| { | |
| "best_metric": 0.13598664104938507, | |
| "best_model_checkpoint": "saves/psy-course-info/Llama-3.1-8B-Instruct/train/fold7/checkpoint-100", | |
| "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": 2.156710386276245, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.9659, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07901234567901234, | |
| "grad_norm": 2.3100240230560303, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 1.2665, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 2.1593027114868164, | |
| "learning_rate": 2.307692307692308e-05, | |
| "loss": 1.0936, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.1580246913580247, | |
| "grad_norm": 2.3243963718414307, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 1.0526, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19753086419753085, | |
| "grad_norm": 2.478224515914917, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 1.0798, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 2.384239673614502, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 1.0236, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.2765432098765432, | |
| "grad_norm": 2.593691349029541, | |
| "learning_rate": 5.384615384615385e-05, | |
| "loss": 0.9135, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3160493827160494, | |
| "grad_norm": 2.4142556190490723, | |
| "learning_rate": 6.153846153846155e-05, | |
| "loss": 0.7278, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 2.4047136306762695, | |
| "learning_rate": 6.923076923076924e-05, | |
| "loss": 0.611, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "grad_norm": 1.5310007333755493, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.5179, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3950617283950617, | |
| "eval_loss": 0.43637895584106445, | |
| "eval_runtime": 22.577, | |
| "eval_samples_per_second": 1.993, | |
| "eval_steps_per_second": 1.993, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.4345679012345679, | |
| "grad_norm": 0.9813531637191772, | |
| "learning_rate": 8.461538461538461e-05, | |
| "loss": 0.4713, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 0.910319447517395, | |
| "learning_rate": 9.230769230769232e-05, | |
| "loss": 0.4306, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.5135802469135803, | |
| "grad_norm": 1.0288078784942627, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4022, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.5530864197530864, | |
| "grad_norm": 1.4431365728378296, | |
| "learning_rate": 9.998033131915266e-05, | |
| "loss": 0.5042, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 1.0542594194412231, | |
| "learning_rate": 9.992134075089084e-05, | |
| "loss": 0.4106, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.6320987654320988, | |
| "grad_norm": 0.9218867421150208, | |
| "learning_rate": 9.982307470588098e-05, | |
| "loss": 0.3652, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.671604938271605, | |
| "grad_norm": 0.7790356874465942, | |
| "learning_rate": 9.968561049466214e-05, | |
| "loss": 0.3193, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 0.7652316093444824, | |
| "learning_rate": 9.950905626682228e-05, | |
| "loss": 0.2857, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7506172839506173, | |
| "grad_norm": 0.6801689863204956, | |
| "learning_rate": 9.92935509259118e-05, | |
| "loss": 0.2022, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "grad_norm": 0.6742631196975708, | |
| "learning_rate": 9.903926402016153e-05, | |
| "loss": 0.258, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.7901234567901234, | |
| "eval_loss": 0.23935669660568237, | |
| "eval_runtime": 22.4807, | |
| "eval_samples_per_second": 2.002, | |
| "eval_steps_per_second": 2.002, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 0.5498079657554626, | |
| "learning_rate": 9.874639560909117e-05, | |
| "loss": 0.2784, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.8691358024691358, | |
| "grad_norm": 0.5437087416648865, | |
| "learning_rate": 9.841517610611309e-05, | |
| "loss": 0.162, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.908641975308642, | |
| "grad_norm": 0.6846128702163696, | |
| "learning_rate": 9.804586609725499e-05, | |
| "loss": 0.235, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 0.6774247288703918, | |
| "learning_rate": 9.763875613614482e-05, | |
| "loss": 0.2233, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 0.8547881841659546, | |
| "learning_rate": 9.719416651541839e-05, | |
| "loss": 0.3232, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 1.0271604938271606, | |
| "grad_norm": 0.9815611243247986, | |
| "learning_rate": 9.671244701472999e-05, | |
| "loss": 0.3709, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 0.6782816648483276, | |
| "learning_rate": 9.619397662556435e-05, | |
| "loss": 0.2099, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.106172839506173, | |
| "grad_norm": 0.6780818700790405, | |
| "learning_rate": 9.563916325306594e-05, | |
| "loss": 0.2107, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.145679012345679, | |
| "grad_norm": 0.7424414753913879, | |
| "learning_rate": 9.504844339512095e-05, | |
| "loss": 0.1588, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "grad_norm": 0.574498176574707, | |
| "learning_rate": 9.442228179894362e-05, | |
| "loss": 0.1759, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.1851851851851851, | |
| "eval_loss": 0.19256526231765747, | |
| "eval_runtime": 22.436, | |
| "eval_samples_per_second": 2.006, | |
| "eval_steps_per_second": 2.006, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.2246913580246914, | |
| "grad_norm": 0.6567896008491516, | |
| "learning_rate": 9.376117109543769e-05, | |
| "loss": 0.1765, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.2641975308641975, | |
| "grad_norm": 0.745578944683075, | |
| "learning_rate": 9.306563141162046e-05, | |
| "loss": 0.202, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.3037037037037038, | |
| "grad_norm": 0.578172504901886, | |
| "learning_rate": 9.233620996141421e-05, | |
| "loss": 0.1826, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.34320987654321, | |
| "grad_norm": 0.5294815301895142, | |
| "learning_rate": 9.157348061512727e-05, | |
| "loss": 0.2087, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.382716049382716, | |
| "grad_norm": 0.5300359725952148, | |
| "learning_rate": 9.077804344796302e-05, | |
| "loss": 0.1655, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.4222222222222223, | |
| "grad_norm": 0.7045089602470398, | |
| "learning_rate": 8.995052426791247e-05, | |
| "loss": 0.2385, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.4617283950617284, | |
| "grad_norm": 0.48318779468536377, | |
| "learning_rate": 8.90915741234015e-05, | |
| "loss": 0.1358, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.5012345679012347, | |
| "grad_norm": 0.45993906259536743, | |
| "learning_rate": 8.820186879108038e-05, | |
| "loss": 0.1462, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.5407407407407407, | |
| "grad_norm": 0.6465219855308533, | |
| "learning_rate": 8.728210824415827e-05, | |
| "loss": 0.1938, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "grad_norm": 0.6309574842453003, | |
| "learning_rate": 8.633301610170135e-05, | |
| "loss": 0.1809, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.5802469135802468, | |
| "eval_loss": 0.17272719740867615, | |
| "eval_runtime": 22.4849, | |
| "eval_samples_per_second": 2.001, | |
| "eval_steps_per_second": 2.001, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.6197530864197531, | |
| "grad_norm": 0.46850496530532837, | |
| "learning_rate": 8.535533905932738e-05, | |
| "loss": 0.1402, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.6592592592592592, | |
| "grad_norm": 0.5316647887229919, | |
| "learning_rate": 8.434984630174509e-05, | |
| "loss": 0.1462, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.6987654320987655, | |
| "grad_norm": 0.5687728524208069, | |
| "learning_rate": 8.33173288976002e-05, | |
| "loss": 0.1327, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.7382716049382716, | |
| "grad_norm": 0.7694992423057556, | |
| "learning_rate": 8.225859917710439e-05, | |
| "loss": 0.1758, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.7777777777777777, | |
| "grad_norm": 0.5915424823760986, | |
| "learning_rate": 8.117449009293668e-05, | |
| "loss": 0.1557, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.817283950617284, | |
| "grad_norm": 0.6176837086677551, | |
| "learning_rate": 8.006585456492029e-05, | |
| "loss": 0.1728, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.8567901234567903, | |
| "grad_norm": 0.5276793241500854, | |
| "learning_rate": 7.89335648089903e-05, | |
| "loss": 0.1459, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.8962962962962964, | |
| "grad_norm": 0.6202795505523682, | |
| "learning_rate": 7.777851165098012e-05, | |
| "loss": 0.1901, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.9358024691358025, | |
| "grad_norm": 0.4415624141693115, | |
| "learning_rate": 7.660160382576683e-05, | |
| "loss": 0.0962, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 0.7787760496139526, | |
| "learning_rate": 7.540376726232648e-05, | |
| "loss": 0.139, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "eval_loss": 0.16411535441875458, | |
| "eval_runtime": 22.4686, | |
| "eval_samples_per_second": 2.003, | |
| "eval_steps_per_second": 2.003, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 2.0148148148148146, | |
| "grad_norm": 0.8991931676864624, | |
| "learning_rate": 7.4185944355262e-05, | |
| "loss": 0.2781, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 2.054320987654321, | |
| "grad_norm": 0.5378757119178772, | |
| "learning_rate": 7.294909322337689e-05, | |
| "loss": 0.1509, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 2.093827160493827, | |
| "grad_norm": 0.5631323456764221, | |
| "learning_rate": 7.169418695587791e-05, | |
| "loss": 0.1317, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.9004403352737427, | |
| "learning_rate": 7.042221284679982e-05, | |
| "loss": 0.207, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.1728395061728394, | |
| "grad_norm": 0.44686245918273926, | |
| "learning_rate": 6.91341716182545e-05, | |
| "loss": 0.0936, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.212345679012346, | |
| "grad_norm": 0.6498768925666809, | |
| "learning_rate": 6.783107663311565e-05, | |
| "loss": 0.1022, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.251851851851852, | |
| "grad_norm": 0.549042284488678, | |
| "learning_rate": 6.651395309775837e-05, | |
| "loss": 0.1409, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.291358024691358, | |
| "grad_norm": 0.5494434237480164, | |
| "learning_rate": 6.518383725548074e-05, | |
| "loss": 0.1308, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.330864197530864, | |
| "grad_norm": 0.7209091782569885, | |
| "learning_rate": 6.384177557124247e-05, | |
| "loss": 0.1569, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "grad_norm": 0.5737520456314087, | |
| "learning_rate": 6.248882390836135e-05, | |
| "loss": 0.0997, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.3703703703703702, | |
| "eval_loss": 0.15064960718154907, | |
| "eval_runtime": 22.4744, | |
| "eval_samples_per_second": 2.002, | |
| "eval_steps_per_second": 2.002, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.4098765432098768, | |
| "grad_norm": 0.604082465171814, | |
| "learning_rate": 6.112604669781572e-05, | |
| "loss": 0.1582, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.449382716049383, | |
| "grad_norm": 0.5986130833625793, | |
| "learning_rate": 5.9754516100806423e-05, | |
| "loss": 0.1267, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.488888888888889, | |
| "grad_norm": 0.5138416886329651, | |
| "learning_rate": 5.837531116523682e-05, | |
| "loss": 0.1065, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.528395061728395, | |
| "grad_norm": 0.4618552625179291, | |
| "learning_rate": 5.698951697677498e-05, | |
| "loss": 0.101, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.567901234567901, | |
| "grad_norm": 0.5696907043457031, | |
| "learning_rate": 5.559822380516539e-05, | |
| "loss": 0.1242, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.6074074074074076, | |
| "grad_norm": 0.6115525960922241, | |
| "learning_rate": 5.420252624646238e-05, | |
| "loss": 0.1281, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.6469135802469137, | |
| "grad_norm": 0.6067851781845093, | |
| "learning_rate": 5.2803522361859594e-05, | |
| "loss": 0.0917, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.68641975308642, | |
| "grad_norm": 0.4724122881889343, | |
| "learning_rate": 5.140231281379345e-05, | |
| "loss": 0.1096, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.725925925925926, | |
| "grad_norm": 0.5594306588172913, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1123, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "grad_norm": 0.5291305780410767, | |
| "learning_rate": 4.859768718620656e-05, | |
| "loss": 0.1006, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.765432098765432, | |
| "eval_loss": 0.14511488378047943, | |
| "eval_runtime": 22.4827, | |
| "eval_samples_per_second": 2.002, | |
| "eval_steps_per_second": 2.002, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.8049382716049385, | |
| "grad_norm": 0.5939591526985168, | |
| "learning_rate": 4.7196477638140404e-05, | |
| "loss": 0.1225, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.8444444444444446, | |
| "grad_norm": 0.7044705152511597, | |
| "learning_rate": 4.579747375353763e-05, | |
| "loss": 0.1193, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.8839506172839506, | |
| "grad_norm": 0.5918217897415161, | |
| "learning_rate": 4.4401776194834613e-05, | |
| "loss": 0.1154, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.9234567901234567, | |
| "grad_norm": 0.511033296585083, | |
| "learning_rate": 4.3010483023225045e-05, | |
| "loss": 0.1204, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 0.601833164691925, | |
| "learning_rate": 4.162468883476319e-05, | |
| "loss": 0.102, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 3.0024691358024693, | |
| "grad_norm": 1.042554497718811, | |
| "learning_rate": 4.0245483899193595e-05, | |
| "loss": 0.1383, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 3.0419753086419754, | |
| "grad_norm": 0.4085560142993927, | |
| "learning_rate": 3.887395330218429e-05, | |
| "loss": 0.08, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 3.0814814814814815, | |
| "grad_norm": 0.3641754686832428, | |
| "learning_rate": 3.7511176091638653e-05, | |
| "loss": 0.0663, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 3.1209876543209876, | |
| "grad_norm": 0.4446675479412079, | |
| "learning_rate": 3.6158224428757535e-05, | |
| "loss": 0.0737, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "grad_norm": 0.48916590213775635, | |
| "learning_rate": 3.4816162744519263e-05, | |
| "loss": 0.0925, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.1604938271604937, | |
| "eval_loss": 0.14041271805763245, | |
| "eval_runtime": 22.4802, | |
| "eval_samples_per_second": 2.002, | |
| "eval_steps_per_second": 2.002, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.42323732376098633, | |
| "learning_rate": 3.3486046902241664e-05, | |
| "loss": 0.0729, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 3.2395061728395063, | |
| "grad_norm": 0.4948965311050415, | |
| "learning_rate": 3.216892336688435e-05, | |
| "loss": 0.1088, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 3.2790123456790123, | |
| "grad_norm": 0.453052282333374, | |
| "learning_rate": 3.086582838174551e-05, | |
| "loss": 0.0863, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 3.3185185185185184, | |
| "grad_norm": 0.5081247687339783, | |
| "learning_rate": 2.9577787153200197e-05, | |
| "loss": 0.0926, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 3.3580246913580245, | |
| "grad_norm": 0.5067690014839172, | |
| "learning_rate": 2.8305813044122097e-05, | |
| "loss": 0.0973, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 3.397530864197531, | |
| "grad_norm": 0.5217188596725464, | |
| "learning_rate": 2.705090677662311e-05, | |
| "loss": 0.078, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 3.437037037037037, | |
| "grad_norm": 0.6190178990364075, | |
| "learning_rate": 2.581405564473801e-05, | |
| "loss": 0.1133, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 3.476543209876543, | |
| "grad_norm": 0.5065380334854126, | |
| "learning_rate": 2.459623273767354e-05, | |
| "loss": 0.077, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 3.5160493827160493, | |
| "grad_norm": 0.5724349021911621, | |
| "learning_rate": 2.3398396174233178e-05, | |
| "loss": 0.094, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "grad_norm": 0.5281288027763367, | |
| "learning_rate": 2.2221488349019903e-05, | |
| "loss": 0.0913, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.5555555555555554, | |
| "eval_loss": 0.1376548707485199, | |
| "eval_runtime": 22.4783, | |
| "eval_samples_per_second": 2.002, | |
| "eval_steps_per_second": 2.002, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 3.595061728395062, | |
| "grad_norm": 0.4375687539577484, | |
| "learning_rate": 2.1066435191009715e-05, | |
| "loss": 0.052, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 3.634567901234568, | |
| "grad_norm": 0.6882292032241821, | |
| "learning_rate": 1.9934145435079702e-05, | |
| "loss": 0.1419, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 3.674074074074074, | |
| "grad_norm": 0.5077499747276306, | |
| "learning_rate": 1.8825509907063327e-05, | |
| "loss": 0.094, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 3.71358024691358, | |
| "grad_norm": 0.5534601211547852, | |
| "learning_rate": 1.774140082289563e-05, | |
| "loss": 0.0997, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 3.753086419753086, | |
| "grad_norm": 0.6414083242416382, | |
| "learning_rate": 1.6682671102399805e-05, | |
| "loss": 0.1133, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 3.7925925925925927, | |
| "grad_norm": 0.5925371050834656, | |
| "learning_rate": 1.5650153698254916e-05, | |
| "loss": 0.0803, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 3.832098765432099, | |
| "grad_norm": 0.5218459963798523, | |
| "learning_rate": 1.4644660940672627e-05, | |
| "loss": 0.0662, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 3.871604938271605, | |
| "grad_norm": 0.6515846848487854, | |
| "learning_rate": 1.3666983898298657e-05, | |
| "loss": 0.1118, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 3.911111111111111, | |
| "grad_norm": 0.6535831093788147, | |
| "learning_rate": 1.2717891755841722e-05, | |
| "loss": 0.1061, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "grad_norm": 0.43857115507125854, | |
| "learning_rate": 1.1798131208919627e-05, | |
| "loss": 0.0819, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.950617283950617, | |
| "eval_loss": 0.13598664104938507, | |
| "eval_runtime": 22.4976, | |
| "eval_samples_per_second": 2.0, | |
| "eval_steps_per_second": 2.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.9901234567901236, | |
| "grad_norm": 0.938236653804779, | |
| "learning_rate": 1.090842587659851e-05, | |
| "loss": 0.1168, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 4.029629629629629, | |
| "grad_norm": 0.6755374073982239, | |
| "learning_rate": 1.004947573208756e-05, | |
| "loss": 0.1473, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 4.069135802469136, | |
| "grad_norm": 0.47335633635520935, | |
| "learning_rate": 9.221956552036992e-06, | |
| "loss": 0.0745, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 4.108641975308642, | |
| "grad_norm": 0.4839572608470917, | |
| "learning_rate": 8.426519384872733e-06, | |
| "loss": 0.0621, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 4.148148148148148, | |
| "grad_norm": 0.5066091418266296, | |
| "learning_rate": 7.663790038585793e-06, | |
| "loss": 0.0839, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 4.187654320987654, | |
| "grad_norm": 0.5020496845245361, | |
| "learning_rate": 6.934368588379553e-06, | |
| "loss": 0.0732, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 4.22716049382716, | |
| "grad_norm": 0.4842373728752136, | |
| "learning_rate": 6.238828904562316e-06, | |
| "loss": 0.0738, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 4.266666666666667, | |
| "grad_norm": 0.4646170437335968, | |
| "learning_rate": 5.577718201056392e-06, | |
| "loss": 0.084, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 4.306172839506173, | |
| "grad_norm": 0.6061789393424988, | |
| "learning_rate": 4.951556604879048e-06, | |
| "loss": 0.1015, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "grad_norm": 0.4986622929573059, | |
| "learning_rate": 4.360836746934055e-06, | |
| "loss": 0.0913, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.345679012345679, | |
| "eval_loss": 0.13764117658138275, | |
| "eval_runtime": 22.4738, | |
| "eval_samples_per_second": 2.002, | |
| "eval_steps_per_second": 2.002, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 4.385185185185185, | |
| "grad_norm": 0.4486829936504364, | |
| "learning_rate": 3.8060233744356633e-06, | |
| "loss": 0.0635, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 4.424691358024692, | |
| "grad_norm": 0.711368203163147, | |
| "learning_rate": 3.2875529852700147e-06, | |
| "loss": 0.1398, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 4.4641975308641975, | |
| "grad_norm": 0.5507614612579346, | |
| "learning_rate": 2.8058334845816213e-06, | |
| "loss": 0.0785, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 4.503703703703704, | |
| "grad_norm": 0.559026837348938, | |
| "learning_rate": 2.361243863855184e-06, | |
| "loss": 0.0741, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 4.54320987654321, | |
| "grad_norm": 0.4491055905818939, | |
| "learning_rate": 1.9541339027450256e-06, | |
| "loss": 0.0561, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 4.582716049382716, | |
| "grad_norm": 0.408561110496521, | |
| "learning_rate": 1.584823893886933e-06, | |
| "loss": 0.0607, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 4.622222222222222, | |
| "grad_norm": 0.46006232500076294, | |
| "learning_rate": 1.2536043909088191e-06, | |
| "loss": 0.0539, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 4.661728395061728, | |
| "grad_norm": 0.45019295811653137, | |
| "learning_rate": 9.607359798384785e-07, | |
| "loss": 0.0843, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 4.701234567901235, | |
| "grad_norm": 0.42193466424942017, | |
| "learning_rate": 7.064490740882057e-07, | |
| "loss": 0.0503, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "grad_norm": 0.38761746883392334, | |
| "learning_rate": 4.909437331777179e-07, | |
| "loss": 0.0575, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.7407407407407405, | |
| "eval_loss": 0.13801008462905884, | |
| "eval_runtime": 22.4665, | |
| "eval_samples_per_second": 2.003, | |
| "eval_steps_per_second": 2.003, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 4.780246913580247, | |
| "grad_norm": 0.6245124936103821, | |
| "learning_rate": 3.143895053378698e-07, | |
| "loss": 0.0712, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 4.8197530864197535, | |
| "grad_norm": 0.47812676429748535, | |
| "learning_rate": 1.7692529411904578e-07, | |
| "loss": 0.0852, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 4.859259259259259, | |
| "grad_norm": 0.46076539158821106, | |
| "learning_rate": 7.865924910916977e-08, | |
| "loss": 0.0887, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 4.898765432098766, | |
| "grad_norm": 0.5903412699699402, | |
| "learning_rate": 1.9668680847356735e-08, | |
| "loss": 0.0903, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "grad_norm": 0.4551125168800354, | |
| "learning_rate": 0.0, | |
| "loss": 0.0669, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 4.938271604938271, | |
| "step": 125, | |
| "total_flos": 1.252407866106839e+17, | |
| "train_loss": 0.2091505795121193, | |
| "train_runtime": 3501.4737, | |
| "train_samples_per_second": 0.578, | |
| "train_steps_per_second": 0.036 | |
| } | |
| ], | |
| "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": 1.252407866106839e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |