| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 150, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.012237500657249524, | |
| "learning_rate": 5.333333333333334e-06, | |
| "loss": 0.1887, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.013758871137983355, | |
| "learning_rate": 1.0666666666666667e-05, | |
| "loss": 0.1945, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.011349998941864321, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.1832, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.013225692918533592, | |
| "learning_rate": 2.1333333333333335e-05, | |
| "loss": 0.1822, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.014633753032591124, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.1864, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.012886110760787857, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 0.1906, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.01575467608089797, | |
| "learning_rate": 3.733333333333334e-05, | |
| "loss": 0.1779, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.015244681368252343, | |
| "learning_rate": 4.266666666666667e-05, | |
| "loss": 0.1763, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.017441120490292387, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.1888, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.01839628534430788, | |
| "learning_rate": 5.333333333333333e-05, | |
| "loss": 0.1706, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.019896354204463063, | |
| "learning_rate": 5.8666666666666665e-05, | |
| "loss": 0.1674, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.018531171870309435, | |
| "learning_rate": 6.400000000000001e-05, | |
| "loss": 0.1647, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.022257142418860426, | |
| "learning_rate": 6.933333333333334e-05, | |
| "loss": 0.1593, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.021518159730842663, | |
| "learning_rate": 7.466666666666667e-05, | |
| "loss": 0.16, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.02127869723166498, | |
| "learning_rate": 8e-05, | |
| "loss": 0.1433, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.022172522219789725, | |
| "learning_rate": 7.998916964717848e-05, | |
| "loss": 0.151, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.01815588204918508, | |
| "learning_rate": 7.9956684453541e-05, | |
| "loss": 0.1495, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.02013260960932998, | |
| "learning_rate": 7.990256201039297e-05, | |
| "loss": 0.1336, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.016575669646089617, | |
| "learning_rate": 7.982683162599218e-05, | |
| "loss": 0.1249, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.02547366804753914, | |
| "learning_rate": 7.972953430967773e-05, | |
| "loss": 0.1409, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.015144092430494334, | |
| "learning_rate": 7.961072274966282e-05, | |
| "loss": 0.1259, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.020151032915714337, | |
| "learning_rate": 7.947046128450319e-05, | |
| "loss": 0.1277, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.015765828096125922, | |
| "learning_rate": 7.930882586825653e-05, | |
| "loss": 0.1098, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.015601871150105974, | |
| "learning_rate": 7.912590402935223e-05, | |
| "loss": 0.1064, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.013633088996436046, | |
| "learning_rate": 7.892179482319297e-05, | |
| "loss": 0.1087, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.011569423463621626, | |
| "learning_rate": 7.869660877851456e-05, | |
| "loss": 0.1023, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.01182774181852502, | |
| "learning_rate": 7.845046783753276e-05, | |
| "loss": 0.1036, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.012875055811951518, | |
| "learning_rate": 7.818350528990929e-05, | |
| "loss": 0.0889, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.01329208569615103, | |
| "learning_rate": 7.789586570057317e-05, | |
| "loss": 0.0821, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.01309821455787008, | |
| "learning_rate": 7.758770483143634e-05, | |
| "loss": 0.0854, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.01455130042398818, | |
| "learning_rate": 7.72591895570457e-05, | |
| "loss": 0.0724, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.011788724272831109, | |
| "learning_rate": 7.69104977742177e-05, | |
| "loss": 0.0853, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.013071130537087054, | |
| "learning_rate": 7.654181830570404e-05, | |
| "loss": 0.0801, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.010218011987115048, | |
| "learning_rate": 7.615335079794083e-05, | |
| "loss": 0.0737, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.01242807630778626, | |
| "learning_rate": 7.57453056129365e-05, | |
| "loss": 0.0722, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.009925629053550735, | |
| "learning_rate": 7.531790371435709e-05, | |
| "loss": 0.0745, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.012506084491127016, | |
| "learning_rate": 7.48713765478705e-05, | |
| "loss": 0.0665, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.011686858735662982, | |
| "learning_rate": 7.440596591581463e-05, | |
| "loss": 0.0633, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.012134790605612606, | |
| "learning_rate": 7.392192384625704e-05, | |
| "loss": 0.066, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.010735630457768281, | |
| "learning_rate": 7.341951245651747e-05, | |
| "loss": 0.0594, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.010874158886472111, | |
| "learning_rate": 7.28990038112265e-05, | |
| "loss": 0.0579, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.011166942930346552, | |
| "learning_rate": 7.236067977499791e-05, | |
| "loss": 0.0613, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.013078295852833188, | |
| "learning_rate": 7.180483185979392e-05, | |
| "loss": 0.0584, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.012804445869326084, | |
| "learning_rate": 7.123176106706638e-05, | |
| "loss": 0.0557, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.012628185317957133, | |
| "learning_rate": 7.064177772475912e-05, | |
| "loss": 0.0514, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.012502491887169885, | |
| "learning_rate": 7.003520131925997e-05, | |
| "loss": 0.054, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.012176293638632393, | |
| "learning_rate": 6.941236032239316e-05, | |
| "loss": 0.0516, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.013934398535720833, | |
| "learning_rate": 6.877359201354606e-05, | |
| "loss": 0.0475, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.013133107256863872, | |
| "learning_rate": 6.811924229702648e-05, | |
| "loss": 0.0515, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.01226898304197433, | |
| "learning_rate": 6.744966551474936e-05, | |
| "loss": 0.0507, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 0.012618473219449911, | |
| "learning_rate": 6.676522425435433e-05, | |
| "loss": 0.055, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.012825805660824328, | |
| "learning_rate": 6.606628915285822e-05, | |
| "loss": 0.0453, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 0.011142990437036586, | |
| "learning_rate": 6.53532386959484e-05, | |
| "loss": 0.0477, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 0.01266850407577348, | |
| "learning_rate": 6.462645901302633e-05, | |
| "loss": 0.0494, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.013249576231026749, | |
| "learning_rate": 6.388634366811146e-05, | |
| "loss": 0.0404, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.012814883287226344, | |
| "learning_rate": 6.313329344671946e-05, | |
| "loss": 0.039, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.1400000000000001, | |
| "grad_norm": 0.013319898218695845, | |
| "learning_rate": 6.236771613882987e-05, | |
| "loss": 0.0422, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 0.013851374171237484, | |
| "learning_rate": 6.159002631806052e-05, | |
| "loss": 0.0432, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.014387494932894856, | |
| "learning_rate": 6.0800645117168616e-05, | |
| "loss": 0.0403, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.012618423251405733, | |
| "learning_rate": 6.000000000000001e-05, | |
| "loss": 0.036, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.012079171161322068, | |
| "learning_rate": 5.918852453000986e-05, | |
| "loss": 0.0481, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.011873936477737802, | |
| "learning_rate": 5.836665813548047e-05, | |
| "loss": 0.0358, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.010723510041581473, | |
| "learning_rate": 5.75348458715631e-05, | |
| "loss": 0.0433, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.011110077015669406, | |
| "learning_rate": 5.669353817927272e-05, | |
| "loss": 0.0324, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.01380773024903358, | |
| "learning_rate": 5.584319064156628e-05, | |
| "loss": 0.0471, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.011361260131747103, | |
| "learning_rate": 5.4984263736636494e-05, | |
| "loss": 0.0315, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.012788680896460226, | |
| "learning_rate": 5.4117222588554756e-05, | |
| "loss": 0.0358, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.3599999999999999, | |
| "grad_norm": 0.012417672274468905, | |
| "learning_rate": 5.324253671539833e-05, | |
| "loss": 0.0459, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.010528260019435217, | |
| "learning_rate": 5.23606797749979e-05, | |
| "loss": 0.0328, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.01025356850505523, | |
| "learning_rate": 5.1472129308443616e-05, | |
| "loss": 0.0319, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.013088138569134062, | |
| "learning_rate": 5.05773664814881e-05, | |
| "loss": 0.0315, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.009363535817017942, | |
| "learning_rate": 4.967687582398671e-05, | |
| "loss": 0.0438, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.00939624667592842, | |
| "learning_rate": 4.877114496751613e-05, | |
| "loss": 0.0331, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.009827443198688923, | |
| "learning_rate": 4.786066438131321e-05, | |
| "loss": 0.0361, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.012694630631627115, | |
| "learning_rate": 4.694592710667723e-05, | |
| "loss": 0.0385, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.008742117644516437, | |
| "learning_rate": 4.602742848997933e-05, | |
| "loss": 0.048, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.010975746014214796, | |
| "learning_rate": 4.51056659144238e-05, | |
| "loss": 0.034, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 0.022245727520080187, | |
| "learning_rate": 4.418113853070614e-05, | |
| "loss": 0.0351, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 0.009081513417101492, | |
| "learning_rate": 4.3254346986714334e-05, | |
| "loss": 0.0419, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.01267456794560121, | |
| "learning_rate": 4.2325793156419035e-05, | |
| "loss": 0.0365, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 0.010517509931952058, | |
| "learning_rate": 4.139597986810005e-05, | |
| "loss": 0.039, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 0.009258174599692184, | |
| "learning_rate": 4.046541063205589e-05, | |
| "loss": 0.0452, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.6600000000000001, | |
| "grad_norm": 0.00813465138189148, | |
| "learning_rate": 3.953458936794413e-05, | |
| "loss": 0.03, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.010337958651513985, | |
| "learning_rate": 3.860402013189998e-05, | |
| "loss": 0.0321, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.009994491661604508, | |
| "learning_rate": 3.767420684358097e-05, | |
| "loss": 0.0519, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 0.01217132440089011, | |
| "learning_rate": 3.674565301328568e-05, | |
| "loss": 0.0396, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 0.011825874271809117, | |
| "learning_rate": 3.581886146929387e-05, | |
| "loss": 0.0384, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.007978828192023736, | |
| "learning_rate": 3.4894334085576215e-05, | |
| "loss": 0.0289, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 0.009590714053429032, | |
| "learning_rate": 3.397257151002068e-05, | |
| "loss": 0.0296, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.011747419714407405, | |
| "learning_rate": 3.305407289332279e-05, | |
| "loss": 0.0365, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.8199999999999998, | |
| "grad_norm": 0.009242524058805289, | |
| "learning_rate": 3.213933561868679e-05, | |
| "loss": 0.0382, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.8399999999999999, | |
| "grad_norm": 0.012770223485225133, | |
| "learning_rate": 3.122885503248386e-05, | |
| "loss": 0.0371, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.8599999999999999, | |
| "grad_norm": 0.009400573052688613, | |
| "learning_rate": 3.0323124176013297e-05, | |
| "loss": 0.0304, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.010154781511472893, | |
| "learning_rate": 2.9422633518511926e-05, | |
| "loss": 0.0348, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.012452492157804102, | |
| "learning_rate": 2.8527870691556404e-05, | |
| "loss": 0.0413, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.009256238629984684, | |
| "learning_rate": 2.7639320225002108e-05, | |
| "loss": 0.0405, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 0.008937956319737868, | |
| "learning_rate": 2.6757463284601682e-05, | |
| "loss": 0.0373, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.007425718199043812, | |
| "learning_rate": 2.5882777411445254e-05, | |
| "loss": 0.0322, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.00888718132247714, | |
| "learning_rate": 2.501573626336352e-05, | |
| "loss": 0.0404, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.011004974534179684, | |
| "learning_rate": 2.4156809358433728e-05, | |
| "loss": 0.0325, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "grad_norm": 0.008108108682747908, | |
| "learning_rate": 2.330646182072729e-05, | |
| "loss": 0.0372, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 0.008679266162637896, | |
| "learning_rate": 2.24651541284369e-05, | |
| "loss": 0.0291, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "grad_norm": 0.009391693953875733, | |
| "learning_rate": 2.1633341864519526e-05, | |
| "loss": 0.0286, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.007817425089267506, | |
| "learning_rate": 2.0811475469990167e-05, | |
| "loss": 0.0346, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.0080102190651166, | |
| "learning_rate": 2.0000000000000012e-05, | |
| "loss": 0.0332, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 0.00890483133914523, | |
| "learning_rate": 1.9199354882831387e-05, | |
| "loss": 0.047, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "grad_norm": 0.008941924673069203, | |
| "learning_rate": 1.8409973681939498e-05, | |
| "loss": 0.0501, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.007260492753066732, | |
| "learning_rate": 1.7632283861170135e-05, | |
| "loss": 0.0312, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "grad_norm": 0.007542099398977327, | |
| "learning_rate": 1.686670655328054e-05, | |
| "loss": 0.0303, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.009181010886397965, | |
| "learning_rate": 1.6113656331888563e-05, | |
| "loss": 0.0314, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "grad_norm": 0.0082002978570301, | |
| "learning_rate": 1.537354098697367e-05, | |
| "loss": 0.0307, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.008138578552276645, | |
| "learning_rate": 1.4646761304051587e-05, | |
| "loss": 0.0278, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "grad_norm": 0.008437335268852292, | |
| "learning_rate": 1.3933710847141795e-05, | |
| "loss": 0.031, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 2.2800000000000002, | |
| "grad_norm": 0.007102438138007538, | |
| "learning_rate": 1.3234775745645684e-05, | |
| "loss": 0.0298, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.010623691942988046, | |
| "learning_rate": 1.2550334485250661e-05, | |
| "loss": 0.0422, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.008514555564784327, | |
| "learning_rate": 1.1880757702973531e-05, | |
| "loss": 0.0303, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 0.008017793798625748, | |
| "learning_rate": 1.1226407986453963e-05, | |
| "loss": 0.0335, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "grad_norm": 0.008669725250579921, | |
| "learning_rate": 1.0587639677606857e-05, | |
| "loss": 0.034, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 0.008667828433003316, | |
| "learning_rate": 9.964798680740033e-06, | |
| "loss": 0.0424, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.008361863318103638, | |
| "learning_rate": 9.358222275240884e-06, | |
| "loss": 0.032, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "grad_norm": 0.009399913861027212, | |
| "learning_rate": 8.768238932933632e-06, | |
| "loss": 0.0372, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "grad_norm": 0.008350646835726937, | |
| "learning_rate": 8.195168140206084e-06, | |
| "loss": 0.0424, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "grad_norm": 0.007803789131594086, | |
| "learning_rate": 7.639320225002106e-06, | |
| "loss": 0.0326, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.006960247891792648, | |
| "learning_rate": 7.1009961887735075e-06, | |
| "loss": 0.0306, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.007284462400377896, | |
| "learning_rate": 6.58048754348255e-06, | |
| "loss": 0.0306, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 0.008496978964651394, | |
| "learning_rate": 6.078076153742962e-06, | |
| "loss": 0.0385, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "grad_norm": 0.0075599496285588446, | |
| "learning_rate": 5.5940340841853915e-06, | |
| "loss": 0.0361, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.007668826080960316, | |
| "learning_rate": 5.128623452129508e-06, | |
| "loss": 0.0269, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "grad_norm": 0.008734111259065381, | |
| "learning_rate": 4.6820962856429205e-06, | |
| "loss": 0.0287, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.007249536995896214, | |
| "learning_rate": 4.254694387063514e-06, | |
| "loss": 0.0473, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "grad_norm": 0.008501677182613158, | |
| "learning_rate": 3.846649202059181e-06, | |
| "loss": 0.03, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.008397989935636663, | |
| "learning_rate": 3.458181694295961e-06, | |
| "loss": 0.0469, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "grad_norm": 0.008189115845957336, | |
| "learning_rate": 3.0895022257823083e-06, | |
| "loss": 0.0324, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 0.007599070522414446, | |
| "learning_rate": 2.7408104429543025e-06, | |
| "loss": 0.0301, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.00835906536613168, | |
| "learning_rate": 2.4122951685636674e-06, | |
| "loss": 0.0295, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 0.007714432648422336, | |
| "learning_rate": 2.104134299426832e-06, | |
| "loss": 0.0256, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 0.008398811533431756, | |
| "learning_rate": 1.8164947100907238e-06, | |
| "loss": 0.0285, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 0.008699208627645135, | |
| "learning_rate": 1.5495321624672443e-06, | |
| "loss": 0.0365, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.7800000000000002, | |
| "grad_norm": 0.007386695794413173, | |
| "learning_rate": 1.3033912214854482e-06, | |
| "loss": 0.0275, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.010892746607433756, | |
| "learning_rate": 1.0782051768070477e-06, | |
| "loss": 0.0438, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 0.008014040444496067, | |
| "learning_rate": 8.740959706477725e-07, | |
| "loss": 0.0325, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 0.00784352307398519, | |
| "learning_rate": 6.911741317434706e-07, | |
| "loss": 0.0313, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "grad_norm": 0.007885558746174392, | |
| "learning_rate": 5.295387154968312e-07, | |
| "loss": 0.0293, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.00789954826144653, | |
| "learning_rate": 3.8927725033718553e-07, | |
| "loss": 0.0393, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.007685085075355499, | |
| "learning_rate": 2.704656903222791e-07, | |
| "loss": 0.0359, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 0.01001469941559618, | |
| "learning_rate": 1.7316837400782604e-07, | |
| "loss": 0.0322, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "grad_norm": 0.0073776135499800345, | |
| "learning_rate": 9.74379896070321e-08, | |
| "loss": 0.0276, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.007566774696093351, | |
| "learning_rate": 4.331554645901737e-08, | |
| "loss": 0.0304, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "grad_norm": 0.00795910919735732, | |
| "learning_rate": 1.0830352821531442e-08, | |
| "loss": 0.0438, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.00742509450083333, | |
| "learning_rate": 0.0, | |
| "loss": 0.0291, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 150, | |
| "total_flos": 112603042349056.0, | |
| "train_loss": 0.06176747960348924, | |
| "train_runtime": 1316.5553, | |
| "train_samples_per_second": 0.911, | |
| "train_steps_per_second": 0.114 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 150, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 112603042349056.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |