, | |
| "grad_norm": 0.6339702606201172, | |
| "learning_rate": 7.42991706621303e-05, | |
| "loss": 0.2289, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.11742968589611989, | |
| "grad_norm": 0.728171706199646, | |
| "learning_rate": 7.394637949075154e-05, | |
| "loss": 0.3104, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.11825087251077808, | |
| "grad_norm": 0.5731471180915833, | |
| "learning_rate": 7.35920344731241e-05, | |
| "loss": 0.2612, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.11907205912543625, | |
| "grad_norm": 0.7613599896430969, | |
| "learning_rate": 7.323615860218843e-05, | |
| "loss": 0.3024, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.11989324574009444, | |
| "grad_norm": 0.5790632963180542, | |
| "learning_rate": 7.287877497021978e-05, | |
| "loss": 0.2493, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.12071443235475261, | |
| "grad_norm": 0.746548593044281, | |
| "learning_rate": 7.251990676732984e-05, | |
| "loss": 0.2999, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.1215356189694108, | |
| "grad_norm": 0.7720208168029785, | |
| "learning_rate": 7.215957727996207e-05, | |
| "loss": 0.2836, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.12235680558406897, | |
| "grad_norm": 0.6100975275039673, | |
| "learning_rate": 7.179780988938051e-05, | |
| "loss": 0.2187, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.12317799219872716, | |
| "grad_norm": 0.7999358177185059, | |
| "learning_rate": 7.143462807015271e-05, | |
| "loss": 0.2709, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.12317799219872716, | |
| "eval_loss": 0.38091349601745605, | |
| "eval_runtime": 291.0342, | |
| "eval_samples_per_second": 7.047, | |
| "eval_steps_per_second": 1.763, | |
| "step": 150 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 50, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.6081050741230797e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |