| { | |
| "best_metric": 1.1766536235809326, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-400", | |
| "epoch": 0.45610034207525657, | |
| "eval_steps": 100, | |
| "global_step": 500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0009122006841505131, | |
| "grad_norm": 0.25783950090408325, | |
| "learning_rate": 5e-06, | |
| "loss": 1.2112, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0009122006841505131, | |
| "eval_loss": 1.528485894203186, | |
| "eval_runtime": 113.6842, | |
| "eval_samples_per_second": 16.238, | |
| "eval_steps_per_second": 4.064, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0018244013683010262, | |
| "grad_norm": 0.3564816117286682, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0953, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0027366020524515395, | |
| "grad_norm": 0.3273909389972687, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.3012, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0036488027366020524, | |
| "grad_norm": 0.2606115937232971, | |
| "learning_rate": 2e-05, | |
| "loss": 1.2529, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.004561003420752566, | |
| "grad_norm": 0.3063911199569702, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.4867, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.005473204104903079, | |
| "grad_norm": 0.30145400762557983, | |
| "learning_rate": 3e-05, | |
| "loss": 1.4505, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0063854047890535915, | |
| "grad_norm": 0.3016102612018585, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.5596, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.007297605473204105, | |
| "grad_norm": 0.291785329580307, | |
| "learning_rate": 4e-05, | |
| "loss": 2.1331, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.008209806157354617, | |
| "grad_norm": 0.38654470443725586, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.998, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.009122006841505131, | |
| "grad_norm": 0.24303749203681946, | |
| "learning_rate": 5e-05, | |
| "loss": 1.2621, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.010034207525655644, | |
| "grad_norm": 0.4852235019207001, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 2.9854, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.010946408209806158, | |
| "grad_norm": 0.2009701430797577, | |
| "learning_rate": 6e-05, | |
| "loss": 1.2834, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.01185860889395667, | |
| "grad_norm": 0.17731308937072754, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 1.3094, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.012770809578107183, | |
| "grad_norm": 0.1488691121339798, | |
| "learning_rate": 7e-05, | |
| "loss": 1.0258, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.013683010262257697, | |
| "grad_norm": 0.15303461253643036, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 1.3275, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.01459521094640821, | |
| "grad_norm": 0.19198919832706451, | |
| "learning_rate": 8e-05, | |
| "loss": 1.2239, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.015507411630558724, | |
| "grad_norm": 0.18219693005084991, | |
| "learning_rate": 8.5e-05, | |
| "loss": 1.1367, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.016419612314709234, | |
| "grad_norm": 0.15564540028572083, | |
| "learning_rate": 9e-05, | |
| "loss": 1.1671, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.01733181299885975, | |
| "grad_norm": 0.15285353362560272, | |
| "learning_rate": 9.5e-05, | |
| "loss": 1.2621, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.018244013683010263, | |
| "grad_norm": 0.1607624590396881, | |
| "learning_rate": 0.0001, | |
| "loss": 1.3366, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.019156214367160775, | |
| "grad_norm": 0.1453542858362198, | |
| "learning_rate": 9.999892908320647e-05, | |
| "loss": 1.0341, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.020068415051311288, | |
| "grad_norm": 0.2023877501487732, | |
| "learning_rate": 9.999571637870036e-05, | |
| "loss": 1.3088, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.0209806157354618, | |
| "grad_norm": 0.14624053239822388, | |
| "learning_rate": 9.999036202410325e-05, | |
| "loss": 0.9596, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.021892816419612316, | |
| "grad_norm": 0.18767771124839783, | |
| "learning_rate": 9.998286624877786e-05, | |
| "loss": 1.2299, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.02280501710376283, | |
| "grad_norm": 0.20108367502689362, | |
| "learning_rate": 9.997322937381829e-05, | |
| "loss": 1.2757, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.02371721778791334, | |
| "grad_norm": 0.2668151259422302, | |
| "learning_rate": 9.996145181203615e-05, | |
| "loss": 1.2839, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.024629418472063853, | |
| "grad_norm": 0.21400977671146393, | |
| "learning_rate": 9.994753406794301e-05, | |
| "loss": 1.4506, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.025541619156214366, | |
| "grad_norm": 0.26276835799217224, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 1.2257, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.026453819840364882, | |
| "grad_norm": 0.23972387611865997, | |
| "learning_rate": 9.991328050923581e-05, | |
| "loss": 1.3055, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.027366020524515394, | |
| "grad_norm": 0.19858954846858978, | |
| "learning_rate": 9.989294616193017e-05, | |
| "loss": 1.0788, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.028278221208665907, | |
| "grad_norm": 0.21251460909843445, | |
| "learning_rate": 9.98704745668676e-05, | |
| "loss": 1.0723, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.02919042189281642, | |
| "grad_norm": 0.25747787952423096, | |
| "learning_rate": 9.98458666866564e-05, | |
| "loss": 1.2754, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.03010262257696693, | |
| "grad_norm": 0.18308940529823303, | |
| "learning_rate": 9.981912357541627e-05, | |
| "loss": 1.1309, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.031014823261117448, | |
| "grad_norm": 0.28769150376319885, | |
| "learning_rate": 9.97902463787331e-05, | |
| "loss": 1.1491, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.03192702394526796, | |
| "grad_norm": 0.33303603529930115, | |
| "learning_rate": 9.975923633360985e-05, | |
| "loss": 1.5505, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.03283922462941847, | |
| "grad_norm": 0.19029738008975983, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 1.3271, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.033751425313568985, | |
| "grad_norm": 0.1386023312807083, | |
| "learning_rate": 9.969082310281891e-05, | |
| "loss": 1.2999, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.0346636259977195, | |
| "grad_norm": 0.13166525959968567, | |
| "learning_rate": 9.965342284774632e-05, | |
| "loss": 1.2839, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.03557582668187001, | |
| "grad_norm": 0.1366724967956543, | |
| "learning_rate": 9.961389560529836e-05, | |
| "loss": 1.327, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.036488027366020526, | |
| "grad_norm": 0.15251226723194122, | |
| "learning_rate": 9.957224306869053e-05, | |
| "loss": 1.312, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.037400228050171035, | |
| "grad_norm": 0.13887174427509308, | |
| "learning_rate": 9.952846702217886e-05, | |
| "loss": 1.3777, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.03831242873432155, | |
| "grad_norm": 0.1658075451850891, | |
| "learning_rate": 9.948256934098352e-05, | |
| "loss": 1.3959, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.039224629418472066, | |
| "grad_norm": 0.16137467324733734, | |
| "learning_rate": 9.943455199120837e-05, | |
| "loss": 1.3763, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.040136830102622575, | |
| "grad_norm": 0.17031268775463104, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 1.3507, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.04104903078677309, | |
| "grad_norm": 0.17008689045906067, | |
| "learning_rate": 9.933216660424395e-05, | |
| "loss": 1.3843, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0419612314709236, | |
| "grad_norm": 0.23986858129501343, | |
| "learning_rate": 9.927780295290389e-05, | |
| "loss": 1.3637, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.042873432155074116, | |
| "grad_norm": 0.253842830657959, | |
| "learning_rate": 9.922132840449459e-05, | |
| "loss": 1.4677, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.04378563283922463, | |
| "grad_norm": 0.3024607002735138, | |
| "learning_rate": 9.916274537819775e-05, | |
| "loss": 1.595, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.04469783352337514, | |
| "grad_norm": 0.4269244372844696, | |
| "learning_rate": 9.91020563835152e-05, | |
| "loss": 1.6949, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.04561003420752566, | |
| "grad_norm": 1.273447871208191, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 2.6063, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.046522234891676166, | |
| "grad_norm": 0.25115707516670227, | |
| "learning_rate": 9.897437097795257e-05, | |
| "loss": 1.2458, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.04743443557582668, | |
| "grad_norm": 0.2734687328338623, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 1.1315, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.0483466362599772, | |
| "grad_norm": 0.3112577199935913, | |
| "learning_rate": 9.883829406604363e-05, | |
| "loss": 1.2837, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.04925883694412771, | |
| "grad_norm": 0.2956832945346832, | |
| "learning_rate": 9.876711602542563e-05, | |
| "loss": 1.6056, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.05017103762827822, | |
| "grad_norm": 0.27159854769706726, | |
| "learning_rate": 9.869384896386668e-05, | |
| "loss": 1.4008, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.05108323831242873, | |
| "grad_norm": 0.23057343065738678, | |
| "learning_rate": 9.861849601988383e-05, | |
| "loss": 1.3612, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.05199543899657925, | |
| "grad_norm": 0.22093012928962708, | |
| "learning_rate": 9.854106042134641e-05, | |
| "loss": 1.3945, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.052907639680729764, | |
| "grad_norm": 0.24365180730819702, | |
| "learning_rate": 9.846154548533773e-05, | |
| "loss": 1.4811, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.05381984036488027, | |
| "grad_norm": 0.255241721868515, | |
| "learning_rate": 9.837995461801299e-05, | |
| "loss": 1.3864, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.05473204104903079, | |
| "grad_norm": 0.2505890429019928, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 1.3, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0556442417331813, | |
| "grad_norm": 0.19075927138328552, | |
| "learning_rate": 9.821055915851647e-05, | |
| "loss": 1.1616, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.05655644241733181, | |
| "grad_norm": 0.19343258440494537, | |
| "learning_rate": 9.812276182268236e-05, | |
| "loss": 1.0507, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.05746864310148233, | |
| "grad_norm": 0.14569373428821564, | |
| "learning_rate": 9.803290306789676e-05, | |
| "loss": 1.1448, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.05838084378563284, | |
| "grad_norm": 0.1645849347114563, | |
| "learning_rate": 9.794098674340965e-05, | |
| "loss": 1.2202, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.059293044469783354, | |
| "grad_norm": 0.1657133549451828, | |
| "learning_rate": 9.784701678661045e-05, | |
| "loss": 0.9998, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.06020524515393386, | |
| "grad_norm": 0.14773251116275787, | |
| "learning_rate": 9.775099722285935e-05, | |
| "loss": 1.0601, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.06111744583808438, | |
| "grad_norm": 0.14379927515983582, | |
| "learning_rate": 9.765293216531486e-05, | |
| "loss": 1.232, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.062029646522234895, | |
| "grad_norm": 0.14879746735095978, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 1.011, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.0629418472063854, | |
| "grad_norm": 0.14008712768554688, | |
| "learning_rate": 9.74506824594107e-05, | |
| "loss": 0.9511, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.06385404789053592, | |
| "grad_norm": 0.15255555510520935, | |
| "learning_rate": 9.73465064747553e-05, | |
| "loss": 0.983, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06476624857468644, | |
| "grad_norm": 0.16574011743068695, | |
| "learning_rate": 9.724030232334391e-05, | |
| "loss": 1.0218, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.06567844925883694, | |
| "grad_norm": 0.16048061847686768, | |
| "learning_rate": 9.713207455460894e-05, | |
| "loss": 1.1199, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.06659064994298745, | |
| "grad_norm": 0.18178284168243408, | |
| "learning_rate": 9.702182780466775e-05, | |
| "loss": 1.0187, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.06750285062713797, | |
| "grad_norm": 0.15241661667823792, | |
| "learning_rate": 9.690956679612421e-05, | |
| "loss": 1.2068, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.06841505131128849, | |
| "grad_norm": 0.1547076255083084, | |
| "learning_rate": 9.67952963378663e-05, | |
| "loss": 0.9773, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.069327251995439, | |
| "grad_norm": 0.1607867181301117, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 1.1561, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.0702394526795895, | |
| "grad_norm": 0.17180593311786652, | |
| "learning_rate": 9.656074673794018e-05, | |
| "loss": 1.0215, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.07115165336374002, | |
| "grad_norm": 0.18671779334545135, | |
| "learning_rate": 9.644047764359622e-05, | |
| "loss": 1.0732, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.07206385404789054, | |
| "grad_norm": 0.22109557688236237, | |
| "learning_rate": 9.631821919375591e-05, | |
| "loss": 1.2983, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.07297605473204105, | |
| "grad_norm": 0.16936635971069336, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 1.0403, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.07388825541619157, | |
| "grad_norm": 0.20985794067382812, | |
| "learning_rate": 9.606775526115963e-05, | |
| "loss": 1.1105, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.07480045610034207, | |
| "grad_norm": 0.24890471994876862, | |
| "learning_rate": 9.593956050744492e-05, | |
| "loss": 1.0215, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.07571265678449259, | |
| "grad_norm": 0.204713374376297, | |
| "learning_rate": 9.580939785585681e-05, | |
| "loss": 1.1271, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.0766248574686431, | |
| "grad_norm": 0.2502953112125397, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 1.2075, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.07753705815279362, | |
| "grad_norm": 0.2343108057975769, | |
| "learning_rate": 9.554319124605879e-05, | |
| "loss": 1.2663, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.07844925883694413, | |
| "grad_norm": 0.15296907722949982, | |
| "learning_rate": 9.540715869125407e-05, | |
| "loss": 1.2773, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.07936145952109463, | |
| "grad_norm": 0.13890579342842102, | |
| "learning_rate": 9.526918104489777e-05, | |
| "loss": 1.0958, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.08027366020524515, | |
| "grad_norm": 0.14783431589603424, | |
| "learning_rate": 9.512926421749304e-05, | |
| "loss": 1.1379, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.08118586088939567, | |
| "grad_norm": 0.1457366943359375, | |
| "learning_rate": 9.498741420261108e-05, | |
| "loss": 1.1979, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.08209806157354618, | |
| "grad_norm": 0.14700822532176971, | |
| "learning_rate": 9.484363707663442e-05, | |
| "loss": 1.2511, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0830102622576967, | |
| "grad_norm": 0.14076313376426697, | |
| "learning_rate": 9.469793899849661e-05, | |
| "loss": 1.273, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.0839224629418472, | |
| "grad_norm": 0.1587403565645218, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 1.2475, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.08483466362599772, | |
| "grad_norm": 0.1829884648323059, | |
| "learning_rate": 9.440080503264037e-05, | |
| "loss": 1.3541, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.08574686431014823, | |
| "grad_norm": 0.15041080117225647, | |
| "learning_rate": 9.42493818731521e-05, | |
| "loss": 1.3175, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.08665906499429875, | |
| "grad_norm": 0.16183224320411682, | |
| "learning_rate": 9.409606321741775e-05, | |
| "loss": 1.3011, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.08757126567844926, | |
| "grad_norm": 0.17238490283489227, | |
| "learning_rate": 9.394085563309827e-05, | |
| "loss": 1.441, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.08848346636259977, | |
| "grad_norm": 0.1964457482099533, | |
| "learning_rate": 9.378376576876999e-05, | |
| "loss": 1.3421, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.08939566704675028, | |
| "grad_norm": 0.2645294964313507, | |
| "learning_rate": 9.362480035363986e-05, | |
| "loss": 1.5238, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.0903078677309008, | |
| "grad_norm": 0.4722297191619873, | |
| "learning_rate": 9.34639661972572e-05, | |
| "loss": 1.6759, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.09122006841505131, | |
| "grad_norm": 1.4135944843292236, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 2.271, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09122006841505131, | |
| "eval_loss": 1.2722834348678589, | |
| "eval_runtime": 113.025, | |
| "eval_samples_per_second": 16.333, | |
| "eval_steps_per_second": 4.088, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09213226909920183, | |
| "grad_norm": 0.12369579821825027, | |
| "learning_rate": 9.31367192988896e-05, | |
| "loss": 0.996, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.09304446978335233, | |
| "grad_norm": 0.12814058363437653, | |
| "learning_rate": 9.297032057507264e-05, | |
| "loss": 1.0837, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.09395667046750285, | |
| "grad_norm": 0.14010389149188995, | |
| "learning_rate": 9.280208114573859e-05, | |
| "loss": 1.2772, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.09486887115165336, | |
| "grad_norm": 0.15505646169185638, | |
| "learning_rate": 9.263200821770461e-05, | |
| "loss": 1.3608, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.09578107183580388, | |
| "grad_norm": 0.16221708059310913, | |
| "learning_rate": 9.246010907632895e-05, | |
| "loss": 1.3107, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.0966932725199544, | |
| "grad_norm": 0.15704494714736938, | |
| "learning_rate": 9.228639108519868e-05, | |
| "loss": 1.1564, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.0976054732041049, | |
| "grad_norm": 0.15318067371845245, | |
| "learning_rate": 9.211086168581433e-05, | |
| "loss": 1.2841, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.09851767388825541, | |
| "grad_norm": 0.16971643269062042, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 1.3776, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.09942987457240593, | |
| "grad_norm": 0.17601683735847473, | |
| "learning_rate": 9.175439881593716e-05, | |
| "loss": 1.1342, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.10034207525655645, | |
| "grad_norm": 0.2042451649904251, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 1.3205, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.10125427594070696, | |
| "grad_norm": 0.20541740953922272, | |
| "learning_rate": 9.139078154477512e-05, | |
| "loss": 1.2256, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.10216647662485746, | |
| "grad_norm": 0.1825324296951294, | |
| "learning_rate": 9.120630943110077e-05, | |
| "loss": 1.1943, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.10307867730900798, | |
| "grad_norm": 0.17503520846366882, | |
| "learning_rate": 9.102007217627568e-05, | |
| "loss": 1.0749, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.1039908779931585, | |
| "grad_norm": 0.1890123039484024, | |
| "learning_rate": 9.083207775808396e-05, | |
| "loss": 1.0601, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.10490307867730901, | |
| "grad_norm": 0.15623880922794342, | |
| "learning_rate": 9.064233422958077e-05, | |
| "loss": 0.8245, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.10581527936145953, | |
| "grad_norm": 0.1406368911266327, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 1.0316, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.10672748004561003, | |
| "grad_norm": 0.1501624435186386, | |
| "learning_rate": 9.025763242814291e-05, | |
| "loss": 1.1908, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.10763968072976055, | |
| "grad_norm": 0.13635428249835968, | |
| "learning_rate": 9.006269063455304e-05, | |
| "loss": 0.9685, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.10855188141391106, | |
| "grad_norm": 0.1702578067779541, | |
| "learning_rate": 8.986603268863536e-05, | |
| "loss": 1.0043, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.10946408209806158, | |
| "grad_norm": 0.16892701387405396, | |
| "learning_rate": 8.966766701456177e-05, | |
| "loss": 1.0515, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.11037628278221209, | |
| "grad_norm": 0.1571364402770996, | |
| "learning_rate": 8.94676021096575e-05, | |
| "loss": 0.9528, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.1112884834663626, | |
| "grad_norm": 0.15004131197929382, | |
| "learning_rate": 8.926584654403724e-05, | |
| "loss": 1.0623, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.11220068415051311, | |
| "grad_norm": 0.18099543452262878, | |
| "learning_rate": 8.906240896023794e-05, | |
| "loss": 1.0798, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.11311288483466363, | |
| "grad_norm": 0.16211232542991638, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 0.9695, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.11402508551881414, | |
| "grad_norm": 0.19721828401088715, | |
| "learning_rate": 8.865052266813685e-05, | |
| "loss": 1.0282, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.11493728620296466, | |
| "grad_norm": 0.16071362793445587, | |
| "learning_rate": 8.844209160367299e-05, | |
| "loss": 1.0692, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.11584948688711516, | |
| "grad_norm": 0.20482495427131653, | |
| "learning_rate": 8.823201380795001e-05, | |
| "loss": 1.1321, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.11676168757126568, | |
| "grad_norm": 0.19026656448841095, | |
| "learning_rate": 8.802029828000156e-05, | |
| "loss": 0.9361, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.11767388825541619, | |
| "grad_norm": 0.19571590423583984, | |
| "learning_rate": 8.780695408901613e-05, | |
| "loss": 1.0152, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.11858608893956671, | |
| "grad_norm": 0.2167755663394928, | |
| "learning_rate": 8.759199037394887e-05, | |
| "loss": 1.1986, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11949828962371722, | |
| "grad_norm": 0.20209772884845734, | |
| "learning_rate": 8.737541634312985e-05, | |
| "loss": 0.9061, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.12041049030786773, | |
| "grad_norm": 0.2215677797794342, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 0.8564, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.12132269099201824, | |
| "grad_norm": 0.2394784688949585, | |
| "learning_rate": 8.693747451206232e-05, | |
| "loss": 1.1324, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.12223489167616876, | |
| "grad_norm": 0.2727131247520447, | |
| "learning_rate": 8.671612547178428e-05, | |
| "loss": 1.3143, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.12314709236031927, | |
| "grad_norm": 0.263237327337265, | |
| "learning_rate": 8.649320363489179e-05, | |
| "loss": 1.654, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.12405929304446979, | |
| "grad_norm": 0.14213962852954865, | |
| "learning_rate": 8.626871855061438e-05, | |
| "loss": 1.2624, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.12497149372862029, | |
| "grad_norm": 0.11428536474704742, | |
| "learning_rate": 8.604267983514594e-05, | |
| "loss": 1.1345, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.1258836944127708, | |
| "grad_norm": 0.11425229161977768, | |
| "learning_rate": 8.581509717123273e-05, | |
| "loss": 1.0768, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.12679589509692132, | |
| "grad_norm": 0.14232636988162994, | |
| "learning_rate": 8.558598030775857e-05, | |
| "loss": 1.2241, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.12770809578107184, | |
| "grad_norm": 0.14870019257068634, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 1.2843, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12862029646522236, | |
| "grad_norm": 0.14005614817142487, | |
| "learning_rate": 8.51231833058426e-05, | |
| "loss": 1.348, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.12953249714937287, | |
| "grad_norm": 0.14191389083862305, | |
| "learning_rate": 8.488952299208401e-05, | |
| "loss": 1.3306, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.1304446978335234, | |
| "grad_norm": 0.14700360596179962, | |
| "learning_rate": 8.46543681272818e-05, | |
| "loss": 1.2998, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.13135689851767388, | |
| "grad_norm": 0.17052660882472992, | |
| "learning_rate": 8.44177287846877e-05, | |
| "loss": 1.3757, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.1322690992018244, | |
| "grad_norm": 0.17917169630527496, | |
| "learning_rate": 8.417961510114356e-05, | |
| "loss": 1.276, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1331812998859749, | |
| "grad_norm": 0.1966487616300583, | |
| "learning_rate": 8.39400372766471e-05, | |
| "loss": 1.2529, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.13409350057012542, | |
| "grad_norm": 0.22469089925289154, | |
| "learning_rate": 8.36990055739149e-05, | |
| "loss": 1.4108, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.13500570125427594, | |
| "grad_norm": 0.29152804613113403, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 1.3991, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.13591790193842646, | |
| "grad_norm": 0.44599342346191406, | |
| "learning_rate": 8.321262189556409e-05, | |
| "loss": 1.5375, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.13683010262257697, | |
| "grad_norm": 0.9355916380882263, | |
| "learning_rate": 8.296729075500344e-05, | |
| "loss": 2.2142, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1377423033067275, | |
| "grad_norm": 0.16966940462589264, | |
| "learning_rate": 8.272054740543052e-05, | |
| "loss": 1.0935, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.138654503990878, | |
| "grad_norm": 0.17415517568588257, | |
| "learning_rate": 8.247240241650918e-05, | |
| "loss": 1.069, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.13956670467502852, | |
| "grad_norm": 0.15520510077476501, | |
| "learning_rate": 8.222286641794488e-05, | |
| "loss": 1.1164, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.140478905359179, | |
| "grad_norm": 0.16199630498886108, | |
| "learning_rate": 8.197195009902924e-05, | |
| "loss": 1.2711, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.14139110604332952, | |
| "grad_norm": 0.20093193650245667, | |
| "learning_rate": 8.171966420818228e-05, | |
| "loss": 1.3726, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.14230330672748004, | |
| "grad_norm": 0.17293791472911835, | |
| "learning_rate": 8.146601955249188e-05, | |
| "loss": 1.2781, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.14321550741163055, | |
| "grad_norm": 0.20883025228977203, | |
| "learning_rate": 8.121102699725089e-05, | |
| "loss": 1.3765, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.14412770809578107, | |
| "grad_norm": 0.1882334053516388, | |
| "learning_rate": 8.095469746549172e-05, | |
| "loss": 1.2449, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.1450399087799316, | |
| "grad_norm": 0.20094580948352814, | |
| "learning_rate": 8.069704193751832e-05, | |
| "loss": 1.2535, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.1459521094640821, | |
| "grad_norm": 0.2629421055316925, | |
| "learning_rate": 8.043807145043604e-05, | |
| "loss": 1.3591, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.14686431014823262, | |
| "grad_norm": 0.6893475651741028, | |
| "learning_rate": 8.017779709767858e-05, | |
| "loss": 2.445, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.14777651083238313, | |
| "grad_norm": 0.14671513438224792, | |
| "learning_rate": 7.991623002853296e-05, | |
| "loss": 0.9265, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.14868871151653365, | |
| "grad_norm": 0.17360147833824158, | |
| "learning_rate": 7.965338144766186e-05, | |
| "loss": 0.9184, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.14960091220068414, | |
| "grad_norm": 0.16476592421531677, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 1.0991, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.15051311288483465, | |
| "grad_norm": 0.18807539343833923, | |
| "learning_rate": 7.912388484339012e-05, | |
| "loss": 1.1164, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.15142531356898517, | |
| "grad_norm": 0.14939163625240326, | |
| "learning_rate": 7.88572595018617e-05, | |
| "loss": 1.0408, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.1523375142531357, | |
| "grad_norm": 0.16947825253009796, | |
| "learning_rate": 7.858939801138061e-05, | |
| "loss": 0.937, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.1532497149372862, | |
| "grad_norm": 0.13935978710651398, | |
| "learning_rate": 7.832031184624164e-05, | |
| "loss": 0.9394, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.15416191562143672, | |
| "grad_norm": 0.16934405267238617, | |
| "learning_rate": 7.80500125332005e-05, | |
| "loss": 1.1168, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.15507411630558723, | |
| "grad_norm": 0.16050076484680176, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.8851, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.15598631698973775, | |
| "grad_norm": 0.13356317579746246, | |
| "learning_rate": 7.750582082977467e-05, | |
| "loss": 0.8304, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.15689851767388827, | |
| "grad_norm": 0.14673428237438202, | |
| "learning_rate": 7.723195175075136e-05, | |
| "loss": 1.1082, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.15781071835803878, | |
| "grad_norm": 0.18469220399856567, | |
| "learning_rate": 7.695691614555003e-05, | |
| "loss": 1.314, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.15872291904218927, | |
| "grad_norm": 0.15937888622283936, | |
| "learning_rate": 7.668072579578058e-05, | |
| "loss": 0.917, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.15963511972633979, | |
| "grad_norm": 0.17641136050224304, | |
| "learning_rate": 7.64033925325184e-05, | |
| "loss": 1.0633, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1605473204104903, | |
| "grad_norm": 0.13878133893013, | |
| "learning_rate": 7.612492823579745e-05, | |
| "loss": 0.6571, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.16145952109464082, | |
| "grad_norm": 0.19035881757736206, | |
| "learning_rate": 7.584534483410137e-05, | |
| "loss": 1.0225, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.16237172177879133, | |
| "grad_norm": 0.19843564927577972, | |
| "learning_rate": 7.55646543038526e-05, | |
| "loss": 1.0645, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.16328392246294185, | |
| "grad_norm": 0.18366238474845886, | |
| "learning_rate": 7.528286866889924e-05, | |
| "loss": 0.91, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.16419612314709237, | |
| "grad_norm": 0.2137911468744278, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.9201, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.16510832383124288, | |
| "grad_norm": 0.19734999537467957, | |
| "learning_rate": 7.471606041430723e-05, | |
| "loss": 1.1128, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.1660205245153934, | |
| "grad_norm": 0.21062205731868744, | |
| "learning_rate": 7.443106207484776e-05, | |
| "loss": 1.0556, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.1669327251995439, | |
| "grad_norm": 0.24208199977874756, | |
| "learning_rate": 7.414501719000187e-05, | |
| "loss": 1.1084, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.1678449258836944, | |
| "grad_norm": 0.3003711998462677, | |
| "learning_rate": 7.385793801298042e-05, | |
| "loss": 1.2188, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.16875712656784492, | |
| "grad_norm": 0.12483101338148117, | |
| "learning_rate": 7.35698368412999e-05, | |
| "loss": 1.242, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.16966932725199543, | |
| "grad_norm": 0.10687603801488876, | |
| "learning_rate": 7.328072601625557e-05, | |
| "loss": 1.0969, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.17058152793614595, | |
| "grad_norm": 0.10687136650085449, | |
| "learning_rate": 7.2990617922393e-05, | |
| "loss": 1.2887, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.17149372862029647, | |
| "grad_norm": 0.12314095348119736, | |
| "learning_rate": 7.269952498697734e-05, | |
| "loss": 1.2609, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.17240592930444698, | |
| "grad_norm": 0.1314985156059265, | |
| "learning_rate": 7.240745967946113e-05, | |
| "loss": 1.2694, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.1733181299885975, | |
| "grad_norm": 0.12387102842330933, | |
| "learning_rate": 7.211443451095007e-05, | |
| "loss": 1.2707, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.174230330672748, | |
| "grad_norm": 0.16482983529567719, | |
| "learning_rate": 7.18204620336671e-05, | |
| "loss": 1.1994, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.17514253135689853, | |
| "grad_norm": 0.16161847114562988, | |
| "learning_rate": 7.152555484041476e-05, | |
| "loss": 1.3605, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.17605473204104904, | |
| "grad_norm": 0.16728828847408295, | |
| "learning_rate": 7.122972556403567e-05, | |
| "loss": 1.1442, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.17696693272519953, | |
| "grad_norm": 0.18906857073307037, | |
| "learning_rate": 7.09329868768714e-05, | |
| "loss": 1.1403, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.17787913340935005, | |
| "grad_norm": 0.21785447001457214, | |
| "learning_rate": 7.063535149021973e-05, | |
| "loss": 1.4139, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.17879133409350056, | |
| "grad_norm": 0.21159727871418, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 1.4149, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.17970353477765108, | |
| "grad_norm": 0.2096451222896576, | |
| "learning_rate": 7.003744165515705e-05, | |
| "loss": 1.1846, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.1806157354618016, | |
| "grad_norm": 0.30022695660591125, | |
| "learning_rate": 6.973719281921335e-05, | |
| "loss": 1.5039, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.1815279361459521, | |
| "grad_norm": 0.39245763421058655, | |
| "learning_rate": 6.943609850761979e-05, | |
| "loss": 1.4439, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.18244013683010263, | |
| "grad_norm": 0.7122502326965332, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 1.7626, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.18244013683010263, | |
| "eval_loss": 1.1971056461334229, | |
| "eval_runtime": 113.4472, | |
| "eval_samples_per_second": 16.272, | |
| "eval_steps_per_second": 4.072, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.18335233751425314, | |
| "grad_norm": 0.11290186643600464, | |
| "learning_rate": 6.883142508466054e-05, | |
| "loss": 0.8787, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.18426453819840366, | |
| "grad_norm": 0.12823708355426788, | |
| "learning_rate": 6.852787187549182e-05, | |
| "loss": 0.8172, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.18517673888255415, | |
| "grad_norm": 0.1377316415309906, | |
| "learning_rate": 6.82235249939575e-05, | |
| "loss": 1.0454, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.18608893956670466, | |
| "grad_norm": 0.1476827710866928, | |
| "learning_rate": 6.7918397477265e-05, | |
| "loss": 1.2703, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.18700114025085518, | |
| "grad_norm": 0.14928844571113586, | |
| "learning_rate": 6.761250239606169e-05, | |
| "loss": 1.3711, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.1879133409350057, | |
| "grad_norm": 0.16220971941947937, | |
| "learning_rate": 6.730585285387465e-05, | |
| "loss": 1.1761, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.1888255416191562, | |
| "grad_norm": 0.15594278275966644, | |
| "learning_rate": 6.699846198654971e-05, | |
| "loss": 1.159, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.18973774230330673, | |
| "grad_norm": 0.22083112597465515, | |
| "learning_rate": 6.669034296168855e-05, | |
| "loss": 1.3703, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.19064994298745724, | |
| "grad_norm": 0.2218848466873169, | |
| "learning_rate": 6.638150897808468e-05, | |
| "loss": 1.0984, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.19156214367160776, | |
| "grad_norm": 0.3727222979068756, | |
| "learning_rate": 6.607197326515808e-05, | |
| "loss": 1.4564, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.19247434435575828, | |
| "grad_norm": 0.14719393849372864, | |
| "learning_rate": 6.57617490823885e-05, | |
| "loss": 1.0078, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.1933865450399088, | |
| "grad_norm": 0.13103777170181274, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 0.9745, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.19429874572405928, | |
| "grad_norm": 0.13376685976982117, | |
| "learning_rate": 6.513928849212873e-05, | |
| "loss": 0.9558, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.1952109464082098, | |
| "grad_norm": 0.1713114231824875, | |
| "learning_rate": 6.482707874877854e-05, | |
| "loss": 1.0795, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.1961231470923603, | |
| "grad_norm": 0.14341691136360168, | |
| "learning_rate": 6.451423386272312e-05, | |
| "loss": 1.0165, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.19703534777651083, | |
| "grad_norm": 0.14939330518245697, | |
| "learning_rate": 6.420076723519614e-05, | |
| "loss": 0.9657, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.19794754846066134, | |
| "grad_norm": 0.1301591396331787, | |
| "learning_rate": 6.388669229406462e-05, | |
| "loss": 0.757, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.19885974914481186, | |
| "grad_norm": 0.160866379737854, | |
| "learning_rate": 6.357202249325371e-05, | |
| "loss": 0.965, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.19977194982896238, | |
| "grad_norm": 0.21365958452224731, | |
| "learning_rate": 6.32567713121704e-05, | |
| "loss": 1.1012, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.2006841505131129, | |
| "grad_norm": 0.16345688700675964, | |
| "learning_rate": 6.294095225512603e-05, | |
| "loss": 1.0251, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2015963511972634, | |
| "grad_norm": 0.1504313200712204, | |
| "learning_rate": 6.26245788507579e-05, | |
| "loss": 1.0471, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.20250855188141392, | |
| "grad_norm": 0.15645664930343628, | |
| "learning_rate": 6.230766465144967e-05, | |
| "loss": 0.932, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.2034207525655644, | |
| "grad_norm": 0.16272133588790894, | |
| "learning_rate": 6.199022323275083e-05, | |
| "loss": 0.9476, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.20433295324971493, | |
| "grad_norm": 0.15730346739292145, | |
| "learning_rate": 6.167226819279528e-05, | |
| "loss": 0.9678, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.20524515393386544, | |
| "grad_norm": 0.16858989000320435, | |
| "learning_rate": 6.135381315171867e-05, | |
| "loss": 1.0503, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.20615735461801596, | |
| "grad_norm": 0.18682138621807098, | |
| "learning_rate": 6.103487175107507e-05, | |
| "loss": 1.0477, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.20706955530216647, | |
| "grad_norm": 0.17781555652618408, | |
| "learning_rate": 6.071545765325254e-05, | |
| "loss": 0.9751, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.207981755986317, | |
| "grad_norm": 0.22681044042110443, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 1.1294, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.2088939566704675, | |
| "grad_norm": 0.20905116200447083, | |
| "learning_rate": 6.007526611628086e-05, | |
| "loss": 1.0836, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.20980615735461802, | |
| "grad_norm": 0.17305400967597961, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 1.0323, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.21071835803876854, | |
| "grad_norm": 0.1822892129421234, | |
| "learning_rate": 5.9433348234327765e-05, | |
| "loss": 0.9336, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.21163055872291905, | |
| "grad_norm": 0.2400861382484436, | |
| "learning_rate": 5.911177627460739e-05, | |
| "loss": 1.0643, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.21254275940706954, | |
| "grad_norm": 0.19734802842140198, | |
| "learning_rate": 5.8789813996717736e-05, | |
| "loss": 1.1696, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.21345496009122006, | |
| "grad_norm": 0.24006140232086182, | |
| "learning_rate": 5.8467475192451226e-05, | |
| "loss": 1.3169, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.21436716077537057, | |
| "grad_norm": 0.1660918891429901, | |
| "learning_rate": 5.814477366972945e-05, | |
| "loss": 1.2928, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2152793614595211, | |
| "grad_norm": 0.10857274383306503, | |
| "learning_rate": 5.782172325201155e-05, | |
| "loss": 1.1304, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.2161915621436716, | |
| "grad_norm": 0.09754611551761627, | |
| "learning_rate": 5.749833777770225e-05, | |
| "loss": 1.0516, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.21710376282782212, | |
| "grad_norm": 0.10842633992433548, | |
| "learning_rate": 5.717463109955896e-05, | |
| "loss": 1.2546, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.21801596351197264, | |
| "grad_norm": 0.11293677240610123, | |
| "learning_rate": 5.685061708409841e-05, | |
| "loss": 1.2007, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.21892816419612315, | |
| "grad_norm": 0.11466323584318161, | |
| "learning_rate": 5.6526309611002594e-05, | |
| "loss": 1.2429, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.21984036488027367, | |
| "grad_norm": 0.12489130347967148, | |
| "learning_rate": 5.6201722572524275e-05, | |
| "loss": 1.2082, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.22075256556442419, | |
| "grad_norm": 0.14694885909557343, | |
| "learning_rate": 5.587686987289189e-05, | |
| "loss": 1.2742, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.22166476624857467, | |
| "grad_norm": 0.14593251049518585, | |
| "learning_rate": 5.5551765427713884e-05, | |
| "loss": 1.3337, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.2225769669327252, | |
| "grad_norm": 0.15233193337917328, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 1.1762, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.2234891676168757, | |
| "grad_norm": 0.16144950687885284, | |
| "learning_rate": 5.490085701647805e-05, | |
| "loss": 1.2968, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.22440136830102622, | |
| "grad_norm": 0.16887733340263367, | |
| "learning_rate": 5.457508093317013e-05, | |
| "loss": 1.312, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.22531356898517674, | |
| "grad_norm": 0.24274863302707672, | |
| "learning_rate": 5.4249108868622086e-05, | |
| "loss": 1.4066, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.22622576966932725, | |
| "grad_norm": 0.24290831387043, | |
| "learning_rate": 5.392295478639225e-05, | |
| "loss": 1.3587, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.22713797035347777, | |
| "grad_norm": 0.35980361700057983, | |
| "learning_rate": 5.359663265783598e-05, | |
| "loss": 1.545, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.22805017103762829, | |
| "grad_norm": 1.2798198461532593, | |
| "learning_rate": 5.327015646150716e-05, | |
| "loss": 1.8089, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2289623717217788, | |
| "grad_norm": 0.15012221038341522, | |
| "learning_rate": 5.294354018255945e-05, | |
| "loss": 1.1607, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.22987457240592932, | |
| "grad_norm": 0.14231781661510468, | |
| "learning_rate": 5.26167978121472e-05, | |
| "loss": 1.0741, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.2307867730900798, | |
| "grad_norm": 0.1374557614326477, | |
| "learning_rate": 5.228994334682604e-05, | |
| "loss": 1.1327, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.23169897377423032, | |
| "grad_norm": 0.16241897642612457, | |
| "learning_rate": 5.196299078795344e-05, | |
| "loss": 1.1431, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.23261117445838084, | |
| "grad_norm": 0.19717371463775635, | |
| "learning_rate": 5.1635954141088813e-05, | |
| "loss": 1.2522, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.23352337514253135, | |
| "grad_norm": 0.16350847482681274, | |
| "learning_rate": 5.1308847415393666e-05, | |
| "loss": 1.1139, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.23443557582668187, | |
| "grad_norm": 0.2092381715774536, | |
| "learning_rate": 5.0981684623031415e-05, | |
| "loss": 1.4112, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.23534777651083238, | |
| "grad_norm": 0.1716463714838028, | |
| "learning_rate": 5.0654479778567223e-05, | |
| "loss": 1.0679, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.2362599771949829, | |
| "grad_norm": 0.2410898059606552, | |
| "learning_rate": 5.0327246898367597e-05, | |
| "loss": 1.5, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.23717217787913342, | |
| "grad_norm": 0.21982981264591217, | |
| "learning_rate": 5e-05, | |
| "loss": 1.1005, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.23808437856328393, | |
| "grad_norm": 0.3865964412689209, | |
| "learning_rate": 4.9672753101632415e-05, | |
| "loss": 1.3476, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.23899657924743445, | |
| "grad_norm": 0.2032884657382965, | |
| "learning_rate": 4.934552022143279e-05, | |
| "loss": 0.9498, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.23990877993158494, | |
| "grad_norm": 0.1798202246427536, | |
| "learning_rate": 4.901831537696859e-05, | |
| "loss": 1.153, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.24082098061573545, | |
| "grad_norm": 0.16176198422908783, | |
| "learning_rate": 4.869115258460635e-05, | |
| "loss": 0.9052, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.24173318129988597, | |
| "grad_norm": 0.18311551213264465, | |
| "learning_rate": 4.83640458589112e-05, | |
| "loss": 1.1653, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.24264538198403648, | |
| "grad_norm": 0.1691948026418686, | |
| "learning_rate": 4.8037009212046586e-05, | |
| "loss": 0.9367, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.243557582668187, | |
| "grad_norm": 0.144092857837677, | |
| "learning_rate": 4.7710056653173976e-05, | |
| "loss": 0.9161, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.24446978335233752, | |
| "grad_norm": 0.16229337453842163, | |
| "learning_rate": 4.738320218785281e-05, | |
| "loss": 0.9686, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.24538198403648803, | |
| "grad_norm": 0.14550195634365082, | |
| "learning_rate": 4.7056459817440544e-05, | |
| "loss": 0.7886, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.24629418472063855, | |
| "grad_norm": 0.15839822590351105, | |
| "learning_rate": 4.6729843538492847e-05, | |
| "loss": 1.0488, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.24720638540478906, | |
| "grad_norm": 0.16177891194820404, | |
| "learning_rate": 4.640336734216403e-05, | |
| "loss": 1.0537, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.24811858608893958, | |
| "grad_norm": 0.15026786923408508, | |
| "learning_rate": 4.607704521360776e-05, | |
| "loss": 0.9876, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.24903078677309007, | |
| "grad_norm": 0.16026102006435394, | |
| "learning_rate": 4.575089113137792e-05, | |
| "loss": 0.9325, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.24994298745724058, | |
| "grad_norm": 0.14951908588409424, | |
| "learning_rate": 4.542491906682989e-05, | |
| "loss": 0.9553, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.2508551881413911, | |
| "grad_norm": 0.1985773742198944, | |
| "learning_rate": 4.509914298352197e-05, | |
| "loss": 1.078, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.2517673888255416, | |
| "grad_norm": 0.15566202998161316, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 0.9121, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.25267958950969216, | |
| "grad_norm": 0.15593530237674713, | |
| "learning_rate": 4.444823457228612e-05, | |
| "loss": 0.8146, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.25359179019384265, | |
| "grad_norm": 0.16978751122951508, | |
| "learning_rate": 4.412313012710813e-05, | |
| "loss": 0.8579, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.25450399087799314, | |
| "grad_norm": 0.22708088159561157, | |
| "learning_rate": 4.379827742747575e-05, | |
| "loss": 0.9367, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.2554161915621437, | |
| "grad_norm": 0.18289873003959656, | |
| "learning_rate": 4.347369038899744e-05, | |
| "loss": 1.1297, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.25632839224629417, | |
| "grad_norm": 0.19509181380271912, | |
| "learning_rate": 4.3149382915901606e-05, | |
| "loss": 0.9119, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.2572405929304447, | |
| "grad_norm": 0.19992905855178833, | |
| "learning_rate": 4.282536890044104e-05, | |
| "loss": 1.1763, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.2581527936145952, | |
| "grad_norm": 0.25321832299232483, | |
| "learning_rate": 4.250166222229774e-05, | |
| "loss": 1.1429, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.25906499429874574, | |
| "grad_norm": 0.26218289136886597, | |
| "learning_rate": 4.2178276747988446e-05, | |
| "loss": 1.2399, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.25997719498289623, | |
| "grad_norm": 0.2333952635526657, | |
| "learning_rate": 4.185522633027057e-05, | |
| "loss": 0.8556, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.2608893956670468, | |
| "grad_norm": 0.1518288403749466, | |
| "learning_rate": 4.153252480754877e-05, | |
| "loss": 1.2989, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.26180159635119726, | |
| "grad_norm": 0.1034812331199646, | |
| "learning_rate": 4.1210186003282275e-05, | |
| "loss": 1.168, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.26271379703534775, | |
| "grad_norm": 0.10131867974996567, | |
| "learning_rate": 4.088822372539263e-05, | |
| "loss": 1.0937, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.2636259977194983, | |
| "grad_norm": 0.11592473089694977, | |
| "learning_rate": 4.0566651765672246e-05, | |
| "loss": 1.1637, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.2645381984036488, | |
| "grad_norm": 0.1236988753080368, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 1.2724, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2654503990877993, | |
| "grad_norm": 0.13103049993515015, | |
| "learning_rate": 3.992473388371915e-05, | |
| "loss": 1.1731, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.2663625997719498, | |
| "grad_norm": 0.13575758039951324, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 1.256, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.26727480045610036, | |
| "grad_norm": 0.13911360502243042, | |
| "learning_rate": 3.928454234674747e-05, | |
| "loss": 1.329, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.26818700114025085, | |
| "grad_norm": 0.1553173065185547, | |
| "learning_rate": 3.896512824892495e-05, | |
| "loss": 1.2601, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.2690992018244014, | |
| "grad_norm": 0.22820927202701569, | |
| "learning_rate": 3.864618684828134e-05, | |
| "loss": 1.2263, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.2700114025085519, | |
| "grad_norm": 0.251654714345932, | |
| "learning_rate": 3.832773180720475e-05, | |
| "loss": 1.3326, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.27092360319270237, | |
| "grad_norm": 0.28499627113342285, | |
| "learning_rate": 3.800977676724919e-05, | |
| "loss": 1.673, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.2718358038768529, | |
| "grad_norm": 0.2628072500228882, | |
| "learning_rate": 3.769233534855035e-05, | |
| "loss": 1.4846, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.2727480045610034, | |
| "grad_norm": 0.3170837461948395, | |
| "learning_rate": 3.73754211492421e-05, | |
| "loss": 1.3636, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.27366020524515394, | |
| "grad_norm": 0.8047922253608704, | |
| "learning_rate": 3.705904774487396e-05, | |
| "loss": 1.1731, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.27366020524515394, | |
| "eval_loss": 1.1972854137420654, | |
| "eval_runtime": 113.2028, | |
| "eval_samples_per_second": 16.307, | |
| "eval_steps_per_second": 4.081, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.27457240592930443, | |
| "grad_norm": 0.10662523657083511, | |
| "learning_rate": 3.6743228687829595e-05, | |
| "loss": 1.0097, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.275484606613455, | |
| "grad_norm": 0.12424086779356003, | |
| "learning_rate": 3.642797750674629e-05, | |
| "loss": 1.0579, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.27639680729760546, | |
| "grad_norm": 0.1385432481765747, | |
| "learning_rate": 3.6113307705935396e-05, | |
| "loss": 1.344, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.277309007981756, | |
| "grad_norm": 0.16241592168807983, | |
| "learning_rate": 3.579923276480387e-05, | |
| "loss": 1.4409, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.2782212086659065, | |
| "grad_norm": 0.15419059991836548, | |
| "learning_rate": 3.5485766137276894e-05, | |
| "loss": 1.3336, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.27913340935005704, | |
| "grad_norm": 0.18154077231884003, | |
| "learning_rate": 3.5172921251221455e-05, | |
| "loss": 1.3504, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.2800456100342075, | |
| "grad_norm": 0.1785464584827423, | |
| "learning_rate": 3.486071150787128e-05, | |
| "loss": 1.2279, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.280957810718358, | |
| "grad_norm": 0.20182940363883972, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 1.4907, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.28187001140250856, | |
| "grad_norm": 0.22802571952342987, | |
| "learning_rate": 3.423825091761153e-05, | |
| "loss": 1.4304, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.28278221208665905, | |
| "grad_norm": 0.41071566939353943, | |
| "learning_rate": 3.392802673484193e-05, | |
| "loss": 1.5391, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2836944127708096, | |
| "grad_norm": 0.2370689958333969, | |
| "learning_rate": 3.361849102191533e-05, | |
| "loss": 1.1847, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.2846066134549601, | |
| "grad_norm": 0.18752293288707733, | |
| "learning_rate": 3.330965703831146e-05, | |
| "loss": 1.1175, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.2855188141391106, | |
| "grad_norm": 0.1348540335893631, | |
| "learning_rate": 3.300153801345028e-05, | |
| "loss": 0.9726, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.2864310148232611, | |
| "grad_norm": 0.14316678047180176, | |
| "learning_rate": 3.2694147146125345e-05, | |
| "loss": 1.0138, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.28734321550741165, | |
| "grad_norm": 0.13163286447525024, | |
| "learning_rate": 3.2387497603938326e-05, | |
| "loss": 0.8848, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.28825541619156214, | |
| "grad_norm": 0.19286373257637024, | |
| "learning_rate": 3.2081602522734986e-05, | |
| "loss": 1.105, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.28916761687571263, | |
| "grad_norm": 0.12224922329187393, | |
| "learning_rate": 3.177647500604252e-05, | |
| "loss": 0.6901, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.2900798175598632, | |
| "grad_norm": 0.1634187549352646, | |
| "learning_rate": 3.147212812450819e-05, | |
| "loss": 0.9563, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.29099201824401366, | |
| "grad_norm": 0.15377987921237946, | |
| "learning_rate": 3.116857491533947e-05, | |
| "loss": 0.8784, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.2919042189281642, | |
| "grad_norm": 0.13165193796157837, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.8488, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.2928164196123147, | |
| "grad_norm": 0.16833990812301636, | |
| "learning_rate": 3.056390149238022e-05, | |
| "loss": 1.1677, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.29372862029646524, | |
| "grad_norm": 0.15428762137889862, | |
| "learning_rate": 3.0262807180786647e-05, | |
| "loss": 0.934, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.2946408209806157, | |
| "grad_norm": 0.15865963697433472, | |
| "learning_rate": 2.996255834484296e-05, | |
| "loss": 1.0805, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.29555302166476627, | |
| "grad_norm": 0.14876452088356018, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 0.9789, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.29646522234891676, | |
| "grad_norm": 0.1823773831129074, | |
| "learning_rate": 2.936464850978027e-05, | |
| "loss": 1.0265, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.2973774230330673, | |
| "grad_norm": 0.17371368408203125, | |
| "learning_rate": 2.9067013123128613e-05, | |
| "loss": 0.9221, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.2982896237172178, | |
| "grad_norm": 0.19091551005840302, | |
| "learning_rate": 2.8770274435964355e-05, | |
| "loss": 0.9469, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.2992018244013683, | |
| "grad_norm": 0.18162409961223602, | |
| "learning_rate": 2.8474445159585235e-05, | |
| "loss": 1.0272, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.3001140250855188, | |
| "grad_norm": 0.16869843006134033, | |
| "learning_rate": 2.8179537966332887e-05, | |
| "loss": 1.1221, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.3010262257696693, | |
| "grad_norm": 0.1988871544599533, | |
| "learning_rate": 2.7885565489049946e-05, | |
| "loss": 0.9206, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.30193842645381985, | |
| "grad_norm": 0.17574124038219452, | |
| "learning_rate": 2.759254032053888e-05, | |
| "loss": 1.067, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.30285062713797034, | |
| "grad_norm": 0.20466162264347076, | |
| "learning_rate": 2.7300475013022663e-05, | |
| "loss": 1.0049, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.3037628278221209, | |
| "grad_norm": 0.30088433623313904, | |
| "learning_rate": 2.700938207760701e-05, | |
| "loss": 1.1633, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.3046750285062714, | |
| "grad_norm": 0.31139159202575684, | |
| "learning_rate": 2.671927398374443e-05, | |
| "loss": 1.1374, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.3055872291904219, | |
| "grad_norm": 0.20724207162857056, | |
| "learning_rate": 2.6430163158700115e-05, | |
| "loss": 1.379, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3064994298745724, | |
| "grad_norm": 0.12832605838775635, | |
| "learning_rate": 2.6142061987019577e-05, | |
| "loss": 1.1881, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.3074116305587229, | |
| "grad_norm": 0.10497277975082397, | |
| "learning_rate": 2.5854982809998153e-05, | |
| "loss": 1.0799, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.30832383124287344, | |
| "grad_norm": 0.1079997569322586, | |
| "learning_rate": 2.556893792515227e-05, | |
| "loss": 1.1005, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.3092360319270239, | |
| "grad_norm": 0.10871066898107529, | |
| "learning_rate": 2.5283939585692783e-05, | |
| "loss": 1.0749, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.31014823261117447, | |
| "grad_norm": 0.11225587129592896, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 1.2332, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.31106043329532496, | |
| "grad_norm": 0.12466397881507874, | |
| "learning_rate": 2.471713133110078e-05, | |
| "loss": 1.2856, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.3119726339794755, | |
| "grad_norm": 0.14498235285282135, | |
| "learning_rate": 2.4435345696147403e-05, | |
| "loss": 1.243, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.312884834663626, | |
| "grad_norm": 0.1273733228445053, | |
| "learning_rate": 2.4154655165898627e-05, | |
| "loss": 1.199, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.31379703534777653, | |
| "grad_norm": 0.17925268411636353, | |
| "learning_rate": 2.3875071764202563e-05, | |
| "loss": 1.1705, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.314709236031927, | |
| "grad_norm": 0.18367154896259308, | |
| "learning_rate": 2.3596607467481603e-05, | |
| "loss": 1.412, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.31562143671607756, | |
| "grad_norm": 0.18743032217025757, | |
| "learning_rate": 2.3319274204219428e-05, | |
| "loss": 1.1635, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.31653363740022805, | |
| "grad_norm": 0.197452574968338, | |
| "learning_rate": 2.3043083854449988e-05, | |
| "loss": 1.3102, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.31744583808437854, | |
| "grad_norm": 0.26595744490623474, | |
| "learning_rate": 2.2768048249248648e-05, | |
| "loss": 1.279, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.3183580387685291, | |
| "grad_norm": 0.4088784158229828, | |
| "learning_rate": 2.2494179170225333e-05, | |
| "loss": 1.5259, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.31927023945267957, | |
| "grad_norm": 0.5466248393058777, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 1.5536, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3201824401368301, | |
| "grad_norm": 0.13032464683055878, | |
| "learning_rate": 2.194998746679952e-05, | |
| "loss": 1.1141, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.3210946408209806, | |
| "grad_norm": 0.1586756855249405, | |
| "learning_rate": 2.167968815375837e-05, | |
| "loss": 1.1245, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.32200684150513115, | |
| "grad_norm": 0.1475306749343872, | |
| "learning_rate": 2.1410601988619394e-05, | |
| "loss": 1.0502, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.32291904218928164, | |
| "grad_norm": 0.16401484608650208, | |
| "learning_rate": 2.1142740498138324e-05, | |
| "loss": 1.2346, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.3238312428734322, | |
| "grad_norm": 0.17974554002285004, | |
| "learning_rate": 2.08761151566099e-05, | |
| "loss": 1.1683, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.32474344355758267, | |
| "grad_norm": 0.17584644258022308, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 1.4097, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.32565564424173316, | |
| "grad_norm": 0.17970524728298187, | |
| "learning_rate": 2.034661855233815e-05, | |
| "loss": 1.2119, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.3265678449258837, | |
| "grad_norm": 0.1974216103553772, | |
| "learning_rate": 2.008376997146705e-05, | |
| "loss": 1.1943, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.3274800456100342, | |
| "grad_norm": 0.1703203320503235, | |
| "learning_rate": 1.982220290232143e-05, | |
| "loss": 0.9671, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.32839224629418473, | |
| "grad_norm": 0.2753969132900238, | |
| "learning_rate": 1.9561928549563968e-05, | |
| "loss": 1.3897, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3293044469783352, | |
| "grad_norm": 0.3241826593875885, | |
| "learning_rate": 1.9302958062481673e-05, | |
| "loss": 1.1562, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.33021664766248576, | |
| "grad_norm": 0.2132183462381363, | |
| "learning_rate": 1.9045302534508297e-05, | |
| "loss": 0.7516, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.33112884834663625, | |
| "grad_norm": 0.14898429811000824, | |
| "learning_rate": 1.8788973002749112e-05, | |
| "loss": 1.1512, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.3320410490307868, | |
| "grad_norm": 0.1288880854845047, | |
| "learning_rate": 1.8533980447508137e-05, | |
| "loss": 0.8832, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.3329532497149373, | |
| "grad_norm": 0.15893982350826263, | |
| "learning_rate": 1.8280335791817733e-05, | |
| "loss": 1.1057, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.3338654503990878, | |
| "grad_norm": 0.15993352234363556, | |
| "learning_rate": 1.8028049900970767e-05, | |
| "loss": 0.9927, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.3347776510832383, | |
| "grad_norm": 0.14739666879177094, | |
| "learning_rate": 1.777713358205514e-05, | |
| "loss": 0.8977, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.3356898517673888, | |
| "grad_norm": 0.1271563321352005, | |
| "learning_rate": 1.7527597583490822e-05, | |
| "loss": 0.8456, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.33660205245153935, | |
| "grad_norm": 0.1543639749288559, | |
| "learning_rate": 1.7279452594569483e-05, | |
| "loss": 1.0841, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.33751425313568983, | |
| "grad_norm": 0.13747648894786835, | |
| "learning_rate": 1.703270924499656e-05, | |
| "loss": 0.8802, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3384264538198404, | |
| "grad_norm": 0.1359814554452896, | |
| "learning_rate": 1.678737810443593e-05, | |
| "loss": 0.8192, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.33933865450399087, | |
| "grad_norm": 0.16268639266490936, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 0.8709, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.3402508551881414, | |
| "grad_norm": 0.1843179315328598, | |
| "learning_rate": 1.6300994426085103e-05, | |
| "loss": 0.8591, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.3411630558722919, | |
| "grad_norm": 0.19724664092063904, | |
| "learning_rate": 1.605996272335291e-05, | |
| "loss": 1.1374, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.34207525655644244, | |
| "grad_norm": 0.2107182741165161, | |
| "learning_rate": 1.5820384898856434e-05, | |
| "loss": 1.0738, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.34298745724059293, | |
| "grad_norm": 0.18194270133972168, | |
| "learning_rate": 1.5582271215312294e-05, | |
| "loss": 1.1537, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.3438996579247434, | |
| "grad_norm": 0.14367586374282837, | |
| "learning_rate": 1.5345631872718214e-05, | |
| "loss": 0.981, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.34481185860889396, | |
| "grad_norm": 0.1595829576253891, | |
| "learning_rate": 1.5110477007916001e-05, | |
| "loss": 0.9495, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.34572405929304445, | |
| "grad_norm": 0.16781240701675415, | |
| "learning_rate": 1.4876816694157419e-05, | |
| "loss": 0.8829, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.346636259977195, | |
| "grad_norm": 0.18402887880802155, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 1.0337, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3475484606613455, | |
| "grad_norm": 0.21096819639205933, | |
| "learning_rate": 1.4414019692241437e-05, | |
| "loss": 1.235, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.348460661345496, | |
| "grad_norm": 0.20575056970119476, | |
| "learning_rate": 1.4184902828767287e-05, | |
| "loss": 1.1308, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.3493728620296465, | |
| "grad_norm": 0.23069782555103302, | |
| "learning_rate": 1.3957320164854059e-05, | |
| "loss": 0.7216, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.35028506271379706, | |
| "grad_norm": 0.2631959617137909, | |
| "learning_rate": 1.373128144938563e-05, | |
| "loss": 1.051, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.35119726339794755, | |
| "grad_norm": 0.2909131646156311, | |
| "learning_rate": 1.3506796365108232e-05, | |
| "loss": 1.4182, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.3521094640820981, | |
| "grad_norm": 0.1559658795595169, | |
| "learning_rate": 1.3283874528215733e-05, | |
| "loss": 1.2366, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.3530216647662486, | |
| "grad_norm": 0.09982962161302567, | |
| "learning_rate": 1.3062525487937699e-05, | |
| "loss": 1.1475, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.35393386545039907, | |
| "grad_norm": 0.10416360944509506, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 1.0829, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.3548460661345496, | |
| "grad_norm": 0.10417391359806061, | |
| "learning_rate": 1.2624583656870154e-05, | |
| "loss": 1.0939, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.3557582668187001, | |
| "grad_norm": 0.12138167768716812, | |
| "learning_rate": 1.2408009626051137e-05, | |
| "loss": 1.3434, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.35667046750285064, | |
| "grad_norm": 0.13164609670639038, | |
| "learning_rate": 1.2193045910983863e-05, | |
| "loss": 1.2702, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.35758266818700113, | |
| "grad_norm": 0.1307135969400406, | |
| "learning_rate": 1.1979701719998453e-05, | |
| "loss": 1.2093, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.3584948688711517, | |
| "grad_norm": 0.1524544358253479, | |
| "learning_rate": 1.1767986192049984e-05, | |
| "loss": 1.179, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.35940706955530216, | |
| "grad_norm": 0.16346591711044312, | |
| "learning_rate": 1.1557908396327028e-05, | |
| "loss": 1.2838, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.3603192702394527, | |
| "grad_norm": 0.18474753201007843, | |
| "learning_rate": 1.134947733186315e-05, | |
| "loss": 1.2491, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.3612314709236032, | |
| "grad_norm": 0.19478006660938263, | |
| "learning_rate": 1.1142701927151456e-05, | |
| "loss": 1.4218, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.3621436716077537, | |
| "grad_norm": 0.22533515095710754, | |
| "learning_rate": 1.0937591039762085e-05, | |
| "loss": 1.4451, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.3630558722919042, | |
| "grad_norm": 0.27334368228912354, | |
| "learning_rate": 1.0734153455962765e-05, | |
| "loss": 1.4695, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.3639680729760547, | |
| "grad_norm": 0.31324154138565063, | |
| "learning_rate": 1.0532397890342505e-05, | |
| "loss": 1.4504, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.36488027366020526, | |
| "grad_norm": 0.9359192848205566, | |
| "learning_rate": 1.0332332985438248e-05, | |
| "loss": 1.5549, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.36488027366020526, | |
| "eval_loss": 1.1766536235809326, | |
| "eval_runtime": 112.3781, | |
| "eval_samples_per_second": 16.427, | |
| "eval_steps_per_second": 4.111, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.36579247434435574, | |
| "grad_norm": 0.12196308374404907, | |
| "learning_rate": 1.013396731136465e-05, | |
| "loss": 1.0011, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 0.3667046750285063, | |
| "grad_norm": 0.13182106614112854, | |
| "learning_rate": 9.937309365446973e-06, | |
| "loss": 1.1752, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 0.3676168757126568, | |
| "grad_norm": 0.14486649632453918, | |
| "learning_rate": 9.742367571857091e-06, | |
| "loss": 1.1526, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 0.3685290763968073, | |
| "grad_norm": 0.14698417484760284, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 1.2639, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 0.3694412770809578, | |
| "grad_norm": 0.16532011330127716, | |
| "learning_rate": 9.357665770419244e-06, | |
| "loss": 1.3237, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.3703534777651083, | |
| "grad_norm": 0.17021730542182922, | |
| "learning_rate": 9.167922241916055e-06, | |
| "loss": 1.3305, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 0.37126567844925884, | |
| "grad_norm": 0.1707400232553482, | |
| "learning_rate": 8.97992782372432e-06, | |
| "loss": 1.1334, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 0.37217787913340933, | |
| "grad_norm": 0.1890312284231186, | |
| "learning_rate": 8.793690568899216e-06, | |
| "loss": 1.2203, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 0.37309007981755987, | |
| "grad_norm": 0.2785188555717468, | |
| "learning_rate": 8.609218455224893e-06, | |
| "loss": 1.3662, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 0.37400228050171036, | |
| "grad_norm": 0.3217642307281494, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 1.5266, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.3749144811858609, | |
| "grad_norm": 0.637722909450531, | |
| "learning_rate": 8.245601184062852e-06, | |
| "loss": 1.7853, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 0.3758266818700114, | |
| "grad_norm": 0.15114322304725647, | |
| "learning_rate": 8.066471602728803e-06, | |
| "loss": 0.8234, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 0.37673888255416194, | |
| "grad_norm": 0.15969091653823853, | |
| "learning_rate": 7.889138314185678e-06, | |
| "loss": 1.0191, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 0.3776510832383124, | |
| "grad_norm": 0.16418735682964325, | |
| "learning_rate": 7.71360891480134e-06, | |
| "loss": 1.1389, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 0.37856328392246297, | |
| "grad_norm": 0.15018145740032196, | |
| "learning_rate": 7.539890923671062e-06, | |
| "loss": 1.0788, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.37947548460661346, | |
| "grad_norm": 0.14109903573989868, | |
| "learning_rate": 7.367991782295391e-06, | |
| "loss": 0.9719, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 0.38038768529076394, | |
| "grad_norm": 0.1540416181087494, | |
| "learning_rate": 7.197918854261432e-06, | |
| "loss": 1.0884, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 0.3812998859749145, | |
| "grad_norm": 0.14207249879837036, | |
| "learning_rate": 7.029679424927365e-06, | |
| "loss": 1.1356, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 0.382212086659065, | |
| "grad_norm": 0.14175021648406982, | |
| "learning_rate": 6.863280701110408e-06, | |
| "loss": 0.8289, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 0.3831242873432155, | |
| "grad_norm": 0.1370278298854828, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 0.9297, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.384036488027366, | |
| "grad_norm": 0.2763122618198395, | |
| "learning_rate": 6.536033802742813e-06, | |
| "loss": 1.2414, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 0.38494868871151655, | |
| "grad_norm": 0.15884122252464294, | |
| "learning_rate": 6.375199646360142e-06, | |
| "loss": 1.0608, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 0.38586088939566704, | |
| "grad_norm": 0.14970403909683228, | |
| "learning_rate": 6.216234231230012e-06, | |
| "loss": 0.7528, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 0.3867730900798176, | |
| "grad_norm": 0.15861846506595612, | |
| "learning_rate": 6.059144366901736e-06, | |
| "loss": 0.924, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 0.38768529076396807, | |
| "grad_norm": 0.18967342376708984, | |
| "learning_rate": 5.903936782582253e-06, | |
| "loss": 1.0738, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.38859749144811856, | |
| "grad_norm": 0.1890667825937271, | |
| "learning_rate": 5.750618126847912e-06, | |
| "loss": 1.0767, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 0.3895096921322691, | |
| "grad_norm": 0.18360556662082672, | |
| "learning_rate": 5.599194967359639e-06, | |
| "loss": 1.137, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 0.3904218928164196, | |
| "grad_norm": 0.2094428986310959, | |
| "learning_rate": 5.449673790581611e-06, | |
| "loss": 1.0618, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 0.39133409350057013, | |
| "grad_norm": 0.17591425776481628, | |
| "learning_rate": 5.302061001503394e-06, | |
| "loss": 0.9315, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 0.3922462941847206, | |
| "grad_norm": 0.19204486906528473, | |
| "learning_rate": 5.156362923365588e-06, | |
| "loss": 1.1151, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.39315849486887117, | |
| "grad_norm": 0.19108223915100098, | |
| "learning_rate": 5.012585797388936e-06, | |
| "loss": 0.9007, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 0.39407069555302165, | |
| "grad_norm": 0.24487371742725372, | |
| "learning_rate": 4.87073578250698e-06, | |
| "loss": 1.0956, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 0.3949828962371722, | |
| "grad_norm": 0.36626699566841125, | |
| "learning_rate": 4.730818955102234e-06, | |
| "loss": 1.1817, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 0.3958950969213227, | |
| "grad_norm": 0.20722968876361847, | |
| "learning_rate": 4.592841308745932e-06, | |
| "loss": 1.1693, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 0.39680729760547323, | |
| "grad_norm": 0.18895329535007477, | |
| "learning_rate": 4.456808753941205e-06, | |
| "loss": 1.2137, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.3977194982896237, | |
| "grad_norm": 0.11202345043420792, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 1.1175, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 0.3986316989737742, | |
| "grad_norm": 0.10380493104457855, | |
| "learning_rate": 4.190602144143207e-06, | |
| "loss": 0.9969, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 0.39954389965792475, | |
| "grad_norm": 0.10471412539482117, | |
| "learning_rate": 4.06043949255509e-06, | |
| "loss": 1.1972, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 0.40045610034207524, | |
| "grad_norm": 0.10842544585466385, | |
| "learning_rate": 3.932244738840379e-06, | |
| "loss": 1.233, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 0.4013683010262258, | |
| "grad_norm": 0.1057286411523819, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 1.1387, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.40228050171037627, | |
| "grad_norm": 0.13041551411151886, | |
| "learning_rate": 3.681780806244095e-06, | |
| "loss": 1.3469, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 0.4031927023945268, | |
| "grad_norm": 0.14460065960884094, | |
| "learning_rate": 3.5595223564037884e-06, | |
| "loss": 1.2273, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 0.4041049030786773, | |
| "grad_norm": 0.1310551017522812, | |
| "learning_rate": 3.4392532620598216e-06, | |
| "loss": 1.2728, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 0.40501710376282785, | |
| "grad_norm": 0.15493756532669067, | |
| "learning_rate": 3.3209786751399187e-06, | |
| "loss": 1.2911, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 0.40592930444697833, | |
| "grad_norm": 0.17276257276535034, | |
| "learning_rate": 3.2047036621337236e-06, | |
| "loss": 1.2539, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.4068415051311288, | |
| "grad_norm": 0.19491133093833923, | |
| "learning_rate": 3.0904332038757977e-06, | |
| "loss": 1.3042, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 0.40775370581527937, | |
| "grad_norm": 0.2031175196170807, | |
| "learning_rate": 2.978172195332263e-06, | |
| "loss": 1.3051, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 0.40866590649942985, | |
| "grad_norm": 0.24767906963825226, | |
| "learning_rate": 2.8679254453910785e-06, | |
| "loss": 1.3451, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 0.4095781071835804, | |
| "grad_norm": 0.3026193380355835, | |
| "learning_rate": 2.759697676656098e-06, | |
| "loss": 1.4116, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 0.4104903078677309, | |
| "grad_norm": 0.8892145156860352, | |
| "learning_rate": 2.653493525244721e-06, | |
| "loss": 1.6982, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.41140250855188143, | |
| "grad_norm": 0.13387838006019592, | |
| "learning_rate": 2.549317540589308e-06, | |
| "loss": 1.0503, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 0.4123147092360319, | |
| "grad_norm": 0.12145627290010452, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 0.9952, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 0.41322690992018246, | |
| "grad_norm": 0.14256002008914948, | |
| "learning_rate": 2.3470678346851518e-06, | |
| "loss": 1.0721, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 0.41413911060433295, | |
| "grad_norm": 0.13277588784694672, | |
| "learning_rate": 2.2490027771406687e-06, | |
| "loss": 1.0023, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 0.4150513112884835, | |
| "grad_norm": 0.14663711190223694, | |
| "learning_rate": 2.152983213389559e-06, | |
| "loss": 1.335, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.415963511972634, | |
| "grad_norm": 0.1930977702140808, | |
| "learning_rate": 2.0590132565903476e-06, | |
| "loss": 1.212, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 0.41687571265678447, | |
| "grad_norm": 0.19310270249843597, | |
| "learning_rate": 1.9670969321032407e-06, | |
| "loss": 1.281, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 0.417787913340935, | |
| "grad_norm": 0.20942994952201843, | |
| "learning_rate": 1.8772381773176417e-06, | |
| "loss": 1.5858, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 0.4187001140250855, | |
| "grad_norm": 0.1741369217634201, | |
| "learning_rate": 1.7894408414835362e-06, | |
| "loss": 1.1413, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 0.41961231470923605, | |
| "grad_norm": 0.2890353500843048, | |
| "learning_rate": 1.70370868554659e-06, | |
| "loss": 1.5824, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.42052451539338653, | |
| "grad_norm": 0.38743922114372253, | |
| "learning_rate": 1.620045381987012e-06, | |
| "loss": 1.2338, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 0.4214367160775371, | |
| "grad_norm": 0.4166823923587799, | |
| "learning_rate": 1.5384545146622852e-06, | |
| "loss": 1.4114, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 0.42234891676168757, | |
| "grad_norm": 0.16984394192695618, | |
| "learning_rate": 1.4589395786535953e-06, | |
| "loss": 1.033, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 0.4232611174458381, | |
| "grad_norm": 0.1662890464067459, | |
| "learning_rate": 1.3815039801161721e-06, | |
| "loss": 1.1955, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 0.4241733181299886, | |
| "grad_norm": 0.15266531705856323, | |
| "learning_rate": 1.3061510361333185e-06, | |
| "loss": 1.0023, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.4250855188141391, | |
| "grad_norm": 0.13052251935005188, | |
| "learning_rate": 1.232883974574367e-06, | |
| "loss": 0.7769, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 0.42599771949828963, | |
| "grad_norm": 0.15434938669204712, | |
| "learning_rate": 1.1617059339563807e-06, | |
| "loss": 0.9081, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 0.4269099201824401, | |
| "grad_norm": 0.1454930305480957, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 0.8577, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 0.42782212086659066, | |
| "grad_norm": 0.14353904128074646, | |
| "learning_rate": 1.0256290220474307e-06, | |
| "loss": 0.9633, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 0.42873432155074115, | |
| "grad_norm": 0.136052668094635, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.8172, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.4296465222348917, | |
| "grad_norm": 0.1676524579524994, | |
| "learning_rate": 8.979436164848088e-07, | |
| "loss": 1.0447, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 0.4305587229190422, | |
| "grad_norm": 0.2036374807357788, | |
| "learning_rate": 8.372546218022747e-07, | |
| "loss": 0.9286, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 0.4314709236031927, | |
| "grad_norm": 0.14977683126926422, | |
| "learning_rate": 7.786715955054203e-07, | |
| "loss": 0.9408, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 0.4323831242873432, | |
| "grad_norm": 0.1696736216545105, | |
| "learning_rate": 7.221970470961125e-07, | |
| "loss": 0.9405, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 0.43329532497149376, | |
| "grad_norm": 0.1618494838476181, | |
| "learning_rate": 6.678333957560512e-07, | |
| "loss": 1.0186, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.43420752565564424, | |
| "grad_norm": 0.19542407989501953, | |
| "learning_rate": 6.15582970243117e-07, | |
| "loss": 0.9891, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 0.43511972633979473, | |
| "grad_norm": 0.1792437881231308, | |
| "learning_rate": 5.654480087916303e-07, | |
| "loss": 0.786, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.4360319270239453, | |
| "grad_norm": 0.19154495000839233, | |
| "learning_rate": 5.174306590164879e-07, | |
| "loss": 1.026, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 0.43694412770809576, | |
| "grad_norm": 0.19262973964214325, | |
| "learning_rate": 4.715329778211375e-07, | |
| "loss": 1.1393, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 0.4378563283922463, | |
| "grad_norm": 0.19536390900611877, | |
| "learning_rate": 4.277569313094809e-07, | |
| "loss": 1.0381, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.4387685290763968, | |
| "grad_norm": 0.19628259539604187, | |
| "learning_rate": 3.8610439470164737e-07, | |
| "loss": 0.9981, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 0.43968072976054734, | |
| "grad_norm": 0.19776402413845062, | |
| "learning_rate": 3.465771522536854e-07, | |
| "loss": 1.0142, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 0.44059293044469783, | |
| "grad_norm": 0.21788012981414795, | |
| "learning_rate": 3.09176897181096e-07, | |
| "loss": 0.9793, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 0.44150513112884837, | |
| "grad_norm": 0.2997768521308899, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 1.0695, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 0.44241733181299886, | |
| "grad_norm": 0.27953919768333435, | |
| "learning_rate": 2.407636663901591e-07, | |
| "loss": 1.3182, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.44332953249714935, | |
| "grad_norm": 0.15639732778072357, | |
| "learning_rate": 2.0975362126691712e-07, | |
| "loss": 1.0899, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 0.4442417331812999, | |
| "grad_norm": 0.12085293978452682, | |
| "learning_rate": 1.8087642458373134e-07, | |
| "loss": 1.3242, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 0.4451539338654504, | |
| "grad_norm": 0.105677030980587, | |
| "learning_rate": 1.5413331334360182e-07, | |
| "loss": 1.2878, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 0.4460661345496009, | |
| "grad_norm": 0.12019108980894089, | |
| "learning_rate": 1.2952543313240472e-07, | |
| "loss": 1.076, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 0.4469783352337514, | |
| "grad_norm": 0.12178989499807358, | |
| "learning_rate": 1.0705383806982606e-07, | |
| "loss": 1.2949, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.44789053591790196, | |
| "grad_norm": 0.13249683380126953, | |
| "learning_rate": 8.671949076420882e-08, | |
| "loss": 1.4278, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 0.44880273660205244, | |
| "grad_norm": 0.13811670243740082, | |
| "learning_rate": 6.852326227130834e-08, | |
| "loss": 1.3262, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 0.449714937286203, | |
| "grad_norm": 0.14182856678962708, | |
| "learning_rate": 5.246593205699424e-08, | |
| "loss": 1.3477, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 0.4506271379703535, | |
| "grad_norm": 0.1499001681804657, | |
| "learning_rate": 3.8548187963854956e-08, | |
| "loss": 1.2677, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 0.45153933865450396, | |
| "grad_norm": 0.16916415095329285, | |
| "learning_rate": 2.6770626181715773e-08, | |
| "loss": 1.3193, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.4524515393386545, | |
| "grad_norm": 0.20377494394779205, | |
| "learning_rate": 1.7133751222137007e-08, | |
| "loss": 1.357, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 0.453363740022805, | |
| "grad_norm": 0.2306216061115265, | |
| "learning_rate": 9.637975896759077e-09, | |
| "loss": 1.7053, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 0.45427594070695554, | |
| "grad_norm": 0.2522677481174469, | |
| "learning_rate": 4.2836212996499865e-09, | |
| "loss": 1.3653, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 0.455188141391106, | |
| "grad_norm": 0.33547061681747437, | |
| "learning_rate": 1.0709167935385455e-09, | |
| "loss": 1.4946, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 0.45610034207525657, | |
| "grad_norm": 0.8273468017578125, | |
| "learning_rate": 0.0, | |
| "loss": 1.7554, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.45610034207525657, | |
| "eval_loss": 1.195978045463562, | |
| "eval_runtime": 113.8887, | |
| "eval_samples_per_second": 16.209, | |
| "eval_steps_per_second": 4.057, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 1 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.202037058732032e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |