| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9739884393063583, | |
| "eval_steps": 129, | |
| "global_step": 258, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007707129094412331, | |
| "grad_norm": 27.94101905822754, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 2.3083, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007707129094412331, | |
| "eval_loss": 2.273209810256958, | |
| "eval_runtime": 27.5737, | |
| "eval_samples_per_second": 1.813, | |
| "eval_steps_per_second": 0.471, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.015414258188824663, | |
| "grad_norm": 27.76470184326172, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 2.2483, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.023121387283236993, | |
| "grad_norm": 28.12770652770996, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 2.29, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.030828516377649325, | |
| "grad_norm": 29.105730056762695, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 2.3396, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.038535645472061654, | |
| "grad_norm": 28.90327262878418, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 2.2654, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.046242774566473986, | |
| "grad_norm": 30.192190170288086, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 2.2205, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.05394990366088632, | |
| "grad_norm": 27.94385528564453, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 2.2161, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.06165703275529865, | |
| "grad_norm": 17.948062896728516, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 2.0695, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.06936416184971098, | |
| "grad_norm": 17.047054290771484, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 2.0742, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.07707129094412331, | |
| "grad_norm": 15.427838325500488, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.9962, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08477842003853564, | |
| "grad_norm": 6.273721694946289, | |
| "learning_rate": 2.2e-06, | |
| "loss": 1.859, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.09248554913294797, | |
| "grad_norm": 5.851183891296387, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 1.8696, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1001926782273603, | |
| "grad_norm": 4.672307968139648, | |
| "learning_rate": 2.6e-06, | |
| "loss": 1.8348, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.10789980732177264, | |
| "grad_norm": 2.6562161445617676, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 1.7719, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.11560693641618497, | |
| "grad_norm": 4.172055721282959, | |
| "learning_rate": 3e-06, | |
| "loss": 1.7984, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1233140655105973, | |
| "grad_norm": 4.055249214172363, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 1.8025, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.13102119460500963, | |
| "grad_norm": 3.3719887733459473, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 1.739, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.13872832369942195, | |
| "grad_norm": 2.81038498878479, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 1.7478, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.1464354527938343, | |
| "grad_norm": 2.2064859867095947, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 1.7384, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.15414258188824662, | |
| "grad_norm": 1.940885305404663, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.7135, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.16184971098265896, | |
| "grad_norm": 1.9488122463226318, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 1.7108, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.16955684007707128, | |
| "grad_norm": 1.7049647569656372, | |
| "learning_rate": 4.4e-06, | |
| "loss": 1.6868, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.17726396917148363, | |
| "grad_norm": 1.5429236888885498, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 1.6947, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.18497109826589594, | |
| "grad_norm": 1.5304620265960693, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 1.6846, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.1926782273603083, | |
| "grad_norm": 1.5696897506713867, | |
| "learning_rate": 5e-06, | |
| "loss": 1.6844, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.2003853564547206, | |
| "grad_norm": 1.4362632036209106, | |
| "learning_rate": 5.2e-06, | |
| "loss": 1.6732, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.20809248554913296, | |
| "grad_norm": 1.3416928052902222, | |
| "learning_rate": 5.400000000000001e-06, | |
| "loss": 1.6424, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.21579961464354527, | |
| "grad_norm": 1.3142507076263428, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 1.677, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.22350674373795762, | |
| "grad_norm": 1.342984676361084, | |
| "learning_rate": 5.8e-06, | |
| "loss": 1.6762, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.23121387283236994, | |
| "grad_norm": 1.2972025871276855, | |
| "learning_rate": 6e-06, | |
| "loss": 1.6716, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.23892100192678228, | |
| "grad_norm": 1.2904590368270874, | |
| "learning_rate": 6.200000000000001e-06, | |
| "loss": 1.6234, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2466281310211946, | |
| "grad_norm": 1.1942962408065796, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 1.6533, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2543352601156069, | |
| "grad_norm": 1.124014973640442, | |
| "learning_rate": 6.600000000000001e-06, | |
| "loss": 1.6604, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.26204238921001927, | |
| "grad_norm": 1.2125813961029053, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 1.6335, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.2697495183044316, | |
| "grad_norm": 1.2104367017745972, | |
| "learning_rate": 7e-06, | |
| "loss": 1.6356, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.2774566473988439, | |
| "grad_norm": 1.1877591609954834, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 1.6349, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.28516377649325625, | |
| "grad_norm": 1.2402458190917969, | |
| "learning_rate": 7.4e-06, | |
| "loss": 1.6463, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2928709055876686, | |
| "grad_norm": 1.1922346353530884, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 1.5998, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.30057803468208094, | |
| "grad_norm": 1.197464942932129, | |
| "learning_rate": 7.800000000000002e-06, | |
| "loss": 1.6265, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.30828516377649323, | |
| "grad_norm": 1.291739821434021, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.6077, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.3159922928709056, | |
| "grad_norm": 1.145663857460022, | |
| "learning_rate": 8.2e-06, | |
| "loss": 1.6152, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.3236994219653179, | |
| "grad_norm": 1.1572788953781128, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 1.6058, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.33140655105973027, | |
| "grad_norm": 1.3273899555206299, | |
| "learning_rate": 8.6e-06, | |
| "loss": 1.6223, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.33911368015414256, | |
| "grad_norm": 1.1160943508148193, | |
| "learning_rate": 8.8e-06, | |
| "loss": 1.5969, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.3468208092485549, | |
| "grad_norm": 1.3087902069091797, | |
| "learning_rate": 9e-06, | |
| "loss": 1.6464, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.35452793834296725, | |
| "grad_norm": 1.1589637994766235, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 1.5799, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.3622350674373796, | |
| "grad_norm": 1.159191370010376, | |
| "learning_rate": 9.4e-06, | |
| "loss": 1.6153, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.3699421965317919, | |
| "grad_norm": 1.206766128540039, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 1.5982, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.37764932562620424, | |
| "grad_norm": 1.1924678087234497, | |
| "learning_rate": 9.800000000000001e-06, | |
| "loss": 1.6054, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.3853564547206166, | |
| "grad_norm": 1.2029445171356201, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6205, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3930635838150289, | |
| "grad_norm": 1.1406632661819458, | |
| "learning_rate": 1.02e-05, | |
| "loss": 1.6158, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.4007707129094412, | |
| "grad_norm": 1.1437443494796753, | |
| "learning_rate": 1.04e-05, | |
| "loss": 1.6045, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.40847784200385356, | |
| "grad_norm": 1.127734661102295, | |
| "learning_rate": 1.0600000000000002e-05, | |
| "loss": 1.5968, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.4161849710982659, | |
| "grad_norm": 1.1851099729537964, | |
| "learning_rate": 1.0800000000000002e-05, | |
| "loss": 1.6045, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.4238921001926782, | |
| "grad_norm": 1.1298301219940186, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 1.5908, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.43159922928709055, | |
| "grad_norm": 1.095090627670288, | |
| "learning_rate": 1.1200000000000001e-05, | |
| "loss": 1.5901, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.4393063583815029, | |
| "grad_norm": 1.1739152669906616, | |
| "learning_rate": 1.14e-05, | |
| "loss": 1.6275, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.44701348747591524, | |
| "grad_norm": 1.1687606573104858, | |
| "learning_rate": 1.16e-05, | |
| "loss": 1.5938, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.45472061657032753, | |
| "grad_norm": 1.1895908117294312, | |
| "learning_rate": 1.18e-05, | |
| "loss": 1.6016, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.4624277456647399, | |
| "grad_norm": 1.199129581451416, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.6317, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.4701348747591522, | |
| "grad_norm": 1.2785886526107788, | |
| "learning_rate": 1.22e-05, | |
| "loss": 1.5672, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.47784200385356457, | |
| "grad_norm": 1.2036688327789307, | |
| "learning_rate": 1.2400000000000002e-05, | |
| "loss": 1.5636, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.48554913294797686, | |
| "grad_norm": 1.2586396932601929, | |
| "learning_rate": 1.2600000000000001e-05, | |
| "loss": 1.5806, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.4932562620423892, | |
| "grad_norm": 1.1760581731796265, | |
| "learning_rate": 1.2800000000000001e-05, | |
| "loss": 1.5724, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5009633911368016, | |
| "grad_norm": 1.1171916723251343, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 1.5982, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5086705202312138, | |
| "grad_norm": 1.234012484550476, | |
| "learning_rate": 1.3200000000000002e-05, | |
| "loss": 1.597, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.5163776493256262, | |
| "grad_norm": 1.1812013387680054, | |
| "learning_rate": 1.3400000000000002e-05, | |
| "loss": 1.6064, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.5240847784200385, | |
| "grad_norm": 1.1740922927856445, | |
| "learning_rate": 1.3600000000000002e-05, | |
| "loss": 1.5915, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.5317919075144508, | |
| "grad_norm": 1.277176856994629, | |
| "learning_rate": 1.38e-05, | |
| "loss": 1.5711, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.5394990366088632, | |
| "grad_norm": 1.1419289112091064, | |
| "learning_rate": 1.4e-05, | |
| "loss": 1.5934, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.5472061657032755, | |
| "grad_norm": 1.2002787590026855, | |
| "learning_rate": 1.4200000000000001e-05, | |
| "loss": 1.6021, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.5549132947976878, | |
| "grad_norm": 1.1920689344406128, | |
| "learning_rate": 1.4400000000000001e-05, | |
| "loss": 1.5893, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.5626204238921002, | |
| "grad_norm": 1.2546113729476929, | |
| "learning_rate": 1.46e-05, | |
| "loss": 1.558, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.5703275529865125, | |
| "grad_norm": 1.2610082626342773, | |
| "learning_rate": 1.48e-05, | |
| "loss": 1.5842, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.5780346820809249, | |
| "grad_norm": 1.1725729703903198, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 1.5746, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.5857418111753372, | |
| "grad_norm": 1.1732540130615234, | |
| "learning_rate": 1.5200000000000002e-05, | |
| "loss": 1.5804, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.5934489402697495, | |
| "grad_norm": 1.281145691871643, | |
| "learning_rate": 1.54e-05, | |
| "loss": 1.5884, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.6011560693641619, | |
| "grad_norm": 1.1668535470962524, | |
| "learning_rate": 1.5600000000000003e-05, | |
| "loss": 1.6048, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.6088631984585742, | |
| "grad_norm": 1.2680914402008057, | |
| "learning_rate": 1.58e-05, | |
| "loss": 1.5893, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.6165703275529865, | |
| "grad_norm": 1.1659042835235596, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.5791, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6242774566473989, | |
| "grad_norm": 1.2156031131744385, | |
| "learning_rate": 1.62e-05, | |
| "loss": 1.5916, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.6319845857418112, | |
| "grad_norm": 1.1217319965362549, | |
| "learning_rate": 1.64e-05, | |
| "loss": 1.5725, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.6396917148362236, | |
| "grad_norm": 1.307479977607727, | |
| "learning_rate": 1.66e-05, | |
| "loss": 1.5723, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.6473988439306358, | |
| "grad_norm": 1.1636345386505127, | |
| "learning_rate": 1.6800000000000002e-05, | |
| "loss": 1.5906, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.6551059730250481, | |
| "grad_norm": 1.3260914087295532, | |
| "learning_rate": 1.7e-05, | |
| "loss": 1.5972, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.6628131021194605, | |
| "grad_norm": 1.14360511302948, | |
| "learning_rate": 1.72e-05, | |
| "loss": 1.5621, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.6705202312138728, | |
| "grad_norm": 1.4284840822219849, | |
| "learning_rate": 1.7400000000000003e-05, | |
| "loss": 1.571, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.6782273603082851, | |
| "grad_norm": 1.1513473987579346, | |
| "learning_rate": 1.76e-05, | |
| "loss": 1.6015, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.6859344894026975, | |
| "grad_norm": 1.3102519512176514, | |
| "learning_rate": 1.7800000000000002e-05, | |
| "loss": 1.576, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.6936416184971098, | |
| "grad_norm": 1.2329882383346558, | |
| "learning_rate": 1.8e-05, | |
| "loss": 1.5759, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7013487475915221, | |
| "grad_norm": 1.1875412464141846, | |
| "learning_rate": 1.8200000000000002e-05, | |
| "loss": 1.555, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.7090558766859345, | |
| "grad_norm": 1.1887799501419067, | |
| "learning_rate": 1.8400000000000003e-05, | |
| "loss": 1.5926, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.7167630057803468, | |
| "grad_norm": 1.3002405166625977, | |
| "learning_rate": 1.86e-05, | |
| "loss": 1.5849, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.7244701348747592, | |
| "grad_norm": 1.194841980934143, | |
| "learning_rate": 1.88e-05, | |
| "loss": 1.5724, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.7321772639691715, | |
| "grad_norm": 1.315577745437622, | |
| "learning_rate": 1.9e-05, | |
| "loss": 1.5296, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.7398843930635838, | |
| "grad_norm": 1.239837884902954, | |
| "learning_rate": 1.9200000000000003e-05, | |
| "loss": 1.5845, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.7475915221579962, | |
| "grad_norm": 1.3335014581680298, | |
| "learning_rate": 1.94e-05, | |
| "loss": 1.5843, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.7552986512524085, | |
| "grad_norm": 1.2278801202774048, | |
| "learning_rate": 1.9600000000000002e-05, | |
| "loss": 1.588, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.7630057803468208, | |
| "grad_norm": 1.3168463706970215, | |
| "learning_rate": 1.98e-05, | |
| "loss": 1.5758, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.7707129094412332, | |
| "grad_norm": 1.3854187726974487, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5489, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7784200385356455, | |
| "grad_norm": 1.1883262395858765, | |
| "learning_rate": 1.9998023297700656e-05, | |
| "loss": 1.5707, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.7861271676300579, | |
| "grad_norm": 1.3683229684829712, | |
| "learning_rate": 1.999209397227302e-05, | |
| "loss": 1.5916, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.7938342967244701, | |
| "grad_norm": 1.2290884256362915, | |
| "learning_rate": 1.998221436781933e-05, | |
| "loss": 1.6028, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.8015414258188824, | |
| "grad_norm": 1.422328233718872, | |
| "learning_rate": 1.996838839014696e-05, | |
| "loss": 1.5753, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.8092485549132948, | |
| "grad_norm": 1.2604609727859497, | |
| "learning_rate": 1.9950621505224276e-05, | |
| "loss": 1.5537, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.8169556840077071, | |
| "grad_norm": 1.1725685596466064, | |
| "learning_rate": 1.9928920737019735e-05, | |
| "loss": 1.5745, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.8246628131021194, | |
| "grad_norm": 1.284792423248291, | |
| "learning_rate": 1.9903294664725023e-05, | |
| "loss": 1.5868, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.8323699421965318, | |
| "grad_norm": 1.1779919862747192, | |
| "learning_rate": 1.9873753419363336e-05, | |
| "loss": 1.5824, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.8400770712909441, | |
| "grad_norm": 1.1214483976364136, | |
| "learning_rate": 1.9840308679784207e-05, | |
| "loss": 1.5486, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.8477842003853564, | |
| "grad_norm": 1.2342500686645508, | |
| "learning_rate": 1.9802973668046364e-05, | |
| "loss": 1.5867, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.8554913294797688, | |
| "grad_norm": 1.1915156841278076, | |
| "learning_rate": 1.976176314419051e-05, | |
| "loss": 1.579, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.8631984585741811, | |
| "grad_norm": 1.215820550918579, | |
| "learning_rate": 1.97166934004041e-05, | |
| "loss": 1.6014, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.8709055876685935, | |
| "grad_norm": 1.2331247329711914, | |
| "learning_rate": 1.9667782254580373e-05, | |
| "loss": 1.5653, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.8786127167630058, | |
| "grad_norm": 1.1282511949539185, | |
| "learning_rate": 1.9615049043274207e-05, | |
| "loss": 1.5584, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.8863198458574181, | |
| "grad_norm": 1.2783879041671753, | |
| "learning_rate": 1.955851461405761e-05, | |
| "loss": 1.5956, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.8940269749518305, | |
| "grad_norm": 1.2387332916259766, | |
| "learning_rate": 1.949820131727783e-05, | |
| "loss": 1.5604, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.9017341040462428, | |
| "grad_norm": 1.3010255098342896, | |
| "learning_rate": 1.9434132997221347e-05, | |
| "loss": 1.5557, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.9094412331406551, | |
| "grad_norm": 1.3249139785766602, | |
| "learning_rate": 1.936633498268728e-05, | |
| "loss": 1.561, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.9171483622350675, | |
| "grad_norm": 1.2389734983444214, | |
| "learning_rate": 1.9294834076973872e-05, | |
| "loss": 1.5726, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.9248554913294798, | |
| "grad_norm": 1.258575677871704, | |
| "learning_rate": 1.921965854728207e-05, | |
| "loss": 1.5504, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.9325626204238922, | |
| "grad_norm": 1.2949562072753906, | |
| "learning_rate": 1.9140838113540347e-05, | |
| "loss": 1.576, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.9402697495183044, | |
| "grad_norm": 1.2721818685531616, | |
| "learning_rate": 1.9058403936655235e-05, | |
| "loss": 1.5697, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.9479768786127167, | |
| "grad_norm": 1.198541522026062, | |
| "learning_rate": 1.8972388606192124e-05, | |
| "loss": 1.5672, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.9556840077071291, | |
| "grad_norm": 1.2318319082260132, | |
| "learning_rate": 1.888282612749132e-05, | |
| "loss": 1.5511, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.9633911368015414, | |
| "grad_norm": 1.3235722780227661, | |
| "learning_rate": 1.878975190822434e-05, | |
| "loss": 1.5972, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.9710982658959537, | |
| "grad_norm": 1.2950528860092163, | |
| "learning_rate": 1.869320274439583e-05, | |
| "loss": 1.5696, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.9788053949903661, | |
| "grad_norm": 1.2997064590454102, | |
| "learning_rate": 1.8593216805796612e-05, | |
| "loss": 1.5751, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.9865125240847784, | |
| "grad_norm": 1.429874062538147, | |
| "learning_rate": 1.8489833620913644e-05, | |
| "loss": 1.5706, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.9942196531791907, | |
| "grad_norm": 1.2658491134643555, | |
| "learning_rate": 1.8383094061302767e-05, | |
| "loss": 1.5681, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.9942196531791907, | |
| "eval_loss": 1.590910792350769, | |
| "eval_runtime": 27.4214, | |
| "eval_samples_per_second": 1.823, | |
| "eval_steps_per_second": 0.474, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.001926782273603, | |
| "grad_norm": 1.3823826313018799, | |
| "learning_rate": 1.8273040325430575e-05, | |
| "loss": 1.5453, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.0096339113680155, | |
| "grad_norm": 1.174560546875, | |
| "learning_rate": 1.8159715921991612e-05, | |
| "loss": 1.5485, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.0028901734104045, | |
| "grad_norm": 1.3361918926239014, | |
| "learning_rate": 1.804316565270765e-05, | |
| "loss": 1.511, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.010597302504817, | |
| "grad_norm": 1.5681639909744263, | |
| "learning_rate": 1.7923435594615744e-05, | |
| "loss": 1.3985, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.0183044315992293, | |
| "grad_norm": 1.3742421865463257, | |
| "learning_rate": 1.7800573081852124e-05, | |
| "loss": 1.4214, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.0260115606936415, | |
| "grad_norm": 1.375709056854248, | |
| "learning_rate": 1.767462668693908e-05, | |
| "loss": 1.3857, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.033718689788054, | |
| "grad_norm": 1.5805290937423706, | |
| "learning_rate": 1.7545646201582304e-05, | |
| "loss": 1.3975, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.0414258188824663, | |
| "grad_norm": 1.4905924797058105, | |
| "learning_rate": 1.7413682616986185e-05, | |
| "loss": 1.3663, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.0491329479768785, | |
| "grad_norm": 1.4072234630584717, | |
| "learning_rate": 1.7278788103694944e-05, | |
| "loss": 1.3969, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.056840077071291, | |
| "grad_norm": 1.3707107305526733, | |
| "learning_rate": 1.71410159909675e-05, | |
| "loss": 1.3932, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.0645472061657033, | |
| "grad_norm": 1.376590609550476, | |
| "learning_rate": 1.7000420745694256e-05, | |
| "loss": 1.3656, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.0722543352601157, | |
| "grad_norm": 1.2771958112716675, | |
| "learning_rate": 1.6857057950864134e-05, | |
| "loss": 1.3694, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.079961464354528, | |
| "grad_norm": 1.4592013359069824, | |
| "learning_rate": 1.671098428359037e-05, | |
| "loss": 1.3913, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.0876685934489403, | |
| "grad_norm": 1.3110437393188477, | |
| "learning_rate": 1.6562257492703756e-05, | |
| "loss": 1.3768, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.0953757225433527, | |
| "grad_norm": 1.344575047492981, | |
| "learning_rate": 1.64109363759222e-05, | |
| "loss": 1.3778, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.1030828516377649, | |
| "grad_norm": 1.277384638786316, | |
| "learning_rate": 1.62570807566056e-05, | |
| "loss": 1.3499, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.1107899807321773, | |
| "grad_norm": 1.2886083126068115, | |
| "learning_rate": 1.6100751460105244e-05, | |
| "loss": 1.3669, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.1184971098265897, | |
| "grad_norm": 1.3069369792938232, | |
| "learning_rate": 1.5942010289717108e-05, | |
| "loss": 1.3918, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.1262042389210019, | |
| "grad_norm": 1.2955520153045654, | |
| "learning_rate": 1.5780920002248484e-05, | |
| "loss": 1.3645, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.1339113680154143, | |
| "grad_norm": 1.3005629777908325, | |
| "learning_rate": 1.561754428320771e-05, | |
| "loss": 1.3522, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.1416184971098267, | |
| "grad_norm": 1.413831114768982, | |
| "learning_rate": 1.5451947721626676e-05, | |
| "loss": 1.4064, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.1493256262042388, | |
| "grad_norm": 1.2129186391830444, | |
| "learning_rate": 1.5284195784526196e-05, | |
| "loss": 1.3576, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.1570327552986512, | |
| "grad_norm": 1.3991036415100098, | |
| "learning_rate": 1.5114354791034225e-05, | |
| "loss": 1.3735, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.1647398843930636, | |
| "grad_norm": 1.2813304662704468, | |
| "learning_rate": 1.494249188616723e-05, | |
| "loss": 1.3689, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.1724470134874758, | |
| "grad_norm": 1.3265056610107422, | |
| "learning_rate": 1.4768675014285063e-05, | |
| "loss": 1.3714, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.1801541425818882, | |
| "grad_norm": 1.244061827659607, | |
| "learning_rate": 1.4592972892229779e-05, | |
| "loss": 1.371, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.1878612716763006, | |
| "grad_norm": 1.2477822303771973, | |
| "learning_rate": 1.4415454982159121e-05, | |
| "loss": 1.3705, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.1955684007707128, | |
| "grad_norm": 1.3200701475143433, | |
| "learning_rate": 1.4236191464085286e-05, | |
| "loss": 1.3657, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.2032755298651252, | |
| "grad_norm": 1.237042784690857, | |
| "learning_rate": 1.405525320812994e-05, | |
| "loss": 1.3602, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.2109826589595376, | |
| "grad_norm": 1.30637526512146, | |
| "learning_rate": 1.3872711746506413e-05, | |
| "loss": 1.3758, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.21868978805395, | |
| "grad_norm": 1.3186436891555786, | |
| "learning_rate": 1.3688639245240078e-05, | |
| "loss": 1.3907, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.2263969171483622, | |
| "grad_norm": 1.2071219682693481, | |
| "learning_rate": 1.3503108475638244e-05, | |
| "loss": 1.3698, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.2341040462427746, | |
| "grad_norm": 1.1885581016540527, | |
| "learning_rate": 1.331619278552068e-05, | |
| "loss": 1.3774, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.241811175337187, | |
| "grad_norm": 1.1943105459213257, | |
| "learning_rate": 1.3127966070222273e-05, | |
| "loss": 1.3538, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.2495183044315992, | |
| "grad_norm": 1.1982208490371704, | |
| "learning_rate": 1.2938502743379212e-05, | |
| "loss": 1.3797, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.2572254335260116, | |
| "grad_norm": 1.191636562347412, | |
| "learning_rate": 1.2747877707510252e-05, | |
| "loss": 1.371, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.264932562620424, | |
| "grad_norm": 1.2649930715560913, | |
| "learning_rate": 1.2556166324404747e-05, | |
| "loss": 1.3789, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.2726396917148362, | |
| "grad_norm": 1.206629753112793, | |
| "learning_rate": 1.2363444385329052e-05, | |
| "loss": 1.4232, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.2803468208092486, | |
| "grad_norm": 1.3122280836105347, | |
| "learning_rate": 1.2169788081063181e-05, | |
| "loss": 1.3871, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.288053949903661, | |
| "grad_norm": 1.1735293865203857, | |
| "learning_rate": 1.1975273971779528e-05, | |
| "loss": 1.3741, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.2957610789980731, | |
| "grad_norm": 1.3187175989151, | |
| "learning_rate": 1.1779978956775507e-05, | |
| "loss": 1.3644, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.3034682080924855, | |
| "grad_norm": 1.2720284461975098, | |
| "learning_rate": 1.158398024407215e-05, | |
| "loss": 1.3661, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.311175337186898, | |
| "grad_norm": 1.3094247579574585, | |
| "learning_rate": 1.1387355319890685e-05, | |
| "loss": 1.3617, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.3188824662813103, | |
| "grad_norm": 1.2710013389587402, | |
| "learning_rate": 1.119018191801905e-05, | |
| "loss": 1.373, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.3265895953757225, | |
| "grad_norm": 1.2845216989517212, | |
| "learning_rate": 1.0992537989080618e-05, | |
| "loss": 1.3712, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.334296724470135, | |
| "grad_norm": 1.277942419052124, | |
| "learning_rate": 1.0794501669717146e-05, | |
| "loss": 1.3676, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.342003853564547, | |
| "grad_norm": 1.190983533859253, | |
| "learning_rate": 1.05961512516982e-05, | |
| "loss": 1.3906, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.3497109826589595, | |
| "grad_norm": 1.3649415969848633, | |
| "learning_rate": 1.039756515096926e-05, | |
| "loss": 1.3883, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.357418111753372, | |
| "grad_norm": 1.2454570531845093, | |
| "learning_rate": 1.0198821876650702e-05, | |
| "loss": 1.3581, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.3651252408477843, | |
| "grad_norm": 1.2593861818313599, | |
| "learning_rate": 1e-05, | |
| "loss": 1.3726, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.3728323699421965, | |
| "grad_norm": 1.2473970651626587, | |
| "learning_rate": 9.801178123349298e-06, | |
| "loss": 1.4003, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.3805394990366089, | |
| "grad_norm": 1.210317611694336, | |
| "learning_rate": 9.602434849030747e-06, | |
| "loss": 1.3875, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.388246628131021, | |
| "grad_norm": 1.2112162113189697, | |
| "learning_rate": 9.403848748301802e-06, | |
| "loss": 1.3769, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.3959537572254335, | |
| "grad_norm": 1.1812710762023926, | |
| "learning_rate": 9.205498330282857e-06, | |
| "loss": 1.3521, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.4036608863198459, | |
| "grad_norm": 1.2227439880371094, | |
| "learning_rate": 9.007462010919387e-06, | |
| "loss": 1.3781, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.4113680154142583, | |
| "grad_norm": 1.2190202474594116, | |
| "learning_rate": 8.809818081980954e-06, | |
| "loss": 1.3529, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.4190751445086704, | |
| "grad_norm": 1.1302087306976318, | |
| "learning_rate": 8.61264468010932e-06, | |
| "loss": 1.3937, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.4267822736030829, | |
| "grad_norm": 1.1406745910644531, | |
| "learning_rate": 8.416019755927851e-06, | |
| "loss": 1.3715, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.4344894026974953, | |
| "grad_norm": 1.1421207189559937, | |
| "learning_rate": 8.2200210432245e-06, | |
| "loss": 1.3441, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.4421965317919074, | |
| "grad_norm": 1.1282238960266113, | |
| "learning_rate": 8.024726028220474e-06, | |
| "loss": 1.3484, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.4499036608863198, | |
| "grad_norm": 1.1182270050048828, | |
| "learning_rate": 7.83021191893682e-06, | |
| "loss": 1.3736, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.4576107899807322, | |
| "grad_norm": 1.1618040800094604, | |
| "learning_rate": 7.636555614670953e-06, | |
| "loss": 1.3481, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.4653179190751446, | |
| "grad_norm": 1.1137522459030151, | |
| "learning_rate": 7.443833675595254e-06, | |
| "loss": 1.3523, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.4730250481695568, | |
| "grad_norm": 1.2066893577575684, | |
| "learning_rate": 7.252122292489747e-06, | |
| "loss": 1.3616, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.4807321772639692, | |
| "grad_norm": 1.1276185512542725, | |
| "learning_rate": 7.061497256620793e-06, | |
| "loss": 1.353, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.4884393063583814, | |
| "grad_norm": 1.1631989479064941, | |
| "learning_rate": 6.872033929777731e-06, | |
| "loss": 1.3483, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.4961464354527938, | |
| "grad_norm": 1.1466474533081055, | |
| "learning_rate": 6.683807214479323e-06, | |
| "loss": 1.3678, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.5038535645472062, | |
| "grad_norm": 1.132791519165039, | |
| "learning_rate": 6.496891524361757e-06, | |
| "loss": 1.3576, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.5115606936416186, | |
| "grad_norm": 1.1244217157363892, | |
| "learning_rate": 6.311360754759923e-06, | |
| "loss": 1.3832, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.5192678227360308, | |
| "grad_norm": 1.1384022235870361, | |
| "learning_rate": 6.127288253493591e-06, | |
| "loss": 1.3578, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.5269749518304432, | |
| "grad_norm": 1.1305923461914062, | |
| "learning_rate": 5.944746791870062e-06, | |
| "loss": 1.368, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.5346820809248554, | |
| "grad_norm": 1.1514254808425903, | |
| "learning_rate": 5.7638085359147235e-06, | |
| "loss": 1.3533, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.5423892100192678, | |
| "grad_norm": 1.1174412965774536, | |
| "learning_rate": 5.584545017840886e-06, | |
| "loss": 1.3729, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.5500963391136802, | |
| "grad_norm": 1.0917550325393677, | |
| "learning_rate": 5.40702710777022e-06, | |
| "loss": 1.3539, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.5578034682080926, | |
| "grad_norm": 1.0902245044708252, | |
| "learning_rate": 5.231324985714942e-06, | |
| "loss": 1.3711, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.565510597302505, | |
| "grad_norm": 1.1163016557693481, | |
| "learning_rate": 5.057508113832772e-06, | |
| "loss": 1.3782, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.5732177263969171, | |
| "grad_norm": 1.1419026851654053, | |
| "learning_rate": 4.885645208965779e-06, | |
| "loss": 1.3825, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.5809248554913293, | |
| "grad_norm": 1.1543022394180298, | |
| "learning_rate": 4.7158042154738094e-06, | |
| "loss": 1.3551, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.5886319845857417, | |
| "grad_norm": 1.0950229167938232, | |
| "learning_rate": 4.548052278373327e-06, | |
| "loss": 1.3375, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.5963391136801541, | |
| "grad_norm": 1.1293272972106934, | |
| "learning_rate": 4.382455716792291e-06, | |
| "loss": 1.3498, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.6040462427745665, | |
| "grad_norm": 1.123294472694397, | |
| "learning_rate": 4.219079997751515e-06, | |
| "loss": 1.3519, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.611753371868979, | |
| "grad_norm": 1.114963412284851, | |
| "learning_rate": 4.057989710282897e-06, | |
| "loss": 1.3597, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.6194605009633911, | |
| "grad_norm": 1.0550687313079834, | |
| "learning_rate": 3.899248539894756e-06, | |
| "loss": 1.3594, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.6271676300578035, | |
| "grad_norm": 1.0849530696868896, | |
| "learning_rate": 3.7429192433944016e-06, | |
| "loss": 1.3585, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.6348747591522157, | |
| "grad_norm": 1.0992666482925415, | |
| "learning_rate": 3.589063624077802e-06, | |
| "loss": 1.3765, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.642581888246628, | |
| "grad_norm": 1.1028841733932495, | |
| "learning_rate": 3.4377425072962467e-06, | |
| "loss": 1.3551, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.6502890173410405, | |
| "grad_norm": 1.0943406820297241, | |
| "learning_rate": 3.2890157164096315e-06, | |
| "loss": 1.3426, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.657996146435453, | |
| "grad_norm": 1.0819505453109741, | |
| "learning_rate": 3.1429420491358696e-06, | |
| "loss": 1.37, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.665703275529865, | |
| "grad_norm": 1.0802693367004395, | |
| "learning_rate": 2.999579254305748e-06, | |
| "loss": 1.363, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.6734104046242775, | |
| "grad_norm": 1.0993719100952148, | |
| "learning_rate": 2.8589840090325028e-06, | |
| "loss": 1.373, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.6811175337186897, | |
| "grad_norm": 1.1456190347671509, | |
| "learning_rate": 2.721211896305059e-06, | |
| "loss": 1.337, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.688824662813102, | |
| "grad_norm": 1.1663914918899536, | |
| "learning_rate": 2.5863173830138212e-06, | |
| "loss": 1.3695, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.6965317919075145, | |
| "grad_norm": 1.10584557056427, | |
| "learning_rate": 2.454353798417698e-06, | |
| "loss": 1.336, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.7042389210019269, | |
| "grad_norm": 1.0759963989257812, | |
| "learning_rate": 2.325373313060919e-06, | |
| "loss": 1.3436, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.7119460500963393, | |
| "grad_norm": 1.0870240926742554, | |
| "learning_rate": 2.19942691814788e-06, | |
| "loss": 1.3458, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.7196531791907514, | |
| "grad_norm": 1.101758360862732, | |
| "learning_rate": 2.0765644053842583e-06, | |
| "loss": 1.3562, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.7273603082851636, | |
| "grad_norm": 1.0890095233917236, | |
| "learning_rate": 1.9568343472923524e-06, | |
| "loss": 1.3717, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.735067437379576, | |
| "grad_norm": 1.0680255889892578, | |
| "learning_rate": 1.840284078008393e-06, | |
| "loss": 1.3402, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 1.7427745664739884, | |
| "grad_norm": 1.0939226150512695, | |
| "learning_rate": 1.7269596745694295e-06, | |
| "loss": 1.3688, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 1.7504816955684008, | |
| "grad_norm": 1.0921400785446167, | |
| "learning_rate": 1.6169059386972342e-06, | |
| "loss": 1.3316, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.7581888246628132, | |
| "grad_norm": 1.048248529434204, | |
| "learning_rate": 1.5101663790863597e-06, | |
| "loss": 1.3347, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.7658959537572254, | |
| "grad_norm": 1.080112338066101, | |
| "learning_rate": 1.4067831942033904e-06, | |
| "loss": 1.3476, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.7736030828516378, | |
| "grad_norm": 1.0681742429733276, | |
| "learning_rate": 1.3067972556041753e-06, | |
| "loss": 1.3833, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.78131021194605, | |
| "grad_norm": 1.070648431777954, | |
| "learning_rate": 1.2102480917756632e-06, | |
| "loss": 1.3601, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.7890173410404624, | |
| "grad_norm": 1.0644824504852295, | |
| "learning_rate": 1.1171738725086833e-06, | |
| "loss": 1.3503, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.7967244701348748, | |
| "grad_norm": 1.0739105939865112, | |
| "learning_rate": 1.0276113938078768e-06, | |
| "loss": 1.3686, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.8044315992292872, | |
| "grad_norm": 1.0678924322128296, | |
| "learning_rate": 9.415960633447674e-07, | |
| "loss": 1.348, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.8121387283236994, | |
| "grad_norm": 1.0799516439437866, | |
| "learning_rate": 8.591618864596541e-07, | |
| "loss": 1.3571, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.8198458574181118, | |
| "grad_norm": 1.0634883642196655, | |
| "learning_rate": 7.803414527179343e-07, | |
| "loss": 1.3383, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.827552986512524, | |
| "grad_norm": 1.0771961212158203, | |
| "learning_rate": 7.051659230261299e-07, | |
| "loss": 1.363, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.8352601156069364, | |
| "grad_norm": 1.0874431133270264, | |
| "learning_rate": 6.336650173127224e-07, | |
| "loss": 1.3617, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.8429672447013488, | |
| "grad_norm": 1.0974795818328857, | |
| "learning_rate": 5.658670027786561e-07, | |
| "loss": 1.3685, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.8506743737957612, | |
| "grad_norm": 1.0479196310043335, | |
| "learning_rate": 5.017986827221733e-07, | |
| "loss": 1.3502, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.8583815028901736, | |
| "grad_norm": 1.0492548942565918, | |
| "learning_rate": 4.4148538594239176e-07, | |
| "loss": 1.335, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.8660886319845857, | |
| "grad_norm": 1.0853266716003418, | |
| "learning_rate": 3.8495095672579584e-07, | |
| "loss": 1.374, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.873795761078998, | |
| "grad_norm": 1.0355820655822754, | |
| "learning_rate": 3.322177454196285e-07, | |
| "loss": 1.3624, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.8815028901734103, | |
| "grad_norm": 1.0834838151931763, | |
| "learning_rate": 2.8330659959589944e-07, | |
| "loss": 1.3798, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.8892100192678227, | |
| "grad_norm": 1.0652782917022705, | |
| "learning_rate": 2.3823685580949273e-07, | |
| "loss": 1.3659, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.8969171483622351, | |
| "grad_norm": 1.0404101610183716, | |
| "learning_rate": 1.9702633195363918e-07, | |
| "loss": 1.3512, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.9046242774566475, | |
| "grad_norm": 1.0568671226501465, | |
| "learning_rate": 1.5969132021579347e-07, | |
| "loss": 1.3674, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.9123314065510597, | |
| "grad_norm": 1.0594927072525024, | |
| "learning_rate": 1.262465806366664e-07, | |
| "loss": 1.3562, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.920038535645472, | |
| "grad_norm": 1.0357680320739746, | |
| "learning_rate": 9.670533527498139e-08, | |
| "loss": 1.3609, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.9277456647398843, | |
| "grad_norm": 1.043979525566101, | |
| "learning_rate": 7.10792629802659e-08, | |
| "loss": 1.3224, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.9354527938342967, | |
| "grad_norm": 1.045696496963501, | |
| "learning_rate": 4.937849477572587e-08, | |
| "loss": 1.3764, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.943159922928709, | |
| "grad_norm": 1.0686546564102173, | |
| "learning_rate": 3.161160985304168e-08, | |
| "loss": 1.3563, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.9508670520231215, | |
| "grad_norm": 1.0238714218139648, | |
| "learning_rate": 1.77856321806702e-08, | |
| "loss": 1.3303, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.9585741811175337, | |
| "grad_norm": 1.0332682132720947, | |
| "learning_rate": 7.906027726981568e-09, | |
| "loss": 1.3412, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.966281310211946, | |
| "grad_norm": 1.079695224761963, | |
| "learning_rate": 1.9767022993444353e-09, | |
| "loss": 1.3787, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.9739884393063583, | |
| "grad_norm": 1.0400174856185913, | |
| "learning_rate": 0.0, | |
| "loss": 1.3201, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.9739884393063583, | |
| "eval_loss": 1.5724855661392212, | |
| "eval_runtime": 27.4882, | |
| "eval_samples_per_second": 1.819, | |
| "eval_steps_per_second": 0.473, | |
| "step": 258 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 258, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 43, | |
| "total_flos": 3.045490266539557e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |