| { | |
| "best_metric": 2.29616379737854, | |
| "best_model_checkpoint": "miner_id_24/checkpoint-400", | |
| "epoch": 0.44469149527515284, | |
| "eval_steps": 50, | |
| "global_step": 400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0011117287381878821, | |
| "grad_norm": 2.472998857498169, | |
| "learning_rate": 1e-05, | |
| "loss": 3.6514, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0011117287381878821, | |
| "eval_loss": 4.773703098297119, | |
| "eval_runtime": 70.9519, | |
| "eval_samples_per_second": 21.352, | |
| "eval_steps_per_second": 5.342, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0022234574763757642, | |
| "grad_norm": 2.228417158126831, | |
| "learning_rate": 2e-05, | |
| "loss": 3.5613, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0033351862145636463, | |
| "grad_norm": 2.4302899837493896, | |
| "learning_rate": 3e-05, | |
| "loss": 3.7739, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0044469149527515284, | |
| "grad_norm": 2.527637481689453, | |
| "learning_rate": 4e-05, | |
| "loss": 3.7436, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.005558643690939411, | |
| "grad_norm": 2.517746686935425, | |
| "learning_rate": 5e-05, | |
| "loss": 3.7961, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.006670372429127293, | |
| "grad_norm": 2.257744073867798, | |
| "learning_rate": 6e-05, | |
| "loss": 3.6019, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.007782101167315175, | |
| "grad_norm": 2.2959647178649902, | |
| "learning_rate": 7e-05, | |
| "loss": 3.6773, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.008893829905503057, | |
| "grad_norm": 2.340047597885132, | |
| "learning_rate": 8e-05, | |
| "loss": 3.5736, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.010005558643690939, | |
| "grad_norm": 2.3508639335632324, | |
| "learning_rate": 9e-05, | |
| "loss": 3.5245, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.011117287381878822, | |
| "grad_norm": 2.224466562271118, | |
| "learning_rate": 0.0001, | |
| "loss": 3.399, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.012229016120066704, | |
| "grad_norm": 2.578826904296875, | |
| "learning_rate": 9.99983777858264e-05, | |
| "loss": 3.1011, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.013340744858254585, | |
| "grad_norm": 3.860137701034546, | |
| "learning_rate": 9.999351124856874e-05, | |
| "loss": 3.1844, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.014452473596442469, | |
| "grad_norm": 5.3637518882751465, | |
| "learning_rate": 9.998540070400966e-05, | |
| "loss": 2.8854, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.01556420233463035, | |
| "grad_norm": 5.5368523597717285, | |
| "learning_rate": 9.997404667843075e-05, | |
| "loss": 2.7676, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.016675931072818232, | |
| "grad_norm": 3.3872969150543213, | |
| "learning_rate": 9.995944990857849e-05, | |
| "loss": 2.4751, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.017787659811006114, | |
| "grad_norm": 2.1491751670837402, | |
| "learning_rate": 9.994161134161634e-05, | |
| "loss": 2.7102, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.018899388549193995, | |
| "grad_norm": 2.828312873840332, | |
| "learning_rate": 9.992053213506334e-05, | |
| "loss": 2.4852, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.020011117287381877, | |
| "grad_norm": 2.784764051437378, | |
| "learning_rate": 9.989621365671902e-05, | |
| "loss": 2.2696, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.021122846025569762, | |
| "grad_norm": 3.483478307723999, | |
| "learning_rate": 9.986865748457457e-05, | |
| "loss": 2.4701, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.022234574763757644, | |
| "grad_norm": 2.1025726795196533, | |
| "learning_rate": 9.983786540671051e-05, | |
| "loss": 2.5995, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.023346303501945526, | |
| "grad_norm": 2.0208017826080322, | |
| "learning_rate": 9.980383942118066e-05, | |
| "loss": 2.439, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.024458032240133407, | |
| "grad_norm": 4.046178340911865, | |
| "learning_rate": 9.976658173588244e-05, | |
| "loss": 2.3586, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.02556976097832129, | |
| "grad_norm": 2.4368722438812256, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 2.3624, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.02668148971650917, | |
| "grad_norm": 2.14642071723938, | |
| "learning_rate": 9.968238114591566e-05, | |
| "loss": 2.2844, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.027793218454697052, | |
| "grad_norm": 3.8467893600463867, | |
| "learning_rate": 9.96354437049027e-05, | |
| "loss": 2.5449, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.028904947192884937, | |
| "grad_norm": 3.0803942680358887, | |
| "learning_rate": 9.95852854910781e-05, | |
| "loss": 2.4109, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.03001667593107282, | |
| "grad_norm": 1.9121437072753906, | |
| "learning_rate": 9.953190975913647e-05, | |
| "loss": 2.2835, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0311284046692607, | |
| "grad_norm": 1.6956825256347656, | |
| "learning_rate": 9.947531997255256e-05, | |
| "loss": 2.3913, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.032240133407448586, | |
| "grad_norm": 1.8031660318374634, | |
| "learning_rate": 9.941551980335652e-05, | |
| "loss": 2.3608, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.033351862145636464, | |
| "grad_norm": 1.9508814811706543, | |
| "learning_rate": 9.935251313189564e-05, | |
| "loss": 2.4243, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03446359088382435, | |
| "grad_norm": 1.6861339807510376, | |
| "learning_rate": 9.928630404658255e-05, | |
| "loss": 2.342, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.03557531962201223, | |
| "grad_norm": 1.4689520597457886, | |
| "learning_rate": 9.921689684362989e-05, | |
| "loss": 2.3789, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.03668704836020011, | |
| "grad_norm": 1.4510256052017212, | |
| "learning_rate": 9.914429602677162e-05, | |
| "loss": 2.2404, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.03779877709838799, | |
| "grad_norm": 2.0053555965423584, | |
| "learning_rate": 9.906850630697068e-05, | |
| "loss": 2.4606, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.038910505836575876, | |
| "grad_norm": 1.4402495622634888, | |
| "learning_rate": 9.898953260211338e-05, | |
| "loss": 2.353, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.040022234574763754, | |
| "grad_norm": 1.6246519088745117, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 2.5241, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.04113396331295164, | |
| "grad_norm": 1.7262248992919922, | |
| "learning_rate": 9.882205394146361e-05, | |
| "loss": 2.0709, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.042245692051139525, | |
| "grad_norm": 1.5005686283111572, | |
| "learning_rate": 9.87335598531214e-05, | |
| "loss": 2.2354, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.0433574207893274, | |
| "grad_norm": 1.618937373161316, | |
| "learning_rate": 9.864190351391822e-05, | |
| "loss": 2.2752, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.04446914952751529, | |
| "grad_norm": 3.1042425632476807, | |
| "learning_rate": 9.85470908713026e-05, | |
| "loss": 2.4815, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.045580878265703166, | |
| "grad_norm": 2.9018545150756836, | |
| "learning_rate": 9.844912807753104e-05, | |
| "loss": 2.5847, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.04669260700389105, | |
| "grad_norm": 1.687359094619751, | |
| "learning_rate": 9.834802148926882e-05, | |
| "loss": 2.1867, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.04780433574207893, | |
| "grad_norm": 2.870513439178467, | |
| "learning_rate": 9.824377766717759e-05, | |
| "loss": 2.2868, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.048916064480266815, | |
| "grad_norm": 2.6799087524414062, | |
| "learning_rate": 9.813640337548954e-05, | |
| "loss": 2.3823, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.0500277932184547, | |
| "grad_norm": 2.1397416591644287, | |
| "learning_rate": 9.802590558156862e-05, | |
| "loss": 2.1061, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05113952195664258, | |
| "grad_norm": 1.9821467399597168, | |
| "learning_rate": 9.791229145545831e-05, | |
| "loss": 2.3679, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.05225125069483046, | |
| "grad_norm": 3.3248050212860107, | |
| "learning_rate": 9.779556836941645e-05, | |
| "loss": 2.2956, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.05336297943301834, | |
| "grad_norm": 2.8718130588531494, | |
| "learning_rate": 9.767574389743682e-05, | |
| "loss": 2.2576, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.054474708171206226, | |
| "grad_norm": 2.6526224613189697, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 2.2742, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.055586436909394105, | |
| "grad_norm": 1.908512830734253, | |
| "learning_rate": 9.742682209735727e-05, | |
| "loss": 2.3169, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.055586436909394105, | |
| "eval_loss": 2.8695197105407715, | |
| "eval_runtime": 71.5125, | |
| "eval_samples_per_second": 21.185, | |
| "eval_steps_per_second": 5.3, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05669816564758199, | |
| "grad_norm": 8.565925598144531, | |
| "learning_rate": 9.729774092143627e-05, | |
| "loss": 3.3931, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.057809894385769875, | |
| "grad_norm": 8.781779289245605, | |
| "learning_rate": 9.716559066288715e-05, | |
| "loss": 3.2973, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.05892162312395775, | |
| "grad_norm": 6.783483505249023, | |
| "learning_rate": 9.703037989675087e-05, | |
| "loss": 3.0626, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.06003335186214564, | |
| "grad_norm": 3.6386938095092773, | |
| "learning_rate": 9.689211739666023e-05, | |
| "loss": 2.7669, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.061145080600333516, | |
| "grad_norm": 1.7182066440582275, | |
| "learning_rate": 9.675081213427076e-05, | |
| "loss": 2.6579, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0622568093385214, | |
| "grad_norm": 1.6245131492614746, | |
| "learning_rate": 9.66064732786784e-05, | |
| "loss": 2.4646, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.06336853807670928, | |
| "grad_norm": 1.7475733757019043, | |
| "learning_rate": 9.645911019582467e-05, | |
| "loss": 2.4532, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.06448026681489717, | |
| "grad_norm": 1.7325527667999268, | |
| "learning_rate": 9.630873244788883e-05, | |
| "loss": 2.5395, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.06559199555308505, | |
| "grad_norm": 1.656267523765564, | |
| "learning_rate": 9.615534979266745e-05, | |
| "loss": 2.3968, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.06670372429127293, | |
| "grad_norm": 1.4385294914245605, | |
| "learning_rate": 9.599897218294122e-05, | |
| "loss": 2.4547, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.0678154530294608, | |
| "grad_norm": 1.149308681488037, | |
| "learning_rate": 9.583960976582913e-05, | |
| "loss": 2.4185, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.0689271817676487, | |
| "grad_norm": 1.0372174978256226, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 2.3459, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.07003891050583658, | |
| "grad_norm": 1.227518081665039, | |
| "learning_rate": 9.551197206565173e-05, | |
| "loss": 2.4516, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.07115063924402446, | |
| "grad_norm": 0.9263134002685547, | |
| "learning_rate": 9.534371804252728e-05, | |
| "loss": 2.3193, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.07226236798221233, | |
| "grad_norm": 1.0493077039718628, | |
| "learning_rate": 9.517252173051911e-05, | |
| "loss": 2.475, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07337409672040023, | |
| "grad_norm": 1.0255634784698486, | |
| "learning_rate": 9.49983942383106e-05, | |
| "loss": 2.3088, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.0744858254585881, | |
| "grad_norm": 1.041403889656067, | |
| "learning_rate": 9.482134686478519e-05, | |
| "loss": 2.2648, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.07559755419677598, | |
| "grad_norm": 1.0037230253219604, | |
| "learning_rate": 9.464139109829321e-05, | |
| "loss": 2.2107, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.07670928293496387, | |
| "grad_norm": 1.3062645196914673, | |
| "learning_rate": 9.445853861590647e-05, | |
| "loss": 2.2864, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.07782101167315175, | |
| "grad_norm": 1.1115295886993408, | |
| "learning_rate": 9.42728012826605e-05, | |
| "loss": 2.185, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07893274041133963, | |
| "grad_norm": 1.0353200435638428, | |
| "learning_rate": 9.408419115078471e-05, | |
| "loss": 2.4227, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.08004446914952751, | |
| "grad_norm": 1.0612479448318481, | |
| "learning_rate": 9.389272045892024e-05, | |
| "loss": 2.3398, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.0811561978877154, | |
| "grad_norm": 1.0379235744476318, | |
| "learning_rate": 9.36984016313259e-05, | |
| "loss": 2.2034, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.08226792662590328, | |
| "grad_norm": 0.9927018880844116, | |
| "learning_rate": 9.350124727707197e-05, | |
| "loss": 2.3159, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.08337965536409116, | |
| "grad_norm": 1.063293695449829, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 2.3267, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08449138410227905, | |
| "grad_norm": 1.0034624338150024, | |
| "learning_rate": 9.309848334400246e-05, | |
| "loss": 2.0903, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.08560311284046693, | |
| "grad_norm": 1.027740240097046, | |
| "learning_rate": 9.289289989996133e-05, | |
| "loss": 2.3924, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.0867148415786548, | |
| "grad_norm": 1.3626140356063843, | |
| "learning_rate": 9.268453319711363e-05, | |
| "loss": 2.0867, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.08782657031684268, | |
| "grad_norm": 1.0475863218307495, | |
| "learning_rate": 9.247339675607605e-05, | |
| "loss": 2.159, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.08893829905503058, | |
| "grad_norm": 1.1371270418167114, | |
| "learning_rate": 9.225950427718975e-05, | |
| "loss": 2.4041, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09005002779321845, | |
| "grad_norm": 1.2929434776306152, | |
| "learning_rate": 9.204286963963111e-05, | |
| "loss": 2.2017, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.09116175653140633, | |
| "grad_norm": 1.0671011209487915, | |
| "learning_rate": 9.182350690051133e-05, | |
| "loss": 2.4558, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.09227348526959422, | |
| "grad_norm": 1.121847152709961, | |
| "learning_rate": 9.160143029396422e-05, | |
| "loss": 2.1243, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.0933852140077821, | |
| "grad_norm": 1.2664432525634766, | |
| "learning_rate": 9.13766542302225e-05, | |
| "loss": 2.1224, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.09449694274596998, | |
| "grad_norm": 1.281224250793457, | |
| "learning_rate": 9.114919329468282e-05, | |
| "loss": 2.2792, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.09560867148415786, | |
| "grad_norm": 1.2077322006225586, | |
| "learning_rate": 9.091906224695935e-05, | |
| "loss": 2.202, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.09672040022234575, | |
| "grad_norm": 1.253800392150879, | |
| "learning_rate": 9.068627601992598e-05, | |
| "loss": 2.2172, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.09783212896053363, | |
| "grad_norm": 1.4512909650802612, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 2.0814, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.09894385769872151, | |
| "grad_norm": 1.256651759147644, | |
| "learning_rate": 9.021279861989885e-05, | |
| "loss": 2.2968, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.1000555864369094, | |
| "grad_norm": 1.8698711395263672, | |
| "learning_rate": 8.997213817017507e-05, | |
| "loss": 2.5366, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.10116731517509728, | |
| "grad_norm": 1.381558895111084, | |
| "learning_rate": 8.972888398568772e-05, | |
| "loss": 2.2141, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.10227904391328516, | |
| "grad_norm": 1.271780252456665, | |
| "learning_rate": 8.948305185085225e-05, | |
| "loss": 2.3866, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.10339077265147303, | |
| "grad_norm": 1.3147594928741455, | |
| "learning_rate": 8.92346577173636e-05, | |
| "loss": 2.0375, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.10450250138966093, | |
| "grad_norm": 1.3043659925460815, | |
| "learning_rate": 8.898371770316111e-05, | |
| "loss": 2.3249, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.1056142301278488, | |
| "grad_norm": 1.241226315498352, | |
| "learning_rate": 8.873024809138272e-05, | |
| "loss": 2.4081, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10672595886603668, | |
| "grad_norm": 1.7342743873596191, | |
| "learning_rate": 8.847426532930831e-05, | |
| "loss": 2.1388, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.10783768760422457, | |
| "grad_norm": 1.5813889503479004, | |
| "learning_rate": 8.821578602729242e-05, | |
| "loss": 2.3315, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.10894941634241245, | |
| "grad_norm": 1.7183290719985962, | |
| "learning_rate": 8.795482695768658e-05, | |
| "loss": 2.1037, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.11006114508060033, | |
| "grad_norm": 16.183635711669922, | |
| "learning_rate": 8.769140505375085e-05, | |
| "loss": 2.3144, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.11117287381878821, | |
| "grad_norm": 1.6419098377227783, | |
| "learning_rate": 8.742553740855506e-05, | |
| "loss": 2.1397, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11117287381878821, | |
| "eval_loss": 2.7260963916778564, | |
| "eval_runtime": 71.6154, | |
| "eval_samples_per_second": 21.155, | |
| "eval_steps_per_second": 5.292, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1122846025569761, | |
| "grad_norm": 5.945138454437256, | |
| "learning_rate": 8.715724127386972e-05, | |
| "loss": 3.3042, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.11339633129516398, | |
| "grad_norm": 5.689327239990234, | |
| "learning_rate": 8.688653405904652e-05, | |
| "loss": 3.2821, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.11450806003335186, | |
| "grad_norm": 5.383906364440918, | |
| "learning_rate": 8.661343332988869e-05, | |
| "loss": 3.1464, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.11561978877153975, | |
| "grad_norm": 3.6286027431488037, | |
| "learning_rate": 8.633795680751116e-05, | |
| "loss": 3.0208, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.11673151750972763, | |
| "grad_norm": 2.5231778621673584, | |
| "learning_rate": 8.606012236719073e-05, | |
| "loss": 2.9116, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.1178432462479155, | |
| "grad_norm": 1.8035188913345337, | |
| "learning_rate": 8.577994803720606e-05, | |
| "loss": 2.7812, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.11895497498610338, | |
| "grad_norm": 1.2647838592529297, | |
| "learning_rate": 8.549745199766792e-05, | |
| "loss": 2.5962, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.12006670372429128, | |
| "grad_norm": 1.0100034475326538, | |
| "learning_rate": 8.521265257933948e-05, | |
| "loss": 2.5903, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.12117843246247915, | |
| "grad_norm": 1.5505986213684082, | |
| "learning_rate": 8.492556826244687e-05, | |
| "loss": 2.4289, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.12229016120066703, | |
| "grad_norm": 2.1767804622650146, | |
| "learning_rate": 8.463621767547998e-05, | |
| "loss": 2.5854, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.12340188993885493, | |
| "grad_norm": 2.0668084621429443, | |
| "learning_rate": 8.434461959398376e-05, | |
| "loss": 2.4399, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.1245136186770428, | |
| "grad_norm": 1.511650562286377, | |
| "learning_rate": 8.405079293933986e-05, | |
| "loss": 2.6072, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.12562534741523068, | |
| "grad_norm": 1.1593494415283203, | |
| "learning_rate": 8.375475677753881e-05, | |
| "loss": 2.4045, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.12673707615341856, | |
| "grad_norm": 1.053107738494873, | |
| "learning_rate": 8.345653031794292e-05, | |
| "loss": 2.4522, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.12784880489160644, | |
| "grad_norm": 0.9937801361083984, | |
| "learning_rate": 8.315613291203976e-05, | |
| "loss": 2.3276, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.12896053362979434, | |
| "grad_norm": 0.8721001744270325, | |
| "learning_rate": 8.285358405218655e-05, | |
| "loss": 2.3567, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.13007226236798222, | |
| "grad_norm": 0.9330883026123047, | |
| "learning_rate": 8.25489033703452e-05, | |
| "loss": 2.2585, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.1311839911061701, | |
| "grad_norm": 0.9001569747924805, | |
| "learning_rate": 8.224211063680853e-05, | |
| "loss": 2.297, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.13229571984435798, | |
| "grad_norm": 0.8666402697563171, | |
| "learning_rate": 8.19332257589174e-05, | |
| "loss": 2.2182, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.13340744858254586, | |
| "grad_norm": 1.176416277885437, | |
| "learning_rate": 8.162226877976887e-05, | |
| "loss": 2.3346, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.13451917732073373, | |
| "grad_norm": 0.9645460844039917, | |
| "learning_rate": 8.130925987691569e-05, | |
| "loss": 2.0891, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.1356309060589216, | |
| "grad_norm": 1.3001148700714111, | |
| "learning_rate": 8.099421936105702e-05, | |
| "loss": 2.3923, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.13674263479710952, | |
| "grad_norm": 1.1771594285964966, | |
| "learning_rate": 8.067716767472045e-05, | |
| "loss": 2.3808, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.1378543635352974, | |
| "grad_norm": 0.997420608997345, | |
| "learning_rate": 8.035812539093557e-05, | |
| "loss": 2.4158, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.13896609227348528, | |
| "grad_norm": 0.8972210884094238, | |
| "learning_rate": 8.003711321189895e-05, | |
| "loss": 2.1985, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.14007782101167315, | |
| "grad_norm": 0.8970302939414978, | |
| "learning_rate": 7.971415196763088e-05, | |
| "loss": 1.9446, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.14118954974986103, | |
| "grad_norm": 1.0234975814819336, | |
| "learning_rate": 7.938926261462366e-05, | |
| "loss": 2.2494, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.1423012784880489, | |
| "grad_norm": 1.1858220100402832, | |
| "learning_rate": 7.906246623448183e-05, | |
| "loss": 2.2926, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.1434130072262368, | |
| "grad_norm": 1.04050612449646, | |
| "learning_rate": 7.873378403255419e-05, | |
| "loss": 2.3006, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.14452473596442467, | |
| "grad_norm": 1.20691978931427, | |
| "learning_rate": 7.840323733655778e-05, | |
| "loss": 2.3302, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14563646470261257, | |
| "grad_norm": 1.0528558492660522, | |
| "learning_rate": 7.807084759519405e-05, | |
| "loss": 2.1477, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.14674819344080045, | |
| "grad_norm": 1.085060954093933, | |
| "learning_rate": 7.773663637675694e-05, | |
| "loss": 2.3391, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.14785992217898833, | |
| "grad_norm": 0.9347543120384216, | |
| "learning_rate": 7.740062536773352e-05, | |
| "loss": 2.2107, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.1489716509171762, | |
| "grad_norm": 1.0611324310302734, | |
| "learning_rate": 7.706283637139658e-05, | |
| "loss": 2.2921, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.15008337965536409, | |
| "grad_norm": 1.1190764904022217, | |
| "learning_rate": 7.672329130639005e-05, | |
| "loss": 2.1771, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.15119510839355196, | |
| "grad_norm": 1.021002173423767, | |
| "learning_rate": 7.638201220530665e-05, | |
| "loss": 2.1964, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.15230683713173984, | |
| "grad_norm": 1.2308954000473022, | |
| "learning_rate": 7.603902121325813e-05, | |
| "loss": 2.2279, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.15341856586992775, | |
| "grad_norm": 1.0030035972595215, | |
| "learning_rate": 7.569434058643844e-05, | |
| "loss": 2.1981, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.15453029460811563, | |
| "grad_norm": 1.0752218961715698, | |
| "learning_rate": 7.534799269067953e-05, | |
| "loss": 2.2809, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.1556420233463035, | |
| "grad_norm": 1.1892122030258179, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 2.2486, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.15675375208449138, | |
| "grad_norm": 1.1939011812210083, | |
| "learning_rate": 7.465038509514688e-05, | |
| "loss": 2.1895, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.15786548082267926, | |
| "grad_norm": 1.430174469947815, | |
| "learning_rate": 7.42991706621303e-05, | |
| "loss": 2.3485, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.15897720956086714, | |
| "grad_norm": 1.1184104681015015, | |
| "learning_rate": 7.394637949075154e-05, | |
| "loss": 2.2954, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.16008893829905502, | |
| "grad_norm": 1.0584604740142822, | |
| "learning_rate": 7.35920344731241e-05, | |
| "loss": 2.0915, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.16120066703724292, | |
| "grad_norm": 1.1394598484039307, | |
| "learning_rate": 7.323615860218843e-05, | |
| "loss": 1.9866, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1623123957754308, | |
| "grad_norm": 1.3819961547851562, | |
| "learning_rate": 7.287877497021978e-05, | |
| "loss": 2.229, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.16342412451361868, | |
| "grad_norm": 1.717722773551941, | |
| "learning_rate": 7.251990676732984e-05, | |
| "loss": 2.0977, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.16453585325180656, | |
| "grad_norm": 1.206673502922058, | |
| "learning_rate": 7.215957727996207e-05, | |
| "loss": 2.2358, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.16564758198999444, | |
| "grad_norm": 1.5048532485961914, | |
| "learning_rate": 7.179780988938051e-05, | |
| "loss": 2.2084, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.16675931072818231, | |
| "grad_norm": 1.4663426876068115, | |
| "learning_rate": 7.143462807015271e-05, | |
| "loss": 2.1774, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16675931072818231, | |
| "eval_loss": 2.5343213081359863, | |
| "eval_runtime": 71.4182, | |
| "eval_samples_per_second": 21.213, | |
| "eval_steps_per_second": 5.307, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1678710394663702, | |
| "grad_norm": 3.798626184463501, | |
| "learning_rate": 7.107005538862646e-05, | |
| "loss": 2.8672, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.1689827682045581, | |
| "grad_norm": 3.3820409774780273, | |
| "learning_rate": 7.07041155014006e-05, | |
| "loss": 2.9326, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.17009449694274598, | |
| "grad_norm": 2.7820358276367188, | |
| "learning_rate": 7.033683215379002e-05, | |
| "loss": 2.8474, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.17120622568093385, | |
| "grad_norm": 2.195570707321167, | |
| "learning_rate": 6.996822917828477e-05, | |
| "loss": 2.7765, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.17231795441912173, | |
| "grad_norm": 1.862131953239441, | |
| "learning_rate": 6.959833049300377e-05, | |
| "loss": 2.6993, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.1734296831573096, | |
| "grad_norm": 1.5429936647415161, | |
| "learning_rate": 6.922716010014255e-05, | |
| "loss": 2.697, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.1745414118954975, | |
| "grad_norm": 1.3168162107467651, | |
| "learning_rate": 6.885474208441603e-05, | |
| "loss": 2.5454, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.17565314063368537, | |
| "grad_norm": 1.1143990755081177, | |
| "learning_rate": 6.848110061149556e-05, | |
| "loss": 2.4594, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.17676486937187327, | |
| "grad_norm": 0.9172198176383972, | |
| "learning_rate": 6.810625992644085e-05, | |
| "loss": 2.3292, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.17787659811006115, | |
| "grad_norm": 0.9379021525382996, | |
| "learning_rate": 6.773024435212678e-05, | |
| "loss": 2.4432, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.17898832684824903, | |
| "grad_norm": 0.7792444229125977, | |
| "learning_rate": 6.735307828766515e-05, | |
| "loss": 2.4288, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.1801000555864369, | |
| "grad_norm": 0.8528432846069336, | |
| "learning_rate": 6.697478620682137e-05, | |
| "loss": 2.2895, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.1812117843246248, | |
| "grad_norm": 0.7585402727127075, | |
| "learning_rate": 6.659539265642643e-05, | |
| "loss": 2.4101, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.18232351306281266, | |
| "grad_norm": 0.8885021805763245, | |
| "learning_rate": 6.621492225478414e-05, | |
| "loss": 2.4375, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.18343524180100054, | |
| "grad_norm": 0.784393846988678, | |
| "learning_rate": 6.583339969007363e-05, | |
| "loss": 2.3757, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.18454697053918845, | |
| "grad_norm": 1.024843692779541, | |
| "learning_rate": 6.545084971874738e-05, | |
| "loss": 2.4254, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.18565869927737633, | |
| "grad_norm": 1.151550531387329, | |
| "learning_rate": 6.506729716392481e-05, | |
| "loss": 2.1751, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.1867704280155642, | |
| "grad_norm": 1.0025200843811035, | |
| "learning_rate": 6.468276691378155e-05, | |
| "loss": 2.4144, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.18788215675375208, | |
| "grad_norm": 0.8717769980430603, | |
| "learning_rate": 6.429728391993446e-05, | |
| "loss": 2.3761, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.18899388549193996, | |
| "grad_norm": 0.8901825547218323, | |
| "learning_rate": 6.391087319582264e-05, | |
| "loss": 2.1471, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.19010561423012784, | |
| "grad_norm": 0.988797664642334, | |
| "learning_rate": 6.35235598150842e-05, | |
| "loss": 2.3141, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.19121734296831572, | |
| "grad_norm": 0.9147389531135559, | |
| "learning_rate": 6.313536890992935e-05, | |
| "loss": 2.2724, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.19232907170650362, | |
| "grad_norm": 0.8791763782501221, | |
| "learning_rate": 6.274632566950967e-05, | |
| "loss": 2.2125, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.1934408004446915, | |
| "grad_norm": 0.9586495161056519, | |
| "learning_rate": 6.235645533828349e-05, | |
| "loss": 2.1094, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.19455252918287938, | |
| "grad_norm": 1.0605213642120361, | |
| "learning_rate": 6.19657832143779e-05, | |
| "loss": 2.2509, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.19566425792106726, | |
| "grad_norm": 1.155334234237671, | |
| "learning_rate": 6.157433464794716e-05, | |
| "loss": 2.2831, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.19677598665925514, | |
| "grad_norm": 0.8409937620162964, | |
| "learning_rate": 6.118213503952779e-05, | |
| "loss": 2.1029, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.19788771539744301, | |
| "grad_norm": 0.9668261408805847, | |
| "learning_rate": 6.078920983839031e-05, | |
| "loss": 2.3095, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.1989994441356309, | |
| "grad_norm": 1.3652583360671997, | |
| "learning_rate": 6.0395584540887963e-05, | |
| "loss": 2.3097, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.2001111728738188, | |
| "grad_norm": 1.2160142660140991, | |
| "learning_rate": 6.0001284688802226e-05, | |
| "loss": 2.367, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.20122290161200668, | |
| "grad_norm": 1.1758296489715576, | |
| "learning_rate": 5.960633586768543e-05, | |
| "loss": 2.2655, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.20233463035019456, | |
| "grad_norm": 1.1766093969345093, | |
| "learning_rate": 5.921076370520058e-05, | |
| "loss": 2.2498, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.20344635908838243, | |
| "grad_norm": 0.9961435794830322, | |
| "learning_rate": 5.8814593869458455e-05, | |
| "loss": 2.1775, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.2045580878265703, | |
| "grad_norm": 1.0376254320144653, | |
| "learning_rate": 5.841785206735192e-05, | |
| "loss": 2.2686, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.2056698165647582, | |
| "grad_norm": 0.9666380882263184, | |
| "learning_rate": 5.8020564042888015e-05, | |
| "loss": 2.0915, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.20678154530294607, | |
| "grad_norm": 1.1380987167358398, | |
| "learning_rate": 5.762275557551727e-05, | |
| "loss": 2.2868, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.20789327404113397, | |
| "grad_norm": 1.2181998491287231, | |
| "learning_rate": 5.7224452478461064e-05, | |
| "loss": 2.4482, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.20900500277932185, | |
| "grad_norm": 1.0468593835830688, | |
| "learning_rate": 5.682568059703659e-05, | |
| "loss": 2.1723, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.21011673151750973, | |
| "grad_norm": 0.961937665939331, | |
| "learning_rate": 5.642646580697973e-05, | |
| "loss": 2.2278, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.2112284602556976, | |
| "grad_norm": 1.102487564086914, | |
| "learning_rate": 5.602683401276615e-05, | |
| "loss": 2.339, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2123401889938855, | |
| "grad_norm": 0.9763503670692444, | |
| "learning_rate": 5.562681114593028e-05, | |
| "loss": 2.1248, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.21345191773207337, | |
| "grad_norm": 1.0857285261154175, | |
| "learning_rate": 5.522642316338268e-05, | |
| "loss": 2.3992, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.21456364647026124, | |
| "grad_norm": 1.103461503982544, | |
| "learning_rate": 5.482569604572576e-05, | |
| "loss": 2.1701, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.21567537520844915, | |
| "grad_norm": 1.18154776096344, | |
| "learning_rate": 5.442465579556793e-05, | |
| "loss": 2.2224, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.21678710394663703, | |
| "grad_norm": 1.6554676294326782, | |
| "learning_rate": 5.402332843583631e-05, | |
| "loss": 2.2127, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2178988326848249, | |
| "grad_norm": 1.4385900497436523, | |
| "learning_rate": 5.3621740008088126e-05, | |
| "loss": 2.3457, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.21901056142301278, | |
| "grad_norm": 1.1937305927276611, | |
| "learning_rate": 5.321991657082097e-05, | |
| "loss": 2.1697, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.22012229016120066, | |
| "grad_norm": 1.5197139978408813, | |
| "learning_rate": 5.281788419778187e-05, | |
| "loss": 2.3358, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.22123401889938854, | |
| "grad_norm": 1.266471028327942, | |
| "learning_rate": 5.2415668976275355e-05, | |
| "loss": 2.2003, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.22234574763757642, | |
| "grad_norm": 1.8889713287353516, | |
| "learning_rate": 5.201329700547076e-05, | |
| "loss": 2.3701, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.22234574763757642, | |
| "eval_loss": 2.5627048015594482, | |
| "eval_runtime": 71.6301, | |
| "eval_samples_per_second": 21.15, | |
| "eval_steps_per_second": 5.291, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.22345747637576432, | |
| "grad_norm": 4.871993064880371, | |
| "learning_rate": 5.161079439470866e-05, | |
| "loss": 2.9347, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.2245692051139522, | |
| "grad_norm": 4.960667610168457, | |
| "learning_rate": 5.1208187261806615e-05, | |
| "loss": 2.9743, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.22568093385214008, | |
| "grad_norm": 4.213318824768066, | |
| "learning_rate": 5.080550173136457e-05, | |
| "loss": 2.9916, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.22679266259032796, | |
| "grad_norm": 3.499946355819702, | |
| "learning_rate": 5.0402763933069496e-05, | |
| "loss": 2.7203, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.22790439132851584, | |
| "grad_norm": 2.6555299758911133, | |
| "learning_rate": 5e-05, | |
| "loss": 2.6411, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.22901612006670372, | |
| "grad_norm": 2.1636953353881836, | |
| "learning_rate": 4.9597236066930516e-05, | |
| "loss": 2.7387, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.2301278488048916, | |
| "grad_norm": 1.7059534788131714, | |
| "learning_rate": 4.919449826863544e-05, | |
| "loss": 2.6006, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.2312395775430795, | |
| "grad_norm": 1.8288934230804443, | |
| "learning_rate": 4.87918127381934e-05, | |
| "loss": 2.559, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.23235130628126738, | |
| "grad_norm": 1.66895592212677, | |
| "learning_rate": 4.8389205605291365e-05, | |
| "loss": 2.6273, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.23346303501945526, | |
| "grad_norm": 1.2062842845916748, | |
| "learning_rate": 4.798670299452926e-05, | |
| "loss": 2.6314, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.23457476375764313, | |
| "grad_norm": 1.236952304840088, | |
| "learning_rate": 4.758433102372466e-05, | |
| "loss": 2.2492, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.235686492495831, | |
| "grad_norm": 1.0907583236694336, | |
| "learning_rate": 4.7182115802218126e-05, | |
| "loss": 2.3356, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.2367982212340189, | |
| "grad_norm": 1.1642088890075684, | |
| "learning_rate": 4.678008342917903e-05, | |
| "loss": 2.3069, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.23790994997220677, | |
| "grad_norm": 1.0130016803741455, | |
| "learning_rate": 4.6378259991911886e-05, | |
| "loss": 2.2744, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.23902167871039467, | |
| "grad_norm": 1.0717731714248657, | |
| "learning_rate": 4.597667156416371e-05, | |
| "loss": 2.2989, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.24013340744858255, | |
| "grad_norm": 0.8758122324943542, | |
| "learning_rate": 4.5575344204432084e-05, | |
| "loss": 2.2346, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.24124513618677043, | |
| "grad_norm": 0.7776920199394226, | |
| "learning_rate": 4.5174303954274244e-05, | |
| "loss": 2.2342, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.2423568649249583, | |
| "grad_norm": 0.858988344669342, | |
| "learning_rate": 4.477357683661734e-05, | |
| "loss": 2.3042, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.2434685936631462, | |
| "grad_norm": 0.902452290058136, | |
| "learning_rate": 4.437318885406973e-05, | |
| "loss": 2.2463, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.24458032240133407, | |
| "grad_norm": 0.8730572462081909, | |
| "learning_rate": 4.397316598723385e-05, | |
| "loss": 2.2071, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.24569205113952194, | |
| "grad_norm": 0.8203500509262085, | |
| "learning_rate": 4.3573534193020274e-05, | |
| "loss": 2.2388, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.24680377987770985, | |
| "grad_norm": 1.4247759580612183, | |
| "learning_rate": 4.317431940296343e-05, | |
| "loss": 2.1673, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.24791550861589773, | |
| "grad_norm": 0.8514419794082642, | |
| "learning_rate": 4.277554752153895e-05, | |
| "loss": 2.3399, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.2490272373540856, | |
| "grad_norm": 0.9166375398635864, | |
| "learning_rate": 4.237724442448273e-05, | |
| "loss": 2.1514, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.2501389660922735, | |
| "grad_norm": 0.8139379024505615, | |
| "learning_rate": 4.197943595711198e-05, | |
| "loss": 1.9978, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.25125069483046136, | |
| "grad_norm": 0.8882634043693542, | |
| "learning_rate": 4.1582147932648074e-05, | |
| "loss": 2.2634, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.25236242356864924, | |
| "grad_norm": 0.9146400690078735, | |
| "learning_rate": 4.118540613054156e-05, | |
| "loss": 2.5159, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.2534741523068371, | |
| "grad_norm": 0.9058739542961121, | |
| "learning_rate": 4.078923629479943e-05, | |
| "loss": 2.1295, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.254585881045025, | |
| "grad_norm": 0.8965374231338501, | |
| "learning_rate": 4.039366413231458e-05, | |
| "loss": 2.3373, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.2556976097832129, | |
| "grad_norm": 0.8202263712882996, | |
| "learning_rate": 3.9998715311197785e-05, | |
| "loss": 2.2355, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.25680933852140075, | |
| "grad_norm": 1.061989188194275, | |
| "learning_rate": 3.960441545911204e-05, | |
| "loss": 2.1224, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.2579210672595887, | |
| "grad_norm": 0.9839004874229431, | |
| "learning_rate": 3.92107901616097e-05, | |
| "loss": 2.2109, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.25903279599777657, | |
| "grad_norm": 0.9290953874588013, | |
| "learning_rate": 3.8817864960472236e-05, | |
| "loss": 2.1705, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.26014452473596444, | |
| "grad_norm": 0.8922295570373535, | |
| "learning_rate": 3.842566535205286e-05, | |
| "loss": 2.1385, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.2612562534741523, | |
| "grad_norm": 1.0526587963104248, | |
| "learning_rate": 3.803421678562213e-05, | |
| "loss": 2.3858, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2623679822123402, | |
| "grad_norm": 1.066570520401001, | |
| "learning_rate": 3.764354466171652e-05, | |
| "loss": 2.2403, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.2634797109505281, | |
| "grad_norm": 1.4028855562210083, | |
| "learning_rate": 3.725367433049033e-05, | |
| "loss": 2.2443, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.26459143968871596, | |
| "grad_norm": 1.0442891120910645, | |
| "learning_rate": 3.6864631090070655e-05, | |
| "loss": 2.2474, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.26570316842690384, | |
| "grad_norm": 0.9756484031677246, | |
| "learning_rate": 3.6476440184915815e-05, | |
| "loss": 2.433, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.2668148971650917, | |
| "grad_norm": 1.0391199588775635, | |
| "learning_rate": 3.608912680417737e-05, | |
| "loss": 2.2412, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2679266259032796, | |
| "grad_norm": 1.3023152351379395, | |
| "learning_rate": 3.570271608006555e-05, | |
| "loss": 2.1084, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.26903835464146747, | |
| "grad_norm": 1.0111302137374878, | |
| "learning_rate": 3.531723308621847e-05, | |
| "loss": 2.0772, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.27015008337965535, | |
| "grad_norm": 1.1569526195526123, | |
| "learning_rate": 3.493270283607522e-05, | |
| "loss": 2.1449, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.2712618121178432, | |
| "grad_norm": 1.4179155826568604, | |
| "learning_rate": 3.4549150281252636e-05, | |
| "loss": 2.2884, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.2723735408560311, | |
| "grad_norm": 1.1831965446472168, | |
| "learning_rate": 3.4166600309926387e-05, | |
| "loss": 2.2144, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.27348526959421904, | |
| "grad_norm": 1.2018048763275146, | |
| "learning_rate": 3.3785077745215873e-05, | |
| "loss": 2.1746, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.2745969983324069, | |
| "grad_norm": 1.330165982246399, | |
| "learning_rate": 3.340460734357359e-05, | |
| "loss": 2.0056, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.2757087270705948, | |
| "grad_norm": 1.8214839696884155, | |
| "learning_rate": 3.3025213793178646e-05, | |
| "loss": 2.2728, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.2768204558087827, | |
| "grad_norm": 1.4953075647354126, | |
| "learning_rate": 3.264692171233485e-05, | |
| "loss": 2.2458, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.27793218454697055, | |
| "grad_norm": 2.2444264888763428, | |
| "learning_rate": 3.226975564787322e-05, | |
| "loss": 2.1073, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.27793218454697055, | |
| "eval_loss": 2.419438123703003, | |
| "eval_runtime": 71.5319, | |
| "eval_samples_per_second": 21.179, | |
| "eval_steps_per_second": 5.298, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.27904391328515843, | |
| "grad_norm": 3.008030891418457, | |
| "learning_rate": 3.189374007355917e-05, | |
| "loss": 2.903, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.2801556420233463, | |
| "grad_norm": 2.9889330863952637, | |
| "learning_rate": 3.151889938850445e-05, | |
| "loss": 2.7301, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.2812673707615342, | |
| "grad_norm": 3.3704285621643066, | |
| "learning_rate": 3.114525791558398e-05, | |
| "loss": 2.8056, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.28237909949972206, | |
| "grad_norm": 3.12371563911438, | |
| "learning_rate": 3.0772839899857464e-05, | |
| "loss": 2.6181, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.28349082823790994, | |
| "grad_norm": 2.795156478881836, | |
| "learning_rate": 3.0401669506996256e-05, | |
| "loss": 2.5891, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2846025569760978, | |
| "grad_norm": 2.5517489910125732, | |
| "learning_rate": 3.003177082171523e-05, | |
| "loss": 2.5578, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.2857142857142857, | |
| "grad_norm": 2.131334066390991, | |
| "learning_rate": 2.9663167846209998e-05, | |
| "loss": 2.9474, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.2868260144524736, | |
| "grad_norm": 2.138378381729126, | |
| "learning_rate": 2.9295884498599414e-05, | |
| "loss": 2.5115, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.28793774319066145, | |
| "grad_norm": 1.636126160621643, | |
| "learning_rate": 2.8929944611373554e-05, | |
| "loss": 2.5024, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.28904947192884933, | |
| "grad_norm": 1.2389148473739624, | |
| "learning_rate": 2.8565371929847284e-05, | |
| "loss": 2.2814, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.29016120066703727, | |
| "grad_norm": 1.189886450767517, | |
| "learning_rate": 2.8202190110619493e-05, | |
| "loss": 2.4194, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.29127292940522514, | |
| "grad_norm": 1.0696269273757935, | |
| "learning_rate": 2.784042272003794e-05, | |
| "loss": 2.6107, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.292384658143413, | |
| "grad_norm": 0.9985766410827637, | |
| "learning_rate": 2.7480093232670158e-05, | |
| "loss": 2.5073, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.2934963868816009, | |
| "grad_norm": 0.9005837440490723, | |
| "learning_rate": 2.712122502978024e-05, | |
| "loss": 2.3042, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.2946081156197888, | |
| "grad_norm": 0.9823698997497559, | |
| "learning_rate": 2.6763841397811573e-05, | |
| "loss": 2.2399, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.29571984435797666, | |
| "grad_norm": 0.9442362189292908, | |
| "learning_rate": 2.64079655268759e-05, | |
| "loss": 2.4393, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.29683157309616454, | |
| "grad_norm": 0.831301748752594, | |
| "learning_rate": 2.605362050924848e-05, | |
| "loss": 2.2221, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.2979433018343524, | |
| "grad_norm": 0.9295997619628906, | |
| "learning_rate": 2.57008293378697e-05, | |
| "loss": 2.3975, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.2990550305725403, | |
| "grad_norm": 1.0188276767730713, | |
| "learning_rate": 2.534961490485313e-05, | |
| "loss": 2.4616, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.30016675931072817, | |
| "grad_norm": 0.9668642282485962, | |
| "learning_rate": 2.500000000000001e-05, | |
| "loss": 2.2432, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.30127848804891605, | |
| "grad_norm": 1.0286155939102173, | |
| "learning_rate": 2.4652007309320498e-05, | |
| "loss": 2.3834, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.3023902167871039, | |
| "grad_norm": 1.0591888427734375, | |
| "learning_rate": 2.430565941356157e-05, | |
| "loss": 2.3591, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.3035019455252918, | |
| "grad_norm": 1.0622197389602661, | |
| "learning_rate": 2.3960978786741877e-05, | |
| "loss": 2.4136, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.3046136742634797, | |
| "grad_norm": 0.9972384572029114, | |
| "learning_rate": 2.361798779469336e-05, | |
| "loss": 2.4249, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.3057254030016676, | |
| "grad_norm": 0.8855586647987366, | |
| "learning_rate": 2.3276708693609943e-05, | |
| "loss": 2.1604, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.3068371317398555, | |
| "grad_norm": 0.9017280340194702, | |
| "learning_rate": 2.2937163628603435e-05, | |
| "loss": 2.0953, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.3079488604780434, | |
| "grad_norm": 0.9258705377578735, | |
| "learning_rate": 2.259937463226651e-05, | |
| "loss": 2.1436, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.30906058921623125, | |
| "grad_norm": 0.8871088027954102, | |
| "learning_rate": 2.2263363623243054e-05, | |
| "loss": 2.3352, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.31017231795441913, | |
| "grad_norm": 0.9475242495536804, | |
| "learning_rate": 2.192915240480596e-05, | |
| "loss": 2.2745, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.311284046692607, | |
| "grad_norm": 0.9121392965316772, | |
| "learning_rate": 2.1596762663442218e-05, | |
| "loss": 2.0523, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3123957754307949, | |
| "grad_norm": 0.9183974266052246, | |
| "learning_rate": 2.1266215967445824e-05, | |
| "loss": 2.1313, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.31350750416898276, | |
| "grad_norm": 1.0411345958709717, | |
| "learning_rate": 2.0937533765518187e-05, | |
| "loss": 2.5137, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.31461923290717064, | |
| "grad_norm": 0.9422702193260193, | |
| "learning_rate": 2.061073738537635e-05, | |
| "loss": 2.3598, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.3157309616453585, | |
| "grad_norm": 0.9007568359375, | |
| "learning_rate": 2.0285848032369137e-05, | |
| "loss": 2.2845, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.3168426903835464, | |
| "grad_norm": 0.895376980304718, | |
| "learning_rate": 1.996288678810105e-05, | |
| "loss": 2.1736, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.3179544191217343, | |
| "grad_norm": 0.9709522128105164, | |
| "learning_rate": 1.9641874609064443e-05, | |
| "loss": 2.2274, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.31906614785992216, | |
| "grad_norm": 1.032238245010376, | |
| "learning_rate": 1.932283232527956e-05, | |
| "loss": 2.2555, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.32017787659811003, | |
| "grad_norm": 1.008985161781311, | |
| "learning_rate": 1.9005780638942982e-05, | |
| "loss": 2.3412, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.32128960533629797, | |
| "grad_norm": 0.9746498465538025, | |
| "learning_rate": 1.8690740123084316e-05, | |
| "loss": 2.1347, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.32240133407448585, | |
| "grad_norm": 1.0789058208465576, | |
| "learning_rate": 1.837773122023114e-05, | |
| "loss": 2.2996, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3235130628126737, | |
| "grad_norm": 1.0640016794204712, | |
| "learning_rate": 1.8066774241082612e-05, | |
| "loss": 2.1882, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.3246247915508616, | |
| "grad_norm": 1.2953685522079468, | |
| "learning_rate": 1.7757889363191483e-05, | |
| "loss": 2.3575, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.3257365202890495, | |
| "grad_norm": 1.0919634103775024, | |
| "learning_rate": 1.745109662965481e-05, | |
| "loss": 2.3441, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.32684824902723736, | |
| "grad_norm": 1.0896896123886108, | |
| "learning_rate": 1.714641594781347e-05, | |
| "loss": 2.4013, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.32795997776542524, | |
| "grad_norm": 1.1087121963500977, | |
| "learning_rate": 1.684386708796025e-05, | |
| "loss": 2.1645, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.3290717065036131, | |
| "grad_norm": 1.2468881607055664, | |
| "learning_rate": 1.6543469682057106e-05, | |
| "loss": 2.4229, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.330183435241801, | |
| "grad_norm": 1.247047781944275, | |
| "learning_rate": 1.62452432224612e-05, | |
| "loss": 2.4761, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.33129516397998887, | |
| "grad_norm": 1.446397066116333, | |
| "learning_rate": 1.5949207060660138e-05, | |
| "loss": 2.5412, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.33240689271817675, | |
| "grad_norm": 1.6580612659454346, | |
| "learning_rate": 1.5655380406016235e-05, | |
| "loss": 2.2901, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.33351862145636463, | |
| "grad_norm": 1.7994294166564941, | |
| "learning_rate": 1.536378232452003e-05, | |
| "loss": 2.4478, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.33351862145636463, | |
| "eval_loss": 2.3351998329162598, | |
| "eval_runtime": 71.5628, | |
| "eval_samples_per_second": 21.17, | |
| "eval_steps_per_second": 5.296, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3346303501945525, | |
| "grad_norm": 1.6746742725372314, | |
| "learning_rate": 1.5074431737553157e-05, | |
| "loss": 2.5023, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.3357420789327404, | |
| "grad_norm": 1.6056064367294312, | |
| "learning_rate": 1.4787347420660541e-05, | |
| "loss": 2.5727, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.3368538076709283, | |
| "grad_norm": 1.4742602109909058, | |
| "learning_rate": 1.4502548002332088e-05, | |
| "loss": 2.43, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.3379655364091162, | |
| "grad_norm": 1.530042052268982, | |
| "learning_rate": 1.422005196279395e-05, | |
| "loss": 2.5862, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.3390772651473041, | |
| "grad_norm": 1.5977087020874023, | |
| "learning_rate": 1.3939877632809278e-05, | |
| "loss": 2.5271, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.34018899388549195, | |
| "grad_norm": 1.583370327949524, | |
| "learning_rate": 1.3662043192488849e-05, | |
| "loss": 2.4107, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.34130072262367983, | |
| "grad_norm": 1.6869573593139648, | |
| "learning_rate": 1.338656667011134e-05, | |
| "loss": 2.6216, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.3424124513618677, | |
| "grad_norm": 1.4852731227874756, | |
| "learning_rate": 1.3113465940953495e-05, | |
| "loss": 2.5307, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.3435241801000556, | |
| "grad_norm": 1.2093337774276733, | |
| "learning_rate": 1.2842758726130283e-05, | |
| "loss": 2.2246, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.34463590883824347, | |
| "grad_norm": 1.3381179571151733, | |
| "learning_rate": 1.257446259144494e-05, | |
| "loss": 2.4152, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.34574763757643134, | |
| "grad_norm": 1.136741280555725, | |
| "learning_rate": 1.2308594946249163e-05, | |
| "loss": 2.5018, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.3468593663146192, | |
| "grad_norm": 0.9899182319641113, | |
| "learning_rate": 1.204517304231343e-05, | |
| "loss": 2.3935, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.3479710950528071, | |
| "grad_norm": 1.0916366577148438, | |
| "learning_rate": 1.178421397270758e-05, | |
| "loss": 2.5129, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.349082823790995, | |
| "grad_norm": 0.9697388410568237, | |
| "learning_rate": 1.1525734670691701e-05, | |
| "loss": 2.098, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.35019455252918286, | |
| "grad_norm": 0.952893078327179, | |
| "learning_rate": 1.1269751908617277e-05, | |
| "loss": 2.3895, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.35130628126737073, | |
| "grad_norm": 0.8646156191825867, | |
| "learning_rate": 1.1016282296838887e-05, | |
| "loss": 2.2197, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.35241801000555867, | |
| "grad_norm": 0.8228775858879089, | |
| "learning_rate": 1.0765342282636416e-05, | |
| "loss": 2.3013, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.35352973874374655, | |
| "grad_norm": 0.856402575969696, | |
| "learning_rate": 1.0516948149147754e-05, | |
| "loss": 2.4294, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.3546414674819344, | |
| "grad_norm": 1.03739333152771, | |
| "learning_rate": 1.0271116014312293e-05, | |
| "loss": 2.3878, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.3557531962201223, | |
| "grad_norm": 0.8302180767059326, | |
| "learning_rate": 1.0027861829824952e-05, | |
| "loss": 2.2219, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.3568649249583102, | |
| "grad_norm": 0.7932164072990417, | |
| "learning_rate": 9.787201380101157e-06, | |
| "loss": 2.2596, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.35797665369649806, | |
| "grad_norm": 0.8476454019546509, | |
| "learning_rate": 9.549150281252633e-06, | |
| "loss": 2.1147, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.35908838243468594, | |
| "grad_norm": 0.9964444637298584, | |
| "learning_rate": 9.313723980074018e-06, | |
| "loss": 2.3652, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.3602001111728738, | |
| "grad_norm": 1.0413466691970825, | |
| "learning_rate": 9.080937753040646e-06, | |
| "loss": 2.2266, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.3613118399110617, | |
| "grad_norm": 0.8853463530540466, | |
| "learning_rate": 8.850806705317183e-06, | |
| "loss": 2.2614, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.3624235686492496, | |
| "grad_norm": 0.7970969080924988, | |
| "learning_rate": 8.623345769777514e-06, | |
| "loss": 2.1004, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.36353529738743745, | |
| "grad_norm": 0.9213111996650696, | |
| "learning_rate": 8.398569706035792e-06, | |
| "loss": 2.0849, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.36464702612562533, | |
| "grad_norm": 0.9494043588638306, | |
| "learning_rate": 8.176493099488663e-06, | |
| "loss": 2.4137, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.3657587548638132, | |
| "grad_norm": 0.9899368286132812, | |
| "learning_rate": 7.957130360368898e-06, | |
| "loss": 2.301, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.3668704836020011, | |
| "grad_norm": 0.9502694606781006, | |
| "learning_rate": 7.740495722810271e-06, | |
| "loss": 2.2525, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.367982212340189, | |
| "grad_norm": 1.0422662496566772, | |
| "learning_rate": 7.526603243923957e-06, | |
| "loss": 2.2026, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.3690939410783769, | |
| "grad_norm": 1.1531425714492798, | |
| "learning_rate": 7.315466802886401e-06, | |
| "loss": 2.5027, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.3702056698165648, | |
| "grad_norm": 0.8836955428123474, | |
| "learning_rate": 7.107100100038671e-06, | |
| "loss": 2.1697, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.37131739855475265, | |
| "grad_norm": 0.9281957149505615, | |
| "learning_rate": 6.901516655997536e-06, | |
| "loss": 2.0896, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.37242912729294053, | |
| "grad_norm": 1.1078437566757202, | |
| "learning_rate": 6.698729810778065e-06, | |
| "loss": 2.1164, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3735408560311284, | |
| "grad_norm": 1.1739754676818848, | |
| "learning_rate": 6.498752722928042e-06, | |
| "loss": 2.2655, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.3746525847693163, | |
| "grad_norm": 1.0183756351470947, | |
| "learning_rate": 6.301598368674105e-06, | |
| "loss": 2.1442, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 0.37576431350750417, | |
| "grad_norm": 1.3615624904632568, | |
| "learning_rate": 6.107279541079769e-06, | |
| "loss": 2.1463, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 0.37687604224569204, | |
| "grad_norm": 1.0861144065856934, | |
| "learning_rate": 5.915808849215304e-06, | |
| "loss": 2.2384, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 0.3779877709838799, | |
| "grad_norm": 1.1558014154434204, | |
| "learning_rate": 5.727198717339511e-06, | |
| "loss": 2.1577, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3790994997220678, | |
| "grad_norm": 1.144195318222046, | |
| "learning_rate": 5.54146138409355e-06, | |
| "loss": 2.2819, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 0.3802112284602557, | |
| "grad_norm": 1.2792952060699463, | |
| "learning_rate": 5.358608901706802e-06, | |
| "loss": 2.2607, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 0.38132295719844356, | |
| "grad_norm": 1.2132809162139893, | |
| "learning_rate": 5.178653135214812e-06, | |
| "loss": 2.1884, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 0.38243468593663144, | |
| "grad_norm": 1.3561714887619019, | |
| "learning_rate": 5.001605761689398e-06, | |
| "loss": 2.2476, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 0.38354641467481937, | |
| "grad_norm": 1.3285537958145142, | |
| "learning_rate": 4.827478269480895e-06, | |
| "loss": 2.016, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.38465814341300725, | |
| "grad_norm": 1.4897688627243042, | |
| "learning_rate": 4.65628195747273e-06, | |
| "loss": 2.4268, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 0.3857698721511951, | |
| "grad_norm": 1.846487045288086, | |
| "learning_rate": 4.488027934348271e-06, | |
| "loss": 2.3771, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 0.386881600889383, | |
| "grad_norm": 1.8320105075836182, | |
| "learning_rate": 4.322727117869951e-06, | |
| "loss": 2.2541, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 0.3879933296275709, | |
| "grad_norm": 2.544713258743286, | |
| "learning_rate": 4.16039023417088e-06, | |
| "loss": 2.1886, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 0.38910505836575876, | |
| "grad_norm": 2.9872875213623047, | |
| "learning_rate": 4.001027817058789e-06, | |
| "loss": 2.288, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.38910505836575876, | |
| "eval_loss": 2.29748797416687, | |
| "eval_runtime": 71.5498, | |
| "eval_samples_per_second": 21.174, | |
| "eval_steps_per_second": 5.297, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.39021678710394664, | |
| "grad_norm": 0.9002532362937927, | |
| "learning_rate": 3.844650207332562e-06, | |
| "loss": 2.4953, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 0.3913285158421345, | |
| "grad_norm": 0.9280316829681396, | |
| "learning_rate": 3.691267552111183e-06, | |
| "loss": 2.6538, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 0.3924402445803224, | |
| "grad_norm": 0.9478785395622253, | |
| "learning_rate": 3.54088980417534e-06, | |
| "loss": 2.5276, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 0.3935519733185103, | |
| "grad_norm": 0.9327703714370728, | |
| "learning_rate": 3.393526721321616e-06, | |
| "loss": 2.487, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 0.39466370205669815, | |
| "grad_norm": 0.8927820920944214, | |
| "learning_rate": 3.249187865729264e-06, | |
| "loss": 2.3213, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.39577543079488603, | |
| "grad_norm": 0.9866592288017273, | |
| "learning_rate": 3.1078826033397843e-06, | |
| "loss": 2.4739, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 0.3968871595330739, | |
| "grad_norm": 0.9074956774711609, | |
| "learning_rate": 2.9696201032491434e-06, | |
| "loss": 2.4302, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 0.3979988882712618, | |
| "grad_norm": 0.8887799382209778, | |
| "learning_rate": 2.8344093371128424e-06, | |
| "loss": 2.6652, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 0.3991106170094497, | |
| "grad_norm": 0.9303501844406128, | |
| "learning_rate": 2.70225907856374e-06, | |
| "loss": 2.3618, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 0.4002223457476376, | |
| "grad_norm": 0.8939053416252136, | |
| "learning_rate": 2.573177902642726e-06, | |
| "loss": 2.3751, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.4013340744858255, | |
| "grad_norm": 0.8913921117782593, | |
| "learning_rate": 2.4471741852423237e-06, | |
| "loss": 2.2483, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 0.40244580322401335, | |
| "grad_norm": 0.8503320813179016, | |
| "learning_rate": 2.324256102563188e-06, | |
| "loss": 2.2931, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 0.40355753196220123, | |
| "grad_norm": 0.9348843693733215, | |
| "learning_rate": 2.204431630583548e-06, | |
| "loss": 2.4732, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 0.4046692607003891, | |
| "grad_norm": 0.8464366793632507, | |
| "learning_rate": 2.087708544541689e-06, | |
| "loss": 2.1994, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 0.405780989438577, | |
| "grad_norm": 0.8891509175300598, | |
| "learning_rate": 1.974094418431388e-06, | |
| "loss": 2.3163, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.40689271817676487, | |
| "grad_norm": 0.8961646556854248, | |
| "learning_rate": 1.8635966245104664e-06, | |
| "loss": 2.2609, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 0.40800444691495275, | |
| "grad_norm": 0.8280013799667358, | |
| "learning_rate": 1.7562223328224325e-06, | |
| "loss": 2.1203, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 0.4091161756531406, | |
| "grad_norm": 0.8764577507972717, | |
| "learning_rate": 1.6519785107311891e-06, | |
| "loss": 2.1943, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 0.4102279043913285, | |
| "grad_norm": 0.8844929337501526, | |
| "learning_rate": 1.5508719224689717e-06, | |
| "loss": 2.3813, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 0.4113396331295164, | |
| "grad_norm": 0.9021992087364197, | |
| "learning_rate": 1.4529091286973995e-06, | |
| "loss": 2.2267, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.41245136186770426, | |
| "grad_norm": 0.8152535557746887, | |
| "learning_rate": 1.358096486081778e-06, | |
| "loss": 2.1442, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 0.41356309060589214, | |
| "grad_norm": 0.9210073947906494, | |
| "learning_rate": 1.2664401468786114e-06, | |
| "loss": 2.2703, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 0.41467481934408007, | |
| "grad_norm": 0.7856800556182861, | |
| "learning_rate": 1.1779460585363944e-06, | |
| "loss": 2.187, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 0.41578654808226795, | |
| "grad_norm": 0.9219398498535156, | |
| "learning_rate": 1.0926199633097157e-06, | |
| "loss": 2.538, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 0.4168982768204558, | |
| "grad_norm": 0.8674458265304565, | |
| "learning_rate": 1.0104673978866164e-06, | |
| "loss": 2.3901, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.4180100055586437, | |
| "grad_norm": 0.8765695095062256, | |
| "learning_rate": 9.314936930293283e-07, | |
| "loss": 2.1564, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 0.4191217342968316, | |
| "grad_norm": 0.7773668766021729, | |
| "learning_rate": 8.557039732283944e-07, | |
| "loss": 2.1513, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 0.42023346303501946, | |
| "grad_norm": 0.799339771270752, | |
| "learning_rate": 7.83103156370113e-07, | |
| "loss": 2.0931, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 0.42134519177320734, | |
| "grad_norm": 0.8551070094108582, | |
| "learning_rate": 7.136959534174592e-07, | |
| "loss": 2.0482, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 0.4224569205113952, | |
| "grad_norm": 0.8003242015838623, | |
| "learning_rate": 6.474868681043578e-07, | |
| "loss": 2.0851, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.4235686492495831, | |
| "grad_norm": 0.7685055732727051, | |
| "learning_rate": 5.844801966434832e-07, | |
| "loss": 1.9849, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 0.424680377987771, | |
| "grad_norm": 0.8753073215484619, | |
| "learning_rate": 5.246800274474439e-07, | |
| "loss": 2.1241, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 0.42579210672595885, | |
| "grad_norm": 0.8959077000617981, | |
| "learning_rate": 4.680902408635335e-07, | |
| "loss": 2.2865, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 0.42690383546414673, | |
| "grad_norm": 0.9518845677375793, | |
| "learning_rate": 4.1471450892189846e-07, | |
| "loss": 2.1737, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 0.4280155642023346, | |
| "grad_norm": 0.885158896446228, | |
| "learning_rate": 3.6455629509730136e-07, | |
| "loss": 2.0735, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.4291272929405225, | |
| "grad_norm": 0.9893671274185181, | |
| "learning_rate": 3.1761885408435054e-07, | |
| "loss": 2.2165, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 0.4302390216787104, | |
| "grad_norm": 0.8594589829444885, | |
| "learning_rate": 2.7390523158633554e-07, | |
| "loss": 2.1061, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 0.4313507504168983, | |
| "grad_norm": 0.9702785015106201, | |
| "learning_rate": 2.334182641175686e-07, | |
| "loss": 2.0657, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 0.4324624791550862, | |
| "grad_norm": 1.0460872650146484, | |
| "learning_rate": 1.9616057881935436e-07, | |
| "loss": 2.0916, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 0.43357420789327406, | |
| "grad_norm": 1.0313807725906372, | |
| "learning_rate": 1.6213459328950352e-07, | |
| "loss": 2.2682, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.43468593663146193, | |
| "grad_norm": 1.0293152332305908, | |
| "learning_rate": 1.3134251542544774e-07, | |
| "loss": 2.076, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 0.4357976653696498, | |
| "grad_norm": 1.0680193901062012, | |
| "learning_rate": 1.0378634328099269e-07, | |
| "loss": 2.0364, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 0.4369093941078377, | |
| "grad_norm": 1.0686061382293701, | |
| "learning_rate": 7.946786493666647e-08, | |
| "loss": 2.1607, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 0.43802112284602557, | |
| "grad_norm": 1.0303860902786255, | |
| "learning_rate": 5.838865838366792e-08, | |
| "loss": 2.1599, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 0.43913285158421345, | |
| "grad_norm": 1.2263742685317993, | |
| "learning_rate": 4.055009142152067e-08, | |
| "loss": 2.1796, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.4402445803224013, | |
| "grad_norm": 1.4119532108306885, | |
| "learning_rate": 2.595332156925534e-08, | |
| "loss": 2.2719, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 0.4413563090605892, | |
| "grad_norm": 1.4214668273925781, | |
| "learning_rate": 1.4599295990352924e-08, | |
| "loss": 2.3379, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 0.4424680377987771, | |
| "grad_norm": 1.72279691696167, | |
| "learning_rate": 6.488751431266149e-09, | |
| "loss": 2.2098, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 0.44357976653696496, | |
| "grad_norm": 1.8887008428573608, | |
| "learning_rate": 1.622214173602199e-09, | |
| "loss": 2.1835, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 0.44469149527515284, | |
| "grad_norm": 3.1898610591888428, | |
| "learning_rate": 0.0, | |
| "loss": 2.4357, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.44469149527515284, | |
| "eval_loss": 2.29616379737854, | |
| "eval_runtime": 71.5597, | |
| "eval_samples_per_second": 21.171, | |
| "eval_steps_per_second": 5.296, | |
| "step": 400 | |
| } | |
| ], | |
| "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": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.276333740621824e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |