| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 690, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004347826086956522, | |
| "learning_rate": 0, | |
| "loss": 2.1156, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.008695652173913044, | |
| "learning_rate": 0, | |
| "loss": 2.1067, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.013043478260869565, | |
| "learning_rate": 0, | |
| "loss": 2.1898, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.017391304347826087, | |
| "grad_norm": 2.296112298965454, | |
| "learning_rate": 0.0, | |
| "loss": 2.1665, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.021739130434782608, | |
| "grad_norm": 2.296112298965454, | |
| "learning_rate": 0.0, | |
| "loss": 2.1879, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02608695652173913, | |
| "grad_norm": 2.296112298965454, | |
| "learning_rate": 0.0, | |
| "loss": 2.1444, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.030434782608695653, | |
| "grad_norm": 2.296112298965454, | |
| "learning_rate": 0.0, | |
| "loss": 2.152, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.034782608695652174, | |
| "grad_norm": 2.2356226444244385, | |
| "learning_rate": 3.2741131089043125e-06, | |
| "loss": 2.1847, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0391304347826087, | |
| "grad_norm": 2.2356226444244385, | |
| "learning_rate": 3.2741131089043125e-06, | |
| "loss": 2.169, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.043478260869565216, | |
| "grad_norm": 2.2356226444244385, | |
| "learning_rate": 3.2741131089043125e-06, | |
| "loss": 2.2589, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04782608695652174, | |
| "grad_norm": 2.2356226444244385, | |
| "learning_rate": 3.2741131089043125e-06, | |
| "loss": 2.1925, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.05217391304347826, | |
| "grad_norm": 2.2073488235473633, | |
| "learning_rate": 5.189346500732899e-06, | |
| "loss": 2.2475, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.05652173913043478, | |
| "grad_norm": 2.2073488235473633, | |
| "learning_rate": 5.189346500732899e-06, | |
| "loss": 2.1839, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.06086956521739131, | |
| "grad_norm": 2.2073488235473633, | |
| "learning_rate": 5.189346500732899e-06, | |
| "loss": 2.2041, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.06521739130434782, | |
| "grad_norm": 2.2073488235473633, | |
| "learning_rate": 5.189346500732899e-06, | |
| "loss": 2.1681, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.06956521739130435, | |
| "grad_norm": 2.202537775039673, | |
| "learning_rate": 6.548226217808625e-06, | |
| "loss": 2.1546, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.07391304347826087, | |
| "grad_norm": 2.202537775039673, | |
| "learning_rate": 6.548226217808625e-06, | |
| "loss": 2.2019, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.0782608695652174, | |
| "grad_norm": 2.202537775039673, | |
| "learning_rate": 6.548226217808625e-06, | |
| "loss": 2.1666, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.08260869565217391, | |
| "grad_norm": 2.202537775039673, | |
| "learning_rate": 6.548226217808625e-06, | |
| "loss": 2.1445, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.08695652173913043, | |
| "grad_norm": 2.1643476486206055, | |
| "learning_rate": 7.60225521340393e-06, | |
| "loss": 2.0461, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.09130434782608696, | |
| "grad_norm": 2.1643476486206055, | |
| "learning_rate": 7.60225521340393e-06, | |
| "loss": 2.139, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.09565217391304348, | |
| "grad_norm": 2.1643476486206055, | |
| "learning_rate": 7.60225521340393e-06, | |
| "loss": 2.253, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 2.1643476486206055, | |
| "learning_rate": 7.60225521340393e-06, | |
| "loss": 2.1876, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.10434782608695652, | |
| "grad_norm": 2.2119083404541016, | |
| "learning_rate": 8.463459609637211e-06, | |
| "loss": 2.2883, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.10869565217391304, | |
| "grad_norm": 2.2119083404541016, | |
| "learning_rate": 8.463459609637211e-06, | |
| "loss": 2.2083, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.11304347826086956, | |
| "grad_norm": 2.2119083404541016, | |
| "learning_rate": 8.463459609637211e-06, | |
| "loss": 2.1793, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.11739130434782609, | |
| "grad_norm": 2.2119083404541016, | |
| "learning_rate": 8.463459609637211e-06, | |
| "loss": 2.1871, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.12173913043478261, | |
| "grad_norm": 2.064499855041504, | |
| "learning_rate": 9.191597551655847e-06, | |
| "loss": 2.24, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.12608695652173912, | |
| "grad_norm": 2.064499855041504, | |
| "learning_rate": 9.191597551655847e-06, | |
| "loss": 2.1223, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.13043478260869565, | |
| "grad_norm": 2.064499855041504, | |
| "learning_rate": 9.191597551655847e-06, | |
| "loss": 2.0987, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.13478260869565217, | |
| "grad_norm": 2.064499855041504, | |
| "learning_rate": 9.191597551655847e-06, | |
| "loss": 1.9908, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.1391304347826087, | |
| "grad_norm": 1.9148058891296387, | |
| "learning_rate": 9.822339326712938e-06, | |
| "loss": 2.1102, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.14347826086956522, | |
| "grad_norm": 1.9148058891296387, | |
| "learning_rate": 9.822339326712938e-06, | |
| "loss": 1.9757, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.14782608695652175, | |
| "grad_norm": 1.9148058891296387, | |
| "learning_rate": 9.822339326712938e-06, | |
| "loss": 2.0812, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.15217391304347827, | |
| "grad_norm": 1.9148058891296387, | |
| "learning_rate": 9.822339326712938e-06, | |
| "loss": 2.0167, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.1565217391304348, | |
| "grad_norm": 1.9139795303344727, | |
| "learning_rate": 1.0378693001465798e-05, | |
| "loss": 2.0972, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.1608695652173913, | |
| "grad_norm": 1.9139795303344727, | |
| "learning_rate": 1.0378693001465798e-05, | |
| "loss": 2.0143, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.16521739130434782, | |
| "grad_norm": 1.9139795303344727, | |
| "learning_rate": 1.0378693001465798e-05, | |
| "loss": 2.0257, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.16956521739130434, | |
| "grad_norm": 1.9139795303344727, | |
| "learning_rate": 1.0378693001465798e-05, | |
| "loss": 2.1066, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.17391304347826086, | |
| "grad_norm": 1.8336902856826782, | |
| "learning_rate": 1.0876368322308244e-05, | |
| "loss": 1.9792, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1782608695652174, | |
| "grad_norm": 1.8336902856826782, | |
| "learning_rate": 1.0876368322308244e-05, | |
| "loss": 1.9625, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1826086956521739, | |
| "grad_norm": 1.8336902856826782, | |
| "learning_rate": 1.0876368322308244e-05, | |
| "loss": 2.0084, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.18695652173913044, | |
| "grad_norm": 1.8336902856826782, | |
| "learning_rate": 1.0876368322308244e-05, | |
| "loss": 2.0717, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.19130434782608696, | |
| "grad_norm": 1.9207242727279663, | |
| "learning_rate": 1.1326570411938442e-05, | |
| "loss": 2.0022, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.1956521739130435, | |
| "grad_norm": 1.9207242727279663, | |
| "learning_rate": 1.1326570411938442e-05, | |
| "loss": 1.9515, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 1.9207242727279663, | |
| "learning_rate": 1.1326570411938442e-05, | |
| "loss": 1.8031, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.20434782608695654, | |
| "grad_norm": 1.9207242727279663, | |
| "learning_rate": 1.1326570411938442e-05, | |
| "loss": 1.8205, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.20869565217391303, | |
| "grad_norm": 1.2533749341964722, | |
| "learning_rate": 1.1737572718541526e-05, | |
| "loss": 1.9172, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.21304347826086956, | |
| "grad_norm": 1.2533749341964722, | |
| "learning_rate": 1.1737572718541526e-05, | |
| "loss": 1.8282, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.21739130434782608, | |
| "grad_norm": 1.2533749341964722, | |
| "learning_rate": 1.1737572718541526e-05, | |
| "loss": 1.8121, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.2217391304347826, | |
| "grad_norm": 1.2533749341964722, | |
| "learning_rate": 1.1737572718541526e-05, | |
| "loss": 1.9039, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.22608695652173913, | |
| "grad_norm": 1.106286644935608, | |
| "learning_rate": 1.2115658189875932e-05, | |
| "loss": 1.8711, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.23043478260869565, | |
| "grad_norm": 1.106286644935608, | |
| "learning_rate": 1.2115658189875932e-05, | |
| "loss": 1.9492, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.23478260869565218, | |
| "grad_norm": 1.106286644935608, | |
| "learning_rate": 1.2115658189875932e-05, | |
| "loss": 1.8475, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.2391304347826087, | |
| "grad_norm": 1.106286644935608, | |
| "learning_rate": 1.2115658189875932e-05, | |
| "loss": 1.8875, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.24347826086956523, | |
| "grad_norm": 1.0397955179214478, | |
| "learning_rate": 1.2465710660560159e-05, | |
| "loss": 1.8635, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.24782608695652175, | |
| "grad_norm": 1.0397955179214478, | |
| "learning_rate": 1.2465710660560159e-05, | |
| "loss": 1.8428, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.25217391304347825, | |
| "grad_norm": 1.0397955179214478, | |
| "learning_rate": 1.2465710660560159e-05, | |
| "loss": 1.9786, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.2565217391304348, | |
| "grad_norm": 1.0397955179214478, | |
| "learning_rate": 1.2465710660560159e-05, | |
| "loss": 1.8905, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.2608695652173913, | |
| "grad_norm": 0.9489036202430725, | |
| "learning_rate": 1.279160171413683e-05, | |
| "loss": 1.8285, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.26521739130434785, | |
| "grad_norm": 0.9489036202430725, | |
| "learning_rate": 1.279160171413683e-05, | |
| "loss": 1.887, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.26956521739130435, | |
| "grad_norm": 0.9489036202430725, | |
| "learning_rate": 1.279160171413683e-05, | |
| "loss": 1.8579, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.27391304347826084, | |
| "grad_norm": 0.9489036202430725, | |
| "learning_rate": 1.279160171413683e-05, | |
| "loss": 1.8302, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.2782608695652174, | |
| "grad_norm": 0.884759247303009, | |
| "learning_rate": 1.309645243561725e-05, | |
| "loss": 1.9077, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.2826086956521739, | |
| "grad_norm": 0.884759247303009, | |
| "learning_rate": 1.309645243561725e-05, | |
| "loss": 1.778, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.28695652173913044, | |
| "grad_norm": 0.884759247303009, | |
| "learning_rate": 1.309645243561725e-05, | |
| "loss": 1.7314, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.29130434782608694, | |
| "grad_norm": 0.884759247303009, | |
| "learning_rate": 1.309645243561725e-05, | |
| "loss": 1.8035, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.2956521739130435, | |
| "grad_norm": 0.9762704968452454, | |
| "learning_rate": 1.3382815670697004e-05, | |
| "loss": 1.7613, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.9762704968452454, | |
| "learning_rate": 1.3382815670697004e-05, | |
| "loss": 1.7655, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.30434782608695654, | |
| "grad_norm": 0.9762704968452454, | |
| "learning_rate": 1.3382815670697004e-05, | |
| "loss": 1.6689, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.30869565217391304, | |
| "grad_norm": 0.9762704968452454, | |
| "learning_rate": 1.3382815670697004e-05, | |
| "loss": 1.7682, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.3130434782608696, | |
| "grad_norm": 0.8275449872016907, | |
| "learning_rate": 1.365280611037011e-05, | |
| "loss": 1.8038, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.3173913043478261, | |
| "grad_norm": 0.8275449872016907, | |
| "learning_rate": 1.365280611037011e-05, | |
| "loss": 1.7022, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.3217391304347826, | |
| "grad_norm": 0.8275449872016907, | |
| "learning_rate": 1.365280611037011e-05, | |
| "loss": 1.7887, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.32608695652173914, | |
| "grad_norm": 0.8275449872016907, | |
| "learning_rate": 1.365280611037011e-05, | |
| "loss": 1.7613, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.33043478260869563, | |
| "grad_norm": 0.7482348680496216, | |
| "learning_rate": 1.3908195157440944e-05, | |
| "loss": 1.7867, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.3347826086956522, | |
| "grad_norm": 0.7482348680496216, | |
| "learning_rate": 1.3908195157440944e-05, | |
| "loss": 1.825, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.3391304347826087, | |
| "grad_norm": 0.7482348680496216, | |
| "learning_rate": 1.3908195157440944e-05, | |
| "loss": 1.7778, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.34347826086956523, | |
| "grad_norm": 0.7482348680496216, | |
| "learning_rate": 1.3908195157440944e-05, | |
| "loss": 1.7115, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.34782608695652173, | |
| "grad_norm": 0.6351860761642456, | |
| "learning_rate": 1.4150481431212555e-05, | |
| "loss": 1.7619, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.3521739130434783, | |
| "grad_norm": 0.6351860761642456, | |
| "learning_rate": 1.4150481431212555e-05, | |
| "loss": 1.7002, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.3565217391304348, | |
| "grad_norm": 0.6351860761642456, | |
| "learning_rate": 1.4150481431212555e-05, | |
| "loss": 1.7352, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.36086956521739133, | |
| "grad_norm": 0.6351860761642456, | |
| "learning_rate": 1.4150481431212555e-05, | |
| "loss": 1.7489, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.3652173913043478, | |
| "grad_norm": 0.5535622835159302, | |
| "learning_rate": 1.4380944052388746e-05, | |
| "loss": 1.7328, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.3695652173913043, | |
| "grad_norm": 0.5535622835159302, | |
| "learning_rate": 1.4380944052388746e-05, | |
| "loss": 1.7426, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.3739130434782609, | |
| "grad_norm": 0.5535622835159302, | |
| "learning_rate": 1.4380944052388746e-05, | |
| "loss": 1.742, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.3782608695652174, | |
| "grad_norm": 0.5535622835159302, | |
| "learning_rate": 1.4380944052388746e-05, | |
| "loss": 1.7249, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.3826086956521739, | |
| "grad_norm": 0.5413718819618225, | |
| "learning_rate": 1.4600683520842756e-05, | |
| "loss": 1.6879, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.3869565217391304, | |
| "grad_norm": 0.5413718819618225, | |
| "learning_rate": 1.4600683520842756e-05, | |
| "loss": 1.7437, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.391304347826087, | |
| "grad_norm": 0.5413718819618225, | |
| "learning_rate": 1.4600683520842756e-05, | |
| "loss": 1.7177, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.39565217391304347, | |
| "grad_norm": 0.5413718819618225, | |
| "learning_rate": 1.4600683520842756e-05, | |
| "loss": 1.6507, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.5403290390968323, | |
| "learning_rate": 1.48106534992671e-05, | |
| "loss": 1.7619, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.4043478260869565, | |
| "grad_norm": 0.5403290390968323, | |
| "learning_rate": 1.48106534992671e-05, | |
| "loss": 1.7386, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.40869565217391307, | |
| "grad_norm": 0.5403290390968323, | |
| "learning_rate": 1.48106534992671e-05, | |
| "loss": 1.6804, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.41304347826086957, | |
| "grad_norm": 0.5403290390968323, | |
| "learning_rate": 1.48106534992671e-05, | |
| "loss": 1.6704, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.41739130434782606, | |
| "grad_norm": 0.48830699920654297, | |
| "learning_rate": 1.5011685827445838e-05, | |
| "loss": 1.7522, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.4217391304347826, | |
| "grad_norm": 0.48830699920654297, | |
| "learning_rate": 1.5011685827445838e-05, | |
| "loss": 1.6867, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.4260869565217391, | |
| "grad_norm": 0.48830699920654297, | |
| "learning_rate": 1.5011685827445838e-05, | |
| "loss": 1.6539, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.43043478260869567, | |
| "grad_norm": 0.48830699920654297, | |
| "learning_rate": 1.5011685827445838e-05, | |
| "loss": 1.7037, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 0.45016181468963623, | |
| "learning_rate": 1.520451042680786e-05, | |
| "loss": 1.634, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4391304347826087, | |
| "grad_norm": 0.45016181468963623, | |
| "learning_rate": 1.520451042680786e-05, | |
| "loss": 1.7027, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.4434782608695652, | |
| "grad_norm": 0.45016181468963623, | |
| "learning_rate": 1.520451042680786e-05, | |
| "loss": 1.6883, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.44782608695652176, | |
| "grad_norm": 0.45016181468963623, | |
| "learning_rate": 1.520451042680786e-05, | |
| "loss": 1.6648, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.45217391304347826, | |
| "grad_norm": 0.44704490900039673, | |
| "learning_rate": 1.5389771298780244e-05, | |
| "loss": 1.7269, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.45652173913043476, | |
| "grad_norm": 0.44704490900039673, | |
| "learning_rate": 1.5389771298780244e-05, | |
| "loss": 1.7, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.4608695652173913, | |
| "grad_norm": 0.44704490900039673, | |
| "learning_rate": 1.5389771298780244e-05, | |
| "loss": 1.7126, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.4652173913043478, | |
| "grad_norm": 0.44704490900039673, | |
| "learning_rate": 1.5389771298780244e-05, | |
| "loss": 1.6477, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.46956521739130436, | |
| "grad_norm": 0.4181145429611206, | |
| "learning_rate": 1.5568039502198696e-05, | |
| "loss": 1.6808, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.47391304347826085, | |
| "grad_norm": 0.4181145429611206, | |
| "learning_rate": 1.5568039502198696e-05, | |
| "loss": 1.6378, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.4782608695652174, | |
| "grad_norm": 0.4181145429611206, | |
| "learning_rate": 1.5568039502198696e-05, | |
| "loss": 1.6749, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.4826086956521739, | |
| "grad_norm": 0.4181145429611206, | |
| "learning_rate": 1.5568039502198696e-05, | |
| "loss": 1.6727, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.48695652173913045, | |
| "grad_norm": 0.47532176971435547, | |
| "learning_rate": 1.5739823769464473e-05, | |
| "loss": 1.6508, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.49130434782608695, | |
| "grad_norm": 0.47532176971435547, | |
| "learning_rate": 1.5739823769464473e-05, | |
| "loss": 1.6771, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.4956521739130435, | |
| "grad_norm": 0.47532176971435547, | |
| "learning_rate": 1.5739823769464473e-05, | |
| "loss": 1.6639, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.47532176971435547, | |
| "learning_rate": 1.5739823769464473e-05, | |
| "loss": 1.6892, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.5043478260869565, | |
| "grad_norm": 0.44671422243118286, | |
| "learning_rate": 1.5905579258955202e-05, | |
| "loss": 1.6251, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.508695652173913, | |
| "grad_norm": 0.44671422243118286, | |
| "learning_rate": 1.5905579258955202e-05, | |
| "loss": 1.6916, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.5130434782608696, | |
| "grad_norm": 0.44671422243118286, | |
| "learning_rate": 1.5905579258955202e-05, | |
| "loss": 1.6385, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.5173913043478261, | |
| "grad_norm": 0.44671422243118286, | |
| "learning_rate": 1.5905579258955202e-05, | |
| "loss": 1.6745, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.5217391304347826, | |
| "grad_norm": 0.43519601225852966, | |
| "learning_rate": 1.606571482304114e-05, | |
| "loss": 1.6723, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.5260869565217391, | |
| "grad_norm": 0.43519601225852966, | |
| "learning_rate": 1.606571482304114e-05, | |
| "loss": 1.7032, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.5304347826086957, | |
| "grad_norm": 0.43519601225852966, | |
| "learning_rate": 1.606571482304114e-05, | |
| "loss": 1.649, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.5347826086956522, | |
| "grad_norm": 0.43519601225852966, | |
| "learning_rate": 1.606571482304114e-05, | |
| "loss": 1.5973, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.5391304347826087, | |
| "grad_norm": 0.4354279041290283, | |
| "learning_rate": 1.6220599083923048e-05, | |
| "loss": 1.6496, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.5434782608695652, | |
| "grad_norm": 0.4354279041290283, | |
| "learning_rate": 1.6220599083923048e-05, | |
| "loss": 1.6142, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.5478260869565217, | |
| "grad_norm": 0.4354279041290283, | |
| "learning_rate": 1.6220599083923048e-05, | |
| "loss": 1.6788, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.5521739130434783, | |
| "grad_norm": 0.4354279041290283, | |
| "learning_rate": 1.6220599083923048e-05, | |
| "loss": 1.6921, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.5565217391304348, | |
| "grad_norm": 0.4076354205608368, | |
| "learning_rate": 1.6370565544521564e-05, | |
| "loss": 1.7186, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.5608695652173913, | |
| "grad_norm": 0.4076354205608368, | |
| "learning_rate": 1.6370565544521564e-05, | |
| "loss": 1.5906, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.5652173913043478, | |
| "grad_norm": 0.4076354205608368, | |
| "learning_rate": 1.6370565544521564e-05, | |
| "loss": 1.5466, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.5695652173913044, | |
| "grad_norm": 0.4076354205608368, | |
| "learning_rate": 1.6370565544521564e-05, | |
| "loss": 1.6492, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.5739130434782609, | |
| "grad_norm": 0.3979344964027405, | |
| "learning_rate": 1.651591691267134e-05, | |
| "loss": 1.6666, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.5782608695652174, | |
| "grad_norm": 0.3979344964027405, | |
| "learning_rate": 1.651591691267134e-05, | |
| "loss": 1.6911, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.5826086956521739, | |
| "grad_norm": 0.3979344964027405, | |
| "learning_rate": 1.651591691267134e-05, | |
| "loss": 1.5961, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.5869565217391305, | |
| "grad_norm": 0.3979344964027405, | |
| "learning_rate": 1.651591691267134e-05, | |
| "loss": 1.6683, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.591304347826087, | |
| "grad_norm": 0.3939136862754822, | |
| "learning_rate": 1.6656928779601318e-05, | |
| "loss": 1.6509, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.5956521739130435, | |
| "grad_norm": 0.3939136862754822, | |
| "learning_rate": 1.6656928779601318e-05, | |
| "loss": 1.6463, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.3939136862754822, | |
| "learning_rate": 1.6656928779601318e-05, | |
| "loss": 1.714, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.6043478260869565, | |
| "grad_norm": 0.3939136862754822, | |
| "learning_rate": 1.6656928779601318e-05, | |
| "loss": 1.6111, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.6086956521739131, | |
| "grad_norm": 0.3857693374156952, | |
| "learning_rate": 1.6793852765059776e-05, | |
| "loss": 1.6036, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.6130434782608696, | |
| "grad_norm": 0.3857693374156952, | |
| "learning_rate": 1.6793852765059776e-05, | |
| "loss": 1.5713, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.6173913043478261, | |
| "grad_norm": 0.3857693374156952, | |
| "learning_rate": 1.6793852765059776e-05, | |
| "loss": 1.6435, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.6217391304347826, | |
| "grad_norm": 0.3857693374156952, | |
| "learning_rate": 1.6793852765059776e-05, | |
| "loss": 1.6701, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.6260869565217392, | |
| "grad_norm": 0.35507267713546753, | |
| "learning_rate": 1.6926919219274422e-05, | |
| "loss": 1.6717, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.6304347826086957, | |
| "grad_norm": 0.35507267713546753, | |
| "learning_rate": 1.6926919219274422e-05, | |
| "loss": 1.6163, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.6347826086956522, | |
| "grad_norm": 0.35507267713546753, | |
| "learning_rate": 1.6926919219274422e-05, | |
| "loss": 1.5982, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.6391304347826087, | |
| "grad_norm": 0.35507267713546753, | |
| "learning_rate": 1.6926919219274422e-05, | |
| "loss": 1.6509, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.6434782608695652, | |
| "grad_norm": 0.35680633783340454, | |
| "learning_rate": 1.7056339554631436e-05, | |
| "loss": 1.6332, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.6478260869565218, | |
| "grad_norm": 0.35680633783340454, | |
| "learning_rate": 1.7056339554631436e-05, | |
| "loss": 1.5773, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.6521739130434783, | |
| "grad_norm": 0.35680633783340454, | |
| "learning_rate": 1.7056339554631436e-05, | |
| "loss": 1.562, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.6565217391304348, | |
| "grad_norm": 0.35680633783340454, | |
| "learning_rate": 1.7056339554631436e-05, | |
| "loss": 1.5869, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.6608695652173913, | |
| "grad_norm": 0.34213584661483765, | |
| "learning_rate": 1.7182308266345256e-05, | |
| "loss": 1.637, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.6652173913043479, | |
| "grad_norm": 0.34213584661483765, | |
| "learning_rate": 1.7182308266345256e-05, | |
| "loss": 1.6367, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.6695652173913044, | |
| "grad_norm": 0.34213584661483765, | |
| "learning_rate": 1.7182308266345256e-05, | |
| "loss": 1.6062, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.6739130434782609, | |
| "grad_norm": 0.34213584661483765, | |
| "learning_rate": 1.7182308266345256e-05, | |
| "loss": 1.6147, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.6782608695652174, | |
| "grad_norm": 0.3420649766921997, | |
| "learning_rate": 1.7305004690608827e-05, | |
| "loss": 1.5698, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.6826086956521739, | |
| "grad_norm": 0.3420649766921997, | |
| "learning_rate": 1.7305004690608827e-05, | |
| "loss": 1.5857, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.6869565217391305, | |
| "grad_norm": 0.3420649766921997, | |
| "learning_rate": 1.7305004690608827e-05, | |
| "loss": 1.5172, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.691304347826087, | |
| "grad_norm": 0.3420649766921997, | |
| "learning_rate": 1.7305004690608827e-05, | |
| "loss": 1.5799, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.6956521739130435, | |
| "grad_norm": 0.3389074504375458, | |
| "learning_rate": 1.7424594540116867e-05, | |
| "loss": 1.5744, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.3389074504375458, | |
| "learning_rate": 1.7424594540116867e-05, | |
| "loss": 1.6075, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.7043478260869566, | |
| "grad_norm": 0.3389074504375458, | |
| "learning_rate": 1.7424594540116867e-05, | |
| "loss": 1.5119, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.7086956521739131, | |
| "grad_norm": 0.3389074504375458, | |
| "learning_rate": 1.7424594540116867e-05, | |
| "loss": 1.6134, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.7130434782608696, | |
| "grad_norm": 0.3438399136066437, | |
| "learning_rate": 1.754123124995665e-05, | |
| "loss": 1.565, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.717391304347826, | |
| "grad_norm": 0.3438399136066437, | |
| "learning_rate": 1.754123124995665e-05, | |
| "loss": 1.6115, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.7217391304347827, | |
| "grad_norm": 0.3438399136066437, | |
| "learning_rate": 1.754123124995665e-05, | |
| "loss": 1.6062, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.7260869565217392, | |
| "grad_norm": 0.3438399136066437, | |
| "learning_rate": 1.754123124995665e-05, | |
| "loss": 1.5927, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.7304347826086957, | |
| "grad_norm": 0.3270600438117981, | |
| "learning_rate": 1.765505716129306e-05, | |
| "loss": 1.5475, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.7347826086956522, | |
| "grad_norm": 0.3270600438117981, | |
| "learning_rate": 1.765505716129306e-05, | |
| "loss": 1.5137, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.7391304347826086, | |
| "grad_norm": 0.3270600438117981, | |
| "learning_rate": 1.765505716129306e-05, | |
| "loss": 1.6577, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.7434782608695653, | |
| "grad_norm": 0.3270600438117981, | |
| "learning_rate": 1.765505716129306e-05, | |
| "loss": 1.6101, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.7478260869565218, | |
| "grad_norm": 0.3228694796562195, | |
| "learning_rate": 1.7766204565755586e-05, | |
| "loss": 1.5702, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.7521739130434782, | |
| "grad_norm": 0.3228694796562195, | |
| "learning_rate": 1.7766204565755586e-05, | |
| "loss": 1.6037, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.7565217391304347, | |
| "grad_norm": 0.3228694796562195, | |
| "learning_rate": 1.7766204565755586e-05, | |
| "loss": 1.5628, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.7608695652173914, | |
| "grad_norm": 0.3228694796562195, | |
| "learning_rate": 1.7766204565755586e-05, | |
| "loss": 1.5571, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.7652173913043478, | |
| "grad_norm": 0.3404123783111572, | |
| "learning_rate": 1.7874796629747068e-05, | |
| "loss": 1.6112, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.7695652173913043, | |
| "grad_norm": 0.3404123783111572, | |
| "learning_rate": 1.7874796629747068e-05, | |
| "loss": 1.5407, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.7739130434782608, | |
| "grad_norm": 0.3404123783111572, | |
| "learning_rate": 1.7874796629747068e-05, | |
| "loss": 1.5466, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.7782608695652173, | |
| "grad_norm": 0.3404123783111572, | |
| "learning_rate": 1.7874796629747068e-05, | |
| "loss": 1.6148, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.782608695652174, | |
| "grad_norm": 0.31963732838630676, | |
| "learning_rate": 1.7980948214869728e-05, | |
| "loss": 1.5987, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.7869565217391304, | |
| "grad_norm": 0.31963732838630676, | |
| "learning_rate": 1.7980948214869728e-05, | |
| "loss": 1.5906, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.7913043478260869, | |
| "grad_norm": 0.31963732838630676, | |
| "learning_rate": 1.7980948214869728e-05, | |
| "loss": 1.6133, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.7956521739130434, | |
| "grad_norm": 0.31963732838630676, | |
| "learning_rate": 1.7980948214869728e-05, | |
| "loss": 1.5955, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.3301166296005249, | |
| "learning_rate": 1.8084766608171415e-05, | |
| "loss": 1.5776, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.8043478260869565, | |
| "grad_norm": 0.3301166296005249, | |
| "learning_rate": 1.8084766608171415e-05, | |
| "loss": 1.6057, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.808695652173913, | |
| "grad_norm": 0.3301166296005249, | |
| "learning_rate": 1.8084766608171415e-05, | |
| "loss": 1.553, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.8130434782608695, | |
| "grad_norm": 0.3301166296005249, | |
| "learning_rate": 1.8084766608171415e-05, | |
| "loss": 1.5757, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.8173913043478261, | |
| "grad_norm": 0.3264283537864685, | |
| "learning_rate": 1.8186352173851508e-05, | |
| "loss": 1.6806, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.8217391304347826, | |
| "grad_norm": 0.3264283537864685, | |
| "learning_rate": 1.8186352173851508e-05, | |
| "loss": 1.5758, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.8260869565217391, | |
| "grad_norm": 0.3264283537864685, | |
| "learning_rate": 1.8186352173851508e-05, | |
| "loss": 1.5834, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.8304347826086956, | |
| "grad_norm": 0.3264283537864685, | |
| "learning_rate": 1.8186352173851508e-05, | |
| "loss": 1.5889, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.8347826086956521, | |
| "grad_norm": 0.3216494917869568, | |
| "learning_rate": 1.828579893635015e-05, | |
| "loss": 1.5547, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.8391304347826087, | |
| "grad_norm": 0.3216494917869568, | |
| "learning_rate": 1.828579893635015e-05, | |
| "loss": 1.5952, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.8434782608695652, | |
| "grad_norm": 0.3216494917869568, | |
| "learning_rate": 1.828579893635015e-05, | |
| "loss": 1.5456, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.8478260869565217, | |
| "grad_norm": 0.3216494917869568, | |
| "learning_rate": 1.828579893635015e-05, | |
| "loss": 1.5244, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.8521739130434782, | |
| "grad_norm": 0.3138068914413452, | |
| "learning_rate": 1.8383195103311694e-05, | |
| "loss": 1.6763, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.8565217391304348, | |
| "grad_norm": 0.3138068914413452, | |
| "learning_rate": 1.8383195103311694e-05, | |
| "loss": 1.5649, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.8608695652173913, | |
| "grad_norm": 0.3138068914413452, | |
| "learning_rate": 1.8383195103311694e-05, | |
| "loss": 1.5598, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.8652173913043478, | |
| "grad_norm": 0.3138068914413452, | |
| "learning_rate": 1.8383195103311694e-05, | |
| "loss": 1.6227, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 0.31041961908340454, | |
| "learning_rate": 1.8478623535712173e-05, | |
| "loss": 1.5523, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8739130434782608, | |
| "grad_norm": 0.31041961908340454, | |
| "learning_rate": 1.8478623535712173e-05, | |
| "loss": 1.6512, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.8782608695652174, | |
| "grad_norm": 0.31041961908340454, | |
| "learning_rate": 1.8478623535712173e-05, | |
| "loss": 1.5482, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.8826086956521739, | |
| "grad_norm": 0.31041961908340454, | |
| "learning_rate": 1.8478623535712173e-05, | |
| "loss": 1.5776, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.8869565217391304, | |
| "grad_norm": 0.3067266643047333, | |
| "learning_rate": 1.8572162171429905e-05, | |
| "loss": 1.5401, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.8913043478260869, | |
| "grad_norm": 0.3067266643047333, | |
| "learning_rate": 1.8572162171429905e-05, | |
| "loss": 1.5391, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.8956521739130435, | |
| "grad_norm": 0.3067266643047333, | |
| "learning_rate": 1.8572162171429905e-05, | |
| "loss": 1.6426, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.3067266643047333, | |
| "learning_rate": 1.8572162171429905e-05, | |
| "loss": 1.5383, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.9043478260869565, | |
| "grad_norm": 0.3177148103713989, | |
| "learning_rate": 1.866388440768456e-05, | |
| "loss": 1.5663, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.908695652173913, | |
| "grad_norm": 0.3177148103713989, | |
| "learning_rate": 1.866388440768456e-05, | |
| "loss": 1.4794, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.9130434782608695, | |
| "grad_norm": 0.3177148103713989, | |
| "learning_rate": 1.866388440768456e-05, | |
| "loss": 1.6449, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.9173913043478261, | |
| "grad_norm": 0.3177148103713989, | |
| "learning_rate": 1.866388440768456e-05, | |
| "loss": 1.5473, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.9217391304347826, | |
| "grad_norm": 0.32579195499420166, | |
| "learning_rate": 1.875385944704652e-05, | |
| "loss": 1.5343, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.9260869565217391, | |
| "grad_norm": 0.32579195499420166, | |
| "learning_rate": 1.875385944704652e-05, | |
| "loss": 1.5602, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.9304347826086956, | |
| "grad_norm": 0.32579195499420166, | |
| "learning_rate": 1.875385944704652e-05, | |
| "loss": 1.5161, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.9347826086956522, | |
| "grad_norm": 0.32579195499420166, | |
| "learning_rate": 1.875385944704652e-05, | |
| "loss": 1.5444, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.9391304347826087, | |
| "grad_norm": 0.301691472530365, | |
| "learning_rate": 1.8842152611103012e-05, | |
| "loss": 1.5774, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.9434782608695652, | |
| "grad_norm": 0.301691472530365, | |
| "learning_rate": 1.8842152611103012e-05, | |
| "loss": 1.5541, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.9478260869565217, | |
| "grad_norm": 0.301691472530365, | |
| "learning_rate": 1.8842152611103012e-05, | |
| "loss": 1.5717, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.9521739130434783, | |
| "grad_norm": 0.301691472530365, | |
| "learning_rate": 1.8842152611103012e-05, | |
| "loss": 1.6241, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.9565217391304348, | |
| "grad_norm": 0.3248462975025177, | |
| "learning_rate": 1.8928825625342374e-05, | |
| "loss": 1.5698, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.9608695652173913, | |
| "grad_norm": 0.3248462975025177, | |
| "learning_rate": 1.8928825625342374e-05, | |
| "loss": 1.5808, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.9652173913043478, | |
| "grad_norm": 0.3248462975025177, | |
| "learning_rate": 1.8928825625342374e-05, | |
| "loss": 1.5969, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.9695652173913043, | |
| "grad_norm": 0.3248462975025177, | |
| "learning_rate": 1.8928825625342374e-05, | |
| "loss": 1.5211, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.9739130434782609, | |
| "grad_norm": 0.32285231351852417, | |
| "learning_rate": 1.901393687836879e-05, | |
| "loss": 1.5536, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.9782608695652174, | |
| "grad_norm": 0.32285231351852417, | |
| "learning_rate": 1.901393687836879e-05, | |
| "loss": 1.5651, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.9826086956521739, | |
| "grad_norm": 0.32285231351852417, | |
| "learning_rate": 1.901393687836879e-05, | |
| "loss": 1.5113, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.9869565217391304, | |
| "grad_norm": 0.32285231351852417, | |
| "learning_rate": 1.901393687836879e-05, | |
| "loss": 1.6101, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.991304347826087, | |
| "grad_norm": 0.3327453136444092, | |
| "learning_rate": 1.9097541658173843e-05, | |
| "loss": 1.5708, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.9956521739130435, | |
| "grad_norm": 0.3327453136444092, | |
| "learning_rate": 1.9097541658173843e-05, | |
| "loss": 1.516, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.3327453136444092, | |
| "learning_rate": 1.9097541658173843e-05, | |
| "loss": 1.6279, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.0043478260869565, | |
| "grad_norm": 0.3327453136444092, | |
| "learning_rate": 1.9097541658173843e-05, | |
| "loss": 1.4676, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 1.008695652173913, | |
| "grad_norm": 0.3148215413093567, | |
| "learning_rate": 1.9179692367859514e-05, | |
| "loss": 1.5175, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 1.0130434782608695, | |
| "grad_norm": 0.3148215413093567, | |
| "learning_rate": 1.9179692367859514e-05, | |
| "loss": 1.5148, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 1.017391304347826, | |
| "grad_norm": 0.3148215413093567, | |
| "learning_rate": 1.9179692367859514e-05, | |
| "loss": 1.505, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 1.0217391304347827, | |
| "grad_norm": 0.3148215413093567, | |
| "learning_rate": 1.9179692367859514e-05, | |
| "loss": 1.4426, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.0260869565217392, | |
| "grad_norm": 0.3161942958831787, | |
| "learning_rate": 1.926043872292045e-05, | |
| "loss": 1.5709, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 1.0304347826086957, | |
| "grad_norm": 0.3161942958831787, | |
| "learning_rate": 1.926043872292045e-05, | |
| "loss": 1.5708, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 1.0347826086956522, | |
| "grad_norm": 0.3161942958831787, | |
| "learning_rate": 1.926043872292045e-05, | |
| "loss": 1.4882, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 1.0391304347826087, | |
| "grad_norm": 0.3161942958831787, | |
| "learning_rate": 1.926043872292045e-05, | |
| "loss": 1.5182, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 1.0434782608695652, | |
| "grad_norm": 0.3121671676635742, | |
| "learning_rate": 1.9339827931945454e-05, | |
| "loss": 1.59, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.0478260869565217, | |
| "grad_norm": 0.3121671676635742, | |
| "learning_rate": 1.9339827931945454e-05, | |
| "loss": 1.5309, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 1.0521739130434782, | |
| "grad_norm": 0.3121671676635742, | |
| "learning_rate": 1.9339827931945454e-05, | |
| "loss": 1.4626, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.0565217391304347, | |
| "grad_norm": 0.3121671676635742, | |
| "learning_rate": 1.9339827931945454e-05, | |
| "loss": 1.5466, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 1.0608695652173914, | |
| "grad_norm": 0.3044740557670593, | |
| "learning_rate": 1.941790486238291e-05, | |
| "loss": 1.5257, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 1.065217391304348, | |
| "grad_norm": 0.3044740557670593, | |
| "learning_rate": 1.941790486238291e-05, | |
| "loss": 1.6118, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.0695652173913044, | |
| "grad_norm": 0.3044740557670593, | |
| "learning_rate": 1.941790486238291e-05, | |
| "loss": 1.5306, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 1.0739130434782609, | |
| "grad_norm": 0.3044740557670593, | |
| "learning_rate": 1.941790486238291e-05, | |
| "loss": 1.4952, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 1.0782608695652174, | |
| "grad_norm": 0.3415665626525879, | |
| "learning_rate": 1.949471219282736e-05, | |
| "loss": 1.5271, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 1.0826086956521739, | |
| "grad_norm": 0.3415665626525879, | |
| "learning_rate": 1.949471219282736e-05, | |
| "loss": 1.4936, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 1.0869565217391304, | |
| "grad_norm": 0.3415665626525879, | |
| "learning_rate": 1.949471219282736e-05, | |
| "loss": 1.49, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.0913043478260869, | |
| "grad_norm": 0.3415665626525879, | |
| "learning_rate": 1.949471219282736e-05, | |
| "loss": 1.5655, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 1.0956521739130434, | |
| "grad_norm": 0.3097412586212158, | |
| "learning_rate": 1.9570290553121646e-05, | |
| "loss": 1.4897, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.3097412586212158, | |
| "learning_rate": 1.9570290553121646e-05, | |
| "loss": 1.4897, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 1.1043478260869566, | |
| "grad_norm": 0.3097412586212158, | |
| "learning_rate": 1.9570290553121646e-05, | |
| "loss": 1.5251, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 1.108695652173913, | |
| "grad_norm": 0.3097412586212158, | |
| "learning_rate": 1.9570290553121646e-05, | |
| "loss": 1.451, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.1130434782608696, | |
| "grad_norm": 0.3289230763912201, | |
| "learning_rate": 1.9644678653425876e-05, | |
| "loss": 1.4828, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 1.117391304347826, | |
| "grad_norm": 0.3289230763912201, | |
| "learning_rate": 1.9644678653425876e-05, | |
| "loss": 1.5472, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 1.1217391304347826, | |
| "grad_norm": 0.3289230763912201, | |
| "learning_rate": 1.9644678653425876e-05, | |
| "loss": 1.5391, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 1.126086956521739, | |
| "grad_norm": 0.3289230763912201, | |
| "learning_rate": 1.9644678653425876e-05, | |
| "loss": 1.5493, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 1.1304347826086956, | |
| "grad_norm": 0.3122900128364563, | |
| "learning_rate": 1.971791340327986e-05, | |
| "loss": 1.4891, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.134782608695652, | |
| "grad_norm": 0.3122900128364563, | |
| "learning_rate": 1.971791340327986e-05, | |
| "loss": 1.4754, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 1.1391304347826088, | |
| "grad_norm": 0.3122900128364563, | |
| "learning_rate": 1.971791340327986e-05, | |
| "loss": 1.4819, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 1.1434782608695653, | |
| "grad_norm": 0.3122900128364563, | |
| "learning_rate": 1.971791340327986e-05, | |
| "loss": 1.5694, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 1.1478260869565218, | |
| "grad_norm": 0.3180365562438965, | |
| "learning_rate": 1.979003002157565e-05, | |
| "loss": 1.4982, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 1.1521739130434783, | |
| "grad_norm": 0.3180365562438965, | |
| "learning_rate": 1.979003002157565e-05, | |
| "loss": 1.5122, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.1565217391304348, | |
| "grad_norm": 0.3180365562438965, | |
| "learning_rate": 1.979003002157565e-05, | |
| "loss": 1.5615, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 1.1608695652173913, | |
| "grad_norm": 0.3180365562438965, | |
| "learning_rate": 1.979003002157565e-05, | |
| "loss": 1.5346, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 1.1652173913043478, | |
| "grad_norm": 0.29976481199264526, | |
| "learning_rate": 1.9861062138260542e-05, | |
| "loss": 1.5567, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 1.1695652173913043, | |
| "grad_norm": 0.29976481199264526, | |
| "learning_rate": 1.9861062138260542e-05, | |
| "loss": 1.5182, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 1.1739130434782608, | |
| "grad_norm": 0.29976481199264526, | |
| "learning_rate": 1.9861062138260542e-05, | |
| "loss": 1.4794, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.1782608695652175, | |
| "grad_norm": 0.29976481199264526, | |
| "learning_rate": 1.9861062138260542e-05, | |
| "loss": 1.5981, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 1.182608695652174, | |
| "grad_norm": 0.3091888129711151, | |
| "learning_rate": 1.993104188850563e-05, | |
| "loss": 1.4554, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 1.1869565217391305, | |
| "grad_norm": 0.3091888129711151, | |
| "learning_rate": 1.993104188850563e-05, | |
| "loss": 1.5417, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.191304347826087, | |
| "grad_norm": 0.3091888129711151, | |
| "learning_rate": 1.993104188850563e-05, | |
| "loss": 1.4955, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 1.1956521739130435, | |
| "grad_norm": 0.3091888129711151, | |
| "learning_rate": 1.993104188850563e-05, | |
| "loss": 1.4146, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.3073773682117462, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5044, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 1.2043478260869565, | |
| "grad_norm": 0.3073773682117462, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5285, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 1.208695652173913, | |
| "grad_norm": 0.3073773682117462, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5259, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 1.2130434782608694, | |
| "grad_norm": 0.3073773682117462, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5399, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 1.2173913043478262, | |
| "grad_norm": 0.31173551082611084, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4887, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.2217391304347827, | |
| "grad_norm": 0.31173551082611084, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4436, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 1.2260869565217392, | |
| "grad_norm": 0.31173551082611084, | |
| "learning_rate": 2e-05, | |
| "loss": 1.412, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 1.2304347826086957, | |
| "grad_norm": 0.31173551082611084, | |
| "learning_rate": 2e-05, | |
| "loss": 1.489, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 1.2347826086956522, | |
| "grad_norm": 0.28800562024116516, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5448, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 1.2391304347826086, | |
| "grad_norm": 0.28800562024116516, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4644, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.2434782608695651, | |
| "grad_norm": 0.28800562024116516, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5625, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 1.2478260869565219, | |
| "grad_norm": 0.28800562024116516, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5496, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 1.2521739130434781, | |
| "grad_norm": 0.3107075095176697, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4766, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 1.2565217391304349, | |
| "grad_norm": 0.3107075095176697, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5138, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 1.2608695652173914, | |
| "grad_norm": 0.3107075095176697, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4492, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.2652173913043478, | |
| "grad_norm": 0.3107075095176697, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4896, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 1.2695652173913043, | |
| "grad_norm": 0.29269176721572876, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5382, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 1.2739130434782608, | |
| "grad_norm": 0.29269176721572876, | |
| "learning_rate": 2e-05, | |
| "loss": 1.563, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 1.2782608695652173, | |
| "grad_norm": 0.29269176721572876, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5334, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 1.2826086956521738, | |
| "grad_norm": 0.29269176721572876, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4961, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.2869565217391306, | |
| "grad_norm": 0.32013219594955444, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5529, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.2913043478260868, | |
| "grad_norm": 0.32013219594955444, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4677, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.2956521739130435, | |
| "grad_norm": 0.32013219594955444, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4208, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 0.32013219594955444, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5618, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 1.3043478260869565, | |
| "grad_norm": 0.31480494141578674, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4929, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.308695652173913, | |
| "grad_norm": 0.31480494141578674, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4795, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 1.3130434782608695, | |
| "grad_norm": 0.31480494141578674, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5061, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 1.317391304347826, | |
| "grad_norm": 0.31480494141578674, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4551, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 1.3217391304347825, | |
| "grad_norm": 0.3061749339103699, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5821, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 1.3260869565217392, | |
| "grad_norm": 0.3061749339103699, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5486, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.3304347826086955, | |
| "grad_norm": 0.3061749339103699, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5276, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 1.3347826086956522, | |
| "grad_norm": 0.3061749339103699, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5527, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 1.3391304347826087, | |
| "grad_norm": 0.30773481726646423, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5047, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 1.3434782608695652, | |
| "grad_norm": 0.30773481726646423, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4873, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 1.3478260869565217, | |
| "grad_norm": 0.30773481726646423, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4933, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.3521739130434782, | |
| "grad_norm": 0.30773481726646423, | |
| "learning_rate": 2e-05, | |
| "loss": 1.481, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 1.3565217391304347, | |
| "grad_norm": 0.3024744987487793, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5306, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 1.3608695652173912, | |
| "grad_norm": 0.3024744987487793, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4178, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 1.365217391304348, | |
| "grad_norm": 0.3024744987487793, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4804, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 1.3695652173913042, | |
| "grad_norm": 0.3024744987487793, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4536, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.373913043478261, | |
| "grad_norm": 0.28332576155662537, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4416, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 1.3782608695652174, | |
| "grad_norm": 0.28332576155662537, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4684, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 1.382608695652174, | |
| "grad_norm": 0.28332576155662537, | |
| "learning_rate": 2e-05, | |
| "loss": 1.46, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 1.3869565217391304, | |
| "grad_norm": 0.28332576155662537, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5102, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 1.391304347826087, | |
| "grad_norm": 0.3170158863067627, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5092, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.3956521739130434, | |
| "grad_norm": 0.3170158863067627, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4435, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 0.3170158863067627, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4784, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 1.4043478260869566, | |
| "grad_norm": 0.3170158863067627, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5239, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 1.4086956521739131, | |
| "grad_norm": 0.32871973514556885, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5463, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.4130434782608696, | |
| "grad_norm": 0.32871973514556885, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5111, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.4173913043478261, | |
| "grad_norm": 0.32871973514556885, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5124, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 1.4217391304347826, | |
| "grad_norm": 0.32871973514556885, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5062, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 1.4260869565217391, | |
| "grad_norm": 0.3206692636013031, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4279, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 1.4304347826086956, | |
| "grad_norm": 0.3206692636013031, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5357, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 1.434782608695652, | |
| "grad_norm": 0.3206692636013031, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4907, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.4391304347826086, | |
| "grad_norm": 0.3206692636013031, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5184, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 1.4434782608695653, | |
| "grad_norm": 0.31712856888771057, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4562, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 1.4478260869565218, | |
| "grad_norm": 0.31712856888771057, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4774, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 1.4521739130434783, | |
| "grad_norm": 0.31712856888771057, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5188, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 1.4565217391304348, | |
| "grad_norm": 0.31712856888771057, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4372, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.4608695652173913, | |
| "grad_norm": 0.3141573965549469, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5092, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 1.4652173913043478, | |
| "grad_norm": 0.3141573965549469, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4997, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 1.4695652173913043, | |
| "grad_norm": 0.3141573965549469, | |
| "learning_rate": 2e-05, | |
| "loss": 1.513, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 1.4739130434782608, | |
| "grad_norm": 0.3141573965549469, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4096, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 1.4782608695652173, | |
| "grad_norm": 0.31141531467437744, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4981, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.482608695652174, | |
| "grad_norm": 0.31141531467437744, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4774, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 1.4869565217391305, | |
| "grad_norm": 0.31141531467437744, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4529, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 1.491304347826087, | |
| "grad_norm": 0.31141531467437744, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4369, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 1.4956521739130435, | |
| "grad_norm": 0.32201042771339417, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5313, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.32201042771339417, | |
| "learning_rate": 2e-05, | |
| "loss": 1.508, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.5043478260869565, | |
| "grad_norm": 0.32201042771339417, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5375, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 1.508695652173913, | |
| "grad_norm": 0.32201042771339417, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5526, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 1.5130434782608697, | |
| "grad_norm": 0.3498378098011017, | |
| "learning_rate": 2e-05, | |
| "loss": 1.464, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 1.517391304347826, | |
| "grad_norm": 0.3498378098011017, | |
| "learning_rate": 2e-05, | |
| "loss": 1.498, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 1.5217391304347827, | |
| "grad_norm": 0.3498378098011017, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4936, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.526086956521739, | |
| "grad_norm": 0.3498378098011017, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5393, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 1.5304347826086957, | |
| "grad_norm": 0.31443101167678833, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3899, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 1.5347826086956522, | |
| "grad_norm": 0.31443101167678833, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4724, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 1.5391304347826087, | |
| "grad_norm": 0.31443101167678833, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5721, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 1.5434782608695652, | |
| "grad_norm": 0.31443101167678833, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4112, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.5478260869565217, | |
| "grad_norm": 0.30996161699295044, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4969, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 1.5521739130434784, | |
| "grad_norm": 0.30996161699295044, | |
| "learning_rate": 2e-05, | |
| "loss": 1.442, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 1.5565217391304347, | |
| "grad_norm": 0.30996161699295044, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4836, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 1.5608695652173914, | |
| "grad_norm": 0.30996161699295044, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4341, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 1.5652173913043477, | |
| "grad_norm": 0.31205815076828003, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5243, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.5695652173913044, | |
| "grad_norm": 0.31205815076828003, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5261, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 1.5739130434782609, | |
| "grad_norm": 0.31205815076828003, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5295, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 1.5782608695652174, | |
| "grad_norm": 0.31205815076828003, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4583, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 1.5826086956521739, | |
| "grad_norm": 0.31103307008743286, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4848, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 1.5869565217391304, | |
| "grad_norm": 0.31103307008743286, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4382, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.591304347826087, | |
| "grad_norm": 0.31103307008743286, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4936, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 1.5956521739130434, | |
| "grad_norm": 0.31103307008743286, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4764, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.3369222581386566, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5201, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 1.6043478260869564, | |
| "grad_norm": 0.3369222581386566, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4769, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 1.608695652173913, | |
| "grad_norm": 0.3369222581386566, | |
| "learning_rate": 2e-05, | |
| "loss": 1.459, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.6130434782608696, | |
| "grad_norm": 0.3369222581386566, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4849, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 1.617391304347826, | |
| "grad_norm": 0.32731279730796814, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4742, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 1.6217391304347826, | |
| "grad_norm": 0.32731279730796814, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4444, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 1.626086956521739, | |
| "grad_norm": 0.32731279730796814, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4716, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 1.6304347826086958, | |
| "grad_norm": 0.32731279730796814, | |
| "learning_rate": 2e-05, | |
| "loss": 1.474, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.634782608695652, | |
| "grad_norm": 0.3295840620994568, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4419, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 1.6391304347826088, | |
| "grad_norm": 0.3295840620994568, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4744, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 1.643478260869565, | |
| "grad_norm": 0.3295840620994568, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4461, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 1.6478260869565218, | |
| "grad_norm": 0.3295840620994568, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4591, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 1.6521739130434783, | |
| "grad_norm": 0.326766699552536, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4586, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.6565217391304348, | |
| "grad_norm": 0.326766699552536, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4196, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 1.6608695652173913, | |
| "grad_norm": 0.326766699552536, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4268, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 1.6652173913043478, | |
| "grad_norm": 0.326766699552536, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4626, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 1.6695652173913045, | |
| "grad_norm": 0.31557193398475647, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5136, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 1.6739130434782608, | |
| "grad_norm": 0.31557193398475647, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5269, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.6782608695652175, | |
| "grad_norm": 0.31557193398475647, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4681, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 1.6826086956521737, | |
| "grad_norm": 0.31557193398475647, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4613, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 1.6869565217391305, | |
| "grad_norm": 0.30831843614578247, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5108, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 1.691304347826087, | |
| "grad_norm": 0.30831843614578247, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4337, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 1.6956521739130435, | |
| "grad_norm": 0.30831843614578247, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4706, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 0.30831843614578247, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4333, | |
| "step": 391 | |
| }, | |
| { | |
| "epoch": 1.7043478260869565, | |
| "grad_norm": 0.3329159915447235, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4945, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 1.7086956521739132, | |
| "grad_norm": 0.3329159915447235, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4734, | |
| "step": 393 | |
| }, | |
| { | |
| "epoch": 1.7130434782608694, | |
| "grad_norm": 0.3329159915447235, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4894, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 1.7173913043478262, | |
| "grad_norm": 0.3329159915447235, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5237, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 1.7217391304347827, | |
| "grad_norm": 0.33224427700042725, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4447, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 1.7260869565217392, | |
| "grad_norm": 0.33224427700042725, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4854, | |
| "step": 397 | |
| }, | |
| { | |
| "epoch": 1.7304347826086957, | |
| "grad_norm": 0.33224427700042725, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4955, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 1.7347826086956522, | |
| "grad_norm": 0.33224427700042725, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5115, | |
| "step": 399 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 0.31770649552345276, | |
| "learning_rate": 2e-05, | |
| "loss": 1.53, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.7434782608695651, | |
| "grad_norm": 0.31770649552345276, | |
| "learning_rate": 2e-05, | |
| "loss": 1.49, | |
| "step": 401 | |
| }, | |
| { | |
| "epoch": 1.7478260869565219, | |
| "grad_norm": 0.31770649552345276, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4991, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 1.7521739130434781, | |
| "grad_norm": 0.31770649552345276, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5023, | |
| "step": 403 | |
| }, | |
| { | |
| "epoch": 1.7565217391304349, | |
| "grad_norm": 0.3262755870819092, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4715, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 1.7608695652173914, | |
| "grad_norm": 0.3262755870819092, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4738, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 1.7652173913043478, | |
| "grad_norm": 0.3262755870819092, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4988, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 1.7695652173913043, | |
| "grad_norm": 0.3262755870819092, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4644, | |
| "step": 407 | |
| }, | |
| { | |
| "epoch": 1.7739130434782608, | |
| "grad_norm": 0.3169100880622864, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4858, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 1.7782608695652173, | |
| "grad_norm": 0.3169100880622864, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5205, | |
| "step": 409 | |
| }, | |
| { | |
| "epoch": 1.7826086956521738, | |
| "grad_norm": 0.3169100880622864, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4483, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.7869565217391306, | |
| "grad_norm": 0.3169100880622864, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3957, | |
| "step": 411 | |
| }, | |
| { | |
| "epoch": 1.7913043478260868, | |
| "grad_norm": 0.3154841661453247, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4797, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 1.7956521739130435, | |
| "grad_norm": 0.3154841661453247, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4251, | |
| "step": 413 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.3154841661453247, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4896, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 1.8043478260869565, | |
| "grad_norm": 0.3154841661453247, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4793, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 1.808695652173913, | |
| "grad_norm": 0.33579424023628235, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4919, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 1.8130434782608695, | |
| "grad_norm": 0.33579424023628235, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5001, | |
| "step": 417 | |
| }, | |
| { | |
| "epoch": 1.8173913043478263, | |
| "grad_norm": 0.33579424023628235, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5017, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 1.8217391304347825, | |
| "grad_norm": 0.33579424023628235, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4573, | |
| "step": 419 | |
| }, | |
| { | |
| "epoch": 1.8260869565217392, | |
| "grad_norm": 0.3174879550933838, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5409, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.8304347826086955, | |
| "grad_norm": 0.3174879550933838, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4664, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.8347826086956522, | |
| "grad_norm": 0.3174879550933838, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4139, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 1.8391304347826087, | |
| "grad_norm": 0.3174879550933838, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5241, | |
| "step": 423 | |
| }, | |
| { | |
| "epoch": 1.8434782608695652, | |
| "grad_norm": 0.32846924662590027, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4632, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 1.8478260869565217, | |
| "grad_norm": 0.32846924662590027, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4625, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 1.8521739130434782, | |
| "grad_norm": 0.32846924662590027, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4093, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 1.856521739130435, | |
| "grad_norm": 0.32846924662590027, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4844, | |
| "step": 427 | |
| }, | |
| { | |
| "epoch": 1.8608695652173912, | |
| "grad_norm": 0.31925633549690247, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4778, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 1.865217391304348, | |
| "grad_norm": 0.31925633549690247, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4719, | |
| "step": 429 | |
| }, | |
| { | |
| "epoch": 1.8695652173913042, | |
| "grad_norm": 0.31925633549690247, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4404, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.873913043478261, | |
| "grad_norm": 0.31925633549690247, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4333, | |
| "step": 431 | |
| }, | |
| { | |
| "epoch": 1.8782608695652174, | |
| "grad_norm": 0.31410935521125793, | |
| "learning_rate": 2e-05, | |
| "loss": 1.452, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 1.882608695652174, | |
| "grad_norm": 0.31410935521125793, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4822, | |
| "step": 433 | |
| }, | |
| { | |
| "epoch": 1.8869565217391304, | |
| "grad_norm": 0.31410935521125793, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4667, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 1.891304347826087, | |
| "grad_norm": 0.31410935521125793, | |
| "learning_rate": 2e-05, | |
| "loss": 1.475, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 1.8956521739130436, | |
| "grad_norm": 0.3550223112106323, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3765, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 0.3550223112106323, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4709, | |
| "step": 437 | |
| }, | |
| { | |
| "epoch": 1.9043478260869566, | |
| "grad_norm": 0.3550223112106323, | |
| "learning_rate": 2e-05, | |
| "loss": 1.529, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 1.908695652173913, | |
| "grad_norm": 0.3550223112106323, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4646, | |
| "step": 439 | |
| }, | |
| { | |
| "epoch": 1.9130434782608696, | |
| "grad_norm": 0.31760966777801514, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4878, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.9173913043478261, | |
| "grad_norm": 0.31760966777801514, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4272, | |
| "step": 441 | |
| }, | |
| { | |
| "epoch": 1.9217391304347826, | |
| "grad_norm": 0.31760966777801514, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4009, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 1.9260869565217391, | |
| "grad_norm": 0.31760966777801514, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3982, | |
| "step": 443 | |
| }, | |
| { | |
| "epoch": 1.9304347826086956, | |
| "grad_norm": 0.3307594358921051, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4638, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 1.9347826086956523, | |
| "grad_norm": 0.3307594358921051, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4494, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 1.9391304347826086, | |
| "grad_norm": 0.3307594358921051, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4583, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 1.9434782608695653, | |
| "grad_norm": 0.3307594358921051, | |
| "learning_rate": 2e-05, | |
| "loss": 1.529, | |
| "step": 447 | |
| }, | |
| { | |
| "epoch": 1.9478260869565216, | |
| "grad_norm": 0.38704583048820496, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4085, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 1.9521739130434783, | |
| "grad_norm": 0.38704583048820496, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4487, | |
| "step": 449 | |
| }, | |
| { | |
| "epoch": 1.9565217391304348, | |
| "grad_norm": 0.38704583048820496, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4419, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.9608695652173913, | |
| "grad_norm": 0.38704583048820496, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4364, | |
| "step": 451 | |
| }, | |
| { | |
| "epoch": 1.9652173913043478, | |
| "grad_norm": 0.3177286684513092, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5169, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 1.9695652173913043, | |
| "grad_norm": 0.3177286684513092, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4835, | |
| "step": 453 | |
| }, | |
| { | |
| "epoch": 1.973913043478261, | |
| "grad_norm": 0.3177286684513092, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4178, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 1.9782608695652173, | |
| "grad_norm": 0.3177286684513092, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4891, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 1.982608695652174, | |
| "grad_norm": 0.31212443113327026, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4395, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 1.9869565217391303, | |
| "grad_norm": 0.31212443113327026, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4942, | |
| "step": 457 | |
| }, | |
| { | |
| "epoch": 1.991304347826087, | |
| "grad_norm": 0.31212443113327026, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4559, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 1.9956521739130435, | |
| "grad_norm": 0.31212443113327026, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5684, | |
| "step": 459 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.34978216886520386, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4239, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.0043478260869567, | |
| "grad_norm": 0.34978216886520386, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4397, | |
| "step": 461 | |
| }, | |
| { | |
| "epoch": 2.008695652173913, | |
| "grad_norm": 0.34978216886520386, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4092, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 2.0130434782608697, | |
| "grad_norm": 0.34978216886520386, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3932, | |
| "step": 463 | |
| }, | |
| { | |
| "epoch": 2.017391304347826, | |
| "grad_norm": 0.3233121931552887, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4781, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 2.0217391304347827, | |
| "grad_norm": 0.3233121931552887, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3867, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 2.026086956521739, | |
| "grad_norm": 0.3233121931552887, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4451, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 2.0304347826086957, | |
| "grad_norm": 0.3233121931552887, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4717, | |
| "step": 467 | |
| }, | |
| { | |
| "epoch": 2.034782608695652, | |
| "grad_norm": 0.3593771159648895, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4251, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 2.0391304347826087, | |
| "grad_norm": 0.3593771159648895, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3973, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 2.0434782608695654, | |
| "grad_norm": 0.3593771159648895, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4282, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.0478260869565217, | |
| "grad_norm": 0.3593771159648895, | |
| "learning_rate": 2e-05, | |
| "loss": 1.413, | |
| "step": 471 | |
| }, | |
| { | |
| "epoch": 2.0521739130434784, | |
| "grad_norm": 0.3393575847148895, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3597, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 2.0565217391304347, | |
| "grad_norm": 0.3393575847148895, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4215, | |
| "step": 473 | |
| }, | |
| { | |
| "epoch": 2.0608695652173914, | |
| "grad_norm": 0.3393575847148895, | |
| "learning_rate": 2e-05, | |
| "loss": 1.475, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 2.0652173913043477, | |
| "grad_norm": 0.3393575847148895, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3778, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 2.0695652173913044, | |
| "grad_norm": 0.34234023094177246, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4295, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 2.0739130434782607, | |
| "grad_norm": 0.34234023094177246, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4279, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 2.0782608695652174, | |
| "grad_norm": 0.34234023094177246, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4368, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 2.082608695652174, | |
| "grad_norm": 0.34234023094177246, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3915, | |
| "step": 479 | |
| }, | |
| { | |
| "epoch": 2.0869565217391304, | |
| "grad_norm": 0.34392404556274414, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4547, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.091304347826087, | |
| "grad_norm": 0.34392404556274414, | |
| "learning_rate": 2e-05, | |
| "loss": 1.395, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 2.0956521739130434, | |
| "grad_norm": 0.34392404556274414, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4057, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "grad_norm": 0.34392404556274414, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3388, | |
| "step": 483 | |
| }, | |
| { | |
| "epoch": 2.1043478260869564, | |
| "grad_norm": 0.32572683691978455, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3976, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 2.108695652173913, | |
| "grad_norm": 0.32572683691978455, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3961, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 2.1130434782608694, | |
| "grad_norm": 0.32572683691978455, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4213, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 2.117391304347826, | |
| "grad_norm": 0.32572683691978455, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4081, | |
| "step": 487 | |
| }, | |
| { | |
| "epoch": 2.121739130434783, | |
| "grad_norm": 0.33425086736679077, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3635, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 2.126086956521739, | |
| "grad_norm": 0.33425086736679077, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4042, | |
| "step": 489 | |
| }, | |
| { | |
| "epoch": 2.130434782608696, | |
| "grad_norm": 0.33425086736679077, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3905, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.134782608695652, | |
| "grad_norm": 0.33425086736679077, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4297, | |
| "step": 491 | |
| }, | |
| { | |
| "epoch": 2.139130434782609, | |
| "grad_norm": 0.3359292149543762, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4222, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 2.143478260869565, | |
| "grad_norm": 0.3359292149543762, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4278, | |
| "step": 493 | |
| }, | |
| { | |
| "epoch": 2.1478260869565218, | |
| "grad_norm": 0.3359292149543762, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4651, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 2.1521739130434785, | |
| "grad_norm": 0.3359292149543762, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3972, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.1565217391304348, | |
| "grad_norm": 0.33683329820632935, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4149, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 2.1608695652173915, | |
| "grad_norm": 0.33683329820632935, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4339, | |
| "step": 497 | |
| }, | |
| { | |
| "epoch": 2.1652173913043478, | |
| "grad_norm": 0.33683329820632935, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3726, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 2.1695652173913045, | |
| "grad_norm": 0.33683329820632935, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4655, | |
| "step": 499 | |
| }, | |
| { | |
| "epoch": 2.1739130434782608, | |
| "grad_norm": 0.3291366994380951, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4376, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.1782608695652175, | |
| "grad_norm": 0.3291366994380951, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4483, | |
| "step": 501 | |
| }, | |
| { | |
| "epoch": 2.1826086956521737, | |
| "grad_norm": 0.3291366994380951, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4218, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 2.1869565217391305, | |
| "grad_norm": 0.3291366994380951, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4179, | |
| "step": 503 | |
| }, | |
| { | |
| "epoch": 2.1913043478260867, | |
| "grad_norm": 0.3259764015674591, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4048, | |
| "step": 504 | |
| }, | |
| { | |
| "epoch": 2.1956521739130435, | |
| "grad_norm": 0.3259764015674591, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4301, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.3259764015674591, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3742, | |
| "step": 506 | |
| }, | |
| { | |
| "epoch": 2.2043478260869565, | |
| "grad_norm": 0.3259764015674591, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4225, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 2.208695652173913, | |
| "grad_norm": 0.30931395292282104, | |
| "learning_rate": 2e-05, | |
| "loss": 1.438, | |
| "step": 508 | |
| }, | |
| { | |
| "epoch": 2.2130434782608694, | |
| "grad_norm": 0.30931395292282104, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4397, | |
| "step": 509 | |
| }, | |
| { | |
| "epoch": 2.217391304347826, | |
| "grad_norm": 0.30931395292282104, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4551, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.2217391304347824, | |
| "grad_norm": 0.30931395292282104, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3048, | |
| "step": 511 | |
| }, | |
| { | |
| "epoch": 2.226086956521739, | |
| "grad_norm": 0.3235810399055481, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4198, | |
| "step": 512 | |
| }, | |
| { | |
| "epoch": 2.230434782608696, | |
| "grad_norm": 0.3235810399055481, | |
| "learning_rate": 2e-05, | |
| "loss": 1.359, | |
| "step": 513 | |
| }, | |
| { | |
| "epoch": 2.234782608695652, | |
| "grad_norm": 0.3235810399055481, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3955, | |
| "step": 514 | |
| }, | |
| { | |
| "epoch": 2.239130434782609, | |
| "grad_norm": 0.3235810399055481, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3925, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.243478260869565, | |
| "grad_norm": 0.33742859959602356, | |
| "learning_rate": 2e-05, | |
| "loss": 1.5026, | |
| "step": 516 | |
| }, | |
| { | |
| "epoch": 2.247826086956522, | |
| "grad_norm": 0.33742859959602356, | |
| "learning_rate": 2e-05, | |
| "loss": 1.423, | |
| "step": 517 | |
| }, | |
| { | |
| "epoch": 2.252173913043478, | |
| "grad_norm": 0.33742859959602356, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3906, | |
| "step": 518 | |
| }, | |
| { | |
| "epoch": 2.256521739130435, | |
| "grad_norm": 0.33742859959602356, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3976, | |
| "step": 519 | |
| }, | |
| { | |
| "epoch": 2.260869565217391, | |
| "grad_norm": 0.3243946135044098, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4636, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.265217391304348, | |
| "grad_norm": 0.3243946135044098, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4101, | |
| "step": 521 | |
| }, | |
| { | |
| "epoch": 2.269565217391304, | |
| "grad_norm": 0.3243946135044098, | |
| "learning_rate": 2e-05, | |
| "loss": 1.37, | |
| "step": 522 | |
| }, | |
| { | |
| "epoch": 2.273913043478261, | |
| "grad_norm": 0.3243946135044098, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3731, | |
| "step": 523 | |
| }, | |
| { | |
| "epoch": 2.2782608695652176, | |
| "grad_norm": 0.3233046233654022, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3996, | |
| "step": 524 | |
| }, | |
| { | |
| "epoch": 2.282608695652174, | |
| "grad_norm": 0.3233046233654022, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3471, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.2869565217391306, | |
| "grad_norm": 0.3233046233654022, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4454, | |
| "step": 526 | |
| }, | |
| { | |
| "epoch": 2.291304347826087, | |
| "grad_norm": 0.3233046233654022, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3706, | |
| "step": 527 | |
| }, | |
| { | |
| "epoch": 2.2956521739130435, | |
| "grad_norm": 0.30293363332748413, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4508, | |
| "step": 528 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 0.30293363332748413, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4687, | |
| "step": 529 | |
| }, | |
| { | |
| "epoch": 2.3043478260869565, | |
| "grad_norm": 0.30293363332748413, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4307, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.3086956521739133, | |
| "grad_norm": 0.30293363332748413, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3598, | |
| "step": 531 | |
| }, | |
| { | |
| "epoch": 2.3130434782608695, | |
| "grad_norm": 0.30868127942085266, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4503, | |
| "step": 532 | |
| }, | |
| { | |
| "epoch": 2.3173913043478263, | |
| "grad_norm": 0.30868127942085266, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3985, | |
| "step": 533 | |
| }, | |
| { | |
| "epoch": 2.3217391304347825, | |
| "grad_norm": 0.30868127942085266, | |
| "learning_rate": 2e-05, | |
| "loss": 1.367, | |
| "step": 534 | |
| }, | |
| { | |
| "epoch": 2.3260869565217392, | |
| "grad_norm": 0.30868127942085266, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4399, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.3304347826086955, | |
| "grad_norm": 0.31287699937820435, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3822, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 2.3347826086956522, | |
| "grad_norm": 0.31287699937820435, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4136, | |
| "step": 537 | |
| }, | |
| { | |
| "epoch": 2.3391304347826085, | |
| "grad_norm": 0.31287699937820435, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4002, | |
| "step": 538 | |
| }, | |
| { | |
| "epoch": 2.3434782608695652, | |
| "grad_norm": 0.31287699937820435, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4282, | |
| "step": 539 | |
| }, | |
| { | |
| "epoch": 2.3478260869565215, | |
| "grad_norm": 0.3195163607597351, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4068, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.3521739130434782, | |
| "grad_norm": 0.3195163607597351, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3813, | |
| "step": 541 | |
| }, | |
| { | |
| "epoch": 2.356521739130435, | |
| "grad_norm": 0.3195163607597351, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4368, | |
| "step": 542 | |
| }, | |
| { | |
| "epoch": 2.360869565217391, | |
| "grad_norm": 0.3195163607597351, | |
| "learning_rate": 2e-05, | |
| "loss": 1.393, | |
| "step": 543 | |
| }, | |
| { | |
| "epoch": 2.365217391304348, | |
| "grad_norm": 0.31394919753074646, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4161, | |
| "step": 544 | |
| }, | |
| { | |
| "epoch": 2.369565217391304, | |
| "grad_norm": 0.31394919753074646, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4387, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.373913043478261, | |
| "grad_norm": 0.31394919753074646, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3624, | |
| "step": 546 | |
| }, | |
| { | |
| "epoch": 2.378260869565217, | |
| "grad_norm": 0.31394919753074646, | |
| "learning_rate": 2e-05, | |
| "loss": 1.452, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.382608695652174, | |
| "grad_norm": 0.3255818784236908, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4097, | |
| "step": 548 | |
| }, | |
| { | |
| "epoch": 2.3869565217391306, | |
| "grad_norm": 0.3255818784236908, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3715, | |
| "step": 549 | |
| }, | |
| { | |
| "epoch": 2.391304347826087, | |
| "grad_norm": 0.3255818784236908, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4356, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.3956521739130436, | |
| "grad_norm": 0.3255818784236908, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4727, | |
| "step": 551 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.34206336736679077, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4212, | |
| "step": 552 | |
| }, | |
| { | |
| "epoch": 2.4043478260869566, | |
| "grad_norm": 0.34206336736679077, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3371, | |
| "step": 553 | |
| }, | |
| { | |
| "epoch": 2.408695652173913, | |
| "grad_norm": 0.34206336736679077, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3715, | |
| "step": 554 | |
| }, | |
| { | |
| "epoch": 2.4130434782608696, | |
| "grad_norm": 0.34206336736679077, | |
| "learning_rate": 2e-05, | |
| "loss": 1.407, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.417391304347826, | |
| "grad_norm": 0.32757705450057983, | |
| "learning_rate": 2e-05, | |
| "loss": 1.404, | |
| "step": 556 | |
| }, | |
| { | |
| "epoch": 2.4217391304347826, | |
| "grad_norm": 0.32757705450057983, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3585, | |
| "step": 557 | |
| }, | |
| { | |
| "epoch": 2.426086956521739, | |
| "grad_norm": 0.32757705450057983, | |
| "learning_rate": 2e-05, | |
| "loss": 1.39, | |
| "step": 558 | |
| }, | |
| { | |
| "epoch": 2.4304347826086956, | |
| "grad_norm": 0.32757705450057983, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4128, | |
| "step": 559 | |
| }, | |
| { | |
| "epoch": 2.4347826086956523, | |
| "grad_norm": 0.33430787920951843, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4144, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.4391304347826086, | |
| "grad_norm": 0.33430787920951843, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4206, | |
| "step": 561 | |
| }, | |
| { | |
| "epoch": 2.4434782608695653, | |
| "grad_norm": 0.33430787920951843, | |
| "learning_rate": 2e-05, | |
| "loss": 1.374, | |
| "step": 562 | |
| }, | |
| { | |
| "epoch": 2.4478260869565216, | |
| "grad_norm": 0.33430787920951843, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4283, | |
| "step": 563 | |
| }, | |
| { | |
| "epoch": 2.4521739130434783, | |
| "grad_norm": 0.3351895213127136, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3168, | |
| "step": 564 | |
| }, | |
| { | |
| "epoch": 2.4565217391304346, | |
| "grad_norm": 0.3351895213127136, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3109, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.4608695652173913, | |
| "grad_norm": 0.3351895213127136, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4633, | |
| "step": 566 | |
| }, | |
| { | |
| "epoch": 2.465217391304348, | |
| "grad_norm": 0.3351895213127136, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3754, | |
| "step": 567 | |
| }, | |
| { | |
| "epoch": 2.4695652173913043, | |
| "grad_norm": 0.31671515107154846, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4487, | |
| "step": 568 | |
| }, | |
| { | |
| "epoch": 2.473913043478261, | |
| "grad_norm": 0.31671515107154846, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4277, | |
| "step": 569 | |
| }, | |
| { | |
| "epoch": 2.4782608695652173, | |
| "grad_norm": 0.31671515107154846, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3566, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.482608695652174, | |
| "grad_norm": 0.31671515107154846, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4195, | |
| "step": 571 | |
| }, | |
| { | |
| "epoch": 2.4869565217391303, | |
| "grad_norm": 0.32078683376312256, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4236, | |
| "step": 572 | |
| }, | |
| { | |
| "epoch": 2.491304347826087, | |
| "grad_norm": 0.32078683376312256, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3728, | |
| "step": 573 | |
| }, | |
| { | |
| "epoch": 2.4956521739130437, | |
| "grad_norm": 0.32078683376312256, | |
| "learning_rate": 2e-05, | |
| "loss": 1.367, | |
| "step": 574 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 0.32078683376312256, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3892, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.5043478260869563, | |
| "grad_norm": 0.3213704228401184, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4073, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 2.508695652173913, | |
| "grad_norm": 0.3213704228401184, | |
| "learning_rate": 2e-05, | |
| "loss": 1.38, | |
| "step": 577 | |
| }, | |
| { | |
| "epoch": 2.5130434782608697, | |
| "grad_norm": 0.3213704228401184, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4, | |
| "step": 578 | |
| }, | |
| { | |
| "epoch": 2.517391304347826, | |
| "grad_norm": 0.3213704228401184, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3417, | |
| "step": 579 | |
| }, | |
| { | |
| "epoch": 2.5217391304347827, | |
| "grad_norm": 0.33939409255981445, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4088, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.526086956521739, | |
| "grad_norm": 0.33939409255981445, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3889, | |
| "step": 581 | |
| }, | |
| { | |
| "epoch": 2.5304347826086957, | |
| "grad_norm": 0.33939409255981445, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3547, | |
| "step": 582 | |
| }, | |
| { | |
| "epoch": 2.534782608695652, | |
| "grad_norm": 0.33939409255981445, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3643, | |
| "step": 583 | |
| }, | |
| { | |
| "epoch": 2.5391304347826087, | |
| "grad_norm": 0.362350732088089, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3045, | |
| "step": 584 | |
| }, | |
| { | |
| "epoch": 2.5434782608695654, | |
| "grad_norm": 0.362350732088089, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4226, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 2.5478260869565217, | |
| "grad_norm": 0.362350732088089, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3398, | |
| "step": 586 | |
| }, | |
| { | |
| "epoch": 2.5521739130434784, | |
| "grad_norm": 0.362350732088089, | |
| "learning_rate": 2e-05, | |
| "loss": 1.434, | |
| "step": 587 | |
| }, | |
| { | |
| "epoch": 2.5565217391304347, | |
| "grad_norm": 0.34566718339920044, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3933, | |
| "step": 588 | |
| }, | |
| { | |
| "epoch": 2.5608695652173914, | |
| "grad_norm": 0.34566718339920044, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3881, | |
| "step": 589 | |
| }, | |
| { | |
| "epoch": 2.5652173913043477, | |
| "grad_norm": 0.34566718339920044, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3688, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.5695652173913044, | |
| "grad_norm": 0.34566718339920044, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4052, | |
| "step": 591 | |
| }, | |
| { | |
| "epoch": 2.573913043478261, | |
| "grad_norm": 0.3323229253292084, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4094, | |
| "step": 592 | |
| }, | |
| { | |
| "epoch": 2.5782608695652174, | |
| "grad_norm": 0.3323229253292084, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3552, | |
| "step": 593 | |
| }, | |
| { | |
| "epoch": 2.5826086956521737, | |
| "grad_norm": 0.3323229253292084, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4367, | |
| "step": 594 | |
| }, | |
| { | |
| "epoch": 2.5869565217391304, | |
| "grad_norm": 0.3323229253292084, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3934, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 2.591304347826087, | |
| "grad_norm": 0.32952815294265747, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3602, | |
| "step": 596 | |
| }, | |
| { | |
| "epoch": 2.5956521739130434, | |
| "grad_norm": 0.32952815294265747, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3772, | |
| "step": 597 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.32952815294265747, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4432, | |
| "step": 598 | |
| }, | |
| { | |
| "epoch": 2.6043478260869564, | |
| "grad_norm": 0.32952815294265747, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4353, | |
| "step": 599 | |
| }, | |
| { | |
| "epoch": 2.608695652173913, | |
| "grad_norm": 0.3243506848812103, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3708, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.6130434782608694, | |
| "grad_norm": 0.3243506848812103, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3407, | |
| "step": 601 | |
| }, | |
| { | |
| "epoch": 2.617391304347826, | |
| "grad_norm": 0.3243506848812103, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3312, | |
| "step": 602 | |
| }, | |
| { | |
| "epoch": 2.621739130434783, | |
| "grad_norm": 0.3243506848812103, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3443, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 2.626086956521739, | |
| "grad_norm": 0.31402692198753357, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3297, | |
| "step": 604 | |
| }, | |
| { | |
| "epoch": 2.630434782608696, | |
| "grad_norm": 0.31402692198753357, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3604, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 2.634782608695652, | |
| "grad_norm": 0.31402692198753357, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3888, | |
| "step": 606 | |
| }, | |
| { | |
| "epoch": 2.639130434782609, | |
| "grad_norm": 0.31402692198753357, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3901, | |
| "step": 607 | |
| }, | |
| { | |
| "epoch": 2.643478260869565, | |
| "grad_norm": 0.3442557156085968, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3481, | |
| "step": 608 | |
| }, | |
| { | |
| "epoch": 2.6478260869565218, | |
| "grad_norm": 0.3442557156085968, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3451, | |
| "step": 609 | |
| }, | |
| { | |
| "epoch": 2.6521739130434785, | |
| "grad_norm": 0.3442557156085968, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3634, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.6565217391304348, | |
| "grad_norm": 0.3442557156085968, | |
| "learning_rate": 2e-05, | |
| "loss": 1.322, | |
| "step": 611 | |
| }, | |
| { | |
| "epoch": 2.660869565217391, | |
| "grad_norm": 0.33002713322639465, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3747, | |
| "step": 612 | |
| }, | |
| { | |
| "epoch": 2.6652173913043478, | |
| "grad_norm": 0.33002713322639465, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4328, | |
| "step": 613 | |
| }, | |
| { | |
| "epoch": 2.6695652173913045, | |
| "grad_norm": 0.33002713322639465, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3775, | |
| "step": 614 | |
| }, | |
| { | |
| "epoch": 2.6739130434782608, | |
| "grad_norm": 0.33002713322639465, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3485, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 2.6782608695652175, | |
| "grad_norm": 0.32511940598487854, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4626, | |
| "step": 616 | |
| }, | |
| { | |
| "epoch": 2.6826086956521737, | |
| "grad_norm": 0.32511940598487854, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3966, | |
| "step": 617 | |
| }, | |
| { | |
| "epoch": 2.6869565217391305, | |
| "grad_norm": 0.32511940598487854, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4073, | |
| "step": 618 | |
| }, | |
| { | |
| "epoch": 2.6913043478260867, | |
| "grad_norm": 0.32511940598487854, | |
| "learning_rate": 2e-05, | |
| "loss": 1.312, | |
| "step": 619 | |
| }, | |
| { | |
| "epoch": 2.6956521739130435, | |
| "grad_norm": 0.3344216048717499, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4085, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "grad_norm": 0.3344216048717499, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4103, | |
| "step": 621 | |
| }, | |
| { | |
| "epoch": 2.7043478260869565, | |
| "grad_norm": 0.3344216048717499, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3982, | |
| "step": 622 | |
| }, | |
| { | |
| "epoch": 2.708695652173913, | |
| "grad_norm": 0.3344216048717499, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4319, | |
| "step": 623 | |
| }, | |
| { | |
| "epoch": 2.7130434782608694, | |
| "grad_norm": 0.33260592818260193, | |
| "learning_rate": 2e-05, | |
| "loss": 1.407, | |
| "step": 624 | |
| }, | |
| { | |
| "epoch": 2.717391304347826, | |
| "grad_norm": 0.33260592818260193, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3397, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 2.7217391304347824, | |
| "grad_norm": 0.33260592818260193, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3808, | |
| "step": 626 | |
| }, | |
| { | |
| "epoch": 2.726086956521739, | |
| "grad_norm": 0.33260592818260193, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4175, | |
| "step": 627 | |
| }, | |
| { | |
| "epoch": 2.730434782608696, | |
| "grad_norm": 0.33184948563575745, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3604, | |
| "step": 628 | |
| }, | |
| { | |
| "epoch": 2.734782608695652, | |
| "grad_norm": 0.33184948563575745, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4572, | |
| "step": 629 | |
| }, | |
| { | |
| "epoch": 2.7391304347826084, | |
| "grad_norm": 0.33184948563575745, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3895, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.743478260869565, | |
| "grad_norm": 0.33184948563575745, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4055, | |
| "step": 631 | |
| }, | |
| { | |
| "epoch": 2.747826086956522, | |
| "grad_norm": 0.32070547342300415, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3849, | |
| "step": 632 | |
| }, | |
| { | |
| "epoch": 2.752173913043478, | |
| "grad_norm": 0.32070547342300415, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3835, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 2.756521739130435, | |
| "grad_norm": 0.32070547342300415, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4008, | |
| "step": 634 | |
| }, | |
| { | |
| "epoch": 2.7608695652173916, | |
| "grad_norm": 0.32070547342300415, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3944, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 2.765217391304348, | |
| "grad_norm": 0.32883745431900024, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3374, | |
| "step": 636 | |
| }, | |
| { | |
| "epoch": 2.769565217391304, | |
| "grad_norm": 0.32883745431900024, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3797, | |
| "step": 637 | |
| }, | |
| { | |
| "epoch": 2.773913043478261, | |
| "grad_norm": 0.32883745431900024, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3135, | |
| "step": 638 | |
| }, | |
| { | |
| "epoch": 2.7782608695652176, | |
| "grad_norm": 0.32883745431900024, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3744, | |
| "step": 639 | |
| }, | |
| { | |
| "epoch": 2.782608695652174, | |
| "grad_norm": 0.3163747489452362, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3676, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.7869565217391306, | |
| "grad_norm": 0.3163747489452362, | |
| "learning_rate": 2e-05, | |
| "loss": 1.392, | |
| "step": 641 | |
| }, | |
| { | |
| "epoch": 2.791304347826087, | |
| "grad_norm": 0.3163747489452362, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3777, | |
| "step": 642 | |
| }, | |
| { | |
| "epoch": 2.7956521739130435, | |
| "grad_norm": 0.3163747489452362, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4129, | |
| "step": 643 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.33478814363479614, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3628, | |
| "step": 644 | |
| }, | |
| { | |
| "epoch": 2.8043478260869565, | |
| "grad_norm": 0.33478814363479614, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3612, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 2.8086956521739133, | |
| "grad_norm": 0.33478814363479614, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4229, | |
| "step": 646 | |
| }, | |
| { | |
| "epoch": 2.8130434782608695, | |
| "grad_norm": 0.33478814363479614, | |
| "learning_rate": 2e-05, | |
| "loss": 1.355, | |
| "step": 647 | |
| }, | |
| { | |
| "epoch": 2.8173913043478263, | |
| "grad_norm": 0.3139325678348541, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3561, | |
| "step": 648 | |
| }, | |
| { | |
| "epoch": 2.8217391304347825, | |
| "grad_norm": 0.3139325678348541, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4152, | |
| "step": 649 | |
| }, | |
| { | |
| "epoch": 2.8260869565217392, | |
| "grad_norm": 0.3139325678348541, | |
| "learning_rate": 2e-05, | |
| "loss": 1.423, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.8304347826086955, | |
| "grad_norm": 0.3139325678348541, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3525, | |
| "step": 651 | |
| }, | |
| { | |
| "epoch": 2.8347826086956522, | |
| "grad_norm": 0.3335968554019928, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3864, | |
| "step": 652 | |
| }, | |
| { | |
| "epoch": 2.839130434782609, | |
| "grad_norm": 0.3335968554019928, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4314, | |
| "step": 653 | |
| }, | |
| { | |
| "epoch": 2.8434782608695652, | |
| "grad_norm": 0.3335968554019928, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4349, | |
| "step": 654 | |
| }, | |
| { | |
| "epoch": 2.8478260869565215, | |
| "grad_norm": 0.3335968554019928, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3767, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 2.8521739130434782, | |
| "grad_norm": 0.3227713406085968, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3651, | |
| "step": 656 | |
| }, | |
| { | |
| "epoch": 2.856521739130435, | |
| "grad_norm": 0.3227713406085968, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3297, | |
| "step": 657 | |
| }, | |
| { | |
| "epoch": 2.860869565217391, | |
| "grad_norm": 0.3227713406085968, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4148, | |
| "step": 658 | |
| }, | |
| { | |
| "epoch": 2.865217391304348, | |
| "grad_norm": 0.3227713406085968, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4048, | |
| "step": 659 | |
| }, | |
| { | |
| "epoch": 2.869565217391304, | |
| "grad_norm": 0.3265250325202942, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3643, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.873913043478261, | |
| "grad_norm": 0.3265250325202942, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4123, | |
| "step": 661 | |
| }, | |
| { | |
| "epoch": 2.878260869565217, | |
| "grad_norm": 0.3265250325202942, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4146, | |
| "step": 662 | |
| }, | |
| { | |
| "epoch": 2.882608695652174, | |
| "grad_norm": 0.3265250325202942, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4065, | |
| "step": 663 | |
| }, | |
| { | |
| "epoch": 2.8869565217391306, | |
| "grad_norm": 0.30972209572792053, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3875, | |
| "step": 664 | |
| }, | |
| { | |
| "epoch": 2.891304347826087, | |
| "grad_norm": 0.30972209572792053, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4828, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 2.8956521739130436, | |
| "grad_norm": 0.30972209572792053, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4318, | |
| "step": 666 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.30972209572792053, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4493, | |
| "step": 667 | |
| }, | |
| { | |
| "epoch": 2.9043478260869566, | |
| "grad_norm": 0.31576821208000183, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3041, | |
| "step": 668 | |
| }, | |
| { | |
| "epoch": 2.908695652173913, | |
| "grad_norm": 0.31576821208000183, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4027, | |
| "step": 669 | |
| }, | |
| { | |
| "epoch": 2.9130434782608696, | |
| "grad_norm": 0.31576821208000183, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4331, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.9173913043478263, | |
| "grad_norm": 0.31576821208000183, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3375, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 2.9217391304347826, | |
| "grad_norm": 0.3335813581943512, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3935, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 2.926086956521739, | |
| "grad_norm": 0.3335813581943512, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3927, | |
| "step": 673 | |
| }, | |
| { | |
| "epoch": 2.9304347826086956, | |
| "grad_norm": 0.3335813581943512, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3892, | |
| "step": 674 | |
| }, | |
| { | |
| "epoch": 2.9347826086956523, | |
| "grad_norm": 0.3335813581943512, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3488, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 2.9391304347826086, | |
| "grad_norm": 0.3342477083206177, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3527, | |
| "step": 676 | |
| }, | |
| { | |
| "epoch": 2.9434782608695653, | |
| "grad_norm": 0.3342477083206177, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4115, | |
| "step": 677 | |
| }, | |
| { | |
| "epoch": 2.9478260869565216, | |
| "grad_norm": 0.3342477083206177, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3586, | |
| "step": 678 | |
| }, | |
| { | |
| "epoch": 2.9521739130434783, | |
| "grad_norm": 0.3342477083206177, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4366, | |
| "step": 679 | |
| }, | |
| { | |
| "epoch": 2.9565217391304346, | |
| "grad_norm": 0.32498446106910706, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3317, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.9608695652173913, | |
| "grad_norm": 0.32498446106910706, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4396, | |
| "step": 681 | |
| }, | |
| { | |
| "epoch": 2.965217391304348, | |
| "grad_norm": 0.32498446106910706, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3385, | |
| "step": 682 | |
| }, | |
| { | |
| "epoch": 2.9695652173913043, | |
| "grad_norm": 0.32498446106910706, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3554, | |
| "step": 683 | |
| }, | |
| { | |
| "epoch": 2.973913043478261, | |
| "grad_norm": 0.3209493160247803, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3452, | |
| "step": 684 | |
| }, | |
| { | |
| "epoch": 2.9782608695652173, | |
| "grad_norm": 0.3209493160247803, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3912, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 2.982608695652174, | |
| "grad_norm": 0.3209493160247803, | |
| "learning_rate": 2e-05, | |
| "loss": 1.4286, | |
| "step": 686 | |
| }, | |
| { | |
| "epoch": 2.9869565217391303, | |
| "grad_norm": 0.3209493160247803, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3302, | |
| "step": 687 | |
| }, | |
| { | |
| "epoch": 2.991304347826087, | |
| "grad_norm": 0.3104861378669739, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3402, | |
| "step": 688 | |
| }, | |
| { | |
| "epoch": 2.9956521739130437, | |
| "grad_norm": 0.3104861378669739, | |
| "learning_rate": 2e-05, | |
| "loss": 1.328, | |
| "step": 689 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.3104861378669739, | |
| "learning_rate": 2e-05, | |
| "loss": 1.3663, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 690, | |
| "total_flos": 3.864388766479155e+17, | |
| "train_loss": 1.544883234086244, | |
| "train_runtime": 425.3662, | |
| "train_samples_per_second": 363.082, | |
| "train_steps_per_second": 1.622 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 690, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 3000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.864388766479155e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |