| { | |
| "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": 6.866188206001596, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 1.0986, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 6.1679557287931965, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 0.9717, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 5.940701490572247, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 0.9758, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 6.081750599571613, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0631, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 5.224647493462944, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 1.0185, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 3.7186566658265856, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.9678, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 2.3373002640664273, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.9368, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 3.601608042283404, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.932, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.441093548136363, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.9135, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 4.4504281693588865, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.8487, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 4.442723298630125, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.876, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 3.8407128042000944, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.8309, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 3.3457448726694903, | |
| "learning_rate": 1e-05, | |
| "loss": 0.834, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 2.668813889422515, | |
| "learning_rate": 9.998197638354428e-06, | |
| "loss": 0.8246, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 2.499685424405167, | |
| "learning_rate": 9.992791852820709e-06, | |
| "loss": 0.7954, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 2.291195119605541, | |
| "learning_rate": 9.983786540671052e-06, | |
| "loss": 0.8142, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 2.073789366044732, | |
| "learning_rate": 9.971188194237141e-06, | |
| "loss": 0.7615, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 2.1244992171104835, | |
| "learning_rate": 9.955005896229543e-06, | |
| "loss": 0.7439, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 1.6996982874919078, | |
| "learning_rate": 9.935251313189564e-06, | |
| "loss": 0.7967, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.5667160210105031, | |
| "learning_rate": 9.911938687078324e-06, | |
| "loss": 0.7318, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 1.5420951812620658, | |
| "learning_rate": 9.885084825009085e-06, | |
| "loss": 0.6477, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 1.4842143674245734, | |
| "learning_rate": 9.854709087130261e-06, | |
| "loss": 0.6659, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 1.1918914434425512, | |
| "learning_rate": 9.820833372667813e-06, | |
| "loss": 0.6362, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.1436804722860308, | |
| "learning_rate": 9.783482104137127e-06, | |
| "loss": 0.7064, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.2089961899994688, | |
| "learning_rate": 9.742682209735727e-06, | |
| "loss": 0.6708, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 1.1813990670824117, | |
| "learning_rate": 9.698463103929542e-06, | |
| "loss": 0.6378, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.9426949209168561, | |
| "learning_rate": 9.650856666246693e-06, | |
| "loss": 0.5774, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.9467859575349933, | |
| "learning_rate": 9.599897218294122e-06, | |
| "loss": 0.6551, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.9747126817004351, | |
| "learning_rate": 9.54562149901362e-06, | |
| "loss": 0.6403, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 1.0088836127230056, | |
| "learning_rate": 9.488068638195072e-06, | |
| "loss": 0.5717, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.1, | |
| "grad_norm": 1.010492876273852, | |
| "learning_rate": 9.427280128266049e-06, | |
| "loss": 0.5395, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.7676283658490435, | |
| "learning_rate": 9.363299794378072e-06, | |
| "loss": 0.504, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "grad_norm": 0.9668774216375602, | |
| "learning_rate": 9.296173762811084e-06, | |
| "loss": 0.5019, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "grad_norm": 1.2053442133469203, | |
| "learning_rate": 9.225950427718974e-06, | |
| "loss": 0.4797, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 3.5, | |
| "grad_norm": 1.0226998507674792, | |
| "learning_rate": 9.152680416240059e-06, | |
| "loss": 0.5129, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 1.0002036497286182, | |
| "learning_rate": 9.076416551997721e-06, | |
| "loss": 0.487, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "grad_norm": 1.0498062757693174, | |
| "learning_rate": 8.997213817017508e-06, | |
| "loss": 0.4711, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "grad_norm": 0.9230235172320319, | |
| "learning_rate": 8.915129312088112e-06, | |
| "loss": 0.4848, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 3.9, | |
| "grad_norm": 0.9719529298769463, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.4679, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.9616182651103187, | |
| "learning_rate": 8.742553740855507e-06, | |
| "loss": 0.4783, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 0.9280575825646215, | |
| "learning_rate": 8.652187091988516e-06, | |
| "loss": 0.3915, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.8746551283210525, | |
| "learning_rate": 8.559187418346703e-06, | |
| "loss": 0.3675, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.3, | |
| "grad_norm": 0.8673005956813418, | |
| "learning_rate": 8.463621767547998e-06, | |
| "loss": 0.3823, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 1.1435696543702474, | |
| "learning_rate": 8.36555903713785e-06, | |
| "loss": 0.3895, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.9162365521244533, | |
| "learning_rate": 8.265069924917925e-06, | |
| "loss": 0.346, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "grad_norm": 0.9173251221647998, | |
| "learning_rate": 8.162226877976886e-06, | |
| "loss": 0.3712, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 4.7, | |
| "grad_norm": 0.8868873761806556, | |
| "learning_rate": 8.057104040460062e-06, | |
| "loss": 0.3657, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.9006113924130832, | |
| "learning_rate": 7.949777200115617e-06, | |
| "loss": 0.3936, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.8935268740114803, | |
| "learning_rate": 7.84032373365578e-06, | |
| "loss": 0.3021, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 1.0779450407316271, | |
| "learning_rate": 7.728822550972523e-06, | |
| "loss": 0.3162, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.8554686145182366, | |
| "learning_rate": 7.615354038247889e-06, | |
| "loss": 0.3209, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 0.803445231180753, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.2398, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 5.3, | |
| "grad_norm": 0.8476945434033398, | |
| "learning_rate": 7.382843600106539e-06, | |
| "loss": 0.2602, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "grad_norm": 0.9889249137178309, | |
| "learning_rate": 7.263969301848188e-06, | |
| "loss": 0.23, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 5.5, | |
| "grad_norm": 0.8591261419213184, | |
| "learning_rate": 7.143462807015271e-06, | |
| "loss": 0.2776, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 0.8121772627021023, | |
| "learning_rate": 7.021410994121525e-06, | |
| "loss": 0.2153, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 5.7, | |
| "grad_norm": 0.9044380043866544, | |
| "learning_rate": 6.897901855769483e-06, | |
| "loss": 0.2596, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "grad_norm": 0.8145950419430453, | |
| "learning_rate": 6.773024435212678e-06, | |
| "loss": 0.2436, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 5.9, | |
| "grad_norm": 0.9979249920005759, | |
| "learning_rate": 6.646868762160399e-06, | |
| "loss": 0.2875, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.706828913121961, | |
| "learning_rate": 6.519525787871235e-06, | |
| "loss": 0.2573, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.1, | |
| "grad_norm": 0.81140052569879, | |
| "learning_rate": 6.391087319582264e-06, | |
| "loss": 0.2225, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "grad_norm": 0.7601057259387739, | |
| "learning_rate": 6.261645954321109e-06, | |
| "loss": 0.1941, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "grad_norm": 0.7973776009721454, | |
| "learning_rate": 6.131295012148613e-06, | |
| "loss": 0.1731, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.9745052233778466, | |
| "learning_rate": 6.000128468880223e-06, | |
| "loss": 0.1507, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "grad_norm": 0.7037997704157211, | |
| "learning_rate": 5.8682408883346535e-06, | |
| "loss": 0.1774, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "grad_norm": 0.8222440619566254, | |
| "learning_rate": 5.735727354158581e-06, | |
| "loss": 0.1692, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 6.7, | |
| "grad_norm": 0.8220966119387922, | |
| "learning_rate": 5.6026834012766155e-06, | |
| "loss": 0.1752, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 0.7389064896690672, | |
| "learning_rate": 5.469204947015897e-06, | |
| "loss": 0.1519, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 6.9, | |
| "grad_norm": 0.8366808613318866, | |
| "learning_rate": 5.335388221955012e-06, | |
| "loss": 0.1391, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.7785989756728257, | |
| "learning_rate": 5.201329700547077e-06, | |
| "loss": 0.16, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.1, | |
| "grad_norm": 0.6482762209389364, | |
| "learning_rate": 5.067126031566988e-06, | |
| "loss": 0.1113, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.6459453572182646, | |
| "learning_rate": 4.932873968433014e-06, | |
| "loss": 0.1243, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 7.3, | |
| "grad_norm": 0.6743885461877015, | |
| "learning_rate": 4.798670299452926e-06, | |
| "loss": 0.1259, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 0.7729053383239757, | |
| "learning_rate": 4.664611778044988e-06, | |
| "loss": 0.128, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.6968402360100354, | |
| "learning_rate": 4.530795052984104e-06, | |
| "loss": 0.1272, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.7297297249354789, | |
| "learning_rate": 4.397316598723385e-06, | |
| "loss": 0.1067, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 7.7, | |
| "grad_norm": 0.7307417535081763, | |
| "learning_rate": 4.264272645841419e-06, | |
| "loss": 0.0882, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "grad_norm": 0.8958859305793883, | |
| "learning_rate": 4.131759111665349e-06, | |
| "loss": 0.1408, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 7.9, | |
| "grad_norm": 0.7237139999385375, | |
| "learning_rate": 3.999871531119779e-06, | |
| "loss": 0.1122, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.6069340414371904, | |
| "learning_rate": 3.86870498785139e-06, | |
| "loss": 0.1039, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.1, | |
| "grad_norm": 0.599406845295279, | |
| "learning_rate": 3.7383540456788915e-06, | |
| "loss": 0.0858, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "grad_norm": 0.7212541160055512, | |
| "learning_rate": 3.6089126804177373e-06, | |
| "loss": 0.0715, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 8.3, | |
| "grad_norm": 0.9924189875413985, | |
| "learning_rate": 3.480474212128766e-06, | |
| "loss": 0.0817, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "grad_norm": 0.7098328935669028, | |
| "learning_rate": 3.3531312378396026e-06, | |
| "loss": 0.0874, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 8.5, | |
| "grad_norm": 0.8254175409760984, | |
| "learning_rate": 3.226975564787322e-06, | |
| "loss": 0.0597, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "grad_norm": 0.766621873071457, | |
| "learning_rate": 3.1020981442305187e-06, | |
| "loss": 0.1019, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 8.7, | |
| "grad_norm": 0.6297553608952663, | |
| "learning_rate": 2.978589005878476e-06, | |
| "loss": 0.0761, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "grad_norm": 0.590791149166634, | |
| "learning_rate": 2.8565371929847286e-06, | |
| "loss": 0.0757, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 8.9, | |
| "grad_norm": 0.5951940493812242, | |
| "learning_rate": 2.736030698151815e-06, | |
| "loss": 0.0616, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.619384386727814, | |
| "learning_rate": 2.6171563998934605e-06, | |
| "loss": 0.0928, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.1, | |
| "grad_norm": 0.8767367800274541, | |
| "learning_rate": 2.5000000000000015e-06, | |
| "loss": 0.0897, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "grad_norm": 0.6518150791784254, | |
| "learning_rate": 2.384645961752113e-06, | |
| "loss": 0.0495, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 9.3, | |
| "grad_norm": 0.5387056076408027, | |
| "learning_rate": 2.2711774490274767e-06, | |
| "loss": 0.0685, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "grad_norm": 0.4863421257869079, | |
| "learning_rate": 2.159676266344222e-06, | |
| "loss": 0.0531, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 9.5, | |
| "grad_norm": 0.5980183187058193, | |
| "learning_rate": 2.050222799884387e-06, | |
| "loss": 0.0562, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "grad_norm": 0.6568138946124423, | |
| "learning_rate": 1.942895959539939e-06, | |
| "loss": 0.0551, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 9.7, | |
| "grad_norm": 0.6801946941749343, | |
| "learning_rate": 1.8377731220231144e-06, | |
| "loss": 0.0598, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "grad_norm": 0.49464004712731985, | |
| "learning_rate": 1.7349300750820758e-06, | |
| "loss": 0.0435, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "grad_norm": 0.6676402317413358, | |
| "learning_rate": 1.6344409628621482e-06, | |
| "loss": 0.0677, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.5796038934392613, | |
| "learning_rate": 1.5363782324520033e-06, | |
| "loss": 0.0658, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.1, | |
| "grad_norm": 0.42194095427095796, | |
| "learning_rate": 1.4408125816532981e-06, | |
| "loss": 0.0309, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "grad_norm": 0.4487069838841247, | |
| "learning_rate": 1.347812908011485e-06, | |
| "loss": 0.0486, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 10.3, | |
| "grad_norm": 0.4896641689264063, | |
| "learning_rate": 1.257446259144494e-06, | |
| "loss": 0.055, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "grad_norm": 0.5075883178470356, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.0512, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "grad_norm": 0.43101740253026166, | |
| "learning_rate": 1.0848706879118893e-06, | |
| "loss": 0.0359, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "grad_norm": 0.373626440958998, | |
| "learning_rate": 1.0027861829824953e-06, | |
| "loss": 0.0507, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 10.7, | |
| "grad_norm": 0.4162748845286249, | |
| "learning_rate": 9.235834480022788e-07, | |
| "loss": 0.0468, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "grad_norm": 0.46300801449368284, | |
| "learning_rate": 8.473195837599419e-07, | |
| "loss": 0.0588, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "grad_norm": 0.43638275818915895, | |
| "learning_rate": 7.740495722810271e-07, | |
| "loss": 0.0519, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.3507038488091641, | |
| "learning_rate": 7.03826237188916e-07, | |
| "loss": 0.0351, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.1, | |
| "grad_norm": 0.2958259093596939, | |
| "learning_rate": 6.367002056219285e-07, | |
| "loss": 0.0285, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "grad_norm": 0.41622339644383466, | |
| "learning_rate": 5.727198717339511e-07, | |
| "loss": 0.05, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 11.3, | |
| "grad_norm": 0.39427618162169226, | |
| "learning_rate": 5.119313618049309e-07, | |
| "loss": 0.0407, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "grad_norm": 0.41398018057062874, | |
| "learning_rate": 4.54378500986381e-07, | |
| "loss": 0.0502, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 11.5, | |
| "grad_norm": 0.2865344034223786, | |
| "learning_rate": 4.001027817058789e-07, | |
| "loss": 0.0214, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "grad_norm": 0.3155056731840797, | |
| "learning_rate": 3.49143333753309e-07, | |
| "loss": 0.0338, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 11.7, | |
| "grad_norm": 0.3591185873967975, | |
| "learning_rate": 3.015368960704584e-07, | |
| "loss": 0.0488, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "grad_norm": 0.30231205469402667, | |
| "learning_rate": 2.573177902642726e-07, | |
| "loss": 0.0173, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 11.9, | |
| "grad_norm": 0.36923353877844073, | |
| "learning_rate": 2.1651789586287442e-07, | |
| "loss": 0.0567, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.34166978254594366, | |
| "learning_rate": 1.7916662733218848e-07, | |
| "loss": 0.0437, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.1, | |
| "grad_norm": 0.34011094678996845, | |
| "learning_rate": 1.4529091286973994e-07, | |
| "loss": 0.0584, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "grad_norm": 0.3321270133559747, | |
| "learning_rate": 1.1491517499091498e-07, | |
| "loss": 0.0494, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 12.3, | |
| "grad_norm": 0.30546270768143635, | |
| "learning_rate": 8.80613129216762e-08, | |
| "loss": 0.029, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "grad_norm": 0.2918865089883478, | |
| "learning_rate": 6.474868681043578e-08, | |
| "loss": 0.0349, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "grad_norm": 0.2876897674738445, | |
| "learning_rate": 4.499410377045765e-08, | |
| "loss": 0.0286, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "grad_norm": 0.33240251786172226, | |
| "learning_rate": 2.8811805762860578e-08, | |
| "loss": 0.0359, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 12.7, | |
| "grad_norm": 0.3086690432531669, | |
| "learning_rate": 1.6213459328950355e-08, | |
| "loss": 0.043, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 0.2583489032115073, | |
| "learning_rate": 7.2081471792911914e-09, | |
| "loss": 0.0187, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 12.9, | |
| "grad_norm": 0.32739419200915676, | |
| "learning_rate": 1.8023616455731253e-09, | |
| "loss": 0.0408, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.2630820852068032, | |
| "learning_rate": 0.0, | |
| "loss": 0.0262, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 130, | |
| "total_flos": 4.186952645056922e+16, | |
| "train_loss": 0.313810807839036, | |
| "train_runtime": 1554.834, | |
| "train_samples_per_second": 2.642, | |
| "train_steps_per_second": 0.084 | |
| } | |
| ], | |
| "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": 4.186952645056922e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |