| { | |
| "best_metric": 0.13747499883174896, | |
| "best_model_checkpoint": "saves/psy-course-info/MentaLLaMA-chat-7B/train/fold9/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": 0.9155487418174744, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 1.0668, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07901234567901234, | |
| "grad_norm": 0.9948328137397766, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 1.0333, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 0.912940263748169, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 1.1047, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1580246913580247, | |
| "grad_norm": 0.9225950837135315, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 1.0671, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 0.9346031546592712, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 1.1099, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 1.0832064151763916, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 1.0688, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2765432098765432, | |
| "grad_norm": 1.0419225692749023, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 1.0821, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3160493827160494, | |
| "grad_norm": 1.1886940002441406, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.8545, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 1.154585599899292, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.8622, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 1.147648572921753, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.7417, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "eval_loss": 0.6983788013458252, | |
| "eval_runtime": 23.0857, | |
| "eval_samples_per_second": 1.949, | |
| "eval_steps_per_second": 1.949, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4345679012345679, | |
| "grad_norm": 1.2759935855865479, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.6613, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 0.9723585844039917, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.6029, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5135802469135803, | |
| "grad_norm": 0.8145802617073059, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4839, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5530864197530864, | |
| "grad_norm": 0.5184298157691956, | |
| "learning_rate": 9.998033131915266e-05, | |
| "loss": 0.4228, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.5669529438018799, | |
| "learning_rate": 9.992134075089084e-05, | |
| "loss": 0.3549, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6320987654320988, | |
| "grad_norm": 0.5457549095153809, | |
| "learning_rate": 9.982307470588098e-05, | |
| "loss": 0.3167, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.671604938271605, | |
| "grad_norm": 0.5821004509925842, | |
| "learning_rate": 9.968561049466214e-05, | |
| "loss": 0.3728, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.5539308190345764, | |
| "learning_rate": 9.950905626682228e-05, | |
| "loss": 0.2675, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7506172839506173, | |
| "grad_norm": 0.5098723769187927, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.3253, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.49655741453170776, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.3034, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "eval_loss": 0.30354806780815125, | |
| "eval_runtime": 23.1096, | |
| "eval_samples_per_second": 1.947, | |
| "eval_steps_per_second": 1.947, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 0.44552484154701233, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.3147, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8691358024691358, | |
| "grad_norm": 0.4713558852672577, | |
| "learning_rate": 9.841517610611309e-05, | |
| "loss": 0.2556, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.908641975308642, | |
| "grad_norm": 0.4724642038345337, | |
| "learning_rate": 9.804586609725499e-05, | |
| "loss": 0.2348, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 0.49854525923728943, | |
| "learning_rate": 9.763875613614482e-05, | |
| "loss": 0.3008, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.4593442976474762, | |
| "learning_rate": 9.719416651541839e-05, | |
| "loss": 0.205, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0271604938271606, | |
| "grad_norm": 0.8938180804252625, | |
| "learning_rate": 9.671244701472999e-05, | |
| "loss": 0.407, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.37950292229652405, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.1968, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.106172839506173, | |
| "grad_norm": 0.33900079131126404, | |
| "learning_rate": 9.563916325306594e-05, | |
| "loss": 0.2109, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.145679012345679, | |
| "grad_norm": 0.3047938048839569, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.1823, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.35635441541671753, | |
| "learning_rate": 9.442228179894362e-05, | |
| "loss": 0.2155, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "eval_loss": 0.22373822331428528, | |
| "eval_runtime": 23.095, | |
| "eval_samples_per_second": 1.948, | |
| "eval_steps_per_second": 1.948, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.2246913580246914, | |
| "grad_norm": 0.4316452443599701, | |
| "learning_rate": 9.376117109543769e-05, | |
| "loss": 0.2239, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2641975308641975, | |
| "grad_norm": 0.46580347418785095, | |
| "learning_rate": 9.306563141162046e-05, | |
| "loss": 0.2611, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 0.35615524649620056, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.1407, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.34320987654321, | |
| "grad_norm": 0.40242788195610046, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.2517, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.33610764145851135, | |
| "learning_rate": 9.077804344796302e-05, | |
| "loss": 0.1893, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.32025980949401855, | |
| "learning_rate": 8.995052426791247e-05, | |
| "loss": 0.1453, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4617283950617284, | |
| "grad_norm": 0.3557215929031372, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.1863, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.5012345679012347, | |
| "grad_norm": 0.40879037976264954, | |
| "learning_rate": 8.820186879108038e-05, | |
| "loss": 0.1903, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 0.39379844069480896, | |
| "learning_rate": 8.728210824415827e-05, | |
| "loss": 0.1668, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.3312002122402191, | |
| "learning_rate": 8.633301610170135e-05, | |
| "loss": 0.154, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "eval_loss": 0.17933771014213562, | |
| "eval_runtime": 23.1017, | |
| "eval_samples_per_second": 1.948, | |
| "eval_steps_per_second": 1.948, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6197530864197531, | |
| "grad_norm": 0.3797035217285156, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.1686, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 0.29529720544815063, | |
| "learning_rate": 8.434984630174509e-05, | |
| "loss": 0.135, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6987654320987655, | |
| "grad_norm": 0.43647223711013794, | |
| "learning_rate": 8.33173288976002e-05, | |
| "loss": 0.2429, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.7382716049382716, | |
| "grad_norm": 0.35793933272361755, | |
| "learning_rate": 8.225859917710439e-05, | |
| "loss": 0.1746, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.39779555797576904, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.1486, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.817283950617284, | |
| "grad_norm": 0.3191227316856384, | |
| "learning_rate": 8.006585456492029e-05, | |
| "loss": 0.119, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8567901234567903, | |
| "grad_norm": 0.429025799036026, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.1293, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 0.35993286967277527, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.175, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.9358024691358025, | |
| "grad_norm": 0.3280028700828552, | |
| "learning_rate": 7.660160382576683e-05, | |
| "loss": 0.1404, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.269521027803421, | |
| "learning_rate": 7.540376726232648e-05, | |
| "loss": 0.1459, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "eval_loss": 0.16235865652561188, | |
| "eval_runtime": 23.0869, | |
| "eval_samples_per_second": 1.949, | |
| "eval_steps_per_second": 1.949, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 0.5833605527877808, | |
| "learning_rate": 7.4185944355262e-05, | |
| "loss": 0.2696, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.054320987654321, | |
| "grad_norm": 0.327554851770401, | |
| "learning_rate": 7.294909322337689e-05, | |
| "loss": 0.1523, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.093827160493827, | |
| "grad_norm": 0.3388730585575104, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.1553, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.2606907784938812, | |
| "learning_rate": 7.042221284679982e-05, | |
| "loss": 0.1233, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.2667919993400574, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.1037, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.212345679012346, | |
| "grad_norm": 0.31768813729286194, | |
| "learning_rate": 6.783107663311565e-05, | |
| "loss": 0.1251, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 0.3257889747619629, | |
| "learning_rate": 6.651395309775837e-05, | |
| "loss": 0.1393, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.291358024691358, | |
| "grad_norm": 0.4400660991668701, | |
| "learning_rate": 6.518383725548074e-05, | |
| "loss": 0.1393, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.330864197530864, | |
| "grad_norm": 0.299668550491333, | |
| "learning_rate": 6.384177557124247e-05, | |
| "loss": 0.1389, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.32180729508399963, | |
| "learning_rate": 6.248882390836135e-05, | |
| "loss": 0.1307, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "eval_loss": 0.1563589721918106, | |
| "eval_runtime": 23.0718, | |
| "eval_samples_per_second": 1.95, | |
| "eval_steps_per_second": 1.95, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.4098765432098768, | |
| "grad_norm": 0.33588045835494995, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.1136, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.449382716049383, | |
| "grad_norm": 0.3659997284412384, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 0.1284, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.4478165805339813, | |
| "learning_rate": 5.837531116523682e-05, | |
| "loss": 0.132, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.528395061728395, | |
| "grad_norm": 0.3690505027770996, | |
| "learning_rate": 5.698951697677498e-05, | |
| "loss": 0.1975, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 0.32860100269317627, | |
| "learning_rate": 5.559822380516539e-05, | |
| "loss": 0.1393, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 0.37297317385673523, | |
| "learning_rate": 5.420252624646238e-05, | |
| "loss": 0.1498, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.6469135802469137, | |
| "grad_norm": 0.489502876996994, | |
| "learning_rate": 5.2803522361859594e-05, | |
| "loss": 0.1468, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.68641975308642, | |
| "grad_norm": 0.42096662521362305, | |
| "learning_rate": 5.140231281379345e-05, | |
| "loss": 0.1434, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 0.32388579845428467, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1181, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.35497477650642395, | |
| "learning_rate": 4.859768718620656e-05, | |
| "loss": 0.118, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "eval_loss": 0.14812791347503662, | |
| "eval_runtime": 23.0928, | |
| "eval_samples_per_second": 1.949, | |
| "eval_steps_per_second": 1.949, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8049382716049385, | |
| "grad_norm": 0.3953438699245453, | |
| "learning_rate": 4.7196477638140404e-05, | |
| "loss": 0.1143, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.3005009591579437, | |
| "learning_rate": 4.579747375353763e-05, | |
| "loss": 0.1166, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.8839506172839506, | |
| "grad_norm": 0.3109400272369385, | |
| "learning_rate": 4.4401776194834613e-05, | |
| "loss": 0.1355, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.9234567901234567, | |
| "grad_norm": 0.31005656719207764, | |
| "learning_rate": 4.3010483023225045e-05, | |
| "loss": 0.0987, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.4532912075519562, | |
| "learning_rate": 4.162468883476319e-05, | |
| "loss": 0.1418, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.0024691358024693, | |
| "grad_norm": 0.5759482383728027, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 0.1754, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 3.0419753086419754, | |
| "grad_norm": 0.2855320870876312, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.0709, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 3.0814814814814815, | |
| "grad_norm": 0.34833818674087524, | |
| "learning_rate": 3.7511176091638653e-05, | |
| "loss": 0.1211, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.1209876543209876, | |
| "grad_norm": 0.3245695233345032, | |
| "learning_rate": 3.6158224428757535e-05, | |
| "loss": 0.1239, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "grad_norm": 0.30690744519233704, | |
| "learning_rate": 3.4816162744519263e-05, | |
| "loss": 0.1127, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "eval_loss": 0.14236007630825043, | |
| "eval_runtime": 23.0372, | |
| "eval_samples_per_second": 1.953, | |
| "eval_steps_per_second": 1.953, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.3772599995136261, | |
| "learning_rate": 3.3486046902241664e-05, | |
| "loss": 0.1271, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 3.2395061728395063, | |
| "grad_norm": 0.2677428424358368, | |
| "learning_rate": 3.216892336688435e-05, | |
| "loss": 0.0626, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 3.2790123456790123, | |
| "grad_norm": 0.3292672038078308, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0984, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 3.3185185185185184, | |
| "grad_norm": 0.36312535405158997, | |
| "learning_rate": 2.9577787153200197e-05, | |
| "loss": 0.1117, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 3.3580246913580245, | |
| "grad_norm": 0.35500839352607727, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.1014, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.397530864197531, | |
| "grad_norm": 0.2993975579738617, | |
| "learning_rate": 2.705090677662311e-05, | |
| "loss": 0.1036, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.437037037037037, | |
| "grad_norm": 0.39413362741470337, | |
| "learning_rate": 2.581405564473801e-05, | |
| "loss": 0.1366, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.476543209876543, | |
| "grad_norm": 0.29331687092781067, | |
| "learning_rate": 2.459623273767354e-05, | |
| "loss": 0.103, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.5160493827160493, | |
| "grad_norm": 0.34595251083374023, | |
| "learning_rate": 2.3398396174233178e-05, | |
| "loss": 0.0778, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.32554715871810913, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.0948, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "eval_loss": 0.13890397548675537, | |
| "eval_runtime": 23.0516, | |
| "eval_samples_per_second": 1.952, | |
| "eval_steps_per_second": 1.952, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.595061728395062, | |
| "grad_norm": 0.3748563528060913, | |
| "learning_rate": 2.1066435191009715e-05, | |
| "loss": 0.1479, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.634567901234568, | |
| "grad_norm": 0.4145781397819519, | |
| "learning_rate": 1.9934145435079702e-05, | |
| "loss": 0.1198, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.674074074074074, | |
| "grad_norm": 0.3355666399002075, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.1082, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.71358024691358, | |
| "grad_norm": 0.3941483497619629, | |
| "learning_rate": 1.774140082289563e-05, | |
| "loss": 0.1351, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.753086419753086, | |
| "grad_norm": 0.39614933729171753, | |
| "learning_rate": 1.6682671102399805e-05, | |
| "loss": 0.135, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.7925925925925927, | |
| "grad_norm": 0.2684767544269562, | |
| "learning_rate": 1.5650153698254916e-05, | |
| "loss": 0.0631, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.832098765432099, | |
| "grad_norm": 0.2957228720188141, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.0982, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.871604938271605, | |
| "grad_norm": 0.36905595660209656, | |
| "learning_rate": 1.3666983898298657e-05, | |
| "loss": 0.1093, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 0.37501490116119385, | |
| "learning_rate": 1.2717891755841722e-05, | |
| "loss": 0.1176, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "grad_norm": 0.4031924903392792, | |
| "learning_rate": 1.1798131208919627e-05, | |
| "loss": 0.1181, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "eval_loss": 0.13871604204177856, | |
| "eval_runtime": 23.0409, | |
| "eval_samples_per_second": 1.953, | |
| "eval_steps_per_second": 1.953, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.9901234567901236, | |
| "grad_norm": 0.5714593529701233, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 0.1484, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 4.029629629629629, | |
| "grad_norm": 0.6697483062744141, | |
| "learning_rate": 1.004947573208756e-05, | |
| "loss": 0.1785, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 4.069135802469136, | |
| "grad_norm": 0.28352972865104675, | |
| "learning_rate": 9.221956552036992e-06, | |
| "loss": 0.0781, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 4.108641975308642, | |
| "grad_norm": 0.2981833219528198, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 0.0901, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 4.148148148148148, | |
| "grad_norm": 0.3071352243423462, | |
| "learning_rate": 7.663790038585793e-06, | |
| "loss": 0.0846, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.187654320987654, | |
| "grad_norm": 0.3345410227775574, | |
| "learning_rate": 6.934368588379553e-06, | |
| "loss": 0.0859, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 4.22716049382716, | |
| "grad_norm": 0.37498584389686584, | |
| "learning_rate": 6.238828904562316e-06, | |
| "loss": 0.0859, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.36547139286994934, | |
| "learning_rate": 5.577718201056392e-06, | |
| "loss": 0.1256, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 4.306172839506173, | |
| "grad_norm": 0.3687337040901184, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.1225, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "grad_norm": 0.3142211139202118, | |
| "learning_rate": 4.360836746934055e-06, | |
| "loss": 0.0853, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "eval_loss": 0.13770605623722076, | |
| "eval_runtime": 23.0343, | |
| "eval_samples_per_second": 1.954, | |
| "eval_steps_per_second": 1.954, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.385185185185185, | |
| "grad_norm": 0.3303620517253876, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.1099, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 4.424691358024692, | |
| "grad_norm": 0.3397717773914337, | |
| "learning_rate": 3.2875529852700147e-06, | |
| "loss": 0.1137, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 4.4641975308641975, | |
| "grad_norm": 0.3000347316265106, | |
| "learning_rate": 2.8058334845816213e-06, | |
| "loss": 0.0778, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 4.503703703703704, | |
| "grad_norm": 0.3576681315898895, | |
| "learning_rate": 2.361243863855184e-06, | |
| "loss": 0.1174, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 4.54320987654321, | |
| "grad_norm": 0.3489622175693512, | |
| "learning_rate": 1.9541339027450256e-06, | |
| "loss": 0.0888, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.582716049382716, | |
| "grad_norm": 0.3765661120414734, | |
| "learning_rate": 1.584823893886933e-06, | |
| "loss": 0.1099, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.622222222222222, | |
| "grad_norm": 0.2720748782157898, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.0751, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.661728395061728, | |
| "grad_norm": 0.35195329785346985, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.1105, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.701234567901235, | |
| "grad_norm": 0.3578346371650696, | |
| "learning_rate": 7.064490740882057e-07, | |
| "loss": 0.1285, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "grad_norm": 0.3291870057582855, | |
| "learning_rate": 4.909437331777179e-07, | |
| "loss": 0.0862, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "eval_loss": 0.13747499883174896, | |
| "eval_runtime": 23.0197, | |
| "eval_samples_per_second": 1.955, | |
| "eval_steps_per_second": 1.955, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.780246913580247, | |
| "grad_norm": 0.37761127948760986, | |
| "learning_rate": 3.143895053378698e-07, | |
| "loss": 0.128, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.8197530864197535, | |
| "grad_norm": 0.3761325776576996, | |
| "learning_rate": 1.7692529411904578e-07, | |
| "loss": 0.0913, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.859259259259259, | |
| "grad_norm": 0.31441816687583923, | |
| "learning_rate": 7.865924910916977e-08, | |
| "loss": 0.0805, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.898765432098766, | |
| "grad_norm": 0.36645349860191345, | |
| "learning_rate": 1.9668680847356735e-08, | |
| "loss": 0.1151, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "grad_norm": 0.2833465039730072, | |
| "learning_rate": 0.0, | |
| "loss": 0.1047, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "step": 125, | |
| "total_flos": 1.381954616866898e+17, | |
| "train_loss": 0.23185506176948548, | |
| "train_runtime": 4051.8599, | |
| "train_samples_per_second": 0.5, | |
| "train_steps_per_second": 0.031 | |
| } | |
| ], | |
| "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": 1.381954616866898e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |