| { | |
| "best_metric": 0.13323988020420074, | |
| "best_model_checkpoint": "saves/psy-course-info/MentaLLaMA-chat-7B/train/fold2/checkpoint-120", | |
| "epoch": 4.938271604938271, | |
| "eval_steps": 10, | |
| "global_step": 125, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03950617283950617, | |
| "grad_norm": 0.8806350827217102, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 1.0031, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07901234567901234, | |
| "grad_norm": 0.9527279138565063, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 1.1037, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 0.8519999980926514, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 1.061, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1580246913580247, | |
| "grad_norm": 0.9130703806877136, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 1.1008, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 0.8759827017784119, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 1.0445, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 1.0796666145324707, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 1.0161, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2765432098765432, | |
| "grad_norm": 1.0442078113555908, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 1.0382, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3160493827160494, | |
| "grad_norm": 1.1886224746704102, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.9012, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 1.3622387647628784, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.9196, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 1.2611912488937378, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.7586, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "eval_loss": 0.6472154259681702, | |
| "eval_runtime": 22.8448, | |
| "eval_samples_per_second": 1.97, | |
| "eval_steps_per_second": 1.97, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4345679012345679, | |
| "grad_norm": 1.3971184492111206, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.7026, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 0.9292118549346924, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.5575, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5135802469135803, | |
| "grad_norm": 0.7465816140174866, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4488, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5530864197530864, | |
| "grad_norm": 0.5243722796440125, | |
| "learning_rate": 9.998033131915266e-05, | |
| "loss": 0.3914, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.5454760789871216, | |
| "learning_rate": 9.992134075089084e-05, | |
| "loss": 0.3393, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6320987654320988, | |
| "grad_norm": 0.6062592267990112, | |
| "learning_rate": 9.982307470588098e-05, | |
| "loss": 0.3842, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.671604938271605, | |
| "grad_norm": 0.603158712387085, | |
| "learning_rate": 9.968561049466214e-05, | |
| "loss": 0.3509, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.5811333656311035, | |
| "learning_rate": 9.950905626682228e-05, | |
| "loss": 0.2778, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7506172839506173, | |
| "grad_norm": 0.45332807302474976, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.2519, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.4229904115200043, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.225, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "eval_loss": 0.27495014667510986, | |
| "eval_runtime": 22.8715, | |
| "eval_samples_per_second": 1.968, | |
| "eval_steps_per_second": 1.968, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 0.5121337175369263, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.3166, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8691358024691358, | |
| "grad_norm": 0.5223689079284668, | |
| "learning_rate": 9.841517610611309e-05, | |
| "loss": 0.2781, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.908641975308642, | |
| "grad_norm": 0.4765855073928833, | |
| "learning_rate": 9.804586609725499e-05, | |
| "loss": 0.2881, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 0.5430426597595215, | |
| "learning_rate": 9.763875613614482e-05, | |
| "loss": 0.266, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.5117508172988892, | |
| "learning_rate": 9.719416651541839e-05, | |
| "loss": 0.2756, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0271604938271606, | |
| "grad_norm": 0.9222873449325562, | |
| "learning_rate": 9.671244701472999e-05, | |
| "loss": 0.494, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.35180237889289856, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.1692, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.106172839506173, | |
| "grad_norm": 0.38448745012283325, | |
| "learning_rate": 9.563916325306594e-05, | |
| "loss": 0.239, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.145679012345679, | |
| "grad_norm": 0.3721097707748413, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.2319, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.31790003180503845, | |
| "learning_rate": 9.442228179894362e-05, | |
| "loss": 0.2279, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "eval_loss": 0.1985669732093811, | |
| "eval_runtime": 22.8618, | |
| "eval_samples_per_second": 1.968, | |
| "eval_steps_per_second": 1.968, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.2246913580246914, | |
| "grad_norm": 0.3828642666339874, | |
| "learning_rate": 9.376117109543769e-05, | |
| "loss": 0.2201, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2641975308641975, | |
| "grad_norm": 0.3742019534111023, | |
| "learning_rate": 9.306563141162046e-05, | |
| "loss": 0.2139, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 0.38906583189964294, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.1525, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.34320987654321, | |
| "grad_norm": 0.32694849371910095, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.145, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.33549103140830994, | |
| "learning_rate": 9.077804344796302e-05, | |
| "loss": 0.1463, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.3349437713623047, | |
| "learning_rate": 8.995052426791247e-05, | |
| "loss": 0.152, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4617283950617284, | |
| "grad_norm": 0.333011269569397, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.1331, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.5012345679012347, | |
| "grad_norm": 0.6567708253860474, | |
| "learning_rate": 8.820186879108038e-05, | |
| "loss": 0.1832, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 0.34091636538505554, | |
| "learning_rate": 8.728210824415827e-05, | |
| "loss": 0.1829, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.34185412526130676, | |
| "learning_rate": 8.633301610170135e-05, | |
| "loss": 0.1615, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "eval_loss": 0.16907067596912384, | |
| "eval_runtime": 22.8853, | |
| "eval_samples_per_second": 1.966, | |
| "eval_steps_per_second": 1.966, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6197530864197531, | |
| "grad_norm": 0.5085591673851013, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.2214, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 0.3815777897834778, | |
| "learning_rate": 8.434984630174509e-05, | |
| "loss": 0.143, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6987654320987655, | |
| "grad_norm": 0.38787031173706055, | |
| "learning_rate": 8.33173288976002e-05, | |
| "loss": 0.1667, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.7382716049382716, | |
| "grad_norm": 0.31477177143096924, | |
| "learning_rate": 8.225859917710439e-05, | |
| "loss": 0.1517, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.32310497760772705, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.1217, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.817283950617284, | |
| "grad_norm": 0.45286473631858826, | |
| "learning_rate": 8.006585456492029e-05, | |
| "loss": 0.2176, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8567901234567903, | |
| "grad_norm": 0.35797223448753357, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.1355, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 0.42817553877830505, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.1574, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.9358024691358025, | |
| "grad_norm": 0.34028342366218567, | |
| "learning_rate": 7.660160382576683e-05, | |
| "loss": 0.1343, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.43994370102882385, | |
| "learning_rate": 7.540376726232648e-05, | |
| "loss": 0.1891, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "eval_loss": 0.1546129584312439, | |
| "eval_runtime": 22.9025, | |
| "eval_samples_per_second": 1.965, | |
| "eval_steps_per_second": 1.965, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 0.7077121138572693, | |
| "learning_rate": 7.4185944355262e-05, | |
| "loss": 0.3425, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.054320987654321, | |
| "grad_norm": 0.35009080171585083, | |
| "learning_rate": 7.294909322337689e-05, | |
| "loss": 0.1412, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.093827160493827, | |
| "grad_norm": 0.2685275673866272, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.1484, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.31295105814933777, | |
| "learning_rate": 7.042221284679982e-05, | |
| "loss": 0.1662, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.33524706959724426, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.1125, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.212345679012346, | |
| "grad_norm": 0.3442281186580658, | |
| "learning_rate": 6.783107663311565e-05, | |
| "loss": 0.1394, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 0.2823539972305298, | |
| "learning_rate": 6.651395309775837e-05, | |
| "loss": 0.1221, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.291358024691358, | |
| "grad_norm": 0.31198570132255554, | |
| "learning_rate": 6.518383725548074e-05, | |
| "loss": 0.12, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.330864197530864, | |
| "grad_norm": 0.3657529652118683, | |
| "learning_rate": 6.384177557124247e-05, | |
| "loss": 0.1663, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.2720486521720886, | |
| "learning_rate": 6.248882390836135e-05, | |
| "loss": 0.1057, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "eval_loss": 0.1460902988910675, | |
| "eval_runtime": 22.9414, | |
| "eval_samples_per_second": 1.962, | |
| "eval_steps_per_second": 1.962, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.4098765432098768, | |
| "grad_norm": 0.25343137979507446, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.106, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.449382716049383, | |
| "grad_norm": 0.3785964548587799, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 0.1288, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.2758293151855469, | |
| "learning_rate": 5.837531116523682e-05, | |
| "loss": 0.1297, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.528395061728395, | |
| "grad_norm": 0.3753138482570648, | |
| "learning_rate": 5.698951697677498e-05, | |
| "loss": 0.1398, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 0.3024027347564697, | |
| "learning_rate": 5.559822380516539e-05, | |
| "loss": 0.133, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 0.2585088610649109, | |
| "learning_rate": 5.420252624646238e-05, | |
| "loss": 0.12, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.6469135802469137, | |
| "grad_norm": 0.38427460193634033, | |
| "learning_rate": 5.2803522361859594e-05, | |
| "loss": 0.1039, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.68641975308642, | |
| "grad_norm": 0.4874839186668396, | |
| "learning_rate": 5.140231281379345e-05, | |
| "loss": 0.1134, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 0.34231987595558167, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1159, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.4255834221839905, | |
| "learning_rate": 4.859768718620656e-05, | |
| "loss": 0.146, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "eval_loss": 0.14183731377124786, | |
| "eval_runtime": 22.9094, | |
| "eval_samples_per_second": 1.964, | |
| "eval_steps_per_second": 1.964, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8049382716049385, | |
| "grad_norm": 0.419005811214447, | |
| "learning_rate": 4.7196477638140404e-05, | |
| "loss": 0.1097, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.2766801118850708, | |
| "learning_rate": 4.579747375353763e-05, | |
| "loss": 0.1009, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.8839506172839506, | |
| "grad_norm": 0.357248991727829, | |
| "learning_rate": 4.4401776194834613e-05, | |
| "loss": 0.1471, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.9234567901234567, | |
| "grad_norm": 0.3669372797012329, | |
| "learning_rate": 4.3010483023225045e-05, | |
| "loss": 0.1351, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.4319014847278595, | |
| "learning_rate": 4.162468883476319e-05, | |
| "loss": 0.1451, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.0024691358024693, | |
| "grad_norm": 0.5606386661529541, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 0.2044, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 3.0419753086419754, | |
| "grad_norm": 0.37234655022621155, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.0956, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 3.0814814814814815, | |
| "grad_norm": 0.31006428599357605, | |
| "learning_rate": 3.7511176091638653e-05, | |
| "loss": 0.1112, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.1209876543209876, | |
| "grad_norm": 0.3588821589946747, | |
| "learning_rate": 3.6158224428757535e-05, | |
| "loss": 0.1079, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "grad_norm": 0.3152711093425751, | |
| "learning_rate": 3.4816162744519263e-05, | |
| "loss": 0.0949, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "eval_loss": 0.13798634707927704, | |
| "eval_runtime": 23.0191, | |
| "eval_samples_per_second": 1.955, | |
| "eval_steps_per_second": 1.955, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.33591189980506897, | |
| "learning_rate": 3.3486046902241664e-05, | |
| "loss": 0.1191, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 3.2395061728395063, | |
| "grad_norm": 0.32581812143325806, | |
| "learning_rate": 3.216892336688435e-05, | |
| "loss": 0.0843, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 3.2790123456790123, | |
| "grad_norm": 0.373647004365921, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.1334, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 3.3185185185185184, | |
| "grad_norm": 0.33775365352630615, | |
| "learning_rate": 2.9577787153200197e-05, | |
| "loss": 0.1117, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 3.3580246913580245, | |
| "grad_norm": 0.3206593990325928, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.0914, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.397530864197531, | |
| "grad_norm": 0.38170021772384644, | |
| "learning_rate": 2.705090677662311e-05, | |
| "loss": 0.1195, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.437037037037037, | |
| "grad_norm": 0.3029358685016632, | |
| "learning_rate": 2.581405564473801e-05, | |
| "loss": 0.107, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.476543209876543, | |
| "grad_norm": 0.3167051076889038, | |
| "learning_rate": 2.459623273767354e-05, | |
| "loss": 0.0868, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.5160493827160493, | |
| "grad_norm": 0.32960379123687744, | |
| "learning_rate": 2.3398396174233178e-05, | |
| "loss": 0.0846, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.3328462541103363, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.078, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "eval_loss": 0.13549260795116425, | |
| "eval_runtime": 22.971, | |
| "eval_samples_per_second": 1.959, | |
| "eval_steps_per_second": 1.959, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.595061728395062, | |
| "grad_norm": 0.36067718267440796, | |
| "learning_rate": 2.1066435191009715e-05, | |
| "loss": 0.1463, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.634567901234568, | |
| "grad_norm": 0.31329378485679626, | |
| "learning_rate": 1.9934145435079702e-05, | |
| "loss": 0.1209, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.674074074074074, | |
| "grad_norm": 0.23603391647338867, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.0837, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.71358024691358, | |
| "grad_norm": 0.4413388967514038, | |
| "learning_rate": 1.774140082289563e-05, | |
| "loss": 0.1212, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.753086419753086, | |
| "grad_norm": 0.31552475690841675, | |
| "learning_rate": 1.6682671102399805e-05, | |
| "loss": 0.1282, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.7925925925925927, | |
| "grad_norm": 0.4378697872161865, | |
| "learning_rate": 1.5650153698254916e-05, | |
| "loss": 0.1045, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.832098765432099, | |
| "grad_norm": 0.316997230052948, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.0763, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.871604938271605, | |
| "grad_norm": 0.3414573669433594, | |
| "learning_rate": 1.3666983898298657e-05, | |
| "loss": 0.127, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 0.5157877802848816, | |
| "learning_rate": 1.2717891755841722e-05, | |
| "loss": 0.1323, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "grad_norm": 0.3348810374736786, | |
| "learning_rate": 1.1798131208919627e-05, | |
| "loss": 0.0861, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "eval_loss": 0.13381603360176086, | |
| "eval_runtime": 22.9982, | |
| "eval_samples_per_second": 1.957, | |
| "eval_steps_per_second": 1.957, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.9901234567901236, | |
| "grad_norm": 0.6225301623344421, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 0.1821, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 4.029629629629629, | |
| "grad_norm": 0.6404198408126831, | |
| "learning_rate": 1.004947573208756e-05, | |
| "loss": 0.1638, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 4.069135802469136, | |
| "grad_norm": 0.2728753983974457, | |
| "learning_rate": 9.221956552036992e-06, | |
| "loss": 0.0776, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 4.108641975308642, | |
| "grad_norm": 0.28971824049949646, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 0.1135, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 4.148148148148148, | |
| "grad_norm": 0.31094127893447876, | |
| "learning_rate": 7.663790038585793e-06, | |
| "loss": 0.0913, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.187654320987654, | |
| "grad_norm": 0.26354196667671204, | |
| "learning_rate": 6.934368588379553e-06, | |
| "loss": 0.0722, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 4.22716049382716, | |
| "grad_norm": 0.30603542923927307, | |
| "learning_rate": 6.238828904562316e-06, | |
| "loss": 0.0934, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.341634064912796, | |
| "learning_rate": 5.577718201056392e-06, | |
| "loss": 0.1113, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 4.306172839506173, | |
| "grad_norm": 0.35080772638320923, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.1182, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "grad_norm": 0.2861529588699341, | |
| "learning_rate": 4.360836746934055e-06, | |
| "loss": 0.0849, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "eval_loss": 0.1334405094385147, | |
| "eval_runtime": 23.0336, | |
| "eval_samples_per_second": 1.954, | |
| "eval_steps_per_second": 1.954, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.385185185185185, | |
| "grad_norm": 0.35606715083122253, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0789, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 4.424691358024692, | |
| "grad_norm": 0.3928696811199188, | |
| "learning_rate": 3.2875529852700147e-06, | |
| "loss": 0.1082, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 4.4641975308641975, | |
| "grad_norm": 0.2768993377685547, | |
| "learning_rate": 2.8058334845816213e-06, | |
| "loss": 0.0997, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 4.503703703703704, | |
| "grad_norm": 0.4091920554637909, | |
| "learning_rate": 2.361243863855184e-06, | |
| "loss": 0.1468, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 4.54320987654321, | |
| "grad_norm": 0.3549928069114685, | |
| "learning_rate": 1.9541339027450256e-06, | |
| "loss": 0.0803, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.582716049382716, | |
| "grad_norm": 0.41080349683761597, | |
| "learning_rate": 1.584823893886933e-06, | |
| "loss": 0.1158, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.622222222222222, | |
| "grad_norm": 0.33904650807380676, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.0843, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.661728395061728, | |
| "grad_norm": 0.37662726640701294, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.1095, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.701234567901235, | |
| "grad_norm": 0.38046106696128845, | |
| "learning_rate": 7.064490740882057e-07, | |
| "loss": 0.1013, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "grad_norm": 0.2807469964027405, | |
| "learning_rate": 4.909437331777179e-07, | |
| "loss": 0.0928, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "eval_loss": 0.13323988020420074, | |
| "eval_runtime": 23.0604, | |
| "eval_samples_per_second": 1.951, | |
| "eval_steps_per_second": 1.951, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.780246913580247, | |
| "grad_norm": 0.3673052191734314, | |
| "learning_rate": 3.143895053378698e-07, | |
| "loss": 0.1241, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.8197530864197535, | |
| "grad_norm": 0.3105607032775879, | |
| "learning_rate": 1.7692529411904578e-07, | |
| "loss": 0.0803, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.859259259259259, | |
| "grad_norm": 0.31604552268981934, | |
| "learning_rate": 7.865924910916977e-08, | |
| "loss": 0.092, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.898765432098766, | |
| "grad_norm": 0.32336661219596863, | |
| "learning_rate": 1.9668680847356735e-08, | |
| "loss": 0.103, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "grad_norm": 0.30424508452415466, | |
| "learning_rate": 0.0, | |
| "loss": 0.0731, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "step": 125, | |
| "total_flos": 1.3673469165541786e+17, | |
| "train_loss": 0.23027716261148454, | |
| "train_runtime": 3917.2576, | |
| "train_samples_per_second": 0.517, | |
| "train_steps_per_second": 0.032 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 125, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3673469165541786e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |