| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 50, | |
| "global_step": 338, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002962962962962963, | |
| "grad_norm": 18.54292869567871, | |
| "learning_rate": 0.0, | |
| "loss": 1.1026, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005925925925925926, | |
| "grad_norm": 21.237518310546875, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.3128, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.008888888888888889, | |
| "grad_norm": 15.317862510681152, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.0681, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.011851851851851851, | |
| "grad_norm": 12.676899909973145, | |
| "learning_rate": 6e-06, | |
| "loss": 0.8133, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.014814814814814815, | |
| "grad_norm": 10.247825622558594, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.6555, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.017777777777777778, | |
| "grad_norm": 10.220847129821777, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7156, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02074074074074074, | |
| "grad_norm": 8.812250137329102, | |
| "learning_rate": 9.96996996996997e-06, | |
| "loss": 0.5781, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.023703703703703703, | |
| "grad_norm": 5.879333019256592, | |
| "learning_rate": 9.93993993993994e-06, | |
| "loss": 0.3565, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.02666666666666667, | |
| "grad_norm": 4.63343620300293, | |
| "learning_rate": 9.90990990990991e-06, | |
| "loss": 0.3589, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.02962962962962963, | |
| "grad_norm": 3.62245774269104, | |
| "learning_rate": 9.879879879879881e-06, | |
| "loss": 0.3689, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03259259259259259, | |
| "grad_norm": 2.765944242477417, | |
| "learning_rate": 9.849849849849851e-06, | |
| "loss": 0.2733, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.035555555555555556, | |
| "grad_norm": 3.584216594696045, | |
| "learning_rate": 9.81981981981982e-06, | |
| "loss": 0.3713, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03851851851851852, | |
| "grad_norm": 2.7825379371643066, | |
| "learning_rate": 9.78978978978979e-06, | |
| "loss": 0.2065, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.04148148148148148, | |
| "grad_norm": 3.2220797538757324, | |
| "learning_rate": 9.75975975975976e-06, | |
| "loss": 0.339, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.044444444444444446, | |
| "grad_norm": 2.662395477294922, | |
| "learning_rate": 9.729729729729732e-06, | |
| "loss": 0.2139, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.047407407407407405, | |
| "grad_norm": 3.193700075149536, | |
| "learning_rate": 9.699699699699701e-06, | |
| "loss": 0.2818, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.05037037037037037, | |
| "grad_norm": 2.712399482727051, | |
| "learning_rate": 9.669669669669671e-06, | |
| "loss": 0.2334, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.05333333333333334, | |
| "grad_norm": 1.8682827949523926, | |
| "learning_rate": 9.63963963963964e-06, | |
| "loss": 0.1452, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.056296296296296296, | |
| "grad_norm": 2.17755126953125, | |
| "learning_rate": 9.60960960960961e-06, | |
| "loss": 0.2388, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.05925925925925926, | |
| "grad_norm": 2.2480223178863525, | |
| "learning_rate": 9.57957957957958e-06, | |
| "loss": 0.2632, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06222222222222222, | |
| "grad_norm": 1.6088407039642334, | |
| "learning_rate": 9.54954954954955e-06, | |
| "loss": 0.1448, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.06518518518518518, | |
| "grad_norm": 2.2946293354034424, | |
| "learning_rate": 9.51951951951952e-06, | |
| "loss": 0.2657, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.06814814814814815, | |
| "grad_norm": 2.1780478954315186, | |
| "learning_rate": 9.489489489489491e-06, | |
| "loss": 0.2869, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.07111111111111111, | |
| "grad_norm": 2.410952568054199, | |
| "learning_rate": 9.45945945945946e-06, | |
| "loss": 0.2952, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 1.8929845094680786, | |
| "learning_rate": 9.42942942942943e-06, | |
| "loss": 0.1793, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.07703703703703704, | |
| "grad_norm": 1.981755256652832, | |
| "learning_rate": 9.3993993993994e-06, | |
| "loss": 0.2381, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 2.8715898990631104, | |
| "learning_rate": 9.36936936936937e-06, | |
| "loss": 0.2877, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.08296296296296296, | |
| "grad_norm": 2.0459938049316406, | |
| "learning_rate": 9.339339339339341e-06, | |
| "loss": 0.2735, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.08592592592592592, | |
| "grad_norm": 1.8093316555023193, | |
| "learning_rate": 9.30930930930931e-06, | |
| "loss": 0.2119, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.08888888888888889, | |
| "grad_norm": 1.6245890855789185, | |
| "learning_rate": 9.27927927927928e-06, | |
| "loss": 0.1766, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09185185185185185, | |
| "grad_norm": 1.4994266033172607, | |
| "learning_rate": 9.24924924924925e-06, | |
| "loss": 0.1381, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.09481481481481481, | |
| "grad_norm": 1.0899797677993774, | |
| "learning_rate": 9.21921921921922e-06, | |
| "loss": 0.0943, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.09777777777777778, | |
| "grad_norm": 1.266209602355957, | |
| "learning_rate": 9.189189189189191e-06, | |
| "loss": 0.1313, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.10074074074074074, | |
| "grad_norm": 1.7627228498458862, | |
| "learning_rate": 9.15915915915916e-06, | |
| "loss": 0.1761, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1037037037037037, | |
| "grad_norm": 1.859174132347107, | |
| "learning_rate": 9.129129129129129e-06, | |
| "loss": 0.2381, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.10666666666666667, | |
| "grad_norm": 1.6145838499069214, | |
| "learning_rate": 9.0990990990991e-06, | |
| "loss": 0.1604, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.10962962962962963, | |
| "grad_norm": 1.773461103439331, | |
| "learning_rate": 9.06906906906907e-06, | |
| "loss": 0.2261, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.11259259259259259, | |
| "grad_norm": 1.7947049140930176, | |
| "learning_rate": 9.03903903903904e-06, | |
| "loss": 0.1922, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.11555555555555555, | |
| "grad_norm": 1.432085394859314, | |
| "learning_rate": 9.00900900900901e-06, | |
| "loss": 0.1281, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.11851851851851852, | |
| "grad_norm": 1.2836321592330933, | |
| "learning_rate": 8.97897897897898e-06, | |
| "loss": 0.1371, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12148148148148148, | |
| "grad_norm": 1.638980746269226, | |
| "learning_rate": 8.94894894894895e-06, | |
| "loss": 0.1948, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.12444444444444444, | |
| "grad_norm": 2.0078577995300293, | |
| "learning_rate": 8.91891891891892e-06, | |
| "loss": 0.2341, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.1274074074074074, | |
| "grad_norm": 1.2297101020812988, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 0.1087, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.13037037037037036, | |
| "grad_norm": 1.225908875465393, | |
| "learning_rate": 8.85885885885886e-06, | |
| "loss": 0.1254, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.13333333333333333, | |
| "grad_norm": 1.675803780555725, | |
| "learning_rate": 8.82882882882883e-06, | |
| "loss": 0.1913, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1362962962962963, | |
| "grad_norm": 1.935347080230713, | |
| "learning_rate": 8.798798798798799e-06, | |
| "loss": 0.2308, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.13925925925925925, | |
| "grad_norm": 2.3072996139526367, | |
| "learning_rate": 8.768768768768769e-06, | |
| "loss": 0.298, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.14222222222222222, | |
| "grad_norm": 2.1588022708892822, | |
| "learning_rate": 8.738738738738739e-06, | |
| "loss": 0.108, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.1451851851851852, | |
| "grad_norm": 1.4197659492492676, | |
| "learning_rate": 8.70870870870871e-06, | |
| "loss": 0.144, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 1.1387606859207153, | |
| "learning_rate": 8.67867867867868e-06, | |
| "loss": 0.1226, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "eval_loss": 0.16039520502090454, | |
| "eval_runtime": 67.9012, | |
| "eval_samples_per_second": 8.836, | |
| "eval_steps_per_second": 4.418, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1511111111111111, | |
| "grad_norm": 1.8014549016952515, | |
| "learning_rate": 8.64864864864865e-06, | |
| "loss": 0.2243, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.15407407407407409, | |
| "grad_norm": 1.5460069179534912, | |
| "learning_rate": 8.618618618618619e-06, | |
| "loss": 0.1907, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.15703703703703703, | |
| "grad_norm": 1.5753509998321533, | |
| "learning_rate": 8.588588588588589e-06, | |
| "loss": 0.2062, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 1.9047173261642456, | |
| "learning_rate": 8.55855855855856e-06, | |
| "loss": 0.2004, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.16296296296296298, | |
| "grad_norm": 1.1410256624221802, | |
| "learning_rate": 8.52852852852853e-06, | |
| "loss": 0.104, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.16592592592592592, | |
| "grad_norm": 1.496903896331787, | |
| "learning_rate": 8.4984984984985e-06, | |
| "loss": 0.163, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1688888888888889, | |
| "grad_norm": 1.4111011028289795, | |
| "learning_rate": 8.46846846846847e-06, | |
| "loss": 0.145, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.17185185185185184, | |
| "grad_norm": 1.1673048734664917, | |
| "learning_rate": 8.438438438438439e-06, | |
| "loss": 0.1036, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1748148148148148, | |
| "grad_norm": 1.5025025606155396, | |
| "learning_rate": 8.408408408408409e-06, | |
| "loss": 0.19, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.17777777777777778, | |
| "grad_norm": 1.3822792768478394, | |
| "learning_rate": 8.378378378378378e-06, | |
| "loss": 0.167, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18074074074074073, | |
| "grad_norm": 1.4234565496444702, | |
| "learning_rate": 8.348348348348348e-06, | |
| "loss": 0.151, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.1837037037037037, | |
| "grad_norm": 1.163219690322876, | |
| "learning_rate": 8.31831831831832e-06, | |
| "loss": 0.1236, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.18666666666666668, | |
| "grad_norm": 1.770808219909668, | |
| "learning_rate": 8.288288288288289e-06, | |
| "loss": 0.2104, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.18962962962962962, | |
| "grad_norm": 1.2196369171142578, | |
| "learning_rate": 8.258258258258259e-06, | |
| "loss": 0.1512, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.1925925925925926, | |
| "grad_norm": 1.5598896741867065, | |
| "learning_rate": 8.228228228228229e-06, | |
| "loss": 0.1736, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.19555555555555557, | |
| "grad_norm": 1.5021666288375854, | |
| "learning_rate": 8.198198198198198e-06, | |
| "loss": 0.1352, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.1985185185185185, | |
| "grad_norm": 1.6956490278244019, | |
| "learning_rate": 8.16816816816817e-06, | |
| "loss": 0.2092, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.20148148148148148, | |
| "grad_norm": 2.1407744884490967, | |
| "learning_rate": 8.13813813813814e-06, | |
| "loss": 0.254, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.20444444444444446, | |
| "grad_norm": 1.805568814277649, | |
| "learning_rate": 8.108108108108109e-06, | |
| "loss": 0.2041, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2074074074074074, | |
| "grad_norm": 1.0428307056427002, | |
| "learning_rate": 8.078078078078079e-06, | |
| "loss": 0.1089, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.21037037037037037, | |
| "grad_norm": 1.7227048873901367, | |
| "learning_rate": 8.048048048048048e-06, | |
| "loss": 0.1856, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.21333333333333335, | |
| "grad_norm": 1.7150681018829346, | |
| "learning_rate": 8.018018018018018e-06, | |
| "loss": 0.187, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2162962962962963, | |
| "grad_norm": 1.3757710456848145, | |
| "learning_rate": 7.987987987987988e-06, | |
| "loss": 0.1286, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.21925925925925926, | |
| "grad_norm": 1.2104195356369019, | |
| "learning_rate": 7.95795795795796e-06, | |
| "loss": 0.1005, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 1.2945895195007324, | |
| "learning_rate": 7.927927927927929e-06, | |
| "loss": 0.1314, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.22518518518518518, | |
| "grad_norm": 1.520178198814392, | |
| "learning_rate": 7.897897897897899e-06, | |
| "loss": 0.1656, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.22814814814814816, | |
| "grad_norm": 1.8112298250198364, | |
| "learning_rate": 7.867867867867868e-06, | |
| "loss": 0.2028, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.2311111111111111, | |
| "grad_norm": 1.4334383010864258, | |
| "learning_rate": 7.837837837837838e-06, | |
| "loss": 0.159, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.23407407407407407, | |
| "grad_norm": 1.2109181880950928, | |
| "learning_rate": 7.807807807807808e-06, | |
| "loss": 0.1223, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.23703703703703705, | |
| "grad_norm": 1.611818552017212, | |
| "learning_rate": 7.77777777777778e-06, | |
| "loss": 0.1578, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.4566220045089722, | |
| "learning_rate": 7.747747747747749e-06, | |
| "loss": 0.1589, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.24296296296296296, | |
| "grad_norm": 1.4583666324615479, | |
| "learning_rate": 7.717717717717719e-06, | |
| "loss": 0.1323, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.24592592592592594, | |
| "grad_norm": 1.1954914331436157, | |
| "learning_rate": 7.687687687687688e-06, | |
| "loss": 0.1102, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.24888888888888888, | |
| "grad_norm": 1.2586238384246826, | |
| "learning_rate": 7.657657657657658e-06, | |
| "loss": 0.1361, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.2518518518518518, | |
| "grad_norm": 1.4241305589675903, | |
| "learning_rate": 7.6276276276276285e-06, | |
| "loss": 0.1716, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2548148148148148, | |
| "grad_norm": 1.2219030857086182, | |
| "learning_rate": 7.597597597597598e-06, | |
| "loss": 0.1533, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.2577777777777778, | |
| "grad_norm": 1.7697409391403198, | |
| "learning_rate": 7.567567567567569e-06, | |
| "loss": 0.2241, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2607407407407407, | |
| "grad_norm": 1.7404961585998535, | |
| "learning_rate": 7.5375375375375385e-06, | |
| "loss": 0.2045, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2637037037037037, | |
| "grad_norm": 1.314818024635315, | |
| "learning_rate": 7.507507507507507e-06, | |
| "loss": 0.1311, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 1.5654951333999634, | |
| "learning_rate": 7.477477477477479e-06, | |
| "loss": 0.1521, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2696296296296296, | |
| "grad_norm": 1.5849696397781372, | |
| "learning_rate": 7.447447447447448e-06, | |
| "loss": 0.2075, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2725925925925926, | |
| "grad_norm": 1.2052536010742188, | |
| "learning_rate": 7.417417417417418e-06, | |
| "loss": 0.1168, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.27555555555555555, | |
| "grad_norm": 1.0663442611694336, | |
| "learning_rate": 7.387387387387388e-06, | |
| "loss": 0.094, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.2785185185185185, | |
| "grad_norm": 1.6939942836761475, | |
| "learning_rate": 7.3573573573573575e-06, | |
| "loss": 0.234, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2814814814814815, | |
| "grad_norm": 1.3187142610549927, | |
| "learning_rate": 7.327327327327328e-06, | |
| "loss": 0.1529, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.28444444444444444, | |
| "grad_norm": 1.5975914001464844, | |
| "learning_rate": 7.297297297297298e-06, | |
| "loss": 0.1827, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.2874074074074074, | |
| "grad_norm": 1.5395817756652832, | |
| "learning_rate": 7.267267267267268e-06, | |
| "loss": 0.199, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.2903703703703704, | |
| "grad_norm": 1.4313832521438599, | |
| "learning_rate": 7.237237237237238e-06, | |
| "loss": 0.1622, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.29333333333333333, | |
| "grad_norm": 1.6650500297546387, | |
| "learning_rate": 7.207207207207208e-06, | |
| "loss": 0.2505, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 1.4698227643966675, | |
| "learning_rate": 7.177177177177178e-06, | |
| "loss": 0.1819, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "eval_loss": 0.14806436002254486, | |
| "eval_runtime": 66.1291, | |
| "eval_samples_per_second": 9.073, | |
| "eval_steps_per_second": 4.537, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2992592592592593, | |
| "grad_norm": 1.3387997150421143, | |
| "learning_rate": 7.147147147147148e-06, | |
| "loss": 0.1158, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.3022222222222222, | |
| "grad_norm": 1.6370638608932495, | |
| "learning_rate": 7.117117117117117e-06, | |
| "loss": 0.174, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.30518518518518517, | |
| "grad_norm": 1.8165297508239746, | |
| "learning_rate": 7.087087087087087e-06, | |
| "loss": 0.2323, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.30814814814814817, | |
| "grad_norm": 1.3221631050109863, | |
| "learning_rate": 7.057057057057057e-06, | |
| "loss": 0.141, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.3111111111111111, | |
| "grad_norm": 1.619100570678711, | |
| "learning_rate": 7.027027027027028e-06, | |
| "loss": 0.2268, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.31407407407407406, | |
| "grad_norm": 1.1240403652191162, | |
| "learning_rate": 6.996996996996997e-06, | |
| "loss": 0.1265, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.31703703703703706, | |
| "grad_norm": 1.4533915519714355, | |
| "learning_rate": 6.966966966966967e-06, | |
| "loss": 0.134, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.1758527755737305, | |
| "learning_rate": 6.936936936936938e-06, | |
| "loss": 0.1331, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.32296296296296295, | |
| "grad_norm": 1.5709505081176758, | |
| "learning_rate": 6.906906906906907e-06, | |
| "loss": 0.1809, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.32592592592592595, | |
| "grad_norm": 1.4079163074493408, | |
| "learning_rate": 6.876876876876878e-06, | |
| "loss": 0.1723, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3288888888888889, | |
| "grad_norm": 1.6966813802719116, | |
| "learning_rate": 6.846846846846848e-06, | |
| "loss": 0.2069, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.33185185185185184, | |
| "grad_norm": 1.2849105596542358, | |
| "learning_rate": 6.816816816816817e-06, | |
| "loss": 0.1498, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.3348148148148148, | |
| "grad_norm": 1.2226041555404663, | |
| "learning_rate": 6.786786786786788e-06, | |
| "loss": 0.1416, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3377777777777778, | |
| "grad_norm": 1.6544311046600342, | |
| "learning_rate": 6.7567567567567575e-06, | |
| "loss": 0.1963, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.34074074074074073, | |
| "grad_norm": 1.3441311120986938, | |
| "learning_rate": 6.726726726726728e-06, | |
| "loss": 0.1557, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3437037037037037, | |
| "grad_norm": 2.0932984352111816, | |
| "learning_rate": 6.696696696696697e-06, | |
| "loss": 0.2517, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.3466666666666667, | |
| "grad_norm": 1.458679437637329, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 0.1437, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3496296296296296, | |
| "grad_norm": 1.7695969343185425, | |
| "learning_rate": 6.636636636636637e-06, | |
| "loss": 0.2091, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.35259259259259257, | |
| "grad_norm": 1.5610897541046143, | |
| "learning_rate": 6.606606606606607e-06, | |
| "loss": 0.2086, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.35555555555555557, | |
| "grad_norm": 1.24899423122406, | |
| "learning_rate": 6.5765765765765775e-06, | |
| "loss": 0.1626, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3585185185185185, | |
| "grad_norm": 1.473816990852356, | |
| "learning_rate": 6.546546546546547e-06, | |
| "loss": 0.1734, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.36148148148148146, | |
| "grad_norm": 1.0760806798934937, | |
| "learning_rate": 6.516516516516517e-06, | |
| "loss": 0.133, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.36444444444444446, | |
| "grad_norm": 1.4529191255569458, | |
| "learning_rate": 6.486486486486487e-06, | |
| "loss": 0.1462, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.3674074074074074, | |
| "grad_norm": 1.8089873790740967, | |
| "learning_rate": 6.456456456456457e-06, | |
| "loss": 0.2484, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 1.4668419361114502, | |
| "learning_rate": 6.426426426426427e-06, | |
| "loss": 0.1521, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.37333333333333335, | |
| "grad_norm": 1.6595661640167236, | |
| "learning_rate": 6.396396396396397e-06, | |
| "loss": 0.2061, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.3762962962962963, | |
| "grad_norm": 1.1693167686462402, | |
| "learning_rate": 6.366366366366366e-06, | |
| "loss": 0.1306, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.37925925925925924, | |
| "grad_norm": 1.0234718322753906, | |
| "learning_rate": 6.336336336336338e-06, | |
| "loss": 0.1059, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.38222222222222224, | |
| "grad_norm": 1.358637809753418, | |
| "learning_rate": 6.3063063063063065e-06, | |
| "loss": 0.1275, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3851851851851852, | |
| "grad_norm": 1.3581657409667969, | |
| "learning_rate": 6.276276276276276e-06, | |
| "loss": 0.1067, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.38814814814814813, | |
| "grad_norm": 1.5347743034362793, | |
| "learning_rate": 6.246246246246247e-06, | |
| "loss": 0.1767, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.39111111111111113, | |
| "grad_norm": 1.2761399745941162, | |
| "learning_rate": 6.2162162162162164e-06, | |
| "loss": 0.1443, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.3940740740740741, | |
| "grad_norm": 1.6570420265197754, | |
| "learning_rate": 6.186186186186187e-06, | |
| "loss": 0.1906, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.397037037037037, | |
| "grad_norm": 1.6703896522521973, | |
| "learning_rate": 6.156156156156157e-06, | |
| "loss": 0.1852, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.3865188360214233, | |
| "learning_rate": 6.126126126126126e-06, | |
| "loss": 0.1546, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.40296296296296297, | |
| "grad_norm": 1.2905173301696777, | |
| "learning_rate": 6.096096096096097e-06, | |
| "loss": 0.1099, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.4059259259259259, | |
| "grad_norm": 1.0323926210403442, | |
| "learning_rate": 6.066066066066067e-06, | |
| "loss": 0.0864, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.4088888888888889, | |
| "grad_norm": 1.4790617227554321, | |
| "learning_rate": 6.036036036036037e-06, | |
| "loss": 0.1433, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.41185185185185186, | |
| "grad_norm": 1.500624418258667, | |
| "learning_rate": 6.006006006006007e-06, | |
| "loss": 0.1541, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.4148148148148148, | |
| "grad_norm": 2.0234739780426025, | |
| "learning_rate": 5.975975975975976e-06, | |
| "loss": 0.2458, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4177777777777778, | |
| "grad_norm": 1.5450382232666016, | |
| "learning_rate": 5.945945945945947e-06, | |
| "loss": 0.1925, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.42074074074074075, | |
| "grad_norm": 1.3305343389511108, | |
| "learning_rate": 5.915915915915916e-06, | |
| "loss": 0.1208, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.4237037037037037, | |
| "grad_norm": 1.5737042427062988, | |
| "learning_rate": 5.885885885885886e-06, | |
| "loss": 0.2002, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.4266666666666667, | |
| "grad_norm": 1.6246366500854492, | |
| "learning_rate": 5.855855855855856e-06, | |
| "loss": 0.1674, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.42962962962962964, | |
| "grad_norm": 1.372275471687317, | |
| "learning_rate": 5.825825825825826e-06, | |
| "loss": 0.1188, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.4325925925925926, | |
| "grad_norm": 0.9744035005569458, | |
| "learning_rate": 5.7957957957957965e-06, | |
| "loss": 0.0802, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.43555555555555553, | |
| "grad_norm": 1.4807896614074707, | |
| "learning_rate": 5.765765765765766e-06, | |
| "loss": 0.1314, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.43851851851851853, | |
| "grad_norm": 1.3947936296463013, | |
| "learning_rate": 5.735735735735736e-06, | |
| "loss": 0.1594, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.4414814814814815, | |
| "grad_norm": 1.5801756381988525, | |
| "learning_rate": 5.7057057057057065e-06, | |
| "loss": 0.1652, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 1.6099817752838135, | |
| "learning_rate": 5.675675675675676e-06, | |
| "loss": 0.2087, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "eval_loss": 0.1412356048822403, | |
| "eval_runtime": 66.142, | |
| "eval_samples_per_second": 9.071, | |
| "eval_steps_per_second": 4.536, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4474074074074074, | |
| "grad_norm": 1.7930363416671753, | |
| "learning_rate": 5.645645645645647e-06, | |
| "loss": 0.2102, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.45037037037037037, | |
| "grad_norm": 1.2251203060150146, | |
| "learning_rate": 5.615615615615616e-06, | |
| "loss": 0.1164, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.4533333333333333, | |
| "grad_norm": 1.5178476572036743, | |
| "learning_rate": 5.585585585585585e-06, | |
| "loss": 0.1493, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.4562962962962963, | |
| "grad_norm": 1.602507472038269, | |
| "learning_rate": 5.555555555555557e-06, | |
| "loss": 0.1624, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.45925925925925926, | |
| "grad_norm": 1.00004243850708, | |
| "learning_rate": 5.5255255255255255e-06, | |
| "loss": 0.0877, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.4622222222222222, | |
| "grad_norm": 0.6796390414237976, | |
| "learning_rate": 5.495495495495496e-06, | |
| "loss": 0.0413, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.4651851851851852, | |
| "grad_norm": 1.7379957437515259, | |
| "learning_rate": 5.465465465465466e-06, | |
| "loss": 0.1784, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.46814814814814815, | |
| "grad_norm": 1.2618839740753174, | |
| "learning_rate": 5.4354354354354355e-06, | |
| "loss": 0.1277, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.4711111111111111, | |
| "grad_norm": 1.5085928440093994, | |
| "learning_rate": 5.405405405405406e-06, | |
| "loss": 0.1855, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.4740740740740741, | |
| "grad_norm": 1.2010939121246338, | |
| "learning_rate": 5.375375375375376e-06, | |
| "loss": 0.1018, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.47703703703703704, | |
| "grad_norm": 1.4355438947677612, | |
| "learning_rate": 5.345345345345346e-06, | |
| "loss": 0.1591, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.1437758207321167, | |
| "learning_rate": 5.315315315315316e-06, | |
| "loss": 0.097, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.482962962962963, | |
| "grad_norm": 1.3701622486114502, | |
| "learning_rate": 5.285285285285286e-06, | |
| "loss": 0.1409, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.48592592592592593, | |
| "grad_norm": 1.248700499534607, | |
| "learning_rate": 5.255255255255256e-06, | |
| "loss": 0.0986, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.4888888888888889, | |
| "grad_norm": 1.459463119506836, | |
| "learning_rate": 5.225225225225226e-06, | |
| "loss": 0.1638, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4918518518518519, | |
| "grad_norm": 1.261518120765686, | |
| "learning_rate": 5.195195195195195e-06, | |
| "loss": 0.1122, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.4948148148148148, | |
| "grad_norm": 1.450064778327942, | |
| "learning_rate": 5.165165165165165e-06, | |
| "loss": 0.1269, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.49777777777777776, | |
| "grad_norm": 1.9161652326583862, | |
| "learning_rate": 5.135135135135135e-06, | |
| "loss": 0.2482, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.5007407407407407, | |
| "grad_norm": 1.3027966022491455, | |
| "learning_rate": 5.105105105105106e-06, | |
| "loss": 0.1393, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.5037037037037037, | |
| "grad_norm": 1.597838044166565, | |
| "learning_rate": 5.075075075075075e-06, | |
| "loss": 0.1957, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5066666666666667, | |
| "grad_norm": 1.4411420822143555, | |
| "learning_rate": 5.045045045045045e-06, | |
| "loss": 0.1754, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.5096296296296297, | |
| "grad_norm": 1.7300653457641602, | |
| "learning_rate": 5.0150150150150156e-06, | |
| "loss": 0.1928, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.5125925925925926, | |
| "grad_norm": 1.5549505949020386, | |
| "learning_rate": 4.984984984984985e-06, | |
| "loss": 0.1829, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.5155555555555555, | |
| "grad_norm": 1.5301578044891357, | |
| "learning_rate": 4.954954954954955e-06, | |
| "loss": 0.164, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.5185185185185185, | |
| "grad_norm": 1.4430240392684937, | |
| "learning_rate": 4.9249249249249255e-06, | |
| "loss": 0.1562, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5214814814814814, | |
| "grad_norm": 1.5305695533752441, | |
| "learning_rate": 4.894894894894895e-06, | |
| "loss": 0.1797, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.5244444444444445, | |
| "grad_norm": 1.1799602508544922, | |
| "learning_rate": 4.864864864864866e-06, | |
| "loss": 0.1043, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.5274074074074074, | |
| "grad_norm": 1.527823805809021, | |
| "learning_rate": 4.8348348348348355e-06, | |
| "loss": 0.1527, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.5303703703703704, | |
| "grad_norm": 1.513073205947876, | |
| "learning_rate": 4.804804804804805e-06, | |
| "loss": 0.1704, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 1.1295517683029175, | |
| "learning_rate": 4.774774774774775e-06, | |
| "loss": 0.1277, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5362962962962963, | |
| "grad_norm": 1.1660453081130981, | |
| "learning_rate": 4.7447447447447454e-06, | |
| "loss": 0.1311, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.5392592592592592, | |
| "grad_norm": 1.4742252826690674, | |
| "learning_rate": 4.714714714714715e-06, | |
| "loss": 0.1371, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.5422222222222223, | |
| "grad_norm": 1.2694472074508667, | |
| "learning_rate": 4.684684684684685e-06, | |
| "loss": 0.105, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.5451851851851852, | |
| "grad_norm": 1.5899627208709717, | |
| "learning_rate": 4.654654654654655e-06, | |
| "loss": 0.195, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.5481481481481482, | |
| "grad_norm": 1.1580640077590942, | |
| "learning_rate": 4.624624624624625e-06, | |
| "loss": 0.0958, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.5511111111111111, | |
| "grad_norm": 1.6027284860610962, | |
| "learning_rate": 4.594594594594596e-06, | |
| "loss": 0.1785, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.554074074074074, | |
| "grad_norm": 1.580238938331604, | |
| "learning_rate": 4.5645645645645645e-06, | |
| "loss": 0.2108, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.557037037037037, | |
| "grad_norm": 1.200641393661499, | |
| "learning_rate": 4.534534534534535e-06, | |
| "loss": 0.1439, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.4592289924621582, | |
| "learning_rate": 4.504504504504505e-06, | |
| "loss": 0.1786, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.562962962962963, | |
| "grad_norm": 1.479386568069458, | |
| "learning_rate": 4.474474474474475e-06, | |
| "loss": 0.1596, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5659259259259259, | |
| "grad_norm": 1.291646957397461, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 0.1486, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.5688888888888889, | |
| "grad_norm": 1.1819597482681274, | |
| "learning_rate": 4.414414414414415e-06, | |
| "loss": 0.1027, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.5718518518518518, | |
| "grad_norm": 1.493479609489441, | |
| "learning_rate": 4.384384384384384e-06, | |
| "loss": 0.1924, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.5748148148148148, | |
| "grad_norm": 1.3410001993179321, | |
| "learning_rate": 4.354354354354355e-06, | |
| "loss": 0.1797, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.5777777777777777, | |
| "grad_norm": 1.2547590732574463, | |
| "learning_rate": 4.324324324324325e-06, | |
| "loss": 0.1236, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.5807407407407408, | |
| "grad_norm": 1.1719931364059448, | |
| "learning_rate": 4.294294294294294e-06, | |
| "loss": 0.1288, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.5837037037037037, | |
| "grad_norm": 1.2825536727905273, | |
| "learning_rate": 4.264264264264265e-06, | |
| "loss": 0.1492, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.5866666666666667, | |
| "grad_norm": 1.4123318195343018, | |
| "learning_rate": 4.234234234234235e-06, | |
| "loss": 0.1563, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.5896296296296296, | |
| "grad_norm": 1.3128525018692017, | |
| "learning_rate": 4.204204204204204e-06, | |
| "loss": 0.1286, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 1.208724856376648, | |
| "learning_rate": 4.174174174174174e-06, | |
| "loss": 0.1166, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "eval_loss": 0.13714094460010529, | |
| "eval_runtime": 66.3114, | |
| "eval_samples_per_second": 9.048, | |
| "eval_steps_per_second": 4.524, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.5955555555555555, | |
| "grad_norm": 1.5521754026412964, | |
| "learning_rate": 4.1441441441441446e-06, | |
| "loss": 0.1886, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.5985185185185186, | |
| "grad_norm": 1.326633334159851, | |
| "learning_rate": 4.114114114114114e-06, | |
| "loss": 0.1456, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.6014814814814815, | |
| "grad_norm": 1.1393183469772339, | |
| "learning_rate": 4.084084084084085e-06, | |
| "loss": 0.0998, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.6044444444444445, | |
| "grad_norm": 1.8083674907684326, | |
| "learning_rate": 4.0540540540540545e-06, | |
| "loss": 0.2146, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.6074074074074074, | |
| "grad_norm": 1.540759801864624, | |
| "learning_rate": 4.024024024024024e-06, | |
| "loss": 0.1637, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.6103703703703703, | |
| "grad_norm": 1.0481210947036743, | |
| "learning_rate": 3.993993993993994e-06, | |
| "loss": 0.1161, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.6133333333333333, | |
| "grad_norm": 1.0134563446044922, | |
| "learning_rate": 3.9639639639639645e-06, | |
| "loss": 0.0786, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.6162962962962963, | |
| "grad_norm": 1.0966730117797852, | |
| "learning_rate": 3.933933933933934e-06, | |
| "loss": 0.107, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.6192592592592593, | |
| "grad_norm": 1.2660399675369263, | |
| "learning_rate": 3.903903903903904e-06, | |
| "loss": 0.1527, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.6222222222222222, | |
| "grad_norm": 1.3738937377929688, | |
| "learning_rate": 3.8738738738738744e-06, | |
| "loss": 0.1649, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.6251851851851852, | |
| "grad_norm": 1.3413881063461304, | |
| "learning_rate": 3.843843843843844e-06, | |
| "loss": 0.1619, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.6281481481481481, | |
| "grad_norm": 1.2533951997756958, | |
| "learning_rate": 3.8138138138138143e-06, | |
| "loss": 0.1421, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.6311111111111111, | |
| "grad_norm": 1.5089728832244873, | |
| "learning_rate": 3.7837837837837844e-06, | |
| "loss": 0.1616, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.6340740740740741, | |
| "grad_norm": 1.2814654111862183, | |
| "learning_rate": 3.7537537537537537e-06, | |
| "loss": 0.133, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.6370370370370371, | |
| "grad_norm": 1.2312473058700562, | |
| "learning_rate": 3.723723723723724e-06, | |
| "loss": 0.1111, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.2571808099746704, | |
| "learning_rate": 3.693693693693694e-06, | |
| "loss": 0.1357, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.642962962962963, | |
| "grad_norm": 1.2212458848953247, | |
| "learning_rate": 3.663663663663664e-06, | |
| "loss": 0.1339, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.6459259259259259, | |
| "grad_norm": 1.1856861114501953, | |
| "learning_rate": 3.633633633633634e-06, | |
| "loss": 0.1009, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.6488888888888888, | |
| "grad_norm": 1.6730061769485474, | |
| "learning_rate": 3.603603603603604e-06, | |
| "loss": 0.189, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.6518518518518519, | |
| "grad_norm": 1.2098989486694336, | |
| "learning_rate": 3.573573573573574e-06, | |
| "loss": 0.1263, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6548148148148148, | |
| "grad_norm": 1.0369914770126343, | |
| "learning_rate": 3.5435435435435437e-06, | |
| "loss": 0.1339, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.6577777777777778, | |
| "grad_norm": 1.2498979568481445, | |
| "learning_rate": 3.513513513513514e-06, | |
| "loss": 0.1282, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.6607407407407407, | |
| "grad_norm": 1.2711259126663208, | |
| "learning_rate": 3.4834834834834835e-06, | |
| "loss": 0.1197, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.6637037037037037, | |
| "grad_norm": 1.1954965591430664, | |
| "learning_rate": 3.4534534534534537e-06, | |
| "loss": 0.0916, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 1.2048304080963135, | |
| "learning_rate": 3.423423423423424e-06, | |
| "loss": 0.1027, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.6696296296296296, | |
| "grad_norm": 1.491741418838501, | |
| "learning_rate": 3.393393393393394e-06, | |
| "loss": 0.1212, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.6725925925925926, | |
| "grad_norm": 1.684773564338684, | |
| "learning_rate": 3.363363363363364e-06, | |
| "loss": 0.1782, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.6755555555555556, | |
| "grad_norm": 1.0407675504684448, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.0889, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.6785185185185185, | |
| "grad_norm": 1.0543864965438843, | |
| "learning_rate": 3.3033033033033035e-06, | |
| "loss": 0.0861, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.6814814814814815, | |
| "grad_norm": 1.0320396423339844, | |
| "learning_rate": 3.2732732732732736e-06, | |
| "loss": 0.1075, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.6844444444444444, | |
| "grad_norm": 1.6749848127365112, | |
| "learning_rate": 3.2432432432432437e-06, | |
| "loss": 0.1942, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.6874074074074074, | |
| "grad_norm": 1.353992223739624, | |
| "learning_rate": 3.2132132132132134e-06, | |
| "loss": 0.1426, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.6903703703703704, | |
| "grad_norm": 1.274588704109192, | |
| "learning_rate": 3.183183183183183e-06, | |
| "loss": 0.1146, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.6933333333333334, | |
| "grad_norm": 1.1322110891342163, | |
| "learning_rate": 3.1531531531531532e-06, | |
| "loss": 0.0973, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.6962962962962963, | |
| "grad_norm": 1.6855742931365967, | |
| "learning_rate": 3.1231231231231234e-06, | |
| "loss": 0.1314, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.6992592592592592, | |
| "grad_norm": 1.6252344846725464, | |
| "learning_rate": 3.0930930930930935e-06, | |
| "loss": 0.1636, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.7022222222222222, | |
| "grad_norm": 1.4905320405960083, | |
| "learning_rate": 3.063063063063063e-06, | |
| "loss": 0.1532, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.7051851851851851, | |
| "grad_norm": 1.5329716205596924, | |
| "learning_rate": 3.0330330330330333e-06, | |
| "loss": 0.1424, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.7081481481481482, | |
| "grad_norm": 1.178130030632019, | |
| "learning_rate": 3.0030030030030034e-06, | |
| "loss": 0.1055, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.7111111111111111, | |
| "grad_norm": 1.4924639463424683, | |
| "learning_rate": 2.9729729729729736e-06, | |
| "loss": 0.1694, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7140740740740741, | |
| "grad_norm": 1.4013866186141968, | |
| "learning_rate": 2.942942942942943e-06, | |
| "loss": 0.1493, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.717037037037037, | |
| "grad_norm": 1.3777011632919312, | |
| "learning_rate": 2.912912912912913e-06, | |
| "loss": 0.1269, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.5744843482971191, | |
| "learning_rate": 2.882882882882883e-06, | |
| "loss": 0.1912, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.7229629629629629, | |
| "grad_norm": 1.2442989349365234, | |
| "learning_rate": 2.8528528528528532e-06, | |
| "loss": 0.1117, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.725925925925926, | |
| "grad_norm": 1.0887988805770874, | |
| "learning_rate": 2.8228228228228234e-06, | |
| "loss": 0.0852, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.7288888888888889, | |
| "grad_norm": 1.5981248617172241, | |
| "learning_rate": 2.7927927927927926e-06, | |
| "loss": 0.1805, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.7318518518518519, | |
| "grad_norm": 1.2195144891738892, | |
| "learning_rate": 2.7627627627627628e-06, | |
| "loss": 0.0967, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.7348148148148148, | |
| "grad_norm": 1.2072299718856812, | |
| "learning_rate": 2.732732732732733e-06, | |
| "loss": 0.1039, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.7377777777777778, | |
| "grad_norm": 1.4150304794311523, | |
| "learning_rate": 2.702702702702703e-06, | |
| "loss": 0.1638, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 1.5039236545562744, | |
| "learning_rate": 2.672672672672673e-06, | |
| "loss": 0.1587, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "eval_loss": 0.13532617688179016, | |
| "eval_runtime": 68.896, | |
| "eval_samples_per_second": 8.709, | |
| "eval_steps_per_second": 4.354, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.7437037037037038, | |
| "grad_norm": 1.3724570274353027, | |
| "learning_rate": 2.642642642642643e-06, | |
| "loss": 0.1404, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.7466666666666667, | |
| "grad_norm": 1.2078779935836792, | |
| "learning_rate": 2.612612612612613e-06, | |
| "loss": 0.1174, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.7496296296296296, | |
| "grad_norm": 1.530771255493164, | |
| "learning_rate": 2.5825825825825827e-06, | |
| "loss": 0.1573, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.7525925925925926, | |
| "grad_norm": 1.4752556085586548, | |
| "learning_rate": 2.552552552552553e-06, | |
| "loss": 0.1272, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.7555555555555555, | |
| "grad_norm": 1.2959232330322266, | |
| "learning_rate": 2.5225225225225225e-06, | |
| "loss": 0.1409, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.7585185185185185, | |
| "grad_norm": 1.4460132122039795, | |
| "learning_rate": 2.4924924924924926e-06, | |
| "loss": 0.1461, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.7614814814814815, | |
| "grad_norm": 1.3305751085281372, | |
| "learning_rate": 2.4624624624624628e-06, | |
| "loss": 0.1728, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.7644444444444445, | |
| "grad_norm": 1.3664606809616089, | |
| "learning_rate": 2.432432432432433e-06, | |
| "loss": 0.1611, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.7674074074074074, | |
| "grad_norm": 1.6641045808792114, | |
| "learning_rate": 2.4024024024024026e-06, | |
| "loss": 0.162, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.7703703703703704, | |
| "grad_norm": 1.3580279350280762, | |
| "learning_rate": 2.3723723723723727e-06, | |
| "loss": 0.1396, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.7733333333333333, | |
| "grad_norm": 1.7089464664459229, | |
| "learning_rate": 2.3423423423423424e-06, | |
| "loss": 0.1882, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.7762962962962963, | |
| "grad_norm": 1.750114917755127, | |
| "learning_rate": 2.3123123123123125e-06, | |
| "loss": 0.1814, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.7792592592592592, | |
| "grad_norm": 1.7936381101608276, | |
| "learning_rate": 2.2822822822822822e-06, | |
| "loss": 0.2087, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.7822222222222223, | |
| "grad_norm": 1.1869782209396362, | |
| "learning_rate": 2.2522522522522524e-06, | |
| "loss": 0.1005, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.7851851851851852, | |
| "grad_norm": 1.225360631942749, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 0.1002, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.7881481481481482, | |
| "grad_norm": 1.0915745496749878, | |
| "learning_rate": 2.192192192192192e-06, | |
| "loss": 0.1074, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.7911111111111111, | |
| "grad_norm": 1.5699282884597778, | |
| "learning_rate": 2.1621621621621623e-06, | |
| "loss": 0.1746, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.794074074074074, | |
| "grad_norm": 1.3691189289093018, | |
| "learning_rate": 2.1321321321321325e-06, | |
| "loss": 0.1207, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.797037037037037, | |
| "grad_norm": 1.663030743598938, | |
| "learning_rate": 2.102102102102102e-06, | |
| "loss": 0.1728, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.888838291168213, | |
| "learning_rate": 2.0720720720720723e-06, | |
| "loss": 0.2271, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.802962962962963, | |
| "grad_norm": 1.2269537448883057, | |
| "learning_rate": 2.0420420420420424e-06, | |
| "loss": 0.1219, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.8059259259259259, | |
| "grad_norm": 1.6327567100524902, | |
| "learning_rate": 2.012012012012012e-06, | |
| "loss": 0.1634, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.8088888888888889, | |
| "grad_norm": 1.7219854593276978, | |
| "learning_rate": 1.9819819819819822e-06, | |
| "loss": 0.2092, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.8118518518518518, | |
| "grad_norm": 0.8737500309944153, | |
| "learning_rate": 1.951951951951952e-06, | |
| "loss": 0.0768, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.8148148148148148, | |
| "grad_norm": 1.4628630876541138, | |
| "learning_rate": 1.921921921921922e-06, | |
| "loss": 0.1759, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.8177777777777778, | |
| "grad_norm": 1.7518917322158813, | |
| "learning_rate": 1.8918918918918922e-06, | |
| "loss": 0.1988, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.8207407407407408, | |
| "grad_norm": 1.438296914100647, | |
| "learning_rate": 1.861861861861862e-06, | |
| "loss": 0.1469, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.8237037037037037, | |
| "grad_norm": 1.0211504697799683, | |
| "learning_rate": 1.831831831831832e-06, | |
| "loss": 0.0815, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.8266666666666667, | |
| "grad_norm": 1.3015204668045044, | |
| "learning_rate": 1.801801801801802e-06, | |
| "loss": 0.1219, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.8296296296296296, | |
| "grad_norm": 1.3698829412460327, | |
| "learning_rate": 1.7717717717717719e-06, | |
| "loss": 0.1649, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.8325925925925926, | |
| "grad_norm": 0.9183998703956604, | |
| "learning_rate": 1.7417417417417418e-06, | |
| "loss": 0.0829, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.8355555555555556, | |
| "grad_norm": 1.6043810844421387, | |
| "learning_rate": 1.711711711711712e-06, | |
| "loss": 0.1731, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.8385185185185186, | |
| "grad_norm": 1.121024489402771, | |
| "learning_rate": 1.681681681681682e-06, | |
| "loss": 0.111, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.8414814814814815, | |
| "grad_norm": 1.400852918624878, | |
| "learning_rate": 1.6516516516516517e-06, | |
| "loss": 0.1389, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.8444444444444444, | |
| "grad_norm": 1.445923089981079, | |
| "learning_rate": 1.6216216216216219e-06, | |
| "loss": 0.1113, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.8474074074074074, | |
| "grad_norm": 1.0851728916168213, | |
| "learning_rate": 1.5915915915915916e-06, | |
| "loss": 0.0994, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.8503703703703703, | |
| "grad_norm": 1.134529709815979, | |
| "learning_rate": 1.5615615615615617e-06, | |
| "loss": 0.0964, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.8533333333333334, | |
| "grad_norm": 1.1203621625900269, | |
| "learning_rate": 1.5315315315315316e-06, | |
| "loss": 0.0999, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.8562962962962963, | |
| "grad_norm": 1.2996597290039062, | |
| "learning_rate": 1.5015015015015017e-06, | |
| "loss": 0.1421, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.8592592592592593, | |
| "grad_norm": 0.8123877048492432, | |
| "learning_rate": 1.4714714714714714e-06, | |
| "loss": 0.061, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.8622222222222222, | |
| "grad_norm": 1.1721727848052979, | |
| "learning_rate": 1.4414414414414416e-06, | |
| "loss": 0.1086, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.8651851851851852, | |
| "grad_norm": 0.9538038969039917, | |
| "learning_rate": 1.4114114114114117e-06, | |
| "loss": 0.0934, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.8681481481481481, | |
| "grad_norm": 1.2165027856826782, | |
| "learning_rate": 1.3813813813813814e-06, | |
| "loss": 0.0913, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.8711111111111111, | |
| "grad_norm": 1.580078363418579, | |
| "learning_rate": 1.3513513513513515e-06, | |
| "loss": 0.195, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.8740740740740741, | |
| "grad_norm": 0.9435044527053833, | |
| "learning_rate": 1.3213213213213214e-06, | |
| "loss": 0.0918, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.8770370370370371, | |
| "grad_norm": 1.6144812107086182, | |
| "learning_rate": 1.2912912912912913e-06, | |
| "loss": 0.1416, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.0741127729415894, | |
| "learning_rate": 1.2612612612612613e-06, | |
| "loss": 0.091, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.882962962962963, | |
| "grad_norm": 1.3908398151397705, | |
| "learning_rate": 1.2312312312312314e-06, | |
| "loss": 0.158, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.8859259259259259, | |
| "grad_norm": 1.2174276113510132, | |
| "learning_rate": 1.2012012012012013e-06, | |
| "loss": 0.1155, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 1.1871130466461182, | |
| "learning_rate": 1.1711711711711712e-06, | |
| "loss": 0.1013, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "eval_loss": 0.13314028084278107, | |
| "eval_runtime": 66.2791, | |
| "eval_samples_per_second": 9.053, | |
| "eval_steps_per_second": 4.526, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.8918518518518519, | |
| "grad_norm": 1.4601322412490845, | |
| "learning_rate": 1.1411411411411411e-06, | |
| "loss": 0.1597, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.8948148148148148, | |
| "grad_norm": 1.5043681859970093, | |
| "learning_rate": 1.111111111111111e-06, | |
| "loss": 0.1376, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.8977777777777778, | |
| "grad_norm": 1.137039303779602, | |
| "learning_rate": 1.0810810810810812e-06, | |
| "loss": 0.0906, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.9007407407407407, | |
| "grad_norm": 1.4676636457443237, | |
| "learning_rate": 1.051051051051051e-06, | |
| "loss": 0.1826, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.9037037037037037, | |
| "grad_norm": 1.2159793376922607, | |
| "learning_rate": 1.0210210210210212e-06, | |
| "loss": 0.1046, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.9066666666666666, | |
| "grad_norm": 1.645277738571167, | |
| "learning_rate": 9.909909909909911e-07, | |
| "loss": 0.1192, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.9096296296296297, | |
| "grad_norm": 0.9919736981391907, | |
| "learning_rate": 9.60960960960961e-07, | |
| "loss": 0.0926, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.9125925925925926, | |
| "grad_norm": 1.4891765117645264, | |
| "learning_rate": 9.30930930930931e-07, | |
| "loss": 0.181, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.9155555555555556, | |
| "grad_norm": 1.7111001014709473, | |
| "learning_rate": 9.00900900900901e-07, | |
| "loss": 0.1941, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.9185185185185185, | |
| "grad_norm": 1.2775518894195557, | |
| "learning_rate": 8.708708708708709e-07, | |
| "loss": 0.1324, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.9214814814814815, | |
| "grad_norm": 0.8898062705993652, | |
| "learning_rate": 8.40840840840841e-07, | |
| "loss": 0.0764, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.9244444444444444, | |
| "grad_norm": 1.087368130683899, | |
| "learning_rate": 8.108108108108109e-07, | |
| "loss": 0.0962, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.9274074074074075, | |
| "grad_norm": 1.4868751764297485, | |
| "learning_rate": 7.807807807807808e-07, | |
| "loss": 0.1662, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.9303703703703704, | |
| "grad_norm": 1.1045292615890503, | |
| "learning_rate": 7.507507507507509e-07, | |
| "loss": 0.0808, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.9333333333333333, | |
| "grad_norm": 1.3961302042007446, | |
| "learning_rate": 7.207207207207208e-07, | |
| "loss": 0.1349, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.9362962962962963, | |
| "grad_norm": 1.4023386240005493, | |
| "learning_rate": 6.906906906906907e-07, | |
| "loss": 0.1659, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.9392592592592592, | |
| "grad_norm": 1.29899263381958, | |
| "learning_rate": 6.606606606606607e-07, | |
| "loss": 0.1255, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.9422222222222222, | |
| "grad_norm": 1.4591727256774902, | |
| "learning_rate": 6.306306306306306e-07, | |
| "loss": 0.1239, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.9451851851851852, | |
| "grad_norm": 1.175674557685852, | |
| "learning_rate": 6.006006006006006e-07, | |
| "loss": 0.1327, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.9481481481481482, | |
| "grad_norm": 1.2185375690460205, | |
| "learning_rate": 5.705705705705706e-07, | |
| "loss": 0.115, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.9511111111111111, | |
| "grad_norm": 1.2898285388946533, | |
| "learning_rate": 5.405405405405406e-07, | |
| "loss": 0.1243, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.9540740740740741, | |
| "grad_norm": 1.305128574371338, | |
| "learning_rate": 5.105105105105106e-07, | |
| "loss": 0.1354, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.957037037037037, | |
| "grad_norm": 1.1169562339782715, | |
| "learning_rate": 4.804804804804805e-07, | |
| "loss": 0.1196, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.1991010904312134, | |
| "learning_rate": 4.504504504504505e-07, | |
| "loss": 0.0881, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 0.9629629629629629, | |
| "grad_norm": 0.971295952796936, | |
| "learning_rate": 4.204204204204205e-07, | |
| "loss": 0.0878, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.965925925925926, | |
| "grad_norm": 1.4201080799102783, | |
| "learning_rate": 3.903903903903904e-07, | |
| "loss": 0.1603, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 0.9688888888888889, | |
| "grad_norm": 0.9753766655921936, | |
| "learning_rate": 3.603603603603604e-07, | |
| "loss": 0.0848, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 0.9718518518518519, | |
| "grad_norm": 1.396687626838684, | |
| "learning_rate": 3.3033033033033036e-07, | |
| "loss": 0.1522, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 0.9748148148148148, | |
| "grad_norm": 1.3470547199249268, | |
| "learning_rate": 3.003003003003003e-07, | |
| "loss": 0.1384, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 0.9777777777777777, | |
| "grad_norm": 1.1435166597366333, | |
| "learning_rate": 2.702702702702703e-07, | |
| "loss": 0.0816, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.9807407407407407, | |
| "grad_norm": 1.6342765092849731, | |
| "learning_rate": 2.4024024024024026e-07, | |
| "loss": 0.183, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 0.9837037037037037, | |
| "grad_norm": 1.1529721021652222, | |
| "learning_rate": 2.1021021021021025e-07, | |
| "loss": 0.1364, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 0.9866666666666667, | |
| "grad_norm": 1.2728444337844849, | |
| "learning_rate": 1.801801801801802e-07, | |
| "loss": 0.111, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 0.9896296296296296, | |
| "grad_norm": 1.048641324043274, | |
| "learning_rate": 1.5015015015015016e-07, | |
| "loss": 0.0956, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 0.9925925925925926, | |
| "grad_norm": 1.4973434209823608, | |
| "learning_rate": 1.2012012012012013e-07, | |
| "loss": 0.1594, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.9955555555555555, | |
| "grad_norm": 1.555802822113037, | |
| "learning_rate": 9.00900900900901e-08, | |
| "loss": 0.1699, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 0.9985185185185185, | |
| "grad_norm": 1.3605561256408691, | |
| "learning_rate": 6.006006006006006e-08, | |
| "loss": 0.1387, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.945210337638855, | |
| "learning_rate": 3.003003003003003e-08, | |
| "loss": 0.1497, | |
| "step": 338 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 338, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.770443693996503e+17, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |