| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 130, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 6.556158688186968, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.0774, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 6.6584009109105295, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 1.0694, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 6.549131199060914, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 1.0466, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 5.764931770395654, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0096, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 4.777677171543128, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.9852, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.4910784715160155, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.9321, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.7400582021724813, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.9037, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.370245908641049, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9166, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.542234583020508, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.9556, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.382433814293378, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.9526, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 3.4595495230807463, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.9124, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.529996375708356, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.8872, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 2.602948862544701, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8261, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 1.7939245197163485, | |
| "learning_rate": 9.998197638354428e-06, | |
| "loss": 0.8027, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.6285472696385528, | |
| "learning_rate": 9.992791852820709e-06, | |
| "loss": 0.7905, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1.683691773044241, | |
| "learning_rate": 9.983786540671052e-06, | |
| "loss": 0.8253, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 1.4139677483495103, | |
| "learning_rate": 9.971188194237141e-06, | |
| "loss": 0.788, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.0212710531457423, | |
| "learning_rate": 9.955005896229543e-06, | |
| "loss": 0.7587, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.13363754152589, | |
| "learning_rate": 9.935251313189564e-06, | |
| "loss": 0.7751, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.0106454202153003, | |
| "learning_rate": 9.911938687078324e-06, | |
| "loss": 0.7144, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.0491709255489072, | |
| "learning_rate": 9.885084825009085e-06, | |
| "loss": 0.7075, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 1.094946181646856, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 0.7278, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 1.0322442691768936, | |
| "learning_rate": 9.820833372667813e-06, | |
| "loss": 0.7141, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.9311261886289833, | |
| "learning_rate": 9.783482104137127e-06, | |
| "loss": 0.6896, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.0539494988269678, | |
| "learning_rate": 9.742682209735727e-06, | |
| "loss": 0.6866, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.9809268417612523, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.6502, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.973218926151647, | |
| "learning_rate": 9.650856666246693e-06, | |
| "loss": 0.6797, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.9253612622878234, | |
| "learning_rate": 9.599897218294122e-06, | |
| "loss": 0.6333, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.8758021391332752, | |
| "learning_rate": 9.54562149901362e-06, | |
| "loss": 0.6748, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.6914012049554424, | |
| "learning_rate": 9.488068638195072e-06, | |
| "loss": 0.6297, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 0.9621192344285577, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 0.626, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.7613199845026468, | |
| "learning_rate": 9.363299794378072e-06, | |
| "loss": 0.5985, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.6598446387427636, | |
| "learning_rate": 9.296173762811084e-06, | |
| "loss": 0.5862, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 0.7209684159946635, | |
| "learning_rate": 9.225950427718974e-06, | |
| "loss": 0.5973, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 0.7484296557235415, | |
| "learning_rate": 9.152680416240059e-06, | |
| "loss": 0.5636, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 0.6794148266441785, | |
| "learning_rate": 9.076416551997721e-06, | |
| "loss": 0.5678, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 0.7423223393538124, | |
| "learning_rate": 8.997213817017508e-06, | |
| "loss": 0.5584, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.8435523335766211, | |
| "learning_rate": 8.915129312088112e-06, | |
| "loss": 0.5585, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.7238304374045665, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.5846, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.6555512347770042, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.5673, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.6870444557880944, | |
| "learning_rate": 8.652187091988516e-06, | |
| "loss": 0.5488, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.8188624973937584, | |
| "learning_rate": 8.559187418346703e-06, | |
| "loss": 0.5408, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.6274927464696114, | |
| "learning_rate": 8.463621767547998e-06, | |
| "loss": 0.4981, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 0.7092024100494124, | |
| "learning_rate": 8.36555903713785e-06, | |
| "loss": 0.4794, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.7289933240726927, | |
| "learning_rate": 8.265069924917925e-06, | |
| "loss": 0.4988, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.6946462463274233, | |
| "learning_rate": 8.162226877976886e-06, | |
| "loss": 0.4944, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.6303206505384189, | |
| "learning_rate": 8.057104040460062e-06, | |
| "loss": 0.4816, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.6811502885813149, | |
| "learning_rate": 7.949777200115617e-06, | |
| "loss": 0.4963, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.6898592221155474, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 0.4851, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.5955730786933627, | |
| "learning_rate": 7.728822550972523e-06, | |
| "loss": 0.4922, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.7134563994437151, | |
| "learning_rate": 7.615354038247889e-06, | |
| "loss": 0.4316, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 0.6466839579437162, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.4185, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 0.7975186676356013, | |
| "learning_rate": 7.382843600106539e-06, | |
| "loss": 0.4056, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 0.737454321698778, | |
| "learning_rate": 7.263969301848188e-06, | |
| "loss": 0.4413, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.6210444010274546, | |
| "learning_rate": 7.143462807015271e-06, | |
| "loss": 0.4094, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 0.6858330281851588, | |
| "learning_rate": 7.021410994121525e-06, | |
| "loss": 0.4102, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 0.7512953756869382, | |
| "learning_rate": 6.897901855769483e-06, | |
| "loss": 0.4115, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 0.6884028671393371, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 0.4304, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 0.7554646428548176, | |
| "learning_rate": 6.646868762160399e-06, | |
| "loss": 0.4271, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.6746809890091767, | |
| "learning_rate": 6.519525787871235e-06, | |
| "loss": 0.3783, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 0.6844032634802641, | |
| "learning_rate": 6.391087319582264e-06, | |
| "loss": 0.3763, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 0.6077646198385589, | |
| "learning_rate": 6.261645954321109e-06, | |
| "loss": 0.3365, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "grad_norm": 0.7691796182624621, | |
| "learning_rate": 6.131295012148613e-06, | |
| "loss": 0.3411, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.6502974305229597, | |
| "learning_rate": 6.000128468880223e-06, | |
| "loss": 0.3509, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 0.740518347432582, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.4017, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "grad_norm": 0.7565229923355626, | |
| "learning_rate": 5.735727354158581e-06, | |
| "loss": 0.3324, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "grad_norm": 0.6942359139232085, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 0.3366, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 0.6660787037181738, | |
| "learning_rate": 5.469204947015897e-06, | |
| "loss": 0.3314, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 0.7108877837824531, | |
| "learning_rate": 5.335388221955012e-06, | |
| "loss": 0.3393, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.592634685880731, | |
| "learning_rate": 5.201329700547077e-06, | |
| "loss": 0.344, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.742092959477595, | |
| "learning_rate": 5.067126031566988e-06, | |
| "loss": 0.3366, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.7359873804227691, | |
| "learning_rate": 4.932873968433014e-06, | |
| "loss": 0.274, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "grad_norm": 0.7700391660175104, | |
| "learning_rate": 4.798670299452926e-06, | |
| "loss": 0.2926, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 0.8341244765642227, | |
| "learning_rate": 4.664611778044988e-06, | |
| "loss": 0.3106, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.6167863583602791, | |
| "learning_rate": 4.530795052984104e-06, | |
| "loss": 0.2945, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.6511221061845168, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 0.281, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 0.6718375711088242, | |
| "learning_rate": 4.264272645841419e-06, | |
| "loss": 0.2592, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 0.6892162534606975, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.2737, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "grad_norm": 0.6623296611676045, | |
| "learning_rate": 3.999871531119779e-06, | |
| "loss": 0.2696, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.7871912262354472, | |
| "learning_rate": 3.86870498785139e-06, | |
| "loss": 0.2906, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "grad_norm": 0.6955102034190365, | |
| "learning_rate": 3.7383540456788915e-06, | |
| "loss": 0.2425, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "grad_norm": 0.7364566377752025, | |
| "learning_rate": 3.6089126804177373e-06, | |
| "loss": 0.2563, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "grad_norm": 0.5941043555009817, | |
| "learning_rate": 3.480474212128766e-06, | |
| "loss": 0.2457, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 0.7051204976678778, | |
| "learning_rate": 3.3531312378396026e-06, | |
| "loss": 0.2314, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.7499074913389743, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.2503, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 0.5731046496080394, | |
| "learning_rate": 3.1020981442305187e-06, | |
| "loss": 0.2341, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "grad_norm": 0.6769839992258984, | |
| "learning_rate": 2.978589005878476e-06, | |
| "loss": 0.2149, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 0.6245350832356765, | |
| "learning_rate": 2.8565371929847286e-06, | |
| "loss": 0.222, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "grad_norm": 0.5523855275377587, | |
| "learning_rate": 2.736030698151815e-06, | |
| "loss": 0.2334, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.5668258778532823, | |
| "learning_rate": 2.6171563998934605e-06, | |
| "loss": 0.2231, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 0.8642183138763878, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.1963, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.5407519616816591, | |
| "learning_rate": 2.384645961752113e-06, | |
| "loss": 0.1968, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 0.6042801283711674, | |
| "learning_rate": 2.2711774490274767e-06, | |
| "loss": 0.1748, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "grad_norm": 0.7852325826414264, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 0.2283, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 0.7546382253277867, | |
| "learning_rate": 2.050222799884387e-06, | |
| "loss": 0.188, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.5694525524615985, | |
| "learning_rate": 1.942895959539939e-06, | |
| "loss": 0.1997, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "grad_norm": 0.7065383021823802, | |
| "learning_rate": 1.8377731220231144e-06, | |
| "loss": 0.1835, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "grad_norm": 0.6331358340455546, | |
| "learning_rate": 1.7349300750820758e-06, | |
| "loss": 0.2115, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "grad_norm": 0.5588844318444566, | |
| "learning_rate": 1.6344409628621482e-06, | |
| "loss": 0.207, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.5844745994563612, | |
| "learning_rate": 1.5363782324520033e-06, | |
| "loss": 0.1643, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "grad_norm": 0.5956723011431428, | |
| "learning_rate": 1.4408125816532981e-06, | |
| "loss": 0.1875, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "grad_norm": 0.5759126662795956, | |
| "learning_rate": 1.347812908011485e-06, | |
| "loss": 0.1802, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.3, | |
| "grad_norm": 0.5999325276736255, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 0.1839, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 0.51533418206436, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.1448, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 0.6676965897519146, | |
| "learning_rate": 1.0848706879118893e-06, | |
| "loss": 0.17, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "grad_norm": 0.5573753518696763, | |
| "learning_rate": 1.0027861829824953e-06, | |
| "loss": 0.1674, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 10.7, | |
| "grad_norm": 0.5929661048833159, | |
| "learning_rate": 9.235834480022788e-07, | |
| "loss": 0.1869, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 0.6078745070782668, | |
| "learning_rate": 8.473195837599419e-07, | |
| "loss": 0.1502, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "grad_norm": 0.5639316265756421, | |
| "learning_rate": 7.740495722810271e-07, | |
| "loss": 0.1887, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.559085105890312, | |
| "learning_rate": 7.03826237188916e-07, | |
| "loss": 0.1827, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.1, | |
| "grad_norm": 0.6445853876839104, | |
| "learning_rate": 6.367002056219285e-07, | |
| "loss": 0.1454, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 0.5558444273449943, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 0.1622, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "grad_norm": 0.502004154274581, | |
| "learning_rate": 5.119313618049309e-07, | |
| "loss": 0.1526, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "grad_norm": 0.5067385650303362, | |
| "learning_rate": 4.54378500986381e-07, | |
| "loss": 0.1338, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 0.4902579165017933, | |
| "learning_rate": 4.001027817058789e-07, | |
| "loss": 0.165, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "grad_norm": 0.5716095348730388, | |
| "learning_rate": 3.49143333753309e-07, | |
| "loss": 0.151, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "grad_norm": 0.5929705234914512, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.1544, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "grad_norm": 0.5823513634342903, | |
| "learning_rate": 2.573177902642726e-07, | |
| "loss": 0.1505, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 11.9, | |
| "grad_norm": 0.6178374738752495, | |
| "learning_rate": 2.1651789586287442e-07, | |
| "loss": 0.1734, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.5160493081797064, | |
| "learning_rate": 1.7916662733218848e-07, | |
| "loss": 0.179, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "grad_norm": 0.4956673438763341, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 0.1699, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "grad_norm": 0.4623395905330783, | |
| "learning_rate": 1.1491517499091498e-07, | |
| "loss": 0.1423, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 12.3, | |
| "grad_norm": 0.44891843300824946, | |
| "learning_rate": 8.80613129216762e-08, | |
| "loss": 0.1522, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 0.48604767705839247, | |
| "learning_rate": 6.474868681043578e-08, | |
| "loss": 0.1518, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.4500341014376347, | |
| "learning_rate": 4.499410377045765e-08, | |
| "loss": 0.1722, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "grad_norm": 0.47521882593921, | |
| "learning_rate": 2.8811805762860578e-08, | |
| "loss": 0.1749, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 12.7, | |
| "grad_norm": 0.4682825369131877, | |
| "learning_rate": 1.6213459328950355e-08, | |
| "loss": 0.1298, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.48005687898063537, | |
| "learning_rate": 7.2081471792911914e-09, | |
| "loss": 0.1658, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "grad_norm": 0.46813060355802094, | |
| "learning_rate": 1.8023616455731253e-09, | |
| "loss": 0.1396, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.4654487806665012, | |
| "learning_rate": 0.0, | |
| "loss": 0.1708, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 130, | |
| "total_flos": 7.223618985105818e+16, | |
| "train_loss": 0.42511636041677914, | |
| "train_runtime": 2459.7128, | |
| "train_samples_per_second": 1.67, | |
| "train_steps_per_second": 0.053 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 130, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 13, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 7.223618985105818e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |